diff --git "a/speech_lora/trainer_state.json" "b/speech_lora/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/speech_lora/trainer_state.json"
@@ -0,0 +1,97153 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 13873,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 7.208246233691343e-05,
+      "grad_norm": 8.076401867860572,
+      "learning_rate": 4.796163069544364e-07,
+      "loss": 3.8029,
+      "step": 1
+    },
+    {
+      "epoch": 0.00014416492467382686,
+      "grad_norm": 9.477582136019377,
+      "learning_rate": 9.592326139088728e-07,
+      "loss": 3.7361,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002162473870107403,
+      "grad_norm": 9.754172971640974,
+      "learning_rate": 1.4388489208633094e-06,
+      "loss": 3.9542,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002883298493476537,
+      "grad_norm": 9.412299821100705,
+      "learning_rate": 1.9184652278177456e-06,
+      "loss": 3.8848,
+      "step": 4
+    },
+    {
+      "epoch": 0.00036041231168456716,
+      "grad_norm": 9.39521205688099,
+      "learning_rate": 2.3980815347721824e-06,
+      "loss": 3.6531,
+      "step": 5
+    },
+    {
+      "epoch": 0.0004324947740214806,
+      "grad_norm": 7.704079966694358,
+      "learning_rate": 2.877697841726619e-06,
+      "loss": 3.7944,
+      "step": 6
+    },
+    {
+      "epoch": 0.000504577236358394,
+      "grad_norm": 9.885660519582697,
+      "learning_rate": 3.3573141486810552e-06,
+      "loss": 3.6807,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005766596986953074,
+      "grad_norm": 8.526931675646148,
+      "learning_rate": 3.836930455635491e-06,
+      "loss": 3.7186,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006487421610322209,
+      "grad_norm": 8.984983121340788,
+      "learning_rate": 4.316546762589929e-06,
+      "loss": 3.7575,
+      "step": 9
+    },
+    {
+      "epoch": 0.0007208246233691343,
+      "grad_norm": 7.8840109642328855,
+      "learning_rate": 4.796163069544365e-06,
+      "loss": 3.7397,
+      "step": 10
+    },
+    {
+      "epoch": 0.0007929070857060478,
+      "grad_norm": 9.143615379284894,
+      "learning_rate": 5.275779376498801e-06,
+      "loss": 3.6664,
+      "step": 11
+    },
+    {
+      "epoch": 0.0008649895480429612,
+      "grad_norm": 8.348142338716908,
+      "learning_rate": 5.755395683453238e-06,
+      "loss": 3.7504,
+      "step": 12
+    },
+    {
+      "epoch": 0.0009370720103798745,
+      "grad_norm": 6.770722611008024,
+      "learning_rate": 6.2350119904076745e-06,
+      "loss": 3.5191,
+      "step": 13
+    },
+    {
+      "epoch": 0.001009154472716788,
+      "grad_norm": 7.353234736589995,
+      "learning_rate": 6.7146282973621105e-06,
+      "loss": 3.6281,
+      "step": 14
+    },
+    {
+      "epoch": 0.0010812369350537015,
+      "grad_norm": 7.11875753214028,
+      "learning_rate": 7.1942446043165465e-06,
+      "loss": 3.3397,
+      "step": 15
+    },
+    {
+      "epoch": 0.0011533193973906149,
+      "grad_norm": 6.976809808098644,
+      "learning_rate": 7.673860911270982e-06,
+      "loss": 3.3978,
+      "step": 16
+    },
+    {
+      "epoch": 0.0012254018597275282,
+      "grad_norm": 6.271434098292993,
+      "learning_rate": 8.153477218225421e-06,
+      "loss": 3.3922,
+      "step": 17
+    },
+    {
+      "epoch": 0.0012974843220644417,
+      "grad_norm": 4.59740532182426,
+      "learning_rate": 8.633093525179858e-06,
+      "loss": 3.1189,
+      "step": 18
+    },
+    {
+      "epoch": 0.001369566784401355,
+      "grad_norm": 4.132005281950382,
+      "learning_rate": 9.112709832134293e-06,
+      "loss": 3.0424,
+      "step": 19
+    },
+    {
+      "epoch": 0.0014416492467382686,
+      "grad_norm": 4.009564448157672,
+      "learning_rate": 9.59232613908873e-06,
+      "loss": 3.0544,
+      "step": 20
+    },
+    {
+      "epoch": 0.001513731709075182,
+      "grad_norm": 3.7374464178207893,
+      "learning_rate": 1.0071942446043167e-05,
+      "loss": 3.0165,
+      "step": 21
+    },
+    {
+      "epoch": 0.0015858141714120955,
+      "grad_norm": 3.4053591640608545,
+      "learning_rate": 1.0551558752997602e-05,
+      "loss": 3.0463,
+      "step": 22
+    },
+    {
+      "epoch": 0.0016578966337490088,
+      "grad_norm": 3.065212447950743,
+      "learning_rate": 1.1031175059952039e-05,
+      "loss": 2.9404,
+      "step": 23
+    },
+    {
+      "epoch": 0.0017299790960859224,
+      "grad_norm": 3.105829880910362,
+      "learning_rate": 1.1510791366906475e-05,
+      "loss": 2.9071,
+      "step": 24
+    },
+    {
+      "epoch": 0.0018020615584228357,
+      "grad_norm": 2.9579466444901144,
+      "learning_rate": 1.1990407673860912e-05,
+      "loss": 2.9676,
+      "step": 25
+    },
+    {
+      "epoch": 0.001874144020759749,
+      "grad_norm": 2.389734595833505,
+      "learning_rate": 1.2470023980815349e-05,
+      "loss": 2.7008,
+      "step": 26
+    },
+    {
+      "epoch": 0.0019462264830966626,
+      "grad_norm": 2.6119347461513707,
+      "learning_rate": 1.2949640287769784e-05,
+      "loss": 2.72,
+      "step": 27
+    },
+    {
+      "epoch": 0.002018308945433576,
+      "grad_norm": 2.4901955493069963,
+      "learning_rate": 1.3429256594724221e-05,
+      "loss": 2.7133,
+      "step": 28
+    },
+    {
+      "epoch": 0.0020903914077704893,
+      "grad_norm": 2.350277518054029,
+      "learning_rate": 1.3908872901678658e-05,
+      "loss": 2.7372,
+      "step": 29
+    },
+    {
+      "epoch": 0.002162473870107403,
+      "grad_norm": 2.168815983905342,
+      "learning_rate": 1.4388489208633093e-05,
+      "loss": 2.7177,
+      "step": 30
+    },
+    {
+      "epoch": 0.0022345563324443164,
+      "grad_norm": 2.203724914220553,
+      "learning_rate": 1.486810551558753e-05,
+      "loss": 2.7342,
+      "step": 31
+    },
+    {
+      "epoch": 0.0023066387947812297,
+      "grad_norm": 1.8249370054648377,
+      "learning_rate": 1.5347721822541965e-05,
+      "loss": 2.4974,
+      "step": 32
+    },
+    {
+      "epoch": 0.002378721257118143,
+      "grad_norm": 1.8382571314833351,
+      "learning_rate": 1.5827338129496403e-05,
+      "loss": 2.5053,
+      "step": 33
+    },
+    {
+      "epoch": 0.0024508037194550564,
+      "grad_norm": 1.808620626675934,
+      "learning_rate": 1.6306954436450842e-05,
+      "loss": 2.4669,
+      "step": 34
+    },
+    {
+      "epoch": 0.00252288618179197,
+      "grad_norm": 1.8086789625780415,
+      "learning_rate": 1.6786570743405277e-05,
+      "loss": 2.4337,
+      "step": 35
+    },
+    {
+      "epoch": 0.0025949686441288835,
+      "grad_norm": 1.781230853346726,
+      "learning_rate": 1.7266187050359716e-05,
+      "loss": 2.6403,
+      "step": 36
+    },
+    {
+      "epoch": 0.002667051106465797,
+      "grad_norm": 1.5900513472308737,
+      "learning_rate": 1.774580335731415e-05,
+      "loss": 2.6565,
+      "step": 37
+    },
+    {
+      "epoch": 0.00273913356880271,
+      "grad_norm": 1.602381788616175,
+      "learning_rate": 1.8225419664268586e-05,
+      "loss": 2.4653,
+      "step": 38
+    },
+    {
+      "epoch": 0.002811216031139624,
+      "grad_norm": 1.5224618507918921,
+      "learning_rate": 1.8705035971223024e-05,
+      "loss": 2.3917,
+      "step": 39
+    },
+    {
+      "epoch": 0.0028832984934765373,
+      "grad_norm": 1.4102575600065401,
+      "learning_rate": 1.918465227817746e-05,
+      "loss": 2.3662,
+      "step": 40
+    },
+    {
+      "epoch": 0.0029553809558134506,
+      "grad_norm": 1.3894044609827687,
+      "learning_rate": 1.9664268585131895e-05,
+      "loss": 2.3197,
+      "step": 41
+    },
+    {
+      "epoch": 0.003027463418150364,
+      "grad_norm": 1.2808816147028577,
+      "learning_rate": 2.0143884892086333e-05,
+      "loss": 2.236,
+      "step": 42
+    },
+    {
+      "epoch": 0.0030995458804872773,
+      "grad_norm": 1.207895530242993,
+      "learning_rate": 2.062350119904077e-05,
+      "loss": 2.1543,
+      "step": 43
+    },
+    {
+      "epoch": 0.003171628342824191,
+      "grad_norm": 1.4774981105063512,
+      "learning_rate": 2.1103117505995203e-05,
+      "loss": 2.3613,
+      "step": 44
+    },
+    {
+      "epoch": 0.0032437108051611044,
+      "grad_norm": 1.3174620265183248,
+      "learning_rate": 2.1582733812949642e-05,
+      "loss": 2.2706,
+      "step": 45
+    },
+    {
+      "epoch": 0.0033157932674980177,
+      "grad_norm": 1.2435380476151985,
+      "learning_rate": 2.2062350119904077e-05,
+      "loss": 2.2614,
+      "step": 46
+    },
+    {
+      "epoch": 0.003387875729834931,
+      "grad_norm": 1.4943145508531717,
+      "learning_rate": 2.2541966426858512e-05,
+      "loss": 2.3343,
+      "step": 47
+    },
+    {
+      "epoch": 0.003459958192171845,
+      "grad_norm": 1.0888655941860528,
+      "learning_rate": 2.302158273381295e-05,
+      "loss": 2.1615,
+      "step": 48
+    },
+    {
+      "epoch": 0.003532040654508758,
+      "grad_norm": 1.1141696669907306,
+      "learning_rate": 2.350119904076739e-05,
+      "loss": 2.2171,
+      "step": 49
+    },
+    {
+      "epoch": 0.0036041231168456715,
+      "grad_norm": 1.1719972368372877,
+      "learning_rate": 2.3980815347721824e-05,
+      "loss": 2.0201,
+      "step": 50
+    },
+    {
+      "epoch": 0.003676205579182585,
+      "grad_norm": 1.1329430089040866,
+      "learning_rate": 2.446043165467626e-05,
+      "loss": 2.133,
+      "step": 51
+    },
+    {
+      "epoch": 0.003748288041519498,
+      "grad_norm": 1.0897141912502468,
+      "learning_rate": 2.4940047961630698e-05,
+      "loss": 2.0527,
+      "step": 52
+    },
+    {
+      "epoch": 0.003820370503856412,
+      "grad_norm": 0.9082517668220094,
+      "learning_rate": 2.541966426858513e-05,
+      "loss": 2.0409,
+      "step": 53
+    },
+    {
+      "epoch": 0.0038924529661933252,
+      "grad_norm": 1.0951144439190348,
+      "learning_rate": 2.589928057553957e-05,
+      "loss": 1.9687,
+      "step": 54
+    },
+    {
+      "epoch": 0.003964535428530239,
+      "grad_norm": 0.96744008415966,
+      "learning_rate": 2.637889688249401e-05,
+      "loss": 2.1128,
+      "step": 55
+    },
+    {
+      "epoch": 0.004036617890867152,
+      "grad_norm": 1.1004161581494876,
+      "learning_rate": 2.6858513189448442e-05,
+      "loss": 2.0528,
+      "step": 56
+    },
+    {
+      "epoch": 0.004108700353204066,
+      "grad_norm": 1.0147898221123464,
+      "learning_rate": 2.733812949640288e-05,
+      "loss": 2.013,
+      "step": 57
+    },
+    {
+      "epoch": 0.004180782815540979,
+      "grad_norm": 0.9250625295622429,
+      "learning_rate": 2.7817745803357316e-05,
+      "loss": 1.9633,
+      "step": 58
+    },
+    {
+      "epoch": 0.004252865277877892,
+      "grad_norm": 1.0204220998559848,
+      "learning_rate": 2.8297362110311754e-05,
+      "loss": 2.0753,
+      "step": 59
+    },
+    {
+      "epoch": 0.004324947740214806,
+      "grad_norm": 0.8467047333428357,
+      "learning_rate": 2.8776978417266186e-05,
+      "loss": 1.9028,
+      "step": 60
+    },
+    {
+      "epoch": 0.004397030202551719,
+      "grad_norm": 0.8392271765705351,
+      "learning_rate": 2.9256594724220628e-05,
+      "loss": 1.9585,
+      "step": 61
+    },
+    {
+      "epoch": 0.004469112664888633,
+      "grad_norm": 0.7307665593539366,
+      "learning_rate": 2.973621103117506e-05,
+      "loss": 1.8974,
+      "step": 62
+    },
+    {
+      "epoch": 0.004541195127225546,
+      "grad_norm": 1.0976978974803557,
+      "learning_rate": 3.0215827338129498e-05,
+      "loss": 2.1287,
+      "step": 63
+    },
+    {
+      "epoch": 0.0046132775895624594,
+      "grad_norm": 1.0181370987075447,
+      "learning_rate": 3.069544364508393e-05,
+      "loss": 2.138,
+      "step": 64
+    },
+    {
+      "epoch": 0.004685360051899373,
+      "grad_norm": 0.7545334207465467,
+      "learning_rate": 3.117505995203837e-05,
+      "loss": 1.8218,
+      "step": 65
+    },
+    {
+      "epoch": 0.004757442514236286,
+      "grad_norm": 0.9059703596110401,
+      "learning_rate": 3.165467625899281e-05,
+      "loss": 1.8971,
+      "step": 66
+    },
+    {
+      "epoch": 0.0048295249765732,
+      "grad_norm": 0.775033916548142,
+      "learning_rate": 3.2134292565947245e-05,
+      "loss": 1.9129,
+      "step": 67
+    },
+    {
+      "epoch": 0.004901607438910113,
+      "grad_norm": 0.81774413801343,
+      "learning_rate": 3.2613908872901684e-05,
+      "loss": 1.9129,
+      "step": 68
+    },
+    {
+      "epoch": 0.0049736899012470265,
+      "grad_norm": 0.7220989131626805,
+      "learning_rate": 3.3093525179856116e-05,
+      "loss": 1.8794,
+      "step": 69
+    },
+    {
+      "epoch": 0.00504577236358394,
+      "grad_norm": 0.8432029716949566,
+      "learning_rate": 3.3573141486810554e-05,
+      "loss": 1.8648,
+      "step": 70
+    },
+    {
+      "epoch": 0.005117854825920853,
+      "grad_norm": 0.7469658524015326,
+      "learning_rate": 3.4052757793764986e-05,
+      "loss": 1.8349,
+      "step": 71
+    },
+    {
+      "epoch": 0.005189937288257767,
+      "grad_norm": 0.7030578427633872,
+      "learning_rate": 3.453237410071943e-05,
+      "loss": 1.8209,
+      "step": 72
+    },
+    {
+      "epoch": 0.00526201975059468,
+      "grad_norm": 0.8354427441685783,
+      "learning_rate": 3.501199040767386e-05,
+      "loss": 1.909,
+      "step": 73
+    },
+    {
+      "epoch": 0.005334102212931594,
+      "grad_norm": 0.7895358421701308,
+      "learning_rate": 3.54916067146283e-05,
+      "loss": 1.8588,
+      "step": 74
+    },
+    {
+      "epoch": 0.005406184675268507,
+      "grad_norm": 0.8912839001025953,
+      "learning_rate": 3.597122302158273e-05,
+      "loss": 1.8584,
+      "step": 75
+    },
+    {
+      "epoch": 0.00547826713760542,
+      "grad_norm": 1.0158555814141412,
+      "learning_rate": 3.645083932853717e-05,
+      "loss": 1.98,
+      "step": 76
+    },
+    {
+      "epoch": 0.005550349599942334,
+      "grad_norm": 0.735480577270388,
+      "learning_rate": 3.6930455635491604e-05,
+      "loss": 1.7502,
+      "step": 77
+    },
+    {
+      "epoch": 0.005622432062279248,
+      "grad_norm": 0.7004434492495069,
+      "learning_rate": 3.741007194244605e-05,
+      "loss": 1.8191,
+      "step": 78
+    },
+    {
+      "epoch": 0.005694514524616161,
+      "grad_norm": 0.9720370226503946,
+      "learning_rate": 3.788968824940048e-05,
+      "loss": 1.8191,
+      "step": 79
+    },
+    {
+      "epoch": 0.0057665969869530745,
+      "grad_norm": 0.8432984469518545,
+      "learning_rate": 3.836930455635492e-05,
+      "loss": 1.8214,
+      "step": 80
+    },
+    {
+      "epoch": 0.005838679449289987,
+      "grad_norm": 0.8848995893915009,
+      "learning_rate": 3.884892086330936e-05,
+      "loss": 1.8282,
+      "step": 81
+    },
+    {
+      "epoch": 0.005910761911626901,
+      "grad_norm": 0.6739517683251914,
+      "learning_rate": 3.932853717026379e-05,
+      "loss": 1.6473,
+      "step": 82
+    },
+    {
+      "epoch": 0.005982844373963815,
+      "grad_norm": 0.728483190452508,
+      "learning_rate": 3.980815347721823e-05,
+      "loss": 1.7713,
+      "step": 83
+    },
+    {
+      "epoch": 0.006054926836300728,
+      "grad_norm": 0.6173814265445161,
+      "learning_rate": 4.0287769784172666e-05,
+      "loss": 1.8257,
+      "step": 84
+    },
+    {
+      "epoch": 0.006127009298637642,
+      "grad_norm": 0.724172381016365,
+      "learning_rate": 4.0767386091127105e-05,
+      "loss": 1.7667,
+      "step": 85
+    },
+    {
+      "epoch": 0.0061990917609745545,
+      "grad_norm": 0.997507653534826,
+      "learning_rate": 4.124700239808154e-05,
+      "loss": 1.883,
+      "step": 86
+    },
+    {
+      "epoch": 0.006271174223311468,
+      "grad_norm": 0.4786564163789892,
+      "learning_rate": 4.1726618705035975e-05,
+      "loss": 1.6716,
+      "step": 87
+    },
+    {
+      "epoch": 0.006343256685648382,
+      "grad_norm": 0.6148659708163432,
+      "learning_rate": 4.220623501199041e-05,
+      "loss": 1.8086,
+      "step": 88
+    },
+    {
+      "epoch": 0.006415339147985295,
+      "grad_norm": 0.5911877900079417,
+      "learning_rate": 4.2685851318944845e-05,
+      "loss": 1.6994,
+      "step": 89
+    },
+    {
+      "epoch": 0.006487421610322209,
+      "grad_norm": 0.5534799837295095,
+      "learning_rate": 4.3165467625899284e-05,
+      "loss": 1.7953,
+      "step": 90
+    },
+    {
+      "epoch": 0.006559504072659122,
+      "grad_norm": 0.9400611391789584,
+      "learning_rate": 4.364508393285372e-05,
+      "loss": 1.8903,
+      "step": 91
+    },
+    {
+      "epoch": 0.006631586534996035,
+      "grad_norm": 0.5317192169061884,
+      "learning_rate": 4.4124700239808154e-05,
+      "loss": 1.8142,
+      "step": 92
+    },
+    {
+      "epoch": 0.006703668997332949,
+      "grad_norm": 0.5778163376575718,
+      "learning_rate": 4.460431654676259e-05,
+      "loss": 1.6611,
+      "step": 93
+    },
+    {
+      "epoch": 0.006775751459669862,
+      "grad_norm": 0.5809880839698385,
+      "learning_rate": 4.5083932853717024e-05,
+      "loss": 1.7973,
+      "step": 94
+    },
+    {
+      "epoch": 0.006847833922006776,
+      "grad_norm": 0.6281947735548433,
+      "learning_rate": 4.556354916067146e-05,
+      "loss": 1.7713,
+      "step": 95
+    },
+    {
+      "epoch": 0.00691991638434369,
+      "grad_norm": 0.8426561565191419,
+      "learning_rate": 4.60431654676259e-05,
+      "loss": 1.8629,
+      "step": 96
+    },
+    {
+      "epoch": 0.0069919988466806025,
+      "grad_norm": 0.4773704826976112,
+      "learning_rate": 4.652278177458034e-05,
+      "loss": 1.5567,
+      "step": 97
+    },
+    {
+      "epoch": 0.007064081309017516,
+      "grad_norm": 0.5665808307675152,
+      "learning_rate": 4.700239808153478e-05,
+      "loss": 1.7634,
+      "step": 98
+    },
+    {
+      "epoch": 0.007136163771354429,
+      "grad_norm": 0.7131756305455605,
+      "learning_rate": 4.748201438848921e-05,
+      "loss": 1.7444,
+      "step": 99
+    },
+    {
+      "epoch": 0.007208246233691343,
+      "grad_norm": 0.6716668006967341,
+      "learning_rate": 4.796163069544365e-05,
+      "loss": 1.7333,
+      "step": 100
+    },
+    {
+      "epoch": 0.007280328696028257,
+      "grad_norm": 0.5049102572005971,
+      "learning_rate": 4.844124700239808e-05,
+      "loss": 1.5683,
+      "step": 101
+    },
+    {
+      "epoch": 0.00735241115836517,
+      "grad_norm": 0.533672939923429,
+      "learning_rate": 4.892086330935252e-05,
+      "loss": 1.6507,
+      "step": 102
+    },
+    {
+      "epoch": 0.007424493620702083,
+      "grad_norm": 0.6394022197577649,
+      "learning_rate": 4.940047961630696e-05,
+      "loss": 1.6934,
+      "step": 103
+    },
+    {
+      "epoch": 0.007496576083038996,
+      "grad_norm": 0.5498873255354602,
+      "learning_rate": 4.9880095923261396e-05,
+      "loss": 1.5874,
+      "step": 104
+    },
+    {
+      "epoch": 0.00756865854537591,
+      "grad_norm": 0.5282347918421952,
+      "learning_rate": 5.035971223021583e-05,
+      "loss": 1.6525,
+      "step": 105
+    },
+    {
+      "epoch": 0.007640741007712824,
+      "grad_norm": 0.5050590280560159,
+      "learning_rate": 5.083932853717026e-05,
+      "loss": 1.7412,
+      "step": 106
+    },
+    {
+      "epoch": 0.007712823470049737,
+      "grad_norm": 0.4788865678345846,
+      "learning_rate": 5.1318944844124705e-05,
+      "loss": 1.6682,
+      "step": 107
+    },
+    {
+      "epoch": 0.0077849059323866505,
+      "grad_norm": 0.5200878126657371,
+      "learning_rate": 5.179856115107914e-05,
+      "loss": 1.6817,
+      "step": 108
+    },
+    {
+      "epoch": 0.007856988394723563,
+      "grad_norm": 0.6087268345775967,
+      "learning_rate": 5.2278177458033575e-05,
+      "loss": 1.6368,
+      "step": 109
+    },
+    {
+      "epoch": 0.007929070857060478,
+      "grad_norm": 0.48817727679283207,
+      "learning_rate": 5.275779376498802e-05,
+      "loss": 1.6737,
+      "step": 110
+    },
+    {
+      "epoch": 0.008001153319397391,
+      "grad_norm": 0.44833804379793374,
+      "learning_rate": 5.323741007194245e-05,
+      "loss": 1.5857,
+      "step": 111
+    },
+    {
+      "epoch": 0.008073235781734304,
+      "grad_norm": 0.3830105720017215,
+      "learning_rate": 5.3717026378896884e-05,
+      "loss": 1.5728,
+      "step": 112
+    },
+    {
+      "epoch": 0.008145318244071217,
+      "grad_norm": 0.4271528948330281,
+      "learning_rate": 5.4196642685851316e-05,
+      "loss": 1.7058,
+      "step": 113
+    },
+    {
+      "epoch": 0.008217400706408131,
+      "grad_norm": 0.41017647125330414,
+      "learning_rate": 5.467625899280576e-05,
+      "loss": 1.6392,
+      "step": 114
+    },
+    {
+      "epoch": 0.008289483168745044,
+      "grad_norm": 0.4657361121817729,
+      "learning_rate": 5.515587529976019e-05,
+      "loss": 1.621,
+      "step": 115
+    },
+    {
+      "epoch": 0.008361565631081957,
+      "grad_norm": 0.47738260687710526,
+      "learning_rate": 5.563549160671463e-05,
+      "loss": 1.6211,
+      "step": 116
+    },
+    {
+      "epoch": 0.008433648093418872,
+      "grad_norm": 0.5005143846003862,
+      "learning_rate": 5.611510791366906e-05,
+      "loss": 1.6298,
+      "step": 117
+    },
+    {
+      "epoch": 0.008505730555755785,
+      "grad_norm": 0.5144090075789134,
+      "learning_rate": 5.659472422062351e-05,
+      "loss": 1.5987,
+      "step": 118
+    },
+    {
+      "epoch": 0.008577813018092698,
+      "grad_norm": 0.4591038819366309,
+      "learning_rate": 5.707434052757794e-05,
+      "loss": 1.5794,
+      "step": 119
+    },
+    {
+      "epoch": 0.008649895480429612,
+      "grad_norm": 0.41586507916773985,
+      "learning_rate": 5.755395683453237e-05,
+      "loss": 1.6749,
+      "step": 120
+    },
+    {
+      "epoch": 0.008721977942766525,
+      "grad_norm": 0.6651747388237197,
+      "learning_rate": 5.803357314148681e-05,
+      "loss": 1.6832,
+      "step": 121
+    },
+    {
+      "epoch": 0.008794060405103438,
+      "grad_norm": 0.4023050793181856,
+      "learning_rate": 5.8513189448441256e-05,
+      "loss": 1.5733,
+      "step": 122
+    },
+    {
+      "epoch": 0.008866142867440351,
+      "grad_norm": 0.4237385901626671,
+      "learning_rate": 5.899280575539569e-05,
+      "loss": 1.6054,
+      "step": 123
+    },
+    {
+      "epoch": 0.008938225329777266,
+      "grad_norm": 0.3824044104578106,
+      "learning_rate": 5.947242206235012e-05,
+      "loss": 1.5675,
+      "step": 124
+    },
+    {
+      "epoch": 0.009010307792114178,
+      "grad_norm": 0.3602164312713383,
+      "learning_rate": 5.9952038369304564e-05,
+      "loss": 1.5602,
+      "step": 125
+    },
+    {
+      "epoch": 0.009082390254451091,
+      "grad_norm": 0.462636428237488,
+      "learning_rate": 6.0431654676258996e-05,
+      "loss": 1.5768,
+      "step": 126
+    },
+    {
+      "epoch": 0.009154472716788006,
+      "grad_norm": 0.4412034230996857,
+      "learning_rate": 6.091127098321343e-05,
+      "loss": 1.6951,
+      "step": 127
+    },
+    {
+      "epoch": 0.009226555179124919,
+      "grad_norm": 0.3956605337651504,
+      "learning_rate": 6.139088729016786e-05,
+      "loss": 1.5975,
+      "step": 128
+    },
+    {
+      "epoch": 0.009298637641461832,
+      "grad_norm": 0.40083682584627894,
+      "learning_rate": 6.187050359712231e-05,
+      "loss": 1.5276,
+      "step": 129
+    },
+    {
+      "epoch": 0.009370720103798746,
+      "grad_norm": 0.3760981128481325,
+      "learning_rate": 6.235011990407674e-05,
+      "loss": 1.593,
+      "step": 130
+    },
+    {
+      "epoch": 0.00944280256613566,
+      "grad_norm": 0.5285269258161931,
+      "learning_rate": 6.282973621103118e-05,
+      "loss": 1.5425,
+      "step": 131
+    },
+    {
+      "epoch": 0.009514885028472572,
+      "grad_norm": 0.3568808208805113,
+      "learning_rate": 6.330935251798561e-05,
+      "loss": 1.5742,
+      "step": 132
+    },
+    {
+      "epoch": 0.009586967490809487,
+      "grad_norm": 0.43990182968300917,
+      "learning_rate": 6.378896882494005e-05,
+      "loss": 1.6209,
+      "step": 133
+    },
+    {
+      "epoch": 0.0096590499531464,
+      "grad_norm": 0.4381591065847358,
+      "learning_rate": 6.426858513189449e-05,
+      "loss": 1.7517,
+      "step": 134
+    },
+    {
+      "epoch": 0.009731132415483313,
+      "grad_norm": 0.38997883549349394,
+      "learning_rate": 6.474820143884892e-05,
+      "loss": 1.586,
+      "step": 135
+    },
+    {
+      "epoch": 0.009803214877820226,
+      "grad_norm": 0.4051776850936413,
+      "learning_rate": 6.522781774580337e-05,
+      "loss": 1.6197,
+      "step": 136
+    },
+    {
+      "epoch": 0.00987529734015714,
+      "grad_norm": 0.4375931272960552,
+      "learning_rate": 6.570743405275779e-05,
+      "loss": 1.5543,
+      "step": 137
+    },
+    {
+      "epoch": 0.009947379802494053,
+      "grad_norm": 0.4455129892986555,
+      "learning_rate": 6.618705035971223e-05,
+      "loss": 1.6159,
+      "step": 138
+    },
+    {
+      "epoch": 0.010019462264830966,
+      "grad_norm": 0.3395559935366979,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.6063,
+      "step": 139
+    },
+    {
+      "epoch": 0.01009154472716788,
+      "grad_norm": 0.43990143269412385,
+      "learning_rate": 6.714628297362111e-05,
+      "loss": 1.6747,
+      "step": 140
+    },
+    {
+      "epoch": 0.010163627189504794,
+      "grad_norm": 0.30443198270484406,
+      "learning_rate": 6.762589928057555e-05,
+      "loss": 1.5083,
+      "step": 141
+    },
+    {
+      "epoch": 0.010235709651841706,
+      "grad_norm": 0.39116745975799505,
+      "learning_rate": 6.810551558752997e-05,
+      "loss": 1.5347,
+      "step": 142
+    },
+    {
+      "epoch": 0.010307792114178621,
+      "grad_norm": 0.3444770563974035,
+      "learning_rate": 6.858513189448441e-05,
+      "loss": 1.483,
+      "step": 143
+    },
+    {
+      "epoch": 0.010379874576515534,
+      "grad_norm": 0.45319695060719234,
+      "learning_rate": 6.906474820143886e-05,
+      "loss": 1.5344,
+      "step": 144
+    },
+    {
+      "epoch": 0.010451957038852447,
+      "grad_norm": 0.6729848299699345,
+      "learning_rate": 6.954436450839329e-05,
+      "loss": 1.5876,
+      "step": 145
+    },
+    {
+      "epoch": 0.01052403950118936,
+      "grad_norm": 0.3524846673663031,
+      "learning_rate": 7.002398081534773e-05,
+      "loss": 1.5866,
+      "step": 146
+    },
+    {
+      "epoch": 0.010596121963526274,
+      "grad_norm": 0.4182895203494307,
+      "learning_rate": 7.050359712230215e-05,
+      "loss": 1.4838,
+      "step": 147
+    },
+    {
+      "epoch": 0.010668204425863187,
+      "grad_norm": 0.521875625409466,
+      "learning_rate": 7.09832134292566e-05,
+      "loss": 1.5059,
+      "step": 148
+    },
+    {
+      "epoch": 0.0107402868882001,
+      "grad_norm": 0.34750105916611734,
+      "learning_rate": 7.146282973621103e-05,
+      "loss": 1.6133,
+      "step": 149
+    },
+    {
+      "epoch": 0.010812369350537015,
+      "grad_norm": 0.330871527751618,
+      "learning_rate": 7.194244604316547e-05,
+      "loss": 1.5488,
+      "step": 150
+    },
+    {
+      "epoch": 0.010884451812873928,
+      "grad_norm": 0.30911027865969076,
+      "learning_rate": 7.242206235011992e-05,
+      "loss": 1.45,
+      "step": 151
+    },
+    {
+      "epoch": 0.01095653427521084,
+      "grad_norm": 0.3779640542567218,
+      "learning_rate": 7.290167865707434e-05,
+      "loss": 1.4757,
+      "step": 152
+    },
+    {
+      "epoch": 0.011028616737547755,
+      "grad_norm": 0.3668966227463583,
+      "learning_rate": 7.338129496402878e-05,
+      "loss": 1.5708,
+      "step": 153
+    },
+    {
+      "epoch": 0.011100699199884668,
+      "grad_norm": 0.2613162790128224,
+      "learning_rate": 7.386091127098321e-05,
+      "loss": 1.4588,
+      "step": 154
+    },
+    {
+      "epoch": 0.011172781662221581,
+      "grad_norm": 0.444165011572818,
+      "learning_rate": 7.434052757793766e-05,
+      "loss": 1.5087,
+      "step": 155
+    },
+    {
+      "epoch": 0.011244864124558496,
+      "grad_norm": 0.32813866140000303,
+      "learning_rate": 7.48201438848921e-05,
+      "loss": 1.5187,
+      "step": 156
+    },
+    {
+      "epoch": 0.011316946586895409,
+      "grad_norm": 0.4261646448388426,
+      "learning_rate": 7.529976019184652e-05,
+      "loss": 1.5424,
+      "step": 157
+    },
+    {
+      "epoch": 0.011389029049232322,
+      "grad_norm": 0.28246038327970985,
+      "learning_rate": 7.577937649880096e-05,
+      "loss": 1.5721,
+      "step": 158
+    },
+    {
+      "epoch": 0.011461111511569234,
+      "grad_norm": 0.3405491316773894,
+      "learning_rate": 7.62589928057554e-05,
+      "loss": 1.5561,
+      "step": 159
+    },
+    {
+      "epoch": 0.011533193973906149,
+      "grad_norm": 0.347557293455909,
+      "learning_rate": 7.673860911270984e-05,
+      "loss": 1.4782,
+      "step": 160
+    },
+    {
+      "epoch": 0.011605276436243062,
+      "grad_norm": 0.40497333613534015,
+      "learning_rate": 7.721822541966426e-05,
+      "loss": 1.5096,
+      "step": 161
+    },
+    {
+      "epoch": 0.011677358898579975,
+      "grad_norm": 0.5220130740944903,
+      "learning_rate": 7.769784172661872e-05,
+      "loss": 1.5815,
+      "step": 162
+    },
+    {
+      "epoch": 0.01174944136091689,
+      "grad_norm": 0.3269735258478042,
+      "learning_rate": 7.817745803357315e-05,
+      "loss": 1.4575,
+      "step": 163
+    },
+    {
+      "epoch": 0.011821523823253802,
+      "grad_norm": 0.35692983621687485,
+      "learning_rate": 7.865707434052758e-05,
+      "loss": 1.37,
+      "step": 164
+    },
+    {
+      "epoch": 0.011893606285590715,
+      "grad_norm": 0.3420099622529274,
+      "learning_rate": 7.913669064748202e-05,
+      "loss": 1.5169,
+      "step": 165
+    },
+    {
+      "epoch": 0.01196568874792763,
+      "grad_norm": 0.3067495790826786,
+      "learning_rate": 7.961630695443646e-05,
+      "loss": 1.5898,
+      "step": 166
+    },
+    {
+      "epoch": 0.012037771210264543,
+      "grad_norm": 0.4932067654393579,
+      "learning_rate": 8.00959232613909e-05,
+      "loss": 1.5039,
+      "step": 167
+    },
+    {
+      "epoch": 0.012109853672601456,
+      "grad_norm": 0.4376679246110129,
+      "learning_rate": 8.057553956834533e-05,
+      "loss": 1.5502,
+      "step": 168
+    },
+    {
+      "epoch": 0.01218193613493837,
+      "grad_norm": 0.500452950367802,
+      "learning_rate": 8.105515587529976e-05,
+      "loss": 1.5585,
+      "step": 169
+    },
+    {
+      "epoch": 0.012254018597275283,
+      "grad_norm": 0.4472227749539023,
+      "learning_rate": 8.153477218225421e-05,
+      "loss": 1.52,
+      "step": 170
+    },
+    {
+      "epoch": 0.012326101059612196,
+      "grad_norm": 0.2586347948865282,
+      "learning_rate": 8.201438848920863e-05,
+      "loss": 1.4687,
+      "step": 171
+    },
+    {
+      "epoch": 0.012398183521949109,
+      "grad_norm": 0.2470025030802705,
+      "learning_rate": 8.249400479616307e-05,
+      "loss": 1.5412,
+      "step": 172
+    },
+    {
+      "epoch": 0.012470265984286024,
+      "grad_norm": 0.35823668854643254,
+      "learning_rate": 8.29736211031175e-05,
+      "loss": 1.5225,
+      "step": 173
+    },
+    {
+      "epoch": 0.012542348446622937,
+      "grad_norm": 0.303885772183371,
+      "learning_rate": 8.345323741007195e-05,
+      "loss": 1.5717,
+      "step": 174
+    },
+    {
+      "epoch": 0.01261443090895985,
+      "grad_norm": 0.30671657939657576,
+      "learning_rate": 8.393285371702639e-05,
+      "loss": 1.5061,
+      "step": 175
+    },
+    {
+      "epoch": 0.012686513371296764,
+      "grad_norm": 0.30260618604931444,
+      "learning_rate": 8.441247002398081e-05,
+      "loss": 1.5861,
+      "step": 176
+    },
+    {
+      "epoch": 0.012758595833633677,
+      "grad_norm": 0.380138661435745,
+      "learning_rate": 8.489208633093527e-05,
+      "loss": 1.4392,
+      "step": 177
+    },
+    {
+      "epoch": 0.01283067829597059,
+      "grad_norm": 0.23131167081337672,
+      "learning_rate": 8.537170263788969e-05,
+      "loss": 1.5633,
+      "step": 178
+    },
+    {
+      "epoch": 0.012902760758307505,
+      "grad_norm": 0.28351090664971595,
+      "learning_rate": 8.585131894484413e-05,
+      "loss": 1.443,
+      "step": 179
+    },
+    {
+      "epoch": 0.012974843220644417,
+      "grad_norm": 0.28539999275808614,
+      "learning_rate": 8.633093525179857e-05,
+      "loss": 1.5207,
+      "step": 180
+    },
+    {
+      "epoch": 0.01304692568298133,
+      "grad_norm": 0.2656933767635461,
+      "learning_rate": 8.6810551558753e-05,
+      "loss": 1.5004,
+      "step": 181
+    },
+    {
+      "epoch": 0.013119008145318243,
+      "grad_norm": 0.8743347463150967,
+      "learning_rate": 8.729016786570744e-05,
+      "loss": 1.4644,
+      "step": 182
+    },
+    {
+      "epoch": 0.013191090607655158,
+      "grad_norm": 0.3491049258770815,
+      "learning_rate": 8.776978417266187e-05,
+      "loss": 1.5028,
+      "step": 183
+    },
+    {
+      "epoch": 0.01326317306999207,
+      "grad_norm": 0.31718510401108735,
+      "learning_rate": 8.824940047961631e-05,
+      "loss": 1.3921,
+      "step": 184
+    },
+    {
+      "epoch": 0.013335255532328984,
+      "grad_norm": 0.3143600894715496,
+      "learning_rate": 8.872901678657075e-05,
+      "loss": 1.5334,
+      "step": 185
+    },
+    {
+      "epoch": 0.013407337994665898,
+      "grad_norm": 0.3101565454683402,
+      "learning_rate": 8.920863309352519e-05,
+      "loss": 1.4149,
+      "step": 186
+    },
+    {
+      "epoch": 0.013479420457002811,
+      "grad_norm": 0.3113081448322479,
+      "learning_rate": 8.968824940047962e-05,
+      "loss": 1.397,
+      "step": 187
+    },
+    {
+      "epoch": 0.013551502919339724,
+      "grad_norm": 0.7565662451121091,
+      "learning_rate": 9.016786570743405e-05,
+      "loss": 1.3724,
+      "step": 188
+    },
+    {
+      "epoch": 0.013623585381676639,
+      "grad_norm": 0.2688394093580687,
+      "learning_rate": 9.06474820143885e-05,
+      "loss": 1.4319,
+      "step": 189
+    },
+    {
+      "epoch": 0.013695667844013552,
+      "grad_norm": 0.2837556011630947,
+      "learning_rate": 9.112709832134293e-05,
+      "loss": 1.4151,
+      "step": 190
+    },
+    {
+      "epoch": 0.013767750306350465,
+      "grad_norm": 0.2942693266007315,
+      "learning_rate": 9.160671462829736e-05,
+      "loss": 1.4569,
+      "step": 191
+    },
+    {
+      "epoch": 0.01383983276868738,
+      "grad_norm": 0.2429607058426557,
+      "learning_rate": 9.20863309352518e-05,
+      "loss": 1.4938,
+      "step": 192
+    },
+    {
+      "epoch": 0.013911915231024292,
+      "grad_norm": 0.20897085115622685,
+      "learning_rate": 9.256594724220624e-05,
+      "loss": 1.4061,
+      "step": 193
+    },
+    {
+      "epoch": 0.013983997693361205,
+      "grad_norm": 0.3023908413639266,
+      "learning_rate": 9.304556354916068e-05,
+      "loss": 1.4143,
+      "step": 194
+    },
+    {
+      "epoch": 0.014056080155698118,
+      "grad_norm": 0.3008330682431941,
+      "learning_rate": 9.35251798561151e-05,
+      "loss": 1.4634,
+      "step": 195
+    },
+    {
+      "epoch": 0.014128162618035033,
+      "grad_norm": 0.29165458457027116,
+      "learning_rate": 9.400479616306956e-05,
+      "loss": 1.4315,
+      "step": 196
+    },
+    {
+      "epoch": 0.014200245080371945,
+      "grad_norm": 0.22800865510212348,
+      "learning_rate": 9.448441247002398e-05,
+      "loss": 1.3351,
+      "step": 197
+    },
+    {
+      "epoch": 0.014272327542708858,
+      "grad_norm": 0.2656296626198674,
+      "learning_rate": 9.496402877697842e-05,
+      "loss": 1.3752,
+      "step": 198
+    },
+    {
+      "epoch": 0.014344410005045773,
+      "grad_norm": 0.2670660369946238,
+      "learning_rate": 9.544364508393286e-05,
+      "loss": 1.4289,
+      "step": 199
+    },
+    {
+      "epoch": 0.014416492467382686,
+      "grad_norm": 0.27596503311318316,
+      "learning_rate": 9.59232613908873e-05,
+      "loss": 1.5186,
+      "step": 200
+    },
+    {
+      "epoch": 0.014488574929719599,
+      "grad_norm": 0.3377421151458901,
+      "learning_rate": 9.640287769784174e-05,
+      "loss": 1.5112,
+      "step": 201
+    },
+    {
+      "epoch": 0.014560657392056513,
+      "grad_norm": 0.25989449308734686,
+      "learning_rate": 9.688249400479616e-05,
+      "loss": 1.5144,
+      "step": 202
+    },
+    {
+      "epoch": 0.014632739854393426,
+      "grad_norm": 0.2999719224987995,
+      "learning_rate": 9.736211031175061e-05,
+      "loss": 1.5036,
+      "step": 203
+    },
+    {
+      "epoch": 0.01470482231673034,
+      "grad_norm": 0.3331191389272093,
+      "learning_rate": 9.784172661870504e-05,
+      "loss": 1.4394,
+      "step": 204
+    },
+    {
+      "epoch": 0.014776904779067252,
+      "grad_norm": 0.36153008593688973,
+      "learning_rate": 9.832134292565948e-05,
+      "loss": 1.5181,
+      "step": 205
+    },
+    {
+      "epoch": 0.014848987241404167,
+      "grad_norm": 0.20581869658552598,
+      "learning_rate": 9.880095923261392e-05,
+      "loss": 1.5305,
+      "step": 206
+    },
+    {
+      "epoch": 0.01492106970374108,
+      "grad_norm": 0.33057712625816205,
+      "learning_rate": 9.928057553956835e-05,
+      "loss": 1.4555,
+      "step": 207
+    },
+    {
+      "epoch": 0.014993152166077993,
+      "grad_norm": 0.2622853981125225,
+      "learning_rate": 9.976019184652279e-05,
+      "loss": 1.4775,
+      "step": 208
+    },
+    {
+      "epoch": 0.015065234628414907,
+      "grad_norm": 0.22385513196126544,
+      "learning_rate": 0.00010023980815347722,
+      "loss": 1.3477,
+      "step": 209
+    },
+    {
+      "epoch": 0.01513731709075182,
+      "grad_norm": 0.4326204057898977,
+      "learning_rate": 0.00010071942446043166,
+      "loss": 1.5462,
+      "step": 210
+    },
+    {
+      "epoch": 0.015209399553088733,
+      "grad_norm": 0.3092835039063757,
+      "learning_rate": 0.0001011990407673861,
+      "loss": 1.4384,
+      "step": 211
+    },
+    {
+      "epoch": 0.015281482015425648,
+      "grad_norm": 0.4774809136621503,
+      "learning_rate": 0.00010167865707434052,
+      "loss": 1.4737,
+      "step": 212
+    },
+    {
+      "epoch": 0.01535356447776256,
+      "grad_norm": 0.2642896241689678,
+      "learning_rate": 0.00010215827338129497,
+      "loss": 1.4798,
+      "step": 213
+    },
+    {
+      "epoch": 0.015425646940099473,
+      "grad_norm": 0.37459264525435165,
+      "learning_rate": 0.00010263788968824941,
+      "loss": 1.4582,
+      "step": 214
+    },
+    {
+      "epoch": 0.015497729402436388,
+      "grad_norm": 0.2919984216020338,
+      "learning_rate": 0.00010311750599520385,
+      "loss": 1.378,
+      "step": 215
+    },
+    {
+      "epoch": 0.015569811864773301,
+      "grad_norm": 0.33312142243698256,
+      "learning_rate": 0.00010359712230215827,
+      "loss": 1.5913,
+      "step": 216
+    },
+    {
+      "epoch": 0.015641894327110214,
+      "grad_norm": 0.21679447979700256,
+      "learning_rate": 0.00010407673860911271,
+      "loss": 1.4524,
+      "step": 217
+    },
+    {
+      "epoch": 0.015713976789447127,
+      "grad_norm": 0.21428305452619087,
+      "learning_rate": 0.00010455635491606715,
+      "loss": 1.4367,
+      "step": 218
+    },
+    {
+      "epoch": 0.01578605925178404,
+      "grad_norm": 0.7836658125044331,
+      "learning_rate": 0.00010503597122302158,
+      "loss": 1.6119,
+      "step": 219
+    },
+    {
+      "epoch": 0.015858141714120956,
+      "grad_norm": 0.21723786828560498,
+      "learning_rate": 0.00010551558752997604,
+      "loss": 1.5729,
+      "step": 220
+    },
+    {
+      "epoch": 0.01593022417645787,
+      "grad_norm": 0.29033757217879014,
+      "learning_rate": 0.00010599520383693047,
+      "loss": 1.4781,
+      "step": 221
+    },
+    {
+      "epoch": 0.016002306638794782,
+      "grad_norm": 0.2204811636129812,
+      "learning_rate": 0.0001064748201438849,
+      "loss": 1.3839,
+      "step": 222
+    },
+    {
+      "epoch": 0.016074389101131695,
+      "grad_norm": 0.30616681761317915,
+      "learning_rate": 0.00010695443645083933,
+      "loss": 1.4143,
+      "step": 223
+    },
+    {
+      "epoch": 0.016146471563468608,
+      "grad_norm": 0.20873435227139098,
+      "learning_rate": 0.00010743405275779377,
+      "loss": 1.3341,
+      "step": 224
+    },
+    {
+      "epoch": 0.01621855402580552,
+      "grad_norm": 0.5433741980653699,
+      "learning_rate": 0.0001079136690647482,
+      "loss": 1.5558,
+      "step": 225
+    },
+    {
+      "epoch": 0.016290636488142433,
+      "grad_norm": 0.28512128977083506,
+      "learning_rate": 0.00010839328537170263,
+      "loss": 1.4872,
+      "step": 226
+    },
+    {
+      "epoch": 0.01636271895047935,
+      "grad_norm": 0.207135389117889,
+      "learning_rate": 0.00010887290167865707,
+      "loss": 1.4519,
+      "step": 227
+    },
+    {
+      "epoch": 0.016434801412816263,
+      "grad_norm": 0.32957985861769584,
+      "learning_rate": 0.00010935251798561152,
+      "loss": 1.4441,
+      "step": 228
+    },
+    {
+      "epoch": 0.016506883875153176,
+      "grad_norm": 0.31713624709519955,
+      "learning_rate": 0.00010983213429256596,
+      "loss": 1.4207,
+      "step": 229
+    },
+    {
+      "epoch": 0.01657896633749009,
+      "grad_norm": 0.2463805564398517,
+      "learning_rate": 0.00011031175059952039,
+      "loss": 1.3751,
+      "step": 230
+    },
+    {
+      "epoch": 0.016651048799827,
+      "grad_norm": 0.26366281638154715,
+      "learning_rate": 0.00011079136690647482,
+      "loss": 1.42,
+      "step": 231
+    },
+    {
+      "epoch": 0.016723131262163914,
+      "grad_norm": 0.21752509552647759,
+      "learning_rate": 0.00011127098321342926,
+      "loss": 1.3827,
+      "step": 232
+    },
+    {
+      "epoch": 0.01679521372450083,
+      "grad_norm": 0.23841789038582736,
+      "learning_rate": 0.00011175059952038369,
+      "loss": 1.4441,
+      "step": 233
+    },
+    {
+      "epoch": 0.016867296186837744,
+      "grad_norm": 0.3088708926876433,
+      "learning_rate": 0.00011223021582733813,
+      "loss": 1.357,
+      "step": 234
+    },
+    {
+      "epoch": 0.016939378649174656,
+      "grad_norm": 0.22337111504131374,
+      "learning_rate": 0.00011270983213429258,
+      "loss": 1.4001,
+      "step": 235
+    },
+    {
+      "epoch": 0.01701146111151157,
+      "grad_norm": 0.3557025692839835,
+      "learning_rate": 0.00011318944844124702,
+      "loss": 1.4859,
+      "step": 236
+    },
+    {
+      "epoch": 0.017083543573848482,
+      "grad_norm": 0.20371593449595726,
+      "learning_rate": 0.00011366906474820144,
+      "loss": 1.4888,
+      "step": 237
+    },
+    {
+      "epoch": 0.017155626036185395,
+      "grad_norm": 0.1987247824507573,
+      "learning_rate": 0.00011414868105515588,
+      "loss": 1.4517,
+      "step": 238
+    },
+    {
+      "epoch": 0.017227708498522308,
+      "grad_norm": 0.37082692307171156,
+      "learning_rate": 0.00011462829736211032,
+      "loss": 1.5236,
+      "step": 239
+    },
+    {
+      "epoch": 0.017299790960859224,
+      "grad_norm": 0.25331787601945105,
+      "learning_rate": 0.00011510791366906474,
+      "loss": 1.4886,
+      "step": 240
+    },
+    {
+      "epoch": 0.017371873423196137,
+      "grad_norm": 0.18245899364940935,
+      "learning_rate": 0.00011558752997601918,
+      "loss": 1.4435,
+      "step": 241
+    },
+    {
+      "epoch": 0.01744395588553305,
+      "grad_norm": 0.27278940335633206,
+      "learning_rate": 0.00011606714628297362,
+      "loss": 1.4877,
+      "step": 242
+    },
+    {
+      "epoch": 0.017516038347869963,
+      "grad_norm": 0.21938451388101368,
+      "learning_rate": 0.00011654676258992807,
+      "loss": 1.444,
+      "step": 243
+    },
+    {
+      "epoch": 0.017588120810206876,
+      "grad_norm": 0.25417352587437037,
+      "learning_rate": 0.00011702637889688251,
+      "loss": 1.3128,
+      "step": 244
+    },
+    {
+      "epoch": 0.01766020327254379,
+      "grad_norm": 0.17975350450703156,
+      "learning_rate": 0.00011750599520383694,
+      "loss": 1.4573,
+      "step": 245
+    },
+    {
+      "epoch": 0.017732285734880702,
+      "grad_norm": 0.238281238642535,
+      "learning_rate": 0.00011798561151079137,
+      "loss": 1.352,
+      "step": 246
+    },
+    {
+      "epoch": 0.017804368197217618,
+      "grad_norm": 0.20531346057791597,
+      "learning_rate": 0.0001184652278177458,
+      "loss": 1.4325,
+      "step": 247
+    },
+    {
+      "epoch": 0.01787645065955453,
+      "grad_norm": 0.1991902709376777,
+      "learning_rate": 0.00011894484412470024,
+      "loss": 1.4103,
+      "step": 248
+    },
+    {
+      "epoch": 0.017948533121891444,
+      "grad_norm": 0.203914372528537,
+      "learning_rate": 0.00011942446043165468,
+      "loss": 1.4439,
+      "step": 249
+    },
+    {
+      "epoch": 0.018020615584228357,
+      "grad_norm": 0.1998504723251893,
+      "learning_rate": 0.00011990407673860913,
+      "loss": 1.4555,
+      "step": 250
+    },
+    {
+      "epoch": 0.01809269804656527,
+      "grad_norm": 0.5127358895046691,
+      "learning_rate": 0.00012038369304556357,
+      "loss": 1.4463,
+      "step": 251
+    },
+    {
+      "epoch": 0.018164780508902183,
+      "grad_norm": 0.307269647506404,
+      "learning_rate": 0.00012086330935251799,
+      "loss": 1.5401,
+      "step": 252
+    },
+    {
+      "epoch": 0.0182368629712391,
+      "grad_norm": 0.21995271564265081,
+      "learning_rate": 0.00012134292565947243,
+      "loss": 1.3725,
+      "step": 253
+    },
+    {
+      "epoch": 0.018308945433576012,
+      "grad_norm": 0.24462948000056514,
+      "learning_rate": 0.00012182254196642686,
+      "loss": 1.3496,
+      "step": 254
+    },
+    {
+      "epoch": 0.018381027895912925,
+      "grad_norm": 0.23212311722285198,
+      "learning_rate": 0.0001223021582733813,
+      "loss": 1.503,
+      "step": 255
+    },
+    {
+      "epoch": 0.018453110358249838,
+      "grad_norm": 0.22211994471659857,
+      "learning_rate": 0.00012278177458033572,
+      "loss": 1.3826,
+      "step": 256
+    },
+    {
+      "epoch": 0.01852519282058675,
+      "grad_norm": 0.1732336120522755,
+      "learning_rate": 0.00012326139088729017,
+      "loss": 1.3682,
+      "step": 257
+    },
+    {
+      "epoch": 0.018597275282923664,
+      "grad_norm": 0.29749322662756755,
+      "learning_rate": 0.00012374100719424462,
+      "loss": 1.4507,
+      "step": 258
+    },
+    {
+      "epoch": 0.018669357745260576,
+      "grad_norm": 0.4070523739937354,
+      "learning_rate": 0.00012422062350119905,
+      "loss": 1.4867,
+      "step": 259
+    },
+    {
+      "epoch": 0.018741440207597493,
+      "grad_norm": 0.22042580365347972,
+      "learning_rate": 0.00012470023980815347,
+      "loss": 1.4385,
+      "step": 260
+    },
+    {
+      "epoch": 0.018813522669934406,
+      "grad_norm": 0.1754028453758367,
+      "learning_rate": 0.00012517985611510793,
+      "loss": 1.4452,
+      "step": 261
+    },
+    {
+      "epoch": 0.01888560513227132,
+      "grad_norm": 0.19927048183990892,
+      "learning_rate": 0.00012565947242206235,
+      "loss": 1.3726,
+      "step": 262
+    },
+    {
+      "epoch": 0.01895768759460823,
+      "grad_norm": 0.4039085559524575,
+      "learning_rate": 0.00012613908872901678,
+      "loss": 1.3366,
+      "step": 263
+    },
+    {
+      "epoch": 0.019029770056945144,
+      "grad_norm": 0.23352672212989384,
+      "learning_rate": 0.00012661870503597123,
+      "loss": 1.4818,
+      "step": 264
+    },
+    {
+      "epoch": 0.019101852519282057,
+      "grad_norm": 0.2400641183515557,
+      "learning_rate": 0.00012709832134292568,
+      "loss": 1.3737,
+      "step": 265
+    },
+    {
+      "epoch": 0.019173934981618974,
+      "grad_norm": 0.30197623761277764,
+      "learning_rate": 0.0001275779376498801,
+      "loss": 1.4281,
+      "step": 266
+    },
+    {
+      "epoch": 0.019246017443955887,
+      "grad_norm": 0.19754030848079707,
+      "learning_rate": 0.00012805755395683453,
+      "loss": 1.4242,
+      "step": 267
+    },
+    {
+      "epoch": 0.0193180999062928,
+      "grad_norm": 0.16593371533608325,
+      "learning_rate": 0.00012853717026378898,
+      "loss": 1.4091,
+      "step": 268
+    },
+    {
+      "epoch": 0.019390182368629712,
+      "grad_norm": 0.17768173457785777,
+      "learning_rate": 0.0001290167865707434,
+      "loss": 1.3941,
+      "step": 269
+    },
+    {
+      "epoch": 0.019462264830966625,
+      "grad_norm": 0.21560203399813577,
+      "learning_rate": 0.00012949640287769783,
+      "loss": 1.3872,
+      "step": 270
+    },
+    {
+      "epoch": 0.019534347293303538,
+      "grad_norm": 0.2156946373493371,
+      "learning_rate": 0.00012997601918465228,
+      "loss": 1.4546,
+      "step": 271
+    },
+    {
+      "epoch": 0.01960642975564045,
+      "grad_norm": 0.16103045705624283,
+      "learning_rate": 0.00013045563549160674,
+      "loss": 1.4338,
+      "step": 272
+    },
+    {
+      "epoch": 0.019678512217977367,
+      "grad_norm": 0.48340451809447105,
+      "learning_rate": 0.00013093525179856116,
+      "loss": 1.489,
+      "step": 273
+    },
+    {
+      "epoch": 0.01975059468031428,
+      "grad_norm": 0.3085383549213157,
+      "learning_rate": 0.00013141486810551559,
+      "loss": 1.4818,
+      "step": 274
+    },
+    {
+      "epoch": 0.019822677142651193,
+      "grad_norm": 0.23585921832492804,
+      "learning_rate": 0.00013189448441247004,
+      "loss": 1.3797,
+      "step": 275
+    },
+    {
+      "epoch": 0.019894759604988106,
+      "grad_norm": 0.18663739539252747,
+      "learning_rate": 0.00013237410071942446,
+      "loss": 1.4595,
+      "step": 276
+    },
+    {
+      "epoch": 0.01996684206732502,
+      "grad_norm": 0.17885500554315795,
+      "learning_rate": 0.0001328537170263789,
+      "loss": 1.4594,
+      "step": 277
+    },
+    {
+      "epoch": 0.020038924529661932,
+      "grad_norm": 0.19193404551202412,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 1.5061,
+      "step": 278
+    },
+    {
+      "epoch": 0.02011100699199885,
+      "grad_norm": 0.20775742400666208,
+      "learning_rate": 0.00013381294964028776,
+      "loss": 1.4759,
+      "step": 279
+    },
+    {
+      "epoch": 0.02018308945433576,
+      "grad_norm": 0.22540591024800477,
+      "learning_rate": 0.00013429256594724222,
+      "loss": 1.3741,
+      "step": 280
+    },
+    {
+      "epoch": 0.020255171916672674,
+      "grad_norm": 0.2993011922897735,
+      "learning_rate": 0.00013477218225419664,
+      "loss": 1.3311,
+      "step": 281
+    },
+    {
+      "epoch": 0.020327254379009587,
+      "grad_norm": 0.1580695494268975,
+      "learning_rate": 0.0001352517985611511,
+      "loss": 1.5182,
+      "step": 282
+    },
+    {
+      "epoch": 0.0203993368413465,
+      "grad_norm": 0.34854576880996313,
+      "learning_rate": 0.00013573141486810552,
+      "loss": 1.5102,
+      "step": 283
+    },
+    {
+      "epoch": 0.020471419303683413,
+      "grad_norm": 0.2180648189064342,
+      "learning_rate": 0.00013621103117505994,
+      "loss": 1.4787,
+      "step": 284
+    },
+    {
+      "epoch": 0.020543501766020326,
+      "grad_norm": 0.22642093129306717,
+      "learning_rate": 0.0001366906474820144,
+      "loss": 1.4346,
+      "step": 285
+    },
+    {
+      "epoch": 0.020615584228357242,
+      "grad_norm": 0.25893094785487264,
+      "learning_rate": 0.00013717026378896882,
+      "loss": 1.4639,
+      "step": 286
+    },
+    {
+      "epoch": 0.020687666690694155,
+      "grad_norm": 0.21272272222777971,
+      "learning_rate": 0.00013764988009592327,
+      "loss": 1.4164,
+      "step": 287
+    },
+    {
+      "epoch": 0.020759749153031068,
+      "grad_norm": 0.16482184383791296,
+      "learning_rate": 0.00013812949640287772,
+      "loss": 1.408,
+      "step": 288
+    },
+    {
+      "epoch": 0.02083183161536798,
+      "grad_norm": 0.16699646761600762,
+      "learning_rate": 0.00013860911270983215,
+      "loss": 1.3666,
+      "step": 289
+    },
+    {
+      "epoch": 0.020903914077704894,
+      "grad_norm": 0.14754393097758098,
+      "learning_rate": 0.00013908872901678657,
+      "loss": 1.4703,
+      "step": 290
+    },
+    {
+      "epoch": 0.020975996540041807,
+      "grad_norm": 0.3759113182333724,
+      "learning_rate": 0.000139568345323741,
+      "loss": 1.4021,
+      "step": 291
+    },
+    {
+      "epoch": 0.02104807900237872,
+      "grad_norm": 0.20489112743371046,
+      "learning_rate": 0.00014004796163069545,
+      "loss": 1.395,
+      "step": 292
+    },
+    {
+      "epoch": 0.021120161464715636,
+      "grad_norm": 0.19794477828799406,
+      "learning_rate": 0.00014052757793764988,
+      "loss": 1.4019,
+      "step": 293
+    },
+    {
+      "epoch": 0.02119224392705255,
+      "grad_norm": 0.2697990836482631,
+      "learning_rate": 0.0001410071942446043,
+      "loss": 1.4095,
+      "step": 294
+    },
+    {
+      "epoch": 0.02126432638938946,
+      "grad_norm": 0.13498537640969785,
+      "learning_rate": 0.00014148681055155878,
+      "loss": 1.4089,
+      "step": 295
+    },
+    {
+      "epoch": 0.021336408851726375,
+      "grad_norm": 0.26092314061036126,
+      "learning_rate": 0.0001419664268585132,
+      "loss": 1.5229,
+      "step": 296
+    },
+    {
+      "epoch": 0.021408491314063288,
+      "grad_norm": 0.1856836380314397,
+      "learning_rate": 0.00014244604316546763,
+      "loss": 1.5454,
+      "step": 297
+    },
+    {
+      "epoch": 0.0214805737764002,
+      "grad_norm": 0.14935107981339024,
+      "learning_rate": 0.00014292565947242206,
+      "loss": 1.3264,
+      "step": 298
+    },
+    {
+      "epoch": 0.021552656238737117,
+      "grad_norm": 0.1929995749747401,
+      "learning_rate": 0.0001434052757793765,
+      "loss": 1.4123,
+      "step": 299
+    },
+    {
+      "epoch": 0.02162473870107403,
+      "grad_norm": 0.2996593772226965,
+      "learning_rate": 0.00014388489208633093,
+      "loss": 1.4359,
+      "step": 300
+    },
+    {
+      "epoch": 0.021696821163410943,
+      "grad_norm": 0.16822639984212567,
+      "learning_rate": 0.00014436450839328536,
+      "loss": 1.4031,
+      "step": 301
+    },
+    {
+      "epoch": 0.021768903625747855,
+      "grad_norm": 0.2064446891204127,
+      "learning_rate": 0.00014484412470023984,
+      "loss": 1.4137,
+      "step": 302
+    },
+    {
+      "epoch": 0.02184098608808477,
+      "grad_norm": 0.24634110539741785,
+      "learning_rate": 0.00014532374100719426,
+      "loss": 1.4613,
+      "step": 303
+    },
+    {
+      "epoch": 0.02191306855042168,
+      "grad_norm": 0.16530603447880574,
+      "learning_rate": 0.0001458033573141487,
+      "loss": 1.3448,
+      "step": 304
+    },
+    {
+      "epoch": 0.021985151012758594,
+      "grad_norm": 0.18667874998769096,
+      "learning_rate": 0.0001462829736211031,
+      "loss": 1.3749,
+      "step": 305
+    },
+    {
+      "epoch": 0.02205723347509551,
+      "grad_norm": 0.1984959009162098,
+      "learning_rate": 0.00014676258992805756,
+      "loss": 1.3282,
+      "step": 306
+    },
+    {
+      "epoch": 0.022129315937432423,
+      "grad_norm": 0.1800005765942182,
+      "learning_rate": 0.000147242206235012,
+      "loss": 1.436,
+      "step": 307
+    },
+    {
+      "epoch": 0.022201398399769336,
+      "grad_norm": 0.3040264172355369,
+      "learning_rate": 0.00014772182254196641,
+      "loss": 1.4394,
+      "step": 308
+    },
+    {
+      "epoch": 0.02227348086210625,
+      "grad_norm": 0.27139231778750067,
+      "learning_rate": 0.00014820143884892087,
+      "loss": 1.4694,
+      "step": 309
+    },
+    {
+      "epoch": 0.022345563324443162,
+      "grad_norm": 0.2094829868308678,
+      "learning_rate": 0.00014868105515587532,
+      "loss": 1.4481,
+      "step": 310
+    },
+    {
+      "epoch": 0.022417645786780075,
+      "grad_norm": 0.3209277214553109,
+      "learning_rate": 0.00014916067146282974,
+      "loss": 1.4403,
+      "step": 311
+    },
+    {
+      "epoch": 0.02248972824911699,
+      "grad_norm": 0.24834676080829274,
+      "learning_rate": 0.0001496402877697842,
+      "loss": 1.4028,
+      "step": 312
+    },
+    {
+      "epoch": 0.022561810711453904,
+      "grad_norm": 0.23851014448324961,
+      "learning_rate": 0.00015011990407673862,
+      "loss": 1.4075,
+      "step": 313
+    },
+    {
+      "epoch": 0.022633893173790817,
+      "grad_norm": 0.22544192421511822,
+      "learning_rate": 0.00015059952038369305,
+      "loss": 1.3557,
+      "step": 314
+    },
+    {
+      "epoch": 0.02270597563612773,
+      "grad_norm": 0.18120652206128085,
+      "learning_rate": 0.00015107913669064747,
+      "loss": 1.362,
+      "step": 315
+    },
+    {
+      "epoch": 0.022778058098464643,
+      "grad_norm": 0.41805601387242114,
+      "learning_rate": 0.00015155875299760192,
+      "loss": 1.4964,
+      "step": 316
+    },
+    {
+      "epoch": 0.022850140560801556,
+      "grad_norm": 0.16764057308301805,
+      "learning_rate": 0.00015203836930455637,
+      "loss": 1.3949,
+      "step": 317
+    },
+    {
+      "epoch": 0.02292222302313847,
+      "grad_norm": 0.2241880887114489,
+      "learning_rate": 0.0001525179856115108,
+      "loss": 1.4776,
+      "step": 318
+    },
+    {
+      "epoch": 0.022994305485475385,
+      "grad_norm": 0.25008672448149294,
+      "learning_rate": 0.00015299760191846525,
+      "loss": 1.3796,
+      "step": 319
+    },
+    {
+      "epoch": 0.023066387947812298,
+      "grad_norm": 0.26532250490779896,
+      "learning_rate": 0.00015347721822541968,
+      "loss": 1.3618,
+      "step": 320
+    },
+    {
+      "epoch": 0.02313847041014921,
+      "grad_norm": 0.2317185573993854,
+      "learning_rate": 0.0001539568345323741,
+      "loss": 1.5262,
+      "step": 321
+    },
+    {
+      "epoch": 0.023210552872486124,
+      "grad_norm": 0.2248079022647416,
+      "learning_rate": 0.00015443645083932853,
+      "loss": 1.3428,
+      "step": 322
+    },
+    {
+      "epoch": 0.023282635334823037,
+      "grad_norm": 0.14246214648448105,
+      "learning_rate": 0.00015491606714628298,
+      "loss": 1.4276,
+      "step": 323
+    },
+    {
+      "epoch": 0.02335471779715995,
+      "grad_norm": 0.19215800389228962,
+      "learning_rate": 0.00015539568345323743,
+      "loss": 1.3856,
+      "step": 324
+    },
+    {
+      "epoch": 0.023426800259496866,
+      "grad_norm": 0.2815432414831489,
+      "learning_rate": 0.00015587529976019186,
+      "loss": 1.3834,
+      "step": 325
+    },
+    {
+      "epoch": 0.02349888272183378,
+      "grad_norm": 0.2827792049359039,
+      "learning_rate": 0.0001563549160671463,
+      "loss": 1.4272,
+      "step": 326
+    },
+    {
+      "epoch": 0.023570965184170692,
+      "grad_norm": 0.17131844547944175,
+      "learning_rate": 0.00015683453237410073,
+      "loss": 1.379,
+      "step": 327
+    },
+    {
+      "epoch": 0.023643047646507605,
+      "grad_norm": 0.2002634543293578,
+      "learning_rate": 0.00015731414868105516,
+      "loss": 1.4033,
+      "step": 328
+    },
+    {
+      "epoch": 0.023715130108844518,
+      "grad_norm": 0.16793915320150804,
+      "learning_rate": 0.00015779376498800958,
+      "loss": 1.358,
+      "step": 329
+    },
+    {
+      "epoch": 0.02378721257118143,
+      "grad_norm": 0.1653831772614202,
+      "learning_rate": 0.00015827338129496403,
+      "loss": 1.3756,
+      "step": 330
+    },
+    {
+      "epoch": 0.023859295033518343,
+      "grad_norm": 0.15370057464411824,
+      "learning_rate": 0.00015875299760191846,
+      "loss": 1.4037,
+      "step": 331
+    },
+    {
+      "epoch": 0.02393137749585526,
+      "grad_norm": 0.21063123641125006,
+      "learning_rate": 0.0001592326139088729,
+      "loss": 1.4105,
+      "step": 332
+    },
+    {
+      "epoch": 0.024003459958192173,
+      "grad_norm": 0.16592158193683318,
+      "learning_rate": 0.00015971223021582736,
+      "loss": 1.3915,
+      "step": 333
+    },
+    {
+      "epoch": 0.024075542420529086,
+      "grad_norm": 0.23186549237208992,
+      "learning_rate": 0.0001601918465227818,
+      "loss": 1.3891,
+      "step": 334
+    },
+    {
+      "epoch": 0.024147624882866,
+      "grad_norm": 0.20767362856613797,
+      "learning_rate": 0.0001606714628297362,
+      "loss": 1.3232,
+      "step": 335
+    },
+    {
+      "epoch": 0.02421970734520291,
+      "grad_norm": 0.13005043320560059,
+      "learning_rate": 0.00016115107913669067,
+      "loss": 1.3287,
+      "step": 336
+    },
+    {
+      "epoch": 0.024291789807539824,
+      "grad_norm": 0.1489775670687961,
+      "learning_rate": 0.0001616306954436451,
+      "loss": 1.3647,
+      "step": 337
+    },
+    {
+      "epoch": 0.02436387226987674,
+      "grad_norm": 0.24434515666012918,
+      "learning_rate": 0.00016211031175059952,
+      "loss": 1.4228,
+      "step": 338
+    },
+    {
+      "epoch": 0.024435954732213654,
+      "grad_norm": 0.1362400171632745,
+      "learning_rate": 0.00016258992805755397,
+      "loss": 1.4213,
+      "step": 339
+    },
+    {
+      "epoch": 0.024508037194550567,
+      "grad_norm": 0.22750512321480262,
+      "learning_rate": 0.00016306954436450842,
+      "loss": 1.4324,
+      "step": 340
+    },
+    {
+      "epoch": 0.02458011965688748,
+      "grad_norm": 0.3816293003219981,
+      "learning_rate": 0.00016354916067146284,
+      "loss": 1.4842,
+      "step": 341
+    },
+    {
+      "epoch": 0.024652202119224392,
+      "grad_norm": 0.14190843382490267,
+      "learning_rate": 0.00016402877697841727,
+      "loss": 1.4225,
+      "step": 342
+    },
+    {
+      "epoch": 0.024724284581561305,
+      "grad_norm": 0.17984778556331935,
+      "learning_rate": 0.00016450839328537172,
+      "loss": 1.324,
+      "step": 343
+    },
+    {
+      "epoch": 0.024796367043898218,
+      "grad_norm": 0.10990567940476088,
+      "learning_rate": 0.00016498800959232615,
+      "loss": 1.3519,
+      "step": 344
+    },
+    {
+      "epoch": 0.024868449506235134,
+      "grad_norm": 0.1495649557852813,
+      "learning_rate": 0.00016546762589928057,
+      "loss": 1.5465,
+      "step": 345
+    },
+    {
+      "epoch": 0.024940531968572047,
+      "grad_norm": 0.15833043044193237,
+      "learning_rate": 0.000165947242206235,
+      "loss": 1.3618,
+      "step": 346
+    },
+    {
+      "epoch": 0.02501261443090896,
+      "grad_norm": 0.2803016697574425,
+      "learning_rate": 0.00016642685851318948,
+      "loss": 1.4086,
+      "step": 347
+    },
+    {
+      "epoch": 0.025084696893245873,
+      "grad_norm": 0.15132656415561543,
+      "learning_rate": 0.0001669064748201439,
+      "loss": 1.3752,
+      "step": 348
+    },
+    {
+      "epoch": 0.025156779355582786,
+      "grad_norm": 0.16640918225588588,
+      "learning_rate": 0.00016738609112709833,
+      "loss": 1.352,
+      "step": 349
+    },
+    {
+      "epoch": 0.0252288618179197,
+      "grad_norm": 0.14837505222324654,
+      "learning_rate": 0.00016786570743405278,
+      "loss": 1.3221,
+      "step": 350
+    },
+    {
+      "epoch": 0.025300944280256612,
+      "grad_norm": 0.15405662587899005,
+      "learning_rate": 0.0001683453237410072,
+      "loss": 1.3893,
+      "step": 351
+    },
+    {
+      "epoch": 0.025373026742593528,
+      "grad_norm": 0.2542901100176926,
+      "learning_rate": 0.00016882494004796163,
+      "loss": 1.4718,
+      "step": 352
+    },
+    {
+      "epoch": 0.02544510920493044,
+      "grad_norm": 0.15589300381865592,
+      "learning_rate": 0.00016930455635491608,
+      "loss": 1.3781,
+      "step": 353
+    },
+    {
+      "epoch": 0.025517191667267354,
+      "grad_norm": 0.266368848426296,
+      "learning_rate": 0.00016978417266187053,
+      "loss": 1.3599,
+      "step": 354
+    },
+    {
+      "epoch": 0.025589274129604267,
+      "grad_norm": 0.16170900858015505,
+      "learning_rate": 0.00017026378896882496,
+      "loss": 1.3957,
+      "step": 355
+    },
+    {
+      "epoch": 0.02566135659194118,
+      "grad_norm": 0.2606744253402408,
+      "learning_rate": 0.00017074340527577938,
+      "loss": 1.4332,
+      "step": 356
+    },
+    {
+      "epoch": 0.025733439054278093,
+      "grad_norm": 0.12913643689412763,
+      "learning_rate": 0.00017122302158273383,
+      "loss": 1.2619,
+      "step": 357
+    },
+    {
+      "epoch": 0.02580552151661501,
+      "grad_norm": 0.16330059976647973,
+      "learning_rate": 0.00017170263788968826,
+      "loss": 1.4983,
+      "step": 358
+    },
+    {
+      "epoch": 0.025877603978951922,
+      "grad_norm": 0.17035734386296944,
+      "learning_rate": 0.00017218225419664268,
+      "loss": 1.4055,
+      "step": 359
+    },
+    {
+      "epoch": 0.025949686441288835,
+      "grad_norm": 0.13066393233588477,
+      "learning_rate": 0.00017266187050359714,
+      "loss": 1.3139,
+      "step": 360
+    },
+    {
+      "epoch": 0.026021768903625748,
+      "grad_norm": 0.13414662953847595,
+      "learning_rate": 0.00017314148681055156,
+      "loss": 1.4359,
+      "step": 361
+    },
+    {
+      "epoch": 0.02609385136596266,
+      "grad_norm": 0.15068536817391692,
+      "learning_rate": 0.000173621103117506,
+      "loss": 1.4828,
+      "step": 362
+    },
+    {
+      "epoch": 0.026165933828299574,
+      "grad_norm": 0.16050748757779798,
+      "learning_rate": 0.00017410071942446044,
+      "loss": 1.315,
+      "step": 363
+    },
+    {
+      "epoch": 0.026238016290636487,
+      "grad_norm": 0.3498271598958392,
+      "learning_rate": 0.0001745803357314149,
+      "loss": 1.4172,
+      "step": 364
+    },
+    {
+      "epoch": 0.026310098752973403,
+      "grad_norm": 0.1554015133630661,
+      "learning_rate": 0.00017505995203836931,
+      "loss": 1.3409,
+      "step": 365
+    },
+    {
+      "epoch": 0.026382181215310316,
+      "grad_norm": 0.15144835372928747,
+      "learning_rate": 0.00017553956834532374,
+      "loss": 1.4281,
+      "step": 366
+    },
+    {
+      "epoch": 0.02645426367764723,
+      "grad_norm": 0.237275189363755,
+      "learning_rate": 0.0001760191846522782,
+      "loss": 1.4655,
+      "step": 367
+    },
+    {
+      "epoch": 0.02652634613998414,
+      "grad_norm": 0.1525471788375583,
+      "learning_rate": 0.00017649880095923262,
+      "loss": 1.3567,
+      "step": 368
+    },
+    {
+      "epoch": 0.026598428602321054,
+      "grad_norm": 0.3812760835928335,
+      "learning_rate": 0.00017697841726618707,
+      "loss": 1.5034,
+      "step": 369
+    },
+    {
+      "epoch": 0.026670511064657967,
+      "grad_norm": 0.13970473929026084,
+      "learning_rate": 0.0001774580335731415,
+      "loss": 1.3413,
+      "step": 370
+    },
+    {
+      "epoch": 0.026742593526994884,
+      "grad_norm": 0.1299342047807293,
+      "learning_rate": 0.00017793764988009595,
+      "loss": 1.3614,
+      "step": 371
+    },
+    {
+      "epoch": 0.026814675989331797,
+      "grad_norm": 0.13007423088121975,
+      "learning_rate": 0.00017841726618705037,
+      "loss": 1.3527,
+      "step": 372
+    },
+    {
+      "epoch": 0.02688675845166871,
+      "grad_norm": 0.3322571520351329,
+      "learning_rate": 0.0001788968824940048,
+      "loss": 1.4023,
+      "step": 373
+    },
+    {
+      "epoch": 0.026958840914005622,
+      "grad_norm": 0.12601814331326106,
+      "learning_rate": 0.00017937649880095925,
+      "loss": 1.4173,
+      "step": 374
+    },
+    {
+      "epoch": 0.027030923376342535,
+      "grad_norm": 0.1814565665314811,
+      "learning_rate": 0.00017985611510791367,
+      "loss": 1.3944,
+      "step": 375
+    },
+    {
+      "epoch": 0.027103005838679448,
+      "grad_norm": 0.1898521263901962,
+      "learning_rate": 0.0001803357314148681,
+      "loss": 1.3968,
+      "step": 376
+    },
+    {
+      "epoch": 0.02717508830101636,
+      "grad_norm": 0.1796651861167344,
+      "learning_rate": 0.00018081534772182255,
+      "loss": 1.4815,
+      "step": 377
+    },
+    {
+      "epoch": 0.027247170763353278,
+      "grad_norm": 0.1354810672690335,
+      "learning_rate": 0.000181294964028777,
+      "loss": 1.4021,
+      "step": 378
+    },
+    {
+      "epoch": 0.02731925322569019,
+      "grad_norm": 0.15532112902770545,
+      "learning_rate": 0.00018177458033573143,
+      "loss": 1.3862,
+      "step": 379
+    },
+    {
+      "epoch": 0.027391335688027103,
+      "grad_norm": 0.22764646528173044,
+      "learning_rate": 0.00018225419664268585,
+      "loss": 1.4386,
+      "step": 380
+    },
+    {
+      "epoch": 0.027463418150364016,
+      "grad_norm": 0.19181274838226894,
+      "learning_rate": 0.0001827338129496403,
+      "loss": 1.382,
+      "step": 381
+    },
+    {
+      "epoch": 0.02753550061270093,
+      "grad_norm": 0.1470119643552272,
+      "learning_rate": 0.00018321342925659473,
+      "loss": 1.3334,
+      "step": 382
+    },
+    {
+      "epoch": 0.027607583075037842,
+      "grad_norm": 0.11844869722030703,
+      "learning_rate": 0.00018369304556354915,
+      "loss": 1.4247,
+      "step": 383
+    },
+    {
+      "epoch": 0.02767966553737476,
+      "grad_norm": 0.16961063759216014,
+      "learning_rate": 0.0001841726618705036,
+      "loss": 1.4161,
+      "step": 384
+    },
+    {
+      "epoch": 0.02775174799971167,
+      "grad_norm": 0.14673586948538558,
+      "learning_rate": 0.00018465227817745806,
+      "loss": 1.4107,
+      "step": 385
+    },
+    {
+      "epoch": 0.027823830462048584,
+      "grad_norm": 0.45750209845529105,
+      "learning_rate": 0.00018513189448441248,
+      "loss": 1.3855,
+      "step": 386
+    },
+    {
+      "epoch": 0.027895912924385497,
+      "grad_norm": 0.2208575014862173,
+      "learning_rate": 0.0001856115107913669,
+      "loss": 1.377,
+      "step": 387
+    },
+    {
+      "epoch": 0.02796799538672241,
+      "grad_norm": 0.1453861858625522,
+      "learning_rate": 0.00018609112709832136,
+      "loss": 1.3149,
+      "step": 388
+    },
+    {
+      "epoch": 0.028040077849059323,
+      "grad_norm": 0.31345302548269793,
+      "learning_rate": 0.00018657074340527579,
+      "loss": 1.3622,
+      "step": 389
+    },
+    {
+      "epoch": 0.028112160311396236,
+      "grad_norm": 0.19515965433479623,
+      "learning_rate": 0.0001870503597122302,
+      "loss": 1.2857,
+      "step": 390
+    },
+    {
+      "epoch": 0.028184242773733152,
+      "grad_norm": 0.12203296321702718,
+      "learning_rate": 0.00018752997601918466,
+      "loss": 1.3628,
+      "step": 391
+    },
+    {
+      "epoch": 0.028256325236070065,
+      "grad_norm": 0.2984470263182884,
+      "learning_rate": 0.00018800959232613911,
+      "loss": 1.3503,
+      "step": 392
+    },
+    {
+      "epoch": 0.028328407698406978,
+      "grad_norm": 0.21993435952764329,
+      "learning_rate": 0.00018848920863309354,
+      "loss": 1.3735,
+      "step": 393
+    },
+    {
+      "epoch": 0.02840049016074389,
+      "grad_norm": 0.13622092484641032,
+      "learning_rate": 0.00018896882494004796,
+      "loss": 1.3756,
+      "step": 394
+    },
+    {
+      "epoch": 0.028472572623080804,
+      "grad_norm": 0.13710171364592968,
+      "learning_rate": 0.00018944844124700242,
+      "loss": 1.4374,
+      "step": 395
+    },
+    {
+      "epoch": 0.028544655085417717,
+      "grad_norm": 0.15046173554642417,
+      "learning_rate": 0.00018992805755395684,
+      "loss": 1.2698,
+      "step": 396
+    },
+    {
+      "epoch": 0.02861673754775463,
+      "grad_norm": 0.15018576770816022,
+      "learning_rate": 0.00019040767386091127,
+      "loss": 1.3947,
+      "step": 397
+    },
+    {
+      "epoch": 0.028688820010091546,
+      "grad_norm": 0.15077898448222182,
+      "learning_rate": 0.00019088729016786572,
+      "loss": 1.4552,
+      "step": 398
+    },
+    {
+      "epoch": 0.02876090247242846,
+      "grad_norm": 0.1538788236324912,
+      "learning_rate": 0.00019136690647482017,
+      "loss": 1.3146,
+      "step": 399
+    },
+    {
+      "epoch": 0.028832984934765372,
+      "grad_norm": 0.25703254434114786,
+      "learning_rate": 0.0001918465227817746,
+      "loss": 1.372,
+      "step": 400
+    },
+    {
+      "epoch": 0.028905067397102285,
+      "grad_norm": 0.21353830962702838,
+      "learning_rate": 0.00019232613908872902,
+      "loss": 1.3959,
+      "step": 401
+    },
+    {
+      "epoch": 0.028977149859439198,
+      "grad_norm": 0.18105836825128216,
+      "learning_rate": 0.00019280575539568347,
+      "loss": 1.4465,
+      "step": 402
+    },
+    {
+      "epoch": 0.02904923232177611,
+      "grad_norm": 0.1608094471939143,
+      "learning_rate": 0.0001932853717026379,
+      "loss": 1.3536,
+      "step": 403
+    },
+    {
+      "epoch": 0.029121314784113027,
+      "grad_norm": 0.13425376115024684,
+      "learning_rate": 0.00019376498800959232,
+      "loss": 1.3964,
+      "step": 404
+    },
+    {
+      "epoch": 0.02919339724644994,
+      "grad_norm": 0.37068467392556875,
+      "learning_rate": 0.00019424460431654677,
+      "loss": 1.4291,
+      "step": 405
+    },
+    {
+      "epoch": 0.029265479708786853,
+      "grad_norm": 0.1365056654457888,
+      "learning_rate": 0.00019472422062350123,
+      "loss": 1.5029,
+      "step": 406
+    },
+    {
+      "epoch": 0.029337562171123766,
+      "grad_norm": 0.10841763850686444,
+      "learning_rate": 0.00019520383693045565,
+      "loss": 1.3012,
+      "step": 407
+    },
+    {
+      "epoch": 0.02940964463346068,
+      "grad_norm": 0.1271940429684462,
+      "learning_rate": 0.00019568345323741008,
+      "loss": 1.3834,
+      "step": 408
+    },
+    {
+      "epoch": 0.02948172709579759,
+      "grad_norm": 0.19115317985220082,
+      "learning_rate": 0.00019616306954436453,
+      "loss": 1.535,
+      "step": 409
+    },
+    {
+      "epoch": 0.029553809558134504,
+      "grad_norm": 0.12812585289773248,
+      "learning_rate": 0.00019664268585131895,
+      "loss": 1.3818,
+      "step": 410
+    },
+    {
+      "epoch": 0.02962589202047142,
+      "grad_norm": 0.18951744697758063,
+      "learning_rate": 0.00019712230215827338,
+      "loss": 1.4125,
+      "step": 411
+    },
+    {
+      "epoch": 0.029697974482808333,
+      "grad_norm": 0.157652965126673,
+      "learning_rate": 0.00019760191846522783,
+      "loss": 1.3416,
+      "step": 412
+    },
+    {
+      "epoch": 0.029770056945145246,
+      "grad_norm": 0.13687657950953525,
+      "learning_rate": 0.00019808153477218226,
+      "loss": 1.3349,
+      "step": 413
+    },
+    {
+      "epoch": 0.02984213940748216,
+      "grad_norm": 0.2906542215738699,
+      "learning_rate": 0.0001985611510791367,
+      "loss": 1.4167,
+      "step": 414
+    },
+    {
+      "epoch": 0.029914221869819072,
+      "grad_norm": 0.1473929093786515,
+      "learning_rate": 0.00019904076738609113,
+      "loss": 1.3674,
+      "step": 415
+    },
+    {
+      "epoch": 0.029986304332155985,
+      "grad_norm": 0.11820504800829863,
+      "learning_rate": 0.00019952038369304558,
+      "loss": 1.3859,
+      "step": 416
+    },
+    {
+      "epoch": 0.0300583867944929,
+      "grad_norm": 0.12198781022712447,
+      "learning_rate": 0.0002,
+      "loss": 1.4384,
+      "step": 417
+    },
+    {
+      "epoch": 0.030130469256829814,
+      "grad_norm": 0.16952146032801904,
+      "learning_rate": 0.0001999999972745527,
+      "loss": 1.4525,
+      "step": 418
+    },
+    {
+      "epoch": 0.030202551719166727,
+      "grad_norm": 0.1244996576211083,
+      "learning_rate": 0.00019999998909821088,
+      "loss": 1.453,
+      "step": 419
+    },
+    {
+      "epoch": 0.03027463418150364,
+      "grad_norm": 0.12551310526259457,
+      "learning_rate": 0.0001999999754709751,
+      "loss": 1.44,
+      "step": 420
+    },
+    {
+      "epoch": 0.030346716643840553,
+      "grad_norm": 0.1433928850961495,
+      "learning_rate": 0.000199999956392846,
+      "loss": 1.415,
+      "step": 421
+    },
+    {
+      "epoch": 0.030418799106177466,
+      "grad_norm": 0.47032267671817135,
+      "learning_rate": 0.00019999993186382462,
+      "loss": 1.4993,
+      "step": 422
+    },
+    {
+      "epoch": 0.03049088156851438,
+      "grad_norm": 0.17642505587303167,
+      "learning_rate": 0.00019999990188391235,
+      "loss": 1.3921,
+      "step": 423
+    },
+    {
+      "epoch": 0.030562964030851295,
+      "grad_norm": 0.24650116247405743,
+      "learning_rate": 0.0001999998664531108,
+      "loss": 1.3825,
+      "step": 424
+    },
+    {
+      "epoch": 0.030635046493188208,
+      "grad_norm": 0.13901583645794405,
+      "learning_rate": 0.0001999998255714219,
+      "loss": 1.3668,
+      "step": 425
+    },
+    {
+      "epoch": 0.03070712895552512,
+      "grad_norm": 0.12633930988448538,
+      "learning_rate": 0.0001999997792388479,
+      "loss": 1.3581,
+      "step": 426
+    },
+    {
+      "epoch": 0.030779211417862034,
+      "grad_norm": 0.21867671640522726,
+      "learning_rate": 0.0001999997274553913,
+      "loss": 1.3864,
+      "step": 427
+    },
+    {
+      "epoch": 0.030851293880198947,
+      "grad_norm": 0.248392266894441,
+      "learning_rate": 0.00019999967022105492,
+      "loss": 1.3791,
+      "step": 428
+    },
+    {
+      "epoch": 0.03092337634253586,
+      "grad_norm": 0.11580982098796153,
+      "learning_rate": 0.00019999960753584191,
+      "loss": 1.347,
+      "step": 429
+    },
+    {
+      "epoch": 0.030995458804872776,
+      "grad_norm": 0.12446211305591312,
+      "learning_rate": 0.00019999953939975565,
+      "loss": 1.4052,
+      "step": 430
+    },
+    {
+      "epoch": 0.03106754126720969,
+      "grad_norm": 0.13808258694994524,
+      "learning_rate": 0.00019999946581279987,
+      "loss": 1.355,
+      "step": 431
+    },
+    {
+      "epoch": 0.031139623729546602,
+      "grad_norm": 0.731713617817915,
+      "learning_rate": 0.00019999938677497857,
+      "loss": 1.4393,
+      "step": 432
+    },
+    {
+      "epoch": 0.031211706191883515,
+      "grad_norm": 0.17216174577481178,
+      "learning_rate": 0.0001999993022862961,
+      "loss": 1.3108,
+      "step": 433
+    },
+    {
+      "epoch": 0.03128378865422043,
+      "grad_norm": 0.2215314937242399,
+      "learning_rate": 0.00019999921234675703,
+      "loss": 1.4346,
+      "step": 434
+    },
+    {
+      "epoch": 0.03135587111655734,
+      "grad_norm": 0.1165333415358688,
+      "learning_rate": 0.00019999911695636627,
+      "loss": 1.3109,
+      "step": 435
+    },
+    {
+      "epoch": 0.031427953578894254,
+      "grad_norm": 0.2899715529419811,
+      "learning_rate": 0.000199999016115129,
+      "loss": 1.363,
+      "step": 436
+    },
+    {
+      "epoch": 0.031500036041231166,
+      "grad_norm": 0.22243153562723167,
+      "learning_rate": 0.00019999890982305075,
+      "loss": 1.3239,
+      "step": 437
+    },
+    {
+      "epoch": 0.03157211850356808,
+      "grad_norm": 0.1516027801403051,
+      "learning_rate": 0.0001999987980801373,
+      "loss": 1.3932,
+      "step": 438
+    },
+    {
+      "epoch": 0.03164420096590499,
+      "grad_norm": 0.13948422370435432,
+      "learning_rate": 0.00019999868088639472,
+      "loss": 1.4587,
+      "step": 439
+    },
+    {
+      "epoch": 0.03171628342824191,
+      "grad_norm": 0.11854600682115937,
+      "learning_rate": 0.00019999855824182948,
+      "loss": 1.3384,
+      "step": 440
+    },
+    {
+      "epoch": 0.031788365890578825,
+      "grad_norm": 0.20121358738642647,
+      "learning_rate": 0.00019999843014644814,
+      "loss": 1.4168,
+      "step": 441
+    },
+    {
+      "epoch": 0.03186044835291574,
+      "grad_norm": 0.24505644714052335,
+      "learning_rate": 0.00019999829660025776,
+      "loss": 1.4043,
+      "step": 442
+    },
+    {
+      "epoch": 0.03193253081525265,
+      "grad_norm": 0.1851418140300701,
+      "learning_rate": 0.0001999981576032656,
+      "loss": 1.2712,
+      "step": 443
+    },
+    {
+      "epoch": 0.032004613277589564,
+      "grad_norm": 0.1406396703284569,
+      "learning_rate": 0.00019999801315547926,
+      "loss": 1.2954,
+      "step": 444
+    },
+    {
+      "epoch": 0.03207669573992648,
+      "grad_norm": 0.12864704465020463,
+      "learning_rate": 0.00019999786325690657,
+      "loss": 1.3898,
+      "step": 445
+    },
+    {
+      "epoch": 0.03214877820226339,
+      "grad_norm": 0.15972260885657852,
+      "learning_rate": 0.00019999770790755575,
+      "loss": 1.3887,
+      "step": 446
+    },
+    {
+      "epoch": 0.0322208606646003,
+      "grad_norm": 0.19992849467759316,
+      "learning_rate": 0.00019999754710743525,
+      "loss": 1.3635,
+      "step": 447
+    },
+    {
+      "epoch": 0.032292943126937215,
+      "grad_norm": 0.15589520604951113,
+      "learning_rate": 0.0001999973808565538,
+      "loss": 1.2753,
+      "step": 448
+    },
+    {
+      "epoch": 0.03236502558927413,
+      "grad_norm": 0.2299435011795565,
+      "learning_rate": 0.0001999972091549205,
+      "loss": 1.3941,
+      "step": 449
+    },
+    {
+      "epoch": 0.03243710805161104,
+      "grad_norm": 0.1854607595065118,
+      "learning_rate": 0.0001999970320025447,
+      "loss": 1.4335,
+      "step": 450
+    },
+    {
+      "epoch": 0.032509190513947954,
+      "grad_norm": 0.1553026457195574,
+      "learning_rate": 0.00019999684939943607,
+      "loss": 1.4535,
+      "step": 451
+    },
+    {
+      "epoch": 0.03258127297628487,
+      "grad_norm": 0.2830098517247314,
+      "learning_rate": 0.00019999666134560453,
+      "loss": 1.3987,
+      "step": 452
+    },
+    {
+      "epoch": 0.03265335543862179,
+      "grad_norm": 0.21058535664732686,
+      "learning_rate": 0.00019999646784106035,
+      "loss": 1.3802,
+      "step": 453
+    },
+    {
+      "epoch": 0.0327254379009587,
+      "grad_norm": 0.21894609147445598,
+      "learning_rate": 0.00019999626888581414,
+      "loss": 1.4054,
+      "step": 454
+    },
+    {
+      "epoch": 0.03279752036329561,
+      "grad_norm": 0.16955092699803676,
+      "learning_rate": 0.00019999606447987663,
+      "loss": 1.3763,
+      "step": 455
+    },
+    {
+      "epoch": 0.032869602825632525,
+      "grad_norm": 0.12930809958579845,
+      "learning_rate": 0.000199995854623259,
+      "loss": 1.4224,
+      "step": 456
+    },
+    {
+      "epoch": 0.03294168528796944,
+      "grad_norm": 0.13438712985767484,
+      "learning_rate": 0.0001999956393159727,
+      "loss": 1.3606,
+      "step": 457
+    },
+    {
+      "epoch": 0.03301376775030635,
+      "grad_norm": 0.13695012196645162,
+      "learning_rate": 0.00019999541855802953,
+      "loss": 1.4237,
+      "step": 458
+    },
+    {
+      "epoch": 0.033085850212643264,
+      "grad_norm": 0.20606244638700527,
+      "learning_rate": 0.00019999519234944143,
+      "loss": 1.3788,
+      "step": 459
+    },
+    {
+      "epoch": 0.03315793267498018,
+      "grad_norm": 0.13488271314319725,
+      "learning_rate": 0.00019999496069022076,
+      "loss": 1.342,
+      "step": 460
+    },
+    {
+      "epoch": 0.03323001513731709,
+      "grad_norm": 0.14975900326691413,
+      "learning_rate": 0.00019999472358038015,
+      "loss": 1.4471,
+      "step": 461
+    },
+    {
+      "epoch": 0.033302097599654,
+      "grad_norm": 0.13456904917268603,
+      "learning_rate": 0.00019999448101993251,
+      "loss": 1.4296,
+      "step": 462
+    },
+    {
+      "epoch": 0.033374180061990916,
+      "grad_norm": 0.11398309829889225,
+      "learning_rate": 0.00019999423300889112,
+      "loss": 1.4277,
+      "step": 463
+    },
+    {
+      "epoch": 0.03344626252432783,
+      "grad_norm": 0.10455620623616561,
+      "learning_rate": 0.00019999397954726945,
+      "loss": 1.3915,
+      "step": 464
+    },
+    {
+      "epoch": 0.03351834498666474,
+      "grad_norm": 0.166199707796551,
+      "learning_rate": 0.0001999937206350813,
+      "loss": 1.5438,
+      "step": 465
+    },
+    {
+      "epoch": 0.03359042744900166,
+      "grad_norm": 0.11848244390802715,
+      "learning_rate": 0.00019999345627234083,
+      "loss": 1.4492,
+      "step": 466
+    },
+    {
+      "epoch": 0.033662509911338574,
+      "grad_norm": 0.23124975388335825,
+      "learning_rate": 0.00019999318645906242,
+      "loss": 1.4298,
+      "step": 467
+    },
+    {
+      "epoch": 0.03373459237367549,
+      "grad_norm": 0.12358325967116329,
+      "learning_rate": 0.00019999291119526076,
+      "loss": 1.5129,
+      "step": 468
+    },
+    {
+      "epoch": 0.0338066748360124,
+      "grad_norm": 0.1230535606377174,
+      "learning_rate": 0.00019999263048095091,
+      "loss": 1.4138,
+      "step": 469
+    },
+    {
+      "epoch": 0.03387875729834931,
+      "grad_norm": 0.1166516824835502,
+      "learning_rate": 0.00019999234431614812,
+      "loss": 1.3409,
+      "step": 470
+    },
+    {
+      "epoch": 0.033950839760686226,
+      "grad_norm": 0.22941557667605872,
+      "learning_rate": 0.000199992052700868,
+      "loss": 1.3076,
+      "step": 471
+    },
+    {
+      "epoch": 0.03402292222302314,
+      "grad_norm": 0.12675302863722918,
+      "learning_rate": 0.00019999175563512647,
+      "loss": 1.3652,
+      "step": 472
+    },
+    {
+      "epoch": 0.03409500468536005,
+      "grad_norm": 0.11115379282903287,
+      "learning_rate": 0.0001999914531189397,
+      "loss": 1.3117,
+      "step": 473
+    },
+    {
+      "epoch": 0.034167087147696965,
+      "grad_norm": 0.1212816850459076,
+      "learning_rate": 0.0001999911451523242,
+      "loss": 1.3845,
+      "step": 474
+    },
+    {
+      "epoch": 0.03423916961003388,
+      "grad_norm": 0.10995675614311369,
+      "learning_rate": 0.00019999083173529673,
+      "loss": 1.3293,
+      "step": 475
+    },
+    {
+      "epoch": 0.03431125207237079,
+      "grad_norm": 0.1947504386500701,
+      "learning_rate": 0.00019999051286787438,
+      "loss": 1.3642,
+      "step": 476
+    },
+    {
+      "epoch": 0.0343833345347077,
+      "grad_norm": 0.1236530567177871,
+      "learning_rate": 0.00019999018855007455,
+      "loss": 1.3725,
+      "step": 477
+    },
+    {
+      "epoch": 0.034455416997044616,
+      "grad_norm": 0.19775578403246286,
+      "learning_rate": 0.00019998985878191488,
+      "loss": 1.4484,
+      "step": 478
+    },
+    {
+      "epoch": 0.03452749945938153,
+      "grad_norm": 0.19833265744187778,
+      "learning_rate": 0.0001999895235634134,
+      "loss": 1.3608,
+      "step": 479
+    },
+    {
+      "epoch": 0.03459958192171845,
+      "grad_norm": 0.10325087582276218,
+      "learning_rate": 0.00019998918289458833,
+      "loss": 1.4084,
+      "step": 480
+    },
+    {
+      "epoch": 0.03467166438405536,
+      "grad_norm": 0.1057932623819157,
+      "learning_rate": 0.0001999888367754583,
+      "loss": 1.4364,
+      "step": 481
+    },
+    {
+      "epoch": 0.034743746846392275,
+      "grad_norm": 0.12986284166996537,
+      "learning_rate": 0.00019998848520604213,
+      "loss": 1.3803,
+      "step": 482
+    },
+    {
+      "epoch": 0.03481582930872919,
+      "grad_norm": 0.18198574192918032,
+      "learning_rate": 0.00019998812818635897,
+      "loss": 1.3481,
+      "step": 483
+    },
+    {
+      "epoch": 0.0348879117710661,
+      "grad_norm": 0.1038465491323711,
+      "learning_rate": 0.00019998776571642833,
+      "loss": 1.419,
+      "step": 484
+    },
+    {
+      "epoch": 0.03495999423340301,
+      "grad_norm": 0.2340522110108444,
+      "learning_rate": 0.00019998739779626993,
+      "loss": 1.349,
+      "step": 485
+    },
+    {
+      "epoch": 0.035032076695739926,
+      "grad_norm": 0.112084394495797,
+      "learning_rate": 0.0001999870244259038,
+      "loss": 1.4415,
+      "step": 486
+    },
+    {
+      "epoch": 0.03510415915807684,
+      "grad_norm": 0.12229865403316786,
+      "learning_rate": 0.00019998664560535038,
+      "loss": 1.3963,
+      "step": 487
+    },
+    {
+      "epoch": 0.03517624162041375,
+      "grad_norm": 0.18425683875102458,
+      "learning_rate": 0.00019998626133463027,
+      "loss": 1.4547,
+      "step": 488
+    },
+    {
+      "epoch": 0.035248324082750665,
+      "grad_norm": 0.1795140170370395,
+      "learning_rate": 0.00019998587161376442,
+      "loss": 1.3581,
+      "step": 489
+    },
+    {
+      "epoch": 0.03532040654508758,
+      "grad_norm": 0.13874135312092403,
+      "learning_rate": 0.00019998547644277403,
+      "loss": 1.4102,
+      "step": 490
+    },
+    {
+      "epoch": 0.03539248900742449,
+      "grad_norm": 0.11479306111488326,
+      "learning_rate": 0.0001999850758216807,
+      "loss": 1.3699,
+      "step": 491
+    },
+    {
+      "epoch": 0.035464571469761404,
+      "grad_norm": 0.12888045517345884,
+      "learning_rate": 0.00019998466975050625,
+      "loss": 1.3389,
+      "step": 492
+    },
+    {
+      "epoch": 0.035536653932098324,
+      "grad_norm": 0.13896892521729448,
+      "learning_rate": 0.0001999842582292728,
+      "loss": 1.2768,
+      "step": 493
+    },
+    {
+      "epoch": 0.035608736394435236,
+      "grad_norm": 0.11952591050277801,
+      "learning_rate": 0.00019998384125800278,
+      "loss": 1.4673,
+      "step": 494
+    },
+    {
+      "epoch": 0.03568081885677215,
+      "grad_norm": 0.11367020744220863,
+      "learning_rate": 0.00019998341883671893,
+      "loss": 1.3545,
+      "step": 495
+    },
+    {
+      "epoch": 0.03575290131910906,
+      "grad_norm": 0.12345222683684523,
+      "learning_rate": 0.0001999829909654443,
+      "loss": 1.4163,
+      "step": 496
+    },
+    {
+      "epoch": 0.035824983781445975,
+      "grad_norm": 0.10843980540736448,
+      "learning_rate": 0.00019998255764420216,
+      "loss": 1.28,
+      "step": 497
+    },
+    {
+      "epoch": 0.03589706624378289,
+      "grad_norm": 0.1412869137825975,
+      "learning_rate": 0.00019998211887301622,
+      "loss": 1.4261,
+      "step": 498
+    },
+    {
+      "epoch": 0.0359691487061198,
+      "grad_norm": 0.11275150997376084,
+      "learning_rate": 0.00019998167465191027,
+      "loss": 1.4165,
+      "step": 499
+    },
+    {
+      "epoch": 0.036041231168456714,
+      "grad_norm": 0.19576621636570868,
+      "learning_rate": 0.00019998122498090864,
+      "loss": 1.4388,
+      "step": 500
+    },
+    {
+      "epoch": 0.03611331363079363,
+      "grad_norm": 0.11627815612682331,
+      "learning_rate": 0.00019998076986003574,
+      "loss": 1.465,
+      "step": 501
+    },
+    {
+      "epoch": 0.03618539609313054,
+      "grad_norm": 0.1324416261935126,
+      "learning_rate": 0.00019998030928931645,
+      "loss": 1.4665,
+      "step": 502
+    },
+    {
+      "epoch": 0.03625747855546745,
+      "grad_norm": 0.10435637119652807,
+      "learning_rate": 0.00019997984326877588,
+      "loss": 1.3533,
+      "step": 503
+    },
+    {
+      "epoch": 0.036329561017804365,
+      "grad_norm": 0.1301846486145787,
+      "learning_rate": 0.00019997937179843937,
+      "loss": 1.4104,
+      "step": 504
+    },
+    {
+      "epoch": 0.03640164348014128,
+      "grad_norm": 0.14758848944043684,
+      "learning_rate": 0.00019997889487833267,
+      "loss": 1.4049,
+      "step": 505
+    },
+    {
+      "epoch": 0.0364737259424782,
+      "grad_norm": 0.25392965528074657,
+      "learning_rate": 0.00019997841250848176,
+      "loss": 1.3351,
+      "step": 506
+    },
+    {
+      "epoch": 0.03654580840481511,
+      "grad_norm": 0.11065178460904067,
+      "learning_rate": 0.00019997792468891295,
+      "loss": 1.3543,
+      "step": 507
+    },
+    {
+      "epoch": 0.036617890867152024,
+      "grad_norm": 0.11692979897482475,
+      "learning_rate": 0.0001999774314196528,
+      "loss": 1.3827,
+      "step": 508
+    },
+    {
+      "epoch": 0.03668997332948894,
+      "grad_norm": 0.1131099633530901,
+      "learning_rate": 0.00019997693270072824,
+      "loss": 1.3589,
+      "step": 509
+    },
+    {
+      "epoch": 0.03676205579182585,
+      "grad_norm": 0.12280658875278318,
+      "learning_rate": 0.00019997642853216641,
+      "loss": 1.4333,
+      "step": 510
+    },
+    {
+      "epoch": 0.03683413825416276,
+      "grad_norm": 0.12695934752081375,
+      "learning_rate": 0.00019997591891399478,
+      "loss": 1.3788,
+      "step": 511
+    },
+    {
+      "epoch": 0.036906220716499676,
+      "grad_norm": 0.13704262257683641,
+      "learning_rate": 0.0001999754038462412,
+      "loss": 1.378,
+      "step": 512
+    },
+    {
+      "epoch": 0.03697830317883659,
+      "grad_norm": 0.13433642346793712,
+      "learning_rate": 0.00019997488332893371,
+      "loss": 1.4637,
+      "step": 513
+    },
+    {
+      "epoch": 0.0370503856411735,
+      "grad_norm": 0.12494638398623496,
+      "learning_rate": 0.00019997435736210064,
+      "loss": 1.4765,
+      "step": 514
+    },
+    {
+      "epoch": 0.037122468103510414,
+      "grad_norm": 0.13177736702216114,
+      "learning_rate": 0.00019997382594577073,
+      "loss": 1.4244,
+      "step": 515
+    },
+    {
+      "epoch": 0.03719455056584733,
+      "grad_norm": 0.2499081911922163,
+      "learning_rate": 0.0001999732890799729,
+      "loss": 1.4623,
+      "step": 516
+    },
+    {
+      "epoch": 0.03726663302818424,
+      "grad_norm": 0.11492969899303354,
+      "learning_rate": 0.0001999727467647364,
+      "loss": 1.4335,
+      "step": 517
+    },
+    {
+      "epoch": 0.03733871549052115,
+      "grad_norm": 0.11705140496046389,
+      "learning_rate": 0.00019997219900009085,
+      "loss": 1.3678,
+      "step": 518
+    },
+    {
+      "epoch": 0.03741079795285807,
+      "grad_norm": 0.12258196134761369,
+      "learning_rate": 0.00019997164578606608,
+      "loss": 1.3964,
+      "step": 519
+    },
+    {
+      "epoch": 0.037482880415194986,
+      "grad_norm": 0.1354601280495836,
+      "learning_rate": 0.00019997108712269227,
+      "loss": 1.4213,
+      "step": 520
+    },
+    {
+      "epoch": 0.0375549628775319,
+      "grad_norm": 0.1847495080695212,
+      "learning_rate": 0.00019997052300999978,
+      "loss": 1.3815,
+      "step": 521
+    },
+    {
+      "epoch": 0.03762704533986881,
+      "grad_norm": 0.15907282921964908,
+      "learning_rate": 0.0001999699534480195,
+      "loss": 1.4532,
+      "step": 522
+    },
+    {
+      "epoch": 0.037699127802205724,
+      "grad_norm": 0.24111546114027732,
+      "learning_rate": 0.00019996937843678233,
+      "loss": 1.2743,
+      "step": 523
+    },
+    {
+      "epoch": 0.03777121026454264,
+      "grad_norm": 0.10837608633922668,
+      "learning_rate": 0.0001999687979763197,
+      "loss": 1.324,
+      "step": 524
+    },
+    {
+      "epoch": 0.03784329272687955,
+      "grad_norm": 0.17588592967731762,
+      "learning_rate": 0.00019996821206666327,
+      "loss": 1.3152,
+      "step": 525
+    },
+    {
+      "epoch": 0.03791537518921646,
+      "grad_norm": 0.1211022586024491,
+      "learning_rate": 0.0001999676207078449,
+      "loss": 1.3778,
+      "step": 526
+    },
+    {
+      "epoch": 0.037987457651553376,
+      "grad_norm": 0.14093144414790323,
+      "learning_rate": 0.00019996702389989693,
+      "loss": 1.3074,
+      "step": 527
+    },
+    {
+      "epoch": 0.03805954011389029,
+      "grad_norm": 0.13606125222100607,
+      "learning_rate": 0.00019996642164285175,
+      "loss": 1.4,
+      "step": 528
+    },
+    {
+      "epoch": 0.0381316225762272,
+      "grad_norm": 0.10784819550287889,
+      "learning_rate": 0.0001999658139367423,
+      "loss": 1.4033,
+      "step": 529
+    },
+    {
+      "epoch": 0.038203705038564115,
+      "grad_norm": 0.1144025883333083,
+      "learning_rate": 0.00019996520078160166,
+      "loss": 1.4657,
+      "step": 530
+    },
+    {
+      "epoch": 0.03827578750090103,
+      "grad_norm": 0.11791143880464845,
+      "learning_rate": 0.00019996458217746329,
+      "loss": 1.3072,
+      "step": 531
+    },
+    {
+      "epoch": 0.03834786996323795,
+      "grad_norm": 0.2217403583802677,
+      "learning_rate": 0.00019996395812436085,
+      "loss": 1.4089,
+      "step": 532
+    },
+    {
+      "epoch": 0.03841995242557486,
+      "grad_norm": 0.10589832219668156,
+      "learning_rate": 0.0001999633286223284,
+      "loss": 1.4329,
+      "step": 533
+    },
+    {
+      "epoch": 0.03849203488791177,
+      "grad_norm": 0.12753746756023393,
+      "learning_rate": 0.00019996269367140025,
+      "loss": 1.4421,
+      "step": 534
+    },
+    {
+      "epoch": 0.038564117350248686,
+      "grad_norm": 0.11787856483281713,
+      "learning_rate": 0.000199962053271611,
+      "loss": 1.3668,
+      "step": 535
+    },
+    {
+      "epoch": 0.0386361998125856,
+      "grad_norm": 0.2373105958230719,
+      "learning_rate": 0.00019996140742299553,
+      "loss": 1.3395,
+      "step": 536
+    },
+    {
+      "epoch": 0.03870828227492251,
+      "grad_norm": 0.21899678927536834,
+      "learning_rate": 0.00019996075612558915,
+      "loss": 1.3036,
+      "step": 537
+    },
+    {
+      "epoch": 0.038780364737259425,
+      "grad_norm": 0.12113592367984363,
+      "learning_rate": 0.00019996009937942724,
+      "loss": 1.4046,
+      "step": 538
+    },
+    {
+      "epoch": 0.03885244719959634,
+      "grad_norm": 0.13484420619869386,
+      "learning_rate": 0.00019995943718454563,
+      "loss": 1.372,
+      "step": 539
+    },
+    {
+      "epoch": 0.03892452966193325,
+      "grad_norm": 0.11538684464128836,
+      "learning_rate": 0.00019995876954098044,
+      "loss": 1.2801,
+      "step": 540
+    },
+    {
+      "epoch": 0.038996612124270164,
+      "grad_norm": 0.15113256873012712,
+      "learning_rate": 0.00019995809644876805,
+      "loss": 1.3232,
+      "step": 541
+    },
+    {
+      "epoch": 0.039068694586607076,
+      "grad_norm": 0.1543010199617122,
+      "learning_rate": 0.00019995741790794518,
+      "loss": 1.4221,
+      "step": 542
+    },
+    {
+      "epoch": 0.03914077704894399,
+      "grad_norm": 0.08928483597982043,
+      "learning_rate": 0.00019995673391854876,
+      "loss": 1.3199,
+      "step": 543
+    },
+    {
+      "epoch": 0.0392128595112809,
+      "grad_norm": 0.12389629834890897,
+      "learning_rate": 0.00019995604448061612,
+      "loss": 1.4069,
+      "step": 544
+    },
+    {
+      "epoch": 0.03928494197361782,
+      "grad_norm": 0.10993608354863656,
+      "learning_rate": 0.00019995534959418482,
+      "loss": 1.4479,
+      "step": 545
+    },
+    {
+      "epoch": 0.039357024435954735,
+      "grad_norm": 0.1751656464522022,
+      "learning_rate": 0.00019995464925929275,
+      "loss": 1.4479,
+      "step": 546
+    },
+    {
+      "epoch": 0.03942910689829165,
+      "grad_norm": 0.17316254735727335,
+      "learning_rate": 0.00019995394347597804,
+      "loss": 1.367,
+      "step": 547
+    },
+    {
+      "epoch": 0.03950118936062856,
+      "grad_norm": 0.10046629407991384,
+      "learning_rate": 0.00019995323224427923,
+      "loss": 1.3568,
+      "step": 548
+    },
+    {
+      "epoch": 0.039573271822965474,
+      "grad_norm": 0.1418440858393439,
+      "learning_rate": 0.00019995251556423503,
+      "loss": 1.4008,
+      "step": 549
+    },
+    {
+      "epoch": 0.03964535428530239,
+      "grad_norm": 0.17473980824596713,
+      "learning_rate": 0.00019995179343588454,
+      "loss": 1.3238,
+      "step": 550
+    },
+    {
+      "epoch": 0.0397174367476393,
+      "grad_norm": 0.10974426676524919,
+      "learning_rate": 0.00019995106585926715,
+      "loss": 1.3776,
+      "step": 551
+    },
+    {
+      "epoch": 0.03978951920997621,
+      "grad_norm": 0.1234532838526836,
+      "learning_rate": 0.00019995033283442247,
+      "loss": 1.2884,
+      "step": 552
+    },
+    {
+      "epoch": 0.039861601672313125,
+      "grad_norm": 0.1555335423958548,
+      "learning_rate": 0.00019994959436139045,
+      "loss": 1.3353,
+      "step": 553
+    },
+    {
+      "epoch": 0.03993368413465004,
+      "grad_norm": 0.15991595432602648,
+      "learning_rate": 0.00019994885044021134,
+      "loss": 1.3656,
+      "step": 554
+    },
+    {
+      "epoch": 0.04000576659698695,
+      "grad_norm": 0.14863121638763327,
+      "learning_rate": 0.00019994810107092575,
+      "loss": 1.3057,
+      "step": 555
+    },
+    {
+      "epoch": 0.040077849059323864,
+      "grad_norm": 0.10888581050827832,
+      "learning_rate": 0.0001999473462535745,
+      "loss": 1.3443,
+      "step": 556
+    },
+    {
+      "epoch": 0.04014993152166078,
+      "grad_norm": 0.22401062460953067,
+      "learning_rate": 0.0001999465859881987,
+      "loss": 1.434,
+      "step": 557
+    },
+    {
+      "epoch": 0.0402220139839977,
+      "grad_norm": 0.14550680604554764,
+      "learning_rate": 0.0001999458202748398,
+      "loss": 1.371,
+      "step": 558
+    },
+    {
+      "epoch": 0.04029409644633461,
+      "grad_norm": 0.35487209182169055,
+      "learning_rate": 0.0001999450491135396,
+      "loss": 1.397,
+      "step": 559
+    },
+    {
+      "epoch": 0.04036617890867152,
+      "grad_norm": 0.11155932626125913,
+      "learning_rate": 0.00019994427250434005,
+      "loss": 1.3371,
+      "step": 560
+    },
+    {
+      "epoch": 0.040438261371008435,
+      "grad_norm": 0.16780874713957136,
+      "learning_rate": 0.00019994349044728354,
+      "loss": 1.4073,
+      "step": 561
+    },
+    {
+      "epoch": 0.04051034383334535,
+      "grad_norm": 0.11497603756189778,
+      "learning_rate": 0.00019994270294241266,
+      "loss": 1.3572,
+      "step": 562
+    },
+    {
+      "epoch": 0.04058242629568226,
+      "grad_norm": 0.11665604677352001,
+      "learning_rate": 0.0001999419099897704,
+      "loss": 1.2901,
+      "step": 563
+    },
+    {
+      "epoch": 0.040654508758019174,
+      "grad_norm": 0.13678074333418844,
+      "learning_rate": 0.00019994111158939989,
+      "loss": 1.4506,
+      "step": 564
+    },
+    {
+      "epoch": 0.04072659122035609,
+      "grad_norm": 0.17309000418459683,
+      "learning_rate": 0.00019994030774134474,
+      "loss": 1.4223,
+      "step": 565
+    },
+    {
+      "epoch": 0.040798673682693,
+      "grad_norm": 0.1613609423937569,
+      "learning_rate": 0.00019993949844564869,
+      "loss": 1.35,
+      "step": 566
+    },
+    {
+      "epoch": 0.04087075614502991,
+      "grad_norm": 0.14192023246989277,
+      "learning_rate": 0.00019993868370235592,
+      "loss": 1.4162,
+      "step": 567
+    },
+    {
+      "epoch": 0.040942838607366826,
+      "grad_norm": 0.13718811698925795,
+      "learning_rate": 0.00019993786351151077,
+      "loss": 1.3744,
+      "step": 568
+    },
+    {
+      "epoch": 0.04101492106970374,
+      "grad_norm": 0.24923845824299792,
+      "learning_rate": 0.00019993703787315803,
+      "loss": 1.4188,
+      "step": 569
+    },
+    {
+      "epoch": 0.04108700353204065,
+      "grad_norm": 0.10489833321517082,
+      "learning_rate": 0.00019993620678734265,
+      "loss": 1.3848,
+      "step": 570
+    },
+    {
+      "epoch": 0.04115908599437757,
+      "grad_norm": 0.13022950589375049,
+      "learning_rate": 0.00019993537025410997,
+      "loss": 1.4575,
+      "step": 571
+    },
+    {
+      "epoch": 0.041231168456714484,
+      "grad_norm": 0.13664212035302345,
+      "learning_rate": 0.00019993452827350553,
+      "loss": 1.4087,
+      "step": 572
+    },
+    {
+      "epoch": 0.0413032509190514,
+      "grad_norm": 0.1384382843274937,
+      "learning_rate": 0.00019993368084557527,
+      "loss": 1.2804,
+      "step": 573
+    },
+    {
+      "epoch": 0.04137533338138831,
+      "grad_norm": 0.11345713383031641,
+      "learning_rate": 0.0001999328279703654,
+      "loss": 1.3543,
+      "step": 574
+    },
+    {
+      "epoch": 0.04144741584372522,
+      "grad_norm": 0.126989071213864,
+      "learning_rate": 0.00019993196964792235,
+      "loss": 1.3715,
+      "step": 575
+    },
+    {
+      "epoch": 0.041519498306062136,
+      "grad_norm": 0.13636696344093832,
+      "learning_rate": 0.00019993110587829293,
+      "loss": 1.261,
+      "step": 576
+    },
+    {
+      "epoch": 0.04159158076839905,
+      "grad_norm": 0.14210404116247755,
+      "learning_rate": 0.00019993023666152425,
+      "loss": 1.4745,
+      "step": 577
+    },
+    {
+      "epoch": 0.04166366323073596,
+      "grad_norm": 0.1123523235680516,
+      "learning_rate": 0.00019992936199766365,
+      "loss": 1.3849,
+      "step": 578
+    },
+    {
+      "epoch": 0.041735745693072875,
+      "grad_norm": 0.10914270664617051,
+      "learning_rate": 0.00019992848188675884,
+      "loss": 1.3553,
+      "step": 579
+    },
+    {
+      "epoch": 0.04180782815540979,
+      "grad_norm": 0.16847713050327143,
+      "learning_rate": 0.0001999275963288578,
+      "loss": 1.4103,
+      "step": 580
+    },
+    {
+      "epoch": 0.0418799106177467,
+      "grad_norm": 0.17059630958559588,
+      "learning_rate": 0.00019992670532400874,
+      "loss": 1.4781,
+      "step": 581
+    },
+    {
+      "epoch": 0.04195199308008361,
+      "grad_norm": 0.10492890764657087,
+      "learning_rate": 0.0001999258088722603,
+      "loss": 1.3837,
+      "step": 582
+    },
+    {
+      "epoch": 0.042024075542420526,
+      "grad_norm": 0.18281885470966508,
+      "learning_rate": 0.00019992490697366127,
+      "loss": 1.3962,
+      "step": 583
+    },
+    {
+      "epoch": 0.04209615800475744,
+      "grad_norm": 0.3082860460568362,
+      "learning_rate": 0.0001999239996282609,
+      "loss": 1.4622,
+      "step": 584
+    },
+    {
+      "epoch": 0.04216824046709436,
+      "grad_norm": 0.11360451424119485,
+      "learning_rate": 0.0001999230868361086,
+      "loss": 1.4363,
+      "step": 585
+    },
+    {
+      "epoch": 0.04224032292943127,
+      "grad_norm": 0.14014483240142234,
+      "learning_rate": 0.00019992216859725413,
+      "loss": 1.5031,
+      "step": 586
+    },
+    {
+      "epoch": 0.042312405391768185,
+      "grad_norm": 0.11169150766556982,
+      "learning_rate": 0.00019992124491174758,
+      "loss": 1.29,
+      "step": 587
+    },
+    {
+      "epoch": 0.0423844878541051,
+      "grad_norm": 0.2955901908553905,
+      "learning_rate": 0.00019992031577963922,
+      "loss": 1.4583,
+      "step": 588
+    },
+    {
+      "epoch": 0.04245657031644201,
+      "grad_norm": 0.10923986954123346,
+      "learning_rate": 0.00019991938120097973,
+      "loss": 1.4225,
+      "step": 589
+    },
+    {
+      "epoch": 0.04252865277877892,
+      "grad_norm": 0.11984287835170823,
+      "learning_rate": 0.00019991844117582007,
+      "loss": 1.3897,
+      "step": 590
+    },
+    {
+      "epoch": 0.042600735241115836,
+      "grad_norm": 0.23831325769017064,
+      "learning_rate": 0.00019991749570421146,
+      "loss": 1.4627,
+      "step": 591
+    },
+    {
+      "epoch": 0.04267281770345275,
+      "grad_norm": 0.09638615392798987,
+      "learning_rate": 0.00019991654478620547,
+      "loss": 1.2933,
+      "step": 592
+    },
+    {
+      "epoch": 0.04274490016578966,
+      "grad_norm": 0.11515593924499484,
+      "learning_rate": 0.00019991558842185388,
+      "loss": 1.4043,
+      "step": 593
+    },
+    {
+      "epoch": 0.042816982628126575,
+      "grad_norm": 0.12149414300798345,
+      "learning_rate": 0.00019991462661120887,
+      "loss": 1.3763,
+      "step": 594
+    },
+    {
+      "epoch": 0.04288906509046349,
+      "grad_norm": 0.1168875747337534,
+      "learning_rate": 0.00019991365935432284,
+      "loss": 1.4413,
+      "step": 595
+    },
+    {
+      "epoch": 0.0429611475528004,
+      "grad_norm": 0.13727602036089823,
+      "learning_rate": 0.00019991268665124853,
+      "loss": 1.3644,
+      "step": 596
+    },
+    {
+      "epoch": 0.043033230015137314,
+      "grad_norm": 0.10660171717179136,
+      "learning_rate": 0.00019991170850203896,
+      "loss": 1.4129,
+      "step": 597
+    },
+    {
+      "epoch": 0.043105312477474234,
+      "grad_norm": 0.12105994729079271,
+      "learning_rate": 0.00019991072490674742,
+      "loss": 1.3539,
+      "step": 598
+    },
+    {
+      "epoch": 0.043177394939811146,
+      "grad_norm": 0.11473401717783793,
+      "learning_rate": 0.00019990973586542755,
+      "loss": 1.3624,
+      "step": 599
+    },
+    {
+      "epoch": 0.04324947740214806,
+      "grad_norm": 0.13567281355425062,
+      "learning_rate": 0.00019990874137813328,
+      "loss": 1.4585,
+      "step": 600
+    },
+    {
+      "epoch": 0.04332155986448497,
+      "grad_norm": 0.13196715109755913,
+      "learning_rate": 0.00019990774144491876,
+      "loss": 1.5336,
+      "step": 601
+    },
+    {
+      "epoch": 0.043393642326821885,
+      "grad_norm": 0.11534847084449042,
+      "learning_rate": 0.00019990673606583856,
+      "loss": 1.3438,
+      "step": 602
+    },
+    {
+      "epoch": 0.0434657247891588,
+      "grad_norm": 0.1128149094067151,
+      "learning_rate": 0.00019990572524094743,
+      "loss": 1.3128,
+      "step": 603
+    },
+    {
+      "epoch": 0.04353780725149571,
+      "grad_norm": 0.13967589775700126,
+      "learning_rate": 0.0001999047089703005,
+      "loss": 1.4225,
+      "step": 604
+    },
+    {
+      "epoch": 0.043609889713832624,
+      "grad_norm": 0.2081051926412406,
+      "learning_rate": 0.00019990368725395318,
+      "loss": 1.4136,
+      "step": 605
+    },
+    {
+      "epoch": 0.04368197217616954,
+      "grad_norm": 0.10357442053465479,
+      "learning_rate": 0.00019990266009196114,
+      "loss": 1.2616,
+      "step": 606
+    },
+    {
+      "epoch": 0.04375405463850645,
+      "grad_norm": 0.128806728588424,
+      "learning_rate": 0.00019990162748438034,
+      "loss": 1.3424,
+      "step": 607
+    },
+    {
+      "epoch": 0.04382613710084336,
+      "grad_norm": 0.20296335529410922,
+      "learning_rate": 0.0001999005894312671,
+      "loss": 1.4591,
+      "step": 608
+    },
+    {
+      "epoch": 0.043898219563180275,
+      "grad_norm": 0.11744581777474322,
+      "learning_rate": 0.00019989954593267804,
+      "loss": 1.3355,
+      "step": 609
+    },
+    {
+      "epoch": 0.04397030202551719,
+      "grad_norm": 0.11575863455433952,
+      "learning_rate": 0.00019989849698866995,
+      "loss": 1.3219,
+      "step": 610
+    },
+    {
+      "epoch": 0.04404238448785411,
+      "grad_norm": 0.2272415480215067,
+      "learning_rate": 0.0001998974425993001,
+      "loss": 1.3513,
+      "step": 611
+    },
+    {
+      "epoch": 0.04411446695019102,
+      "grad_norm": 0.19139528565269665,
+      "learning_rate": 0.0001998963827646259,
+      "loss": 1.3638,
+      "step": 612
+    },
+    {
+      "epoch": 0.044186549412527934,
+      "grad_norm": 0.12274470570198333,
+      "learning_rate": 0.00019989531748470515,
+      "loss": 1.3711,
+      "step": 613
+    },
+    {
+      "epoch": 0.04425863187486485,
+      "grad_norm": 0.11089174429991616,
+      "learning_rate": 0.00019989424675959588,
+      "loss": 1.3481,
+      "step": 614
+    },
+    {
+      "epoch": 0.04433071433720176,
+      "grad_norm": 0.11975887747656319,
+      "learning_rate": 0.00019989317058935648,
+      "loss": 1.3834,
+      "step": 615
+    },
+    {
+      "epoch": 0.04440279679953867,
+      "grad_norm": 0.14368197981052053,
+      "learning_rate": 0.00019989208897404565,
+      "loss": 1.4234,
+      "step": 616
+    },
+    {
+      "epoch": 0.044474879261875586,
+      "grad_norm": 0.10673699203372436,
+      "learning_rate": 0.0001998910019137223,
+      "loss": 1.3715,
+      "step": 617
+    },
+    {
+      "epoch": 0.0445469617242125,
+      "grad_norm": 0.1345934053599682,
+      "learning_rate": 0.00019988990940844568,
+      "loss": 1.3146,
+      "step": 618
+    },
+    {
+      "epoch": 0.04461904418654941,
+      "grad_norm": 0.1126467462569645,
+      "learning_rate": 0.00019988881145827536,
+      "loss": 1.3458,
+      "step": 619
+    },
+    {
+      "epoch": 0.044691126648886324,
+      "grad_norm": 0.13572014842338803,
+      "learning_rate": 0.0001998877080632712,
+      "loss": 1.3051,
+      "step": 620
+    },
+    {
+      "epoch": 0.04476320911122324,
+      "grad_norm": 0.13293846788889016,
+      "learning_rate": 0.0001998865992234933,
+      "loss": 1.4475,
+      "step": 621
+    },
+    {
+      "epoch": 0.04483529157356015,
+      "grad_norm": 0.1038965919984643,
+      "learning_rate": 0.00019988548493900216,
+      "loss": 1.3315,
+      "step": 622
+    },
+    {
+      "epoch": 0.04490737403589706,
+      "grad_norm": 0.12170689516951189,
+      "learning_rate": 0.0001998843652098585,
+      "loss": 1.3694,
+      "step": 623
+    },
+    {
+      "epoch": 0.04497945649823398,
+      "grad_norm": 0.11487478197507911,
+      "learning_rate": 0.0001998832400361233,
+      "loss": 1.2699,
+      "step": 624
+    },
+    {
+      "epoch": 0.045051538960570896,
+      "grad_norm": 0.14363534966439945,
+      "learning_rate": 0.00019988210941785796,
+      "loss": 1.3185,
+      "step": 625
+    },
+    {
+      "epoch": 0.04512362142290781,
+      "grad_norm": 0.10370169815388791,
+      "learning_rate": 0.0001998809733551241,
+      "loss": 1.2814,
+      "step": 626
+    },
+    {
+      "epoch": 0.04519570388524472,
+      "grad_norm": 0.103025342499429,
+      "learning_rate": 0.00019987983184798362,
+      "loss": 1.3291,
+      "step": 627
+    },
+    {
+      "epoch": 0.045267786347581634,
+      "grad_norm": 0.24138477504497605,
+      "learning_rate": 0.00019987868489649874,
+      "loss": 1.4008,
+      "step": 628
+    },
+    {
+      "epoch": 0.04533986880991855,
+      "grad_norm": 0.12953398261602955,
+      "learning_rate": 0.000199877532500732,
+      "loss": 1.375,
+      "step": 629
+    },
+    {
+      "epoch": 0.04541195127225546,
+      "grad_norm": 0.11127011411977725,
+      "learning_rate": 0.00019987637466074626,
+      "loss": 1.3675,
+      "step": 630
+    },
+    {
+      "epoch": 0.04548403373459237,
+      "grad_norm": 0.12198483877764124,
+      "learning_rate": 0.00019987521137660453,
+      "loss": 1.2428,
+      "step": 631
+    },
+    {
+      "epoch": 0.045556116196929286,
+      "grad_norm": 0.18484866893163826,
+      "learning_rate": 0.00019987404264837025,
+      "loss": 1.4084,
+      "step": 632
+    },
+    {
+      "epoch": 0.0456281986592662,
+      "grad_norm": 0.11391943470869435,
+      "learning_rate": 0.00019987286847610716,
+      "loss": 1.2654,
+      "step": 633
+    },
+    {
+      "epoch": 0.04570028112160311,
+      "grad_norm": 0.11905478324132125,
+      "learning_rate": 0.00019987168885987928,
+      "loss": 1.4094,
+      "step": 634
+    },
+    {
+      "epoch": 0.045772363583940025,
+      "grad_norm": 0.1350916621441763,
+      "learning_rate": 0.00019987050379975085,
+      "loss": 1.3387,
+      "step": 635
+    },
+    {
+      "epoch": 0.04584444604627694,
+      "grad_norm": 0.104553202057287,
+      "learning_rate": 0.0001998693132957865,
+      "loss": 1.3607,
+      "step": 636
+    },
+    {
+      "epoch": 0.04591652850861386,
+      "grad_norm": 0.16592179034308102,
+      "learning_rate": 0.00019986811734805116,
+      "loss": 1.4176,
+      "step": 637
+    },
+    {
+      "epoch": 0.04598861097095077,
+      "grad_norm": 0.10560881051129284,
+      "learning_rate": 0.00019986691595660994,
+      "loss": 1.4022,
+      "step": 638
+    },
+    {
+      "epoch": 0.04606069343328768,
+      "grad_norm": 0.1130870319450636,
+      "learning_rate": 0.00019986570912152837,
+      "loss": 1.277,
+      "step": 639
+    },
+    {
+      "epoch": 0.046132775895624596,
+      "grad_norm": 0.36523992475889466,
+      "learning_rate": 0.00019986449684287222,
+      "loss": 1.3971,
+      "step": 640
+    },
+    {
+      "epoch": 0.04620485835796151,
+      "grad_norm": 0.11727524949872137,
+      "learning_rate": 0.0001998632791207076,
+      "loss": 1.2702,
+      "step": 641
+    },
+    {
+      "epoch": 0.04627694082029842,
+      "grad_norm": 0.1191511241672131,
+      "learning_rate": 0.00019986205595510086,
+      "loss": 1.3233,
+      "step": 642
+    },
+    {
+      "epoch": 0.046349023282635335,
+      "grad_norm": 0.22601841732482184,
+      "learning_rate": 0.00019986082734611867,
+      "loss": 1.327,
+      "step": 643
+    },
+    {
+      "epoch": 0.04642110574497225,
+      "grad_norm": 0.27965743306455204,
+      "learning_rate": 0.000199859593293828,
+      "loss": 1.4001,
+      "step": 644
+    },
+    {
+      "epoch": 0.04649318820730916,
+      "grad_norm": 0.11833395332428068,
+      "learning_rate": 0.00019985835379829616,
+      "loss": 1.4254,
+      "step": 645
+    },
+    {
+      "epoch": 0.046565270669646074,
+      "grad_norm": 0.12375427001451303,
+      "learning_rate": 0.00019985710885959065,
+      "loss": 1.4129,
+      "step": 646
+    },
+    {
+      "epoch": 0.046637353131982986,
+      "grad_norm": 0.14516159165752596,
+      "learning_rate": 0.0001998558584777794,
+      "loss": 1.3734,
+      "step": 647
+    },
+    {
+      "epoch": 0.0467094355943199,
+      "grad_norm": 0.23516834461808211,
+      "learning_rate": 0.00019985460265293048,
+      "loss": 1.3719,
+      "step": 648
+    },
+    {
+      "epoch": 0.04678151805665681,
+      "grad_norm": 0.12006213804205365,
+      "learning_rate": 0.00019985334138511237,
+      "loss": 1.3099,
+      "step": 649
+    },
+    {
+      "epoch": 0.04685360051899373,
+      "grad_norm": 0.1200774832823618,
+      "learning_rate": 0.0001998520746743939,
+      "loss": 1.3919,
+      "step": 650
+    },
+    {
+      "epoch": 0.046925682981330645,
+      "grad_norm": 0.1474806633311758,
+      "learning_rate": 0.00019985080252084406,
+      "loss": 1.2637,
+      "step": 651
+    },
+    {
+      "epoch": 0.04699776544366756,
+      "grad_norm": 0.2367578418357703,
+      "learning_rate": 0.00019984952492453216,
+      "loss": 1.3866,
+      "step": 652
+    },
+    {
+      "epoch": 0.04706984790600447,
+      "grad_norm": 0.1314812059592217,
+      "learning_rate": 0.00019984824188552786,
+      "loss": 1.4029,
+      "step": 653
+    },
+    {
+      "epoch": 0.047141930368341384,
+      "grad_norm": 0.1166672261912521,
+      "learning_rate": 0.00019984695340390113,
+      "loss": 1.3509,
+      "step": 654
+    },
+    {
+      "epoch": 0.0472140128306783,
+      "grad_norm": 0.11849861476556392,
+      "learning_rate": 0.00019984565947972217,
+      "loss": 1.4042,
+      "step": 655
+    },
+    {
+      "epoch": 0.04728609529301521,
+      "grad_norm": 0.12407906842865328,
+      "learning_rate": 0.00019984436011306157,
+      "loss": 1.342,
+      "step": 656
+    },
+    {
+      "epoch": 0.04735817775535212,
+      "grad_norm": 0.2494203644402783,
+      "learning_rate": 0.00019984305530399007,
+      "loss": 1.2821,
+      "step": 657
+    },
+    {
+      "epoch": 0.047430260217689035,
+      "grad_norm": 0.12510333453741287,
+      "learning_rate": 0.00019984174505257882,
+      "loss": 1.3512,
+      "step": 658
+    },
+    {
+      "epoch": 0.04750234268002595,
+      "grad_norm": 0.12499537116216534,
+      "learning_rate": 0.00019984042935889928,
+      "loss": 1.3536,
+      "step": 659
+    },
+    {
+      "epoch": 0.04757442514236286,
+      "grad_norm": 0.3287816446713974,
+      "learning_rate": 0.00019983910822302315,
+      "loss": 1.3977,
+      "step": 660
+    },
+    {
+      "epoch": 0.047646507604699774,
+      "grad_norm": 0.11591546439591649,
+      "learning_rate": 0.0001998377816450224,
+      "loss": 1.4232,
+      "step": 661
+    },
+    {
+      "epoch": 0.04771859006703669,
+      "grad_norm": 0.12974032592810547,
+      "learning_rate": 0.00019983644962496942,
+      "loss": 1.4454,
+      "step": 662
+    },
+    {
+      "epoch": 0.04779067252937361,
+      "grad_norm": 0.11186104961979962,
+      "learning_rate": 0.00019983511216293675,
+      "loss": 1.313,
+      "step": 663
+    },
+    {
+      "epoch": 0.04786275499171052,
+      "grad_norm": 0.13426220845341763,
+      "learning_rate": 0.00019983376925899733,
+      "loss": 1.3293,
+      "step": 664
+    },
+    {
+      "epoch": 0.04793483745404743,
+      "grad_norm": 0.09164975752934408,
+      "learning_rate": 0.00019983242091322433,
+      "loss": 1.3458,
+      "step": 665
+    },
+    {
+      "epoch": 0.048006919916384345,
+      "grad_norm": 0.10878417229432297,
+      "learning_rate": 0.00019983106712569127,
+      "loss": 1.2288,
+      "step": 666
+    },
+    {
+      "epoch": 0.04807900237872126,
+      "grad_norm": 0.11313045464175234,
+      "learning_rate": 0.00019982970789647194,
+      "loss": 1.3829,
+      "step": 667
+    },
+    {
+      "epoch": 0.04815108484105817,
+      "grad_norm": 0.09784757397791267,
+      "learning_rate": 0.0001998283432256404,
+      "loss": 1.3354,
+      "step": 668
+    },
+    {
+      "epoch": 0.048223167303395084,
+      "grad_norm": 0.1125817358693705,
+      "learning_rate": 0.0001998269731132711,
+      "loss": 1.368,
+      "step": 669
+    },
+    {
+      "epoch": 0.048295249765732,
+      "grad_norm": 0.11330323393438912,
+      "learning_rate": 0.0001998255975594387,
+      "loss": 1.3154,
+      "step": 670
+    },
+    {
+      "epoch": 0.04836733222806891,
+      "grad_norm": 0.11031064765089844,
+      "learning_rate": 0.0001998242165642181,
+      "loss": 1.3696,
+      "step": 671
+    },
+    {
+      "epoch": 0.04843941469040582,
+      "grad_norm": 0.11343591572054597,
+      "learning_rate": 0.00019982283012768472,
+      "loss": 1.4054,
+      "step": 672
+    },
+    {
+      "epoch": 0.048511497152742736,
+      "grad_norm": 0.1041482288191251,
+      "learning_rate": 0.00019982143824991402,
+      "loss": 1.2963,
+      "step": 673
+    },
+    {
+      "epoch": 0.04858357961507965,
+      "grad_norm": 0.11375883654196418,
+      "learning_rate": 0.00019982004093098193,
+      "loss": 1.4065,
+      "step": 674
+    },
+    {
+      "epoch": 0.04865566207741656,
+      "grad_norm": 0.12668179324203188,
+      "learning_rate": 0.00019981863817096458,
+      "loss": 1.3476,
+      "step": 675
+    },
+    {
+      "epoch": 0.04872774453975348,
+      "grad_norm": 0.14646171414830425,
+      "learning_rate": 0.00019981722996993844,
+      "loss": 1.3814,
+      "step": 676
+    },
+    {
+      "epoch": 0.048799827002090394,
+      "grad_norm": 0.24513520025423835,
+      "learning_rate": 0.00019981581632798028,
+      "loss": 1.4307,
+      "step": 677
+    },
+    {
+      "epoch": 0.04887190946442731,
+      "grad_norm": 0.13171503122567832,
+      "learning_rate": 0.00019981439724516716,
+      "loss": 1.379,
+      "step": 678
+    },
+    {
+      "epoch": 0.04894399192676422,
+      "grad_norm": 0.11202736721946475,
+      "learning_rate": 0.00019981297272157639,
+      "loss": 1.3916,
+      "step": 679
+    },
+    {
+      "epoch": 0.04901607438910113,
+      "grad_norm": 0.12088115005351471,
+      "learning_rate": 0.0001998115427572857,
+      "loss": 1.399,
+      "step": 680
+    },
+    {
+      "epoch": 0.049088156851438046,
+      "grad_norm": 0.1340866487188867,
+      "learning_rate": 0.00019981010735237295,
+      "loss": 1.3863,
+      "step": 681
+    },
+    {
+      "epoch": 0.04916023931377496,
+      "grad_norm": 0.1505661304983566,
+      "learning_rate": 0.00019980866650691647,
+      "loss": 1.3294,
+      "step": 682
+    },
+    {
+      "epoch": 0.04923232177611187,
+      "grad_norm": 0.1264083958350432,
+      "learning_rate": 0.0001998072202209947,
+      "loss": 1.3372,
+      "step": 683
+    },
+    {
+      "epoch": 0.049304404238448785,
+      "grad_norm": 0.16616982791690074,
+      "learning_rate": 0.00019980576849468656,
+      "loss": 1.348,
+      "step": 684
+    },
+    {
+      "epoch": 0.0493764867007857,
+      "grad_norm": 0.114585214754043,
+      "learning_rate": 0.00019980431132807117,
+      "loss": 1.4465,
+      "step": 685
+    },
+    {
+      "epoch": 0.04944856916312261,
+      "grad_norm": 0.11964148610557963,
+      "learning_rate": 0.00019980284872122788,
+      "loss": 1.2969,
+      "step": 686
+    },
+    {
+      "epoch": 0.04952065162545952,
+      "grad_norm": 0.1392522090491423,
+      "learning_rate": 0.0001998013806742365,
+      "loss": 1.3826,
+      "step": 687
+    },
+    {
+      "epoch": 0.049592734087796436,
+      "grad_norm": 0.13463920533635257,
+      "learning_rate": 0.00019979990718717704,
+      "loss": 1.3967,
+      "step": 688
+    },
+    {
+      "epoch": 0.04966481655013335,
+      "grad_norm": 0.1285870132931408,
+      "learning_rate": 0.00019979842826012978,
+      "loss": 1.4043,
+      "step": 689
+    },
+    {
+      "epoch": 0.04973689901247027,
+      "grad_norm": 0.14775276695900563,
+      "learning_rate": 0.00019979694389317537,
+      "loss": 1.336,
+      "step": 690
+    },
+    {
+      "epoch": 0.04980898147480718,
+      "grad_norm": 0.11968938184167995,
+      "learning_rate": 0.00019979545408639471,
+      "loss": 1.4302,
+      "step": 691
+    },
+    {
+      "epoch": 0.049881063937144095,
+      "grad_norm": 0.20184630536006715,
+      "learning_rate": 0.000199793958839869,
+      "loss": 1.3526,
+      "step": 692
+    },
+    {
+      "epoch": 0.04995314639948101,
+      "grad_norm": 0.1380560713865779,
+      "learning_rate": 0.00019979245815367978,
+      "loss": 1.4276,
+      "step": 693
+    },
+    {
+      "epoch": 0.05002522886181792,
+      "grad_norm": 0.10269166682107261,
+      "learning_rate": 0.00019979095202790877,
+      "loss": 1.3124,
+      "step": 694
+    },
+    {
+      "epoch": 0.05009731132415483,
+      "grad_norm": 0.09825477940510507,
+      "learning_rate": 0.00019978944046263815,
+      "loss": 1.4088,
+      "step": 695
+    },
+    {
+      "epoch": 0.050169393786491746,
+      "grad_norm": 0.10111995898113697,
+      "learning_rate": 0.0001997879234579503,
+      "loss": 1.3746,
+      "step": 696
+    },
+    {
+      "epoch": 0.05024147624882866,
+      "grad_norm": 0.3030073061036525,
+      "learning_rate": 0.00019978640101392785,
+      "loss": 1.3594,
+      "step": 697
+    },
+    {
+      "epoch": 0.05031355871116557,
+      "grad_norm": 0.1158360349522765,
+      "learning_rate": 0.00019978487313065385,
+      "loss": 1.4475,
+      "step": 698
+    },
+    {
+      "epoch": 0.050385641173502485,
+      "grad_norm": 0.10868782904188841,
+      "learning_rate": 0.00019978333980821158,
+      "loss": 1.3088,
+      "step": 699
+    },
+    {
+      "epoch": 0.0504577236358394,
+      "grad_norm": 0.2203730587533625,
+      "learning_rate": 0.00019978180104668458,
+      "loss": 1.4405,
+      "step": 700
+    },
+    {
+      "epoch": 0.05052980609817631,
+      "grad_norm": 0.10889115950046425,
+      "learning_rate": 0.00019978025684615674,
+      "loss": 1.4427,
+      "step": 701
+    },
+    {
+      "epoch": 0.050601888560513224,
+      "grad_norm": 0.12844208463694054,
+      "learning_rate": 0.00019977870720671228,
+      "loss": 1.4299,
+      "step": 702
+    },
+    {
+      "epoch": 0.050673971022850144,
+      "grad_norm": 0.23710555884026027,
+      "learning_rate": 0.0001997771521284356,
+      "loss": 1.4343,
+      "step": 703
+    },
+    {
+      "epoch": 0.050746053485187056,
+      "grad_norm": 0.14891313506561688,
+      "learning_rate": 0.00019977559161141152,
+      "loss": 1.443,
+      "step": 704
+    },
+    {
+      "epoch": 0.05081813594752397,
+      "grad_norm": 0.12212374187379625,
+      "learning_rate": 0.00019977402565572505,
+      "loss": 1.3584,
+      "step": 705
+    },
+    {
+      "epoch": 0.05089021840986088,
+      "grad_norm": 0.1332994688140711,
+      "learning_rate": 0.00019977245426146161,
+      "loss": 1.4204,
+      "step": 706
+    },
+    {
+      "epoch": 0.050962300872197795,
+      "grad_norm": 0.1840825897686932,
+      "learning_rate": 0.0001997708774287068,
+      "loss": 1.3126,
+      "step": 707
+    },
+    {
+      "epoch": 0.05103438333453471,
+      "grad_norm": 0.11680164031419712,
+      "learning_rate": 0.00019976929515754657,
+      "loss": 1.4362,
+      "step": 708
+    },
+    {
+      "epoch": 0.05110646579687162,
+      "grad_norm": 0.10182977993348243,
+      "learning_rate": 0.00019976770744806725,
+      "loss": 1.2829,
+      "step": 709
+    },
+    {
+      "epoch": 0.051178548259208534,
+      "grad_norm": 0.12126246153705827,
+      "learning_rate": 0.0001997661143003553,
+      "loss": 1.3945,
+      "step": 710
+    },
+    {
+      "epoch": 0.05125063072154545,
+      "grad_norm": 0.13648107188975586,
+      "learning_rate": 0.0001997645157144976,
+      "loss": 1.4526,
+      "step": 711
+    },
+    {
+      "epoch": 0.05132271318388236,
+      "grad_norm": 0.13716460351048979,
+      "learning_rate": 0.00019976291169058127,
+      "loss": 1.3527,
+      "step": 712
+    },
+    {
+      "epoch": 0.05139479564621927,
+      "grad_norm": 0.12089357754635356,
+      "learning_rate": 0.00019976130222869375,
+      "loss": 1.2826,
+      "step": 713
+    },
+    {
+      "epoch": 0.051466878108556186,
+      "grad_norm": 0.18441933236894867,
+      "learning_rate": 0.00019975968732892274,
+      "loss": 1.4059,
+      "step": 714
+    },
+    {
+      "epoch": 0.0515389605708931,
+      "grad_norm": 0.10911728120415677,
+      "learning_rate": 0.0001997580669913563,
+      "loss": 1.389,
+      "step": 715
+    },
+    {
+      "epoch": 0.05161104303323002,
+      "grad_norm": 0.1040442616102749,
+      "learning_rate": 0.00019975644121608278,
+      "loss": 1.3341,
+      "step": 716
+    },
+    {
+      "epoch": 0.05168312549556693,
+      "grad_norm": 0.1105898167589363,
+      "learning_rate": 0.00019975481000319073,
+      "loss": 1.3857,
+      "step": 717
+    },
+    {
+      "epoch": 0.051755207957903844,
+      "grad_norm": 0.29701330677992327,
+      "learning_rate": 0.00019975317335276915,
+      "loss": 1.3674,
+      "step": 718
+    },
+    {
+      "epoch": 0.05182729042024076,
+      "grad_norm": 0.11954463610103343,
+      "learning_rate": 0.00019975153126490717,
+      "loss": 1.4257,
+      "step": 719
+    },
+    {
+      "epoch": 0.05189937288257767,
+      "grad_norm": 0.13563501089678262,
+      "learning_rate": 0.00019974988373969434,
+      "loss": 1.4854,
+      "step": 720
+    },
+    {
+      "epoch": 0.05197145534491458,
+      "grad_norm": 0.27486557043460375,
+      "learning_rate": 0.00019974823077722044,
+      "loss": 1.3997,
+      "step": 721
+    },
+    {
+      "epoch": 0.052043537807251496,
+      "grad_norm": 0.12402204709754643,
+      "learning_rate": 0.00019974657237757564,
+      "loss": 1.3036,
+      "step": 722
+    },
+    {
+      "epoch": 0.05211562026958841,
+      "grad_norm": 0.10951960528589343,
+      "learning_rate": 0.00019974490854085026,
+      "loss": 1.349,
+      "step": 723
+    },
+    {
+      "epoch": 0.05218770273192532,
+      "grad_norm": 0.12275564965987101,
+      "learning_rate": 0.000199743239267135,
+      "loss": 1.3059,
+      "step": 724
+    },
+    {
+      "epoch": 0.052259785194262234,
+      "grad_norm": 0.10183756085665789,
+      "learning_rate": 0.0001997415645565209,
+      "loss": 1.3537,
+      "step": 725
+    },
+    {
+      "epoch": 0.05233186765659915,
+      "grad_norm": 0.1038152092000094,
+      "learning_rate": 0.0001997398844090992,
+      "loss": 1.3438,
+      "step": 726
+    },
+    {
+      "epoch": 0.05240395011893606,
+      "grad_norm": 0.13252819368345964,
+      "learning_rate": 0.0001997381988249615,
+      "loss": 1.2745,
+      "step": 727
+    },
+    {
+      "epoch": 0.05247603258127297,
+      "grad_norm": 0.16236215205569884,
+      "learning_rate": 0.00019973650780419968,
+      "loss": 1.3985,
+      "step": 728
+    },
+    {
+      "epoch": 0.05254811504360989,
+      "grad_norm": 0.11787723338841961,
+      "learning_rate": 0.00019973481134690592,
+      "loss": 1.4251,
+      "step": 729
+    },
+    {
+      "epoch": 0.052620197505946806,
+      "grad_norm": 0.12793212932157733,
+      "learning_rate": 0.00019973310945317271,
+      "loss": 1.3226,
+      "step": 730
+    },
+    {
+      "epoch": 0.05269227996828372,
+      "grad_norm": 0.1480516995860252,
+      "learning_rate": 0.00019973140212309277,
+      "loss": 1.3893,
+      "step": 731
+    },
+    {
+      "epoch": 0.05276436243062063,
+      "grad_norm": 0.10322420220612787,
+      "learning_rate": 0.00019972968935675923,
+      "loss": 1.3167,
+      "step": 732
+    },
+    {
+      "epoch": 0.052836444892957544,
+      "grad_norm": 0.10788132278980742,
+      "learning_rate": 0.00019972797115426537,
+      "loss": 1.3311,
+      "step": 733
+    },
+    {
+      "epoch": 0.05290852735529446,
+      "grad_norm": 0.11545045823534629,
+      "learning_rate": 0.0001997262475157049,
+      "loss": 1.3193,
+      "step": 734
+    },
+    {
+      "epoch": 0.05298060981763137,
+      "grad_norm": 0.252925678990466,
+      "learning_rate": 0.00019972451844117178,
+      "loss": 1.45,
+      "step": 735
+    },
+    {
+      "epoch": 0.05305269227996828,
+      "grad_norm": 0.10111921419456175,
+      "learning_rate": 0.00019972278393076023,
+      "loss": 1.3605,
+      "step": 736
+    },
+    {
+      "epoch": 0.053124774742305196,
+      "grad_norm": 0.11520601766035488,
+      "learning_rate": 0.0001997210439845648,
+      "loss": 1.3972,
+      "step": 737
+    },
+    {
+      "epoch": 0.05319685720464211,
+      "grad_norm": 0.11116023603803275,
+      "learning_rate": 0.00019971929860268035,
+      "loss": 1.3718,
+      "step": 738
+    },
+    {
+      "epoch": 0.05326893966697902,
+      "grad_norm": 0.12211073188248074,
+      "learning_rate": 0.00019971754778520202,
+      "loss": 1.4105,
+      "step": 739
+    },
+    {
+      "epoch": 0.053341022129315935,
+      "grad_norm": 0.11556021098726889,
+      "learning_rate": 0.00019971579153222523,
+      "loss": 1.3537,
+      "step": 740
+    },
+    {
+      "epoch": 0.05341310459165285,
+      "grad_norm": 0.1159158308589622,
+      "learning_rate": 0.00019971402984384574,
+      "loss": 1.4349,
+      "step": 741
+    },
+    {
+      "epoch": 0.05348518705398977,
+      "grad_norm": 0.39023027638227614,
+      "learning_rate": 0.00019971226272015954,
+      "loss": 1.4307,
+      "step": 742
+    },
+    {
+      "epoch": 0.05355726951632668,
+      "grad_norm": 0.14154249003961344,
+      "learning_rate": 0.00019971049016126298,
+      "loss": 1.4726,
+      "step": 743
+    },
+    {
+      "epoch": 0.05362935197866359,
+      "grad_norm": 0.16013887524913187,
+      "learning_rate": 0.00019970871216725265,
+      "loss": 1.3276,
+      "step": 744
+    },
+    {
+      "epoch": 0.053701434441000506,
+      "grad_norm": 0.1187826888964526,
+      "learning_rate": 0.0001997069287382255,
+      "loss": 1.3896,
+      "step": 745
+    },
+    {
+      "epoch": 0.05377351690333742,
+      "grad_norm": 0.10126096818165758,
+      "learning_rate": 0.00019970513987427872,
+      "loss": 1.3075,
+      "step": 746
+    },
+    {
+      "epoch": 0.05384559936567433,
+      "grad_norm": 0.2506970991389486,
+      "learning_rate": 0.00019970334557550983,
+      "loss": 1.3632,
+      "step": 747
+    },
+    {
+      "epoch": 0.053917681828011245,
+      "grad_norm": 0.10876041036411495,
+      "learning_rate": 0.00019970154584201665,
+      "loss": 1.3709,
+      "step": 748
+    },
+    {
+      "epoch": 0.05398976429034816,
+      "grad_norm": 0.15667230552172368,
+      "learning_rate": 0.00019969974067389728,
+      "loss": 1.4043,
+      "step": 749
+    },
+    {
+      "epoch": 0.05406184675268507,
+      "grad_norm": 0.10670841140723177,
+      "learning_rate": 0.00019969793007125006,
+      "loss": 1.3942,
+      "step": 750
+    },
+    {
+      "epoch": 0.054133929215021984,
+      "grad_norm": 0.10615139588393528,
+      "learning_rate": 0.00019969611403417375,
+      "loss": 1.3105,
+      "step": 751
+    },
+    {
+      "epoch": 0.054206011677358897,
+      "grad_norm": 0.11637634914058408,
+      "learning_rate": 0.0001996942925627673,
+      "loss": 1.3323,
+      "step": 752
+    },
+    {
+      "epoch": 0.05427809413969581,
+      "grad_norm": 0.34974872992302203,
+      "learning_rate": 0.00019969246565713005,
+      "loss": 1.3886,
+      "step": 753
+    },
+    {
+      "epoch": 0.05435017660203272,
+      "grad_norm": 0.11675045045145058,
+      "learning_rate": 0.00019969063331736153,
+      "loss": 1.3558,
+      "step": 754
+    },
+    {
+      "epoch": 0.05442225906436964,
+      "grad_norm": 0.10651562956928153,
+      "learning_rate": 0.00019968879554356164,
+      "loss": 1.3665,
+      "step": 755
+    },
+    {
+      "epoch": 0.054494341526706555,
+      "grad_norm": 0.26960093744261676,
+      "learning_rate": 0.00019968695233583056,
+      "loss": 1.3988,
+      "step": 756
+    },
+    {
+      "epoch": 0.05456642398904347,
+      "grad_norm": 0.14143909629278564,
+      "learning_rate": 0.0001996851036942687,
+      "loss": 1.364,
+      "step": 757
+    },
+    {
+      "epoch": 0.05463850645138038,
+      "grad_norm": 0.1099505238120949,
+      "learning_rate": 0.00019968324961897694,
+      "loss": 1.347,
+      "step": 758
+    },
+    {
+      "epoch": 0.054710588913717294,
+      "grad_norm": 0.11455672479696219,
+      "learning_rate": 0.00019968139011005627,
+      "loss": 1.4274,
+      "step": 759
+    },
+    {
+      "epoch": 0.05478267137605421,
+      "grad_norm": 0.3629779056649862,
+      "learning_rate": 0.00019967952516760807,
+      "loss": 1.5021,
+      "step": 760
+    },
+    {
+      "epoch": 0.05485475383839112,
+      "grad_norm": 0.11273823520673892,
+      "learning_rate": 0.00019967765479173398,
+      "loss": 1.389,
+      "step": 761
+    },
+    {
+      "epoch": 0.05492683630072803,
+      "grad_norm": 0.15917853953954425,
+      "learning_rate": 0.00019967577898253596,
+      "loss": 1.3132,
+      "step": 762
+    },
+    {
+      "epoch": 0.054998918763064945,
+      "grad_norm": 0.1691765750424334,
+      "learning_rate": 0.00019967389774011625,
+      "loss": 1.3836,
+      "step": 763
+    },
+    {
+      "epoch": 0.05507100122540186,
+      "grad_norm": 0.13441886299139022,
+      "learning_rate": 0.00019967201106457742,
+      "loss": 1.3595,
+      "step": 764
+    },
+    {
+      "epoch": 0.05514308368773877,
+      "grad_norm": 0.17201254069870342,
+      "learning_rate": 0.0001996701189560223,
+      "loss": 1.3292,
+      "step": 765
+    },
+    {
+      "epoch": 0.055215166150075684,
+      "grad_norm": 0.10537120273687611,
+      "learning_rate": 0.000199668221414554,
+      "loss": 1.4851,
+      "step": 766
+    },
+    {
+      "epoch": 0.0552872486124126,
+      "grad_norm": 0.1431206731582436,
+      "learning_rate": 0.000199666318440276,
+      "loss": 1.2942,
+      "step": 767
+    },
+    {
+      "epoch": 0.05535933107474952,
+      "grad_norm": 0.1256391420747347,
+      "learning_rate": 0.00019966441003329197,
+      "loss": 1.271,
+      "step": 768
+    },
+    {
+      "epoch": 0.05543141353708643,
+      "grad_norm": 0.12037142236259782,
+      "learning_rate": 0.000199662496193706,
+      "loss": 1.498,
+      "step": 769
+    },
+    {
+      "epoch": 0.05550349599942334,
+      "grad_norm": 0.13747488468593033,
+      "learning_rate": 0.0001996605769216224,
+      "loss": 1.2972,
+      "step": 770
+    },
+    {
+      "epoch": 0.055575578461760256,
+      "grad_norm": 0.1124941757379595,
+      "learning_rate": 0.00019965865221714573,
+      "loss": 1.3761,
+      "step": 771
+    },
+    {
+      "epoch": 0.05564766092409717,
+      "grad_norm": 0.2581963789042948,
+      "learning_rate": 0.00019965672208038096,
+      "loss": 1.3097,
+      "step": 772
+    },
+    {
+      "epoch": 0.05571974338643408,
+      "grad_norm": 0.13493453720104984,
+      "learning_rate": 0.00019965478651143328,
+      "loss": 1.3798,
+      "step": 773
+    },
+    {
+      "epoch": 0.055791825848770994,
+      "grad_norm": 0.16213915672900125,
+      "learning_rate": 0.00019965284551040817,
+      "loss": 1.4062,
+      "step": 774
+    },
+    {
+      "epoch": 0.05586390831110791,
+      "grad_norm": 0.35935317256962557,
+      "learning_rate": 0.0001996508990774115,
+      "loss": 1.3855,
+      "step": 775
+    },
+    {
+      "epoch": 0.05593599077344482,
+      "grad_norm": 0.10178673000099885,
+      "learning_rate": 0.00019964894721254933,
+      "loss": 1.4029,
+      "step": 776
+    },
+    {
+      "epoch": 0.05600807323578173,
+      "grad_norm": 0.16817166100409756,
+      "learning_rate": 0.00019964698991592803,
+      "loss": 1.3537,
+      "step": 777
+    },
+    {
+      "epoch": 0.056080155698118646,
+      "grad_norm": 0.12558536273919174,
+      "learning_rate": 0.00019964502718765429,
+      "loss": 1.3604,
+      "step": 778
+    },
+    {
+      "epoch": 0.05615223816045556,
+      "grad_norm": 0.16170590067290852,
+      "learning_rate": 0.0001996430590278352,
+      "loss": 1.3797,
+      "step": 779
+    },
+    {
+      "epoch": 0.05622432062279247,
+      "grad_norm": 0.1860023765270928,
+      "learning_rate": 0.00019964108543657786,
+      "loss": 1.4334,
+      "step": 780
+    },
+    {
+      "epoch": 0.05629640308512939,
+      "grad_norm": 0.11323035930872828,
+      "learning_rate": 0.00019963910641399002,
+      "loss": 1.2657,
+      "step": 781
+    },
+    {
+      "epoch": 0.056368485547466304,
+      "grad_norm": 0.1159242995021525,
+      "learning_rate": 0.00019963712196017946,
+      "loss": 1.2387,
+      "step": 782
+    },
+    {
+      "epoch": 0.05644056800980322,
+      "grad_norm": 0.7975196847892553,
+      "learning_rate": 0.00019963513207525436,
+      "loss": 1.3496,
+      "step": 783
+    },
+    {
+      "epoch": 0.05651265047214013,
+      "grad_norm": 0.2823527159814666,
+      "learning_rate": 0.0001996331367593232,
+      "loss": 1.4902,
+      "step": 784
+    },
+    {
+      "epoch": 0.05658473293447704,
+      "grad_norm": 0.15432748799867205,
+      "learning_rate": 0.00019963113601249478,
+      "loss": 1.3725,
+      "step": 785
+    },
+    {
+      "epoch": 0.056656815396813956,
+      "grad_norm": 0.10477948070788747,
+      "learning_rate": 0.0001996291298348781,
+      "loss": 1.4092,
+      "step": 786
+    },
+    {
+      "epoch": 0.05672889785915087,
+      "grad_norm": 0.11774743030143181,
+      "learning_rate": 0.00019962711822658254,
+      "loss": 1.3918,
+      "step": 787
+    },
+    {
+      "epoch": 0.05680098032148778,
+      "grad_norm": 0.12211522857000855,
+      "learning_rate": 0.00019962510118771772,
+      "loss": 1.4713,
+      "step": 788
+    },
+    {
+      "epoch": 0.056873062783824695,
+      "grad_norm": 0.10026268969450791,
+      "learning_rate": 0.00019962307871839366,
+      "loss": 1.2421,
+      "step": 789
+    },
+    {
+      "epoch": 0.05694514524616161,
+      "grad_norm": 0.1074370757907574,
+      "learning_rate": 0.00019962105081872055,
+      "loss": 1.3857,
+      "step": 790
+    },
+    {
+      "epoch": 0.05701722770849852,
+      "grad_norm": 0.1291380866412019,
+      "learning_rate": 0.00019961901748880888,
+      "loss": 1.3151,
+      "step": 791
+    },
+    {
+      "epoch": 0.05708931017083543,
+      "grad_norm": 0.17651962028013102,
+      "learning_rate": 0.0001996169787287696,
+      "loss": 1.3691,
+      "step": 792
+    },
+    {
+      "epoch": 0.057161392633172346,
+      "grad_norm": 0.1287835984747274,
+      "learning_rate": 0.00019961493453871375,
+      "loss": 1.3153,
+      "step": 793
+    },
+    {
+      "epoch": 0.05723347509550926,
+      "grad_norm": 0.12454725136274242,
+      "learning_rate": 0.00019961288491875278,
+      "loss": 1.3169,
+      "step": 794
+    },
+    {
+      "epoch": 0.05730555755784618,
+      "grad_norm": 0.11100207640874914,
+      "learning_rate": 0.00019961082986899844,
+      "loss": 1.4057,
+      "step": 795
+    },
+    {
+      "epoch": 0.05737764002018309,
+      "grad_norm": 0.1460862297448828,
+      "learning_rate": 0.0001996087693895627,
+      "loss": 1.1972,
+      "step": 796
+    },
+    {
+      "epoch": 0.057449722482520005,
+      "grad_norm": 0.1646216750939346,
+      "learning_rate": 0.00019960670348055794,
+      "loss": 1.3355,
+      "step": 797
+    },
+    {
+      "epoch": 0.05752180494485692,
+      "grad_norm": 0.13649057649624347,
+      "learning_rate": 0.00019960463214209673,
+      "loss": 1.3597,
+      "step": 798
+    },
+    {
+      "epoch": 0.05759388740719383,
+      "grad_norm": 0.12477877623511162,
+      "learning_rate": 0.00019960255537429194,
+      "loss": 1.3546,
+      "step": 799
+    },
+    {
+      "epoch": 0.057665969869530743,
+      "grad_norm": 0.2192538007189068,
+      "learning_rate": 0.00019960047317725684,
+      "loss": 1.3908,
+      "step": 800
+    },
+    {
+      "epoch": 0.057738052331867656,
+      "grad_norm": 0.14429650037731107,
+      "learning_rate": 0.00019959838555110493,
+      "loss": 1.2889,
+      "step": 801
+    },
+    {
+      "epoch": 0.05781013479420457,
+      "grad_norm": 0.11678813958436081,
+      "learning_rate": 0.00019959629249594992,
+      "loss": 1.4173,
+      "step": 802
+    },
+    {
+      "epoch": 0.05788221725654148,
+      "grad_norm": 0.1289530737386515,
+      "learning_rate": 0.000199594194011906,
+      "loss": 1.4194,
+      "step": 803
+    },
+    {
+      "epoch": 0.057954299718878395,
+      "grad_norm": 0.12457369688700884,
+      "learning_rate": 0.0001995920900990875,
+      "loss": 1.293,
+      "step": 804
+    },
+    {
+      "epoch": 0.05802638218121531,
+      "grad_norm": 0.11326358460031193,
+      "learning_rate": 0.0001995899807576091,
+      "loss": 1.3743,
+      "step": 805
+    },
+    {
+      "epoch": 0.05809846464355222,
+      "grad_norm": 0.10298104225648243,
+      "learning_rate": 0.0001995878659875858,
+      "loss": 1.3433,
+      "step": 806
+    },
+    {
+      "epoch": 0.058170547105889134,
+      "grad_norm": 0.10496306257701371,
+      "learning_rate": 0.00019958574578913285,
+      "loss": 1.3736,
+      "step": 807
+    },
+    {
+      "epoch": 0.058242629568226054,
+      "grad_norm": 0.09937263077345894,
+      "learning_rate": 0.00019958362016236585,
+      "loss": 1.3398,
+      "step": 808
+    },
+    {
+      "epoch": 0.05831471203056297,
+      "grad_norm": 0.10104285131630987,
+      "learning_rate": 0.00019958148910740063,
+      "loss": 1.3992,
+      "step": 809
+    },
+    {
+      "epoch": 0.05838679449289988,
+      "grad_norm": 0.12806542858295158,
+      "learning_rate": 0.0001995793526243534,
+      "loss": 1.369,
+      "step": 810
+    },
+    {
+      "epoch": 0.05845887695523679,
+      "grad_norm": 0.22901288836956513,
+      "learning_rate": 0.0001995772107133406,
+      "loss": 1.3695,
+      "step": 811
+    },
+    {
+      "epoch": 0.058530959417573705,
+      "grad_norm": 0.1461105328947744,
+      "learning_rate": 0.00019957506337447894,
+      "loss": 1.4027,
+      "step": 812
+    },
+    {
+      "epoch": 0.05860304187991062,
+      "grad_norm": 0.10553082175201439,
+      "learning_rate": 0.00019957291060788553,
+      "loss": 1.3769,
+      "step": 813
+    },
+    {
+      "epoch": 0.05867512434224753,
+      "grad_norm": 0.12489595093398205,
+      "learning_rate": 0.00019957075241367766,
+      "loss": 1.4594,
+      "step": 814
+    },
+    {
+      "epoch": 0.058747206804584444,
+      "grad_norm": 0.21775115693869332,
+      "learning_rate": 0.00019956858879197303,
+      "loss": 1.2663,
+      "step": 815
+    },
+    {
+      "epoch": 0.05881928926692136,
+      "grad_norm": 0.10376503148299801,
+      "learning_rate": 0.0001995664197428895,
+      "loss": 1.358,
+      "step": 816
+    },
+    {
+      "epoch": 0.05889137172925827,
+      "grad_norm": 0.11730333476824067,
+      "learning_rate": 0.00019956424526654538,
+      "loss": 1.3908,
+      "step": 817
+    },
+    {
+      "epoch": 0.05896345419159518,
+      "grad_norm": 0.11838816272587714,
+      "learning_rate": 0.00019956206536305915,
+      "loss": 1.3629,
+      "step": 818
+    },
+    {
+      "epoch": 0.059035536653932096,
+      "grad_norm": 0.13260173481479115,
+      "learning_rate": 0.00019955988003254966,
+      "loss": 1.4018,
+      "step": 819
+    },
+    {
+      "epoch": 0.05910761911626901,
+      "grad_norm": 0.12733475003942482,
+      "learning_rate": 0.00019955768927513602,
+      "loss": 1.3787,
+      "step": 820
+    },
+    {
+      "epoch": 0.05917970157860593,
+      "grad_norm": 0.11957854563802181,
+      "learning_rate": 0.00019955549309093765,
+      "loss": 1.3915,
+      "step": 821
+    },
+    {
+      "epoch": 0.05925178404094284,
+      "grad_norm": 0.32745005213984013,
+      "learning_rate": 0.00019955329148007425,
+      "loss": 1.3424,
+      "step": 822
+    },
+    {
+      "epoch": 0.059323866503279754,
+      "grad_norm": 0.10584163197676781,
+      "learning_rate": 0.00019955108444266585,
+      "loss": 1.404,
+      "step": 823
+    },
+    {
+      "epoch": 0.05939594896561667,
+      "grad_norm": 0.11097848440998075,
+      "learning_rate": 0.0001995488719788327,
+      "loss": 1.3875,
+      "step": 824
+    },
+    {
+      "epoch": 0.05946803142795358,
+      "grad_norm": 0.27485932464584184,
+      "learning_rate": 0.00019954665408869547,
+      "loss": 1.3609,
+      "step": 825
+    },
+    {
+      "epoch": 0.05954011389029049,
+      "grad_norm": 0.2708488672711469,
+      "learning_rate": 0.000199544430772375,
+      "loss": 1.2675,
+      "step": 826
+    },
+    {
+      "epoch": 0.059612196352627406,
+      "grad_norm": 0.19454109032050546,
+      "learning_rate": 0.00019954220202999252,
+      "loss": 1.4019,
+      "step": 827
+    },
+    {
+      "epoch": 0.05968427881496432,
+      "grad_norm": 0.10015558045128736,
+      "learning_rate": 0.0001995399678616695,
+      "loss": 1.3915,
+      "step": 828
+    },
+    {
+      "epoch": 0.05975636127730123,
+      "grad_norm": 0.25854085612219413,
+      "learning_rate": 0.00019953772826752773,
+      "loss": 1.3769,
+      "step": 829
+    },
+    {
+      "epoch": 0.059828443739638144,
+      "grad_norm": 0.09313753534884918,
+      "learning_rate": 0.00019953548324768927,
+      "loss": 1.2896,
+      "step": 830
+    },
+    {
+      "epoch": 0.05990052620197506,
+      "grad_norm": 0.11797504146147504,
+      "learning_rate": 0.00019953323280227648,
+      "loss": 1.3619,
+      "step": 831
+    },
+    {
+      "epoch": 0.05997260866431197,
+      "grad_norm": 0.1150979142246714,
+      "learning_rate": 0.00019953097693141208,
+      "loss": 1.3384,
+      "step": 832
+    },
+    {
+      "epoch": 0.06004469112664888,
+      "grad_norm": 0.10220649212189498,
+      "learning_rate": 0.000199528715635219,
+      "loss": 1.3501,
+      "step": 833
+    },
+    {
+      "epoch": 0.0601167735889858,
+      "grad_norm": 0.1347064606292187,
+      "learning_rate": 0.0001995264489138205,
+      "loss": 1.3103,
+      "step": 834
+    },
+    {
+      "epoch": 0.060188856051322716,
+      "grad_norm": 0.1032186913648923,
+      "learning_rate": 0.00019952417676734015,
+      "loss": 1.3239,
+      "step": 835
+    },
+    {
+      "epoch": 0.06026093851365963,
+      "grad_norm": 0.13855754383912994,
+      "learning_rate": 0.0001995218991959018,
+      "loss": 1.3476,
+      "step": 836
+    },
+    {
+      "epoch": 0.06033302097599654,
+      "grad_norm": 0.10689077292051115,
+      "learning_rate": 0.00019951961619962962,
+      "loss": 1.3442,
+      "step": 837
+    },
+    {
+      "epoch": 0.060405103438333455,
+      "grad_norm": 0.10669043464446007,
+      "learning_rate": 0.000199517327778648,
+      "loss": 1.4286,
+      "step": 838
+    },
+    {
+      "epoch": 0.06047718590067037,
+      "grad_norm": 0.15954385580151073,
+      "learning_rate": 0.0001995150339330817,
+      "loss": 1.4463,
+      "step": 839
+    },
+    {
+      "epoch": 0.06054926836300728,
+      "grad_norm": 0.11153484481977656,
+      "learning_rate": 0.0001995127346630558,
+      "loss": 1.1758,
+      "step": 840
+    },
+    {
+      "epoch": 0.06062135082534419,
+      "grad_norm": 0.12170838294989669,
+      "learning_rate": 0.00019951042996869557,
+      "loss": 1.4161,
+      "step": 841
+    },
+    {
+      "epoch": 0.060693433287681106,
+      "grad_norm": 0.10423461501341155,
+      "learning_rate": 0.00019950811985012666,
+      "loss": 1.4205,
+      "step": 842
+    },
+    {
+      "epoch": 0.06076551575001802,
+      "grad_norm": 0.110913113561229,
+      "learning_rate": 0.000199505804307475,
+      "loss": 1.3898,
+      "step": 843
+    },
+    {
+      "epoch": 0.06083759821235493,
+      "grad_norm": 0.09679203149598288,
+      "learning_rate": 0.0001995034833408668,
+      "loss": 1.3145,
+      "step": 844
+    },
+    {
+      "epoch": 0.060909680674691845,
+      "grad_norm": 0.11384555842296193,
+      "learning_rate": 0.00019950115695042858,
+      "loss": 1.4455,
+      "step": 845
+    },
+    {
+      "epoch": 0.06098176313702876,
+      "grad_norm": 0.10007092513632047,
+      "learning_rate": 0.00019949882513628712,
+      "loss": 1.2487,
+      "step": 846
+    },
+    {
+      "epoch": 0.06105384559936568,
+      "grad_norm": 0.16144510901398074,
+      "learning_rate": 0.0001994964878985696,
+      "loss": 1.3071,
+      "step": 847
+    },
+    {
+      "epoch": 0.06112592806170259,
+      "grad_norm": 0.10213221575128861,
+      "learning_rate": 0.0001994941452374033,
+      "loss": 1.4028,
+      "step": 848
+    },
+    {
+      "epoch": 0.0611980105240395,
+      "grad_norm": 0.09721449972037241,
+      "learning_rate": 0.00019949179715291604,
+      "loss": 1.4306,
+      "step": 849
+    },
+    {
+      "epoch": 0.061270092986376416,
+      "grad_norm": 0.2483672510271683,
+      "learning_rate": 0.00019948944364523572,
+      "loss": 1.3523,
+      "step": 850
+    },
+    {
+      "epoch": 0.06134217544871333,
+      "grad_norm": 0.20431494975324002,
+      "learning_rate": 0.0001994870847144907,
+      "loss": 1.3954,
+      "step": 851
+    },
+    {
+      "epoch": 0.06141425791105024,
+      "grad_norm": 0.16892980480089506,
+      "learning_rate": 0.00019948472036080949,
+      "loss": 1.2627,
+      "step": 852
+    },
+    {
+      "epoch": 0.061486340373387155,
+      "grad_norm": 0.10422366008456731,
+      "learning_rate": 0.000199482350584321,
+      "loss": 1.3357,
+      "step": 853
+    },
+    {
+      "epoch": 0.06155842283572407,
+      "grad_norm": 0.09949241210751333,
+      "learning_rate": 0.00019947997538515445,
+      "loss": 1.4282,
+      "step": 854
+    },
+    {
+      "epoch": 0.06163050529806098,
+      "grad_norm": 0.09901521558951777,
+      "learning_rate": 0.00019947759476343922,
+      "loss": 1.4181,
+      "step": 855
+    },
+    {
+      "epoch": 0.061702587760397894,
+      "grad_norm": 0.09498025606716745,
+      "learning_rate": 0.00019947520871930517,
+      "loss": 1.3047,
+      "step": 856
+    },
+    {
+      "epoch": 0.06177467022273481,
+      "grad_norm": 0.4087804984722178,
+      "learning_rate": 0.00019947281725288227,
+      "loss": 1.3689,
+      "step": 857
+    },
+    {
+      "epoch": 0.06184675268507172,
+      "grad_norm": 0.10768339037566389,
+      "learning_rate": 0.00019947042036430097,
+      "loss": 1.3513,
+      "step": 858
+    },
+    {
+      "epoch": 0.06191883514740863,
+      "grad_norm": 0.10790257351329727,
+      "learning_rate": 0.00019946801805369182,
+      "loss": 1.4325,
+      "step": 859
+    },
+    {
+      "epoch": 0.06199091760974555,
+      "grad_norm": 0.24016085131096154,
+      "learning_rate": 0.00019946561032118586,
+      "loss": 1.2182,
+      "step": 860
+    },
+    {
+      "epoch": 0.062063000072082465,
+      "grad_norm": 0.11272217571710472,
+      "learning_rate": 0.00019946319716691426,
+      "loss": 1.321,
+      "step": 861
+    },
+    {
+      "epoch": 0.06213508253441938,
+      "grad_norm": 0.15806788133733252,
+      "learning_rate": 0.00019946077859100863,
+      "loss": 1.4322,
+      "step": 862
+    },
+    {
+      "epoch": 0.06220716499675629,
+      "grad_norm": 0.15067535390591502,
+      "learning_rate": 0.00019945835459360073,
+      "loss": 1.3061,
+      "step": 863
+    },
+    {
+      "epoch": 0.062279247459093204,
+      "grad_norm": 0.11302740774575164,
+      "learning_rate": 0.00019945592517482273,
+      "loss": 1.2944,
+      "step": 864
+    },
+    {
+      "epoch": 0.06235132992143012,
+      "grad_norm": 0.10159171000273794,
+      "learning_rate": 0.00019945349033480706,
+      "loss": 1.2468,
+      "step": 865
+    },
+    {
+      "epoch": 0.06242341238376703,
+      "grad_norm": 0.10899979468125924,
+      "learning_rate": 0.00019945105007368645,
+      "loss": 1.3145,
+      "step": 866
+    },
+    {
+      "epoch": 0.06249549484610394,
+      "grad_norm": 0.11502774195779347,
+      "learning_rate": 0.00019944860439159387,
+      "loss": 1.3365,
+      "step": 867
+    },
+    {
+      "epoch": 0.06256757730844086,
+      "grad_norm": 0.11598420132715058,
+      "learning_rate": 0.00019944615328866268,
+      "loss": 1.2639,
+      "step": 868
+    },
+    {
+      "epoch": 0.06263965977077778,
+      "grad_norm": 0.30150014013622217,
+      "learning_rate": 0.00019944369676502642,
+      "loss": 1.321,
+      "step": 869
+    },
+    {
+      "epoch": 0.06271174223311468,
+      "grad_norm": 0.3210842168411934,
+      "learning_rate": 0.00019944123482081908,
+      "loss": 1.3783,
+      "step": 870
+    },
+    {
+      "epoch": 0.0627838246954516,
+      "grad_norm": 0.8009676526656323,
+      "learning_rate": 0.0001994387674561748,
+      "loss": 1.3564,
+      "step": 871
+    },
+    {
+      "epoch": 0.06285590715778851,
+      "grad_norm": 0.28074739893124884,
+      "learning_rate": 0.00019943629467122808,
+      "loss": 1.4149,
+      "step": 872
+    },
+    {
+      "epoch": 0.06292798962012543,
+      "grad_norm": 0.16347483523143855,
+      "learning_rate": 0.0001994338164661137,
+      "loss": 1.3466,
+      "step": 873
+    },
+    {
+      "epoch": 0.06300007208246233,
+      "grad_norm": 0.10039182135228704,
+      "learning_rate": 0.00019943133284096683,
+      "loss": 1.3139,
+      "step": 874
+    },
+    {
+      "epoch": 0.06307215454479925,
+      "grad_norm": 0.12412560490258513,
+      "learning_rate": 0.00019942884379592273,
+      "loss": 1.3707,
+      "step": 875
+    },
+    {
+      "epoch": 0.06314423700713616,
+      "grad_norm": 0.12161003796080047,
+      "learning_rate": 0.00019942634933111715,
+      "loss": 1.4069,
+      "step": 876
+    },
+    {
+      "epoch": 0.06321631946947308,
+      "grad_norm": 0.1051911941098551,
+      "learning_rate": 0.000199423849446686,
+      "loss": 1.3683,
+      "step": 877
+    },
+    {
+      "epoch": 0.06328840193180998,
+      "grad_norm": 0.11794154487398928,
+      "learning_rate": 0.00019942134414276566,
+      "loss": 1.3037,
+      "step": 878
+    },
+    {
+      "epoch": 0.0633604843941469,
+      "grad_norm": 0.12340898842880017,
+      "learning_rate": 0.00019941883341949255,
+      "loss": 1.3826,
+      "step": 879
+    },
+    {
+      "epoch": 0.06343256685648382,
+      "grad_norm": 0.18243984114379744,
+      "learning_rate": 0.0001994163172770036,
+      "loss": 1.2956,
+      "step": 880
+    },
+    {
+      "epoch": 0.06350464931882073,
+      "grad_norm": 0.14193472269230373,
+      "learning_rate": 0.00019941379571543596,
+      "loss": 1.368,
+      "step": 881
+    },
+    {
+      "epoch": 0.06357673178115765,
+      "grad_norm": 0.16266749981438056,
+      "learning_rate": 0.00019941126873492708,
+      "loss": 1.2853,
+      "step": 882
+    },
+    {
+      "epoch": 0.06364881424349456,
+      "grad_norm": 0.1223510874624418,
+      "learning_rate": 0.0001994087363356147,
+      "loss": 1.3589,
+      "step": 883
+    },
+    {
+      "epoch": 0.06372089670583148,
+      "grad_norm": 0.12430909830040898,
+      "learning_rate": 0.00019940619851763685,
+      "loss": 1.3755,
+      "step": 884
+    },
+    {
+      "epoch": 0.06379297916816838,
+      "grad_norm": 0.17809867652840636,
+      "learning_rate": 0.00019940365528113186,
+      "loss": 1.3254,
+      "step": 885
+    },
+    {
+      "epoch": 0.0638650616305053,
+      "grad_norm": 0.19512343227636367,
+      "learning_rate": 0.00019940110662623837,
+      "loss": 1.4052,
+      "step": 886
+    },
+    {
+      "epoch": 0.06393714409284221,
+      "grad_norm": 0.21386635930690806,
+      "learning_rate": 0.0001993985525530953,
+      "loss": 1.3957,
+      "step": 887
+    },
+    {
+      "epoch": 0.06400922655517913,
+      "grad_norm": 0.12856420006005426,
+      "learning_rate": 0.00019939599306184188,
+      "loss": 1.3589,
+      "step": 888
+    },
+    {
+      "epoch": 0.06408130901751603,
+      "grad_norm": 0.10625399437176578,
+      "learning_rate": 0.0001993934281526176,
+      "loss": 1.3779,
+      "step": 889
+    },
+    {
+      "epoch": 0.06415339147985295,
+      "grad_norm": 0.1283216657817702,
+      "learning_rate": 0.0001993908578255623,
+      "loss": 1.3002,
+      "step": 890
+    },
+    {
+      "epoch": 0.06422547394218986,
+      "grad_norm": 0.18484931634129903,
+      "learning_rate": 0.00019938828208081606,
+      "loss": 1.3643,
+      "step": 891
+    },
+    {
+      "epoch": 0.06429755640452678,
+      "grad_norm": 0.11418443953624666,
+      "learning_rate": 0.0001993857009185193,
+      "loss": 1.3585,
+      "step": 892
+    },
+    {
+      "epoch": 0.0643696388668637,
+      "grad_norm": 0.11862600802365496,
+      "learning_rate": 0.00019938311433881272,
+      "loss": 1.3568,
+      "step": 893
+    },
+    {
+      "epoch": 0.0644417213292006,
+      "grad_norm": 0.10978926340472589,
+      "learning_rate": 0.0001993805223418373,
+      "loss": 1.4083,
+      "step": 894
+    },
+    {
+      "epoch": 0.06451380379153752,
+      "grad_norm": 0.10185850848032084,
+      "learning_rate": 0.0001993779249277343,
+      "loss": 1.4741,
+      "step": 895
+    },
+    {
+      "epoch": 0.06458588625387443,
+      "grad_norm": 0.09884976432631248,
+      "learning_rate": 0.00019937532209664534,
+      "loss": 1.4427,
+      "step": 896
+    },
+    {
+      "epoch": 0.06465796871621135,
+      "grad_norm": 0.1580372623920786,
+      "learning_rate": 0.00019937271384871233,
+      "loss": 1.3491,
+      "step": 897
+    },
+    {
+      "epoch": 0.06473005117854826,
+      "grad_norm": 0.15433561921118866,
+      "learning_rate": 0.00019937010018407735,
+      "loss": 1.2781,
+      "step": 898
+    },
+    {
+      "epoch": 0.06480213364088518,
+      "grad_norm": 0.12225973615419122,
+      "learning_rate": 0.00019936748110288295,
+      "loss": 1.3856,
+      "step": 899
+    },
+    {
+      "epoch": 0.06487421610322208,
+      "grad_norm": 0.16899079850506643,
+      "learning_rate": 0.00019936485660527187,
+      "loss": 1.43,
+      "step": 900
+    },
+    {
+      "epoch": 0.064946298565559,
+      "grad_norm": 0.10964302741132632,
+      "learning_rate": 0.00019936222669138714,
+      "loss": 1.3307,
+      "step": 901
+    },
+    {
+      "epoch": 0.06501838102789591,
+      "grad_norm": 0.1141045865498986,
+      "learning_rate": 0.00019935959136137214,
+      "loss": 1.3818,
+      "step": 902
+    },
+    {
+      "epoch": 0.06509046349023283,
+      "grad_norm": 0.20000508816363555,
+      "learning_rate": 0.00019935695061537053,
+      "loss": 1.3372,
+      "step": 903
+    },
+    {
+      "epoch": 0.06516254595256973,
+      "grad_norm": 0.11239119422557978,
+      "learning_rate": 0.00019935430445352622,
+      "loss": 1.3628,
+      "step": 904
+    },
+    {
+      "epoch": 0.06523462841490665,
+      "grad_norm": 0.11015550980706222,
+      "learning_rate": 0.00019935165287598348,
+      "loss": 1.3304,
+      "step": 905
+    },
+    {
+      "epoch": 0.06530671087724357,
+      "grad_norm": 0.12641727208893871,
+      "learning_rate": 0.00019934899588288683,
+      "loss": 1.2748,
+      "step": 906
+    },
+    {
+      "epoch": 0.06537879333958048,
+      "grad_norm": 0.14404165073184924,
+      "learning_rate": 0.00019934633347438107,
+      "loss": 1.3797,
+      "step": 907
+    },
+    {
+      "epoch": 0.0654508758019174,
+      "grad_norm": 0.11935717131377009,
+      "learning_rate": 0.0001993436656506114,
+      "loss": 1.2583,
+      "step": 908
+    },
+    {
+      "epoch": 0.0655229582642543,
+      "grad_norm": 0.11058476421063036,
+      "learning_rate": 0.0001993409924117232,
+      "loss": 1.3161,
+      "step": 909
+    },
+    {
+      "epoch": 0.06559504072659122,
+      "grad_norm": 0.13893058581686868,
+      "learning_rate": 0.00019933831375786216,
+      "loss": 1.3543,
+      "step": 910
+    },
+    {
+      "epoch": 0.06566712318892813,
+      "grad_norm": 0.10636286793491304,
+      "learning_rate": 0.0001993356296891743,
+      "loss": 1.3763,
+      "step": 911
+    },
+    {
+      "epoch": 0.06573920565126505,
+      "grad_norm": 0.13071945377779753,
+      "learning_rate": 0.000199332940205806,
+      "loss": 1.3355,
+      "step": 912
+    },
+    {
+      "epoch": 0.06581128811360196,
+      "grad_norm": 0.15195609237639096,
+      "learning_rate": 0.00019933024530790377,
+      "loss": 1.2657,
+      "step": 913
+    },
+    {
+      "epoch": 0.06588337057593888,
+      "grad_norm": 0.24227242172921765,
+      "learning_rate": 0.0001993275449956145,
+      "loss": 1.386,
+      "step": 914
+    },
+    {
+      "epoch": 0.06595545303827578,
+      "grad_norm": 0.112557828929706,
+      "learning_rate": 0.00019932483926908548,
+      "loss": 1.3176,
+      "step": 915
+    },
+    {
+      "epoch": 0.0660275355006127,
+      "grad_norm": 0.19396108790764144,
+      "learning_rate": 0.0001993221281284641,
+      "loss": 1.3656,
+      "step": 916
+    },
+    {
+      "epoch": 0.06609961796294961,
+      "grad_norm": 0.10782342691366167,
+      "learning_rate": 0.00019931941157389817,
+      "loss": 1.2966,
+      "step": 917
+    },
+    {
+      "epoch": 0.06617170042528653,
+      "grad_norm": 0.2307897044923339,
+      "learning_rate": 0.0001993166896055358,
+      "loss": 1.2527,
+      "step": 918
+    },
+    {
+      "epoch": 0.06624378288762345,
+      "grad_norm": 0.11814355586670114,
+      "learning_rate": 0.00019931396222352529,
+      "loss": 1.3482,
+      "step": 919
+    },
+    {
+      "epoch": 0.06631586534996035,
+      "grad_norm": 0.2131719146990173,
+      "learning_rate": 0.00019931122942801538,
+      "loss": 1.3733,
+      "step": 920
+    },
+    {
+      "epoch": 0.06638794781229727,
+      "grad_norm": 0.10925513013192897,
+      "learning_rate": 0.000199308491219155,
+      "loss": 1.3504,
+      "step": 921
+    },
+    {
+      "epoch": 0.06646003027463418,
+      "grad_norm": 0.2664082352760664,
+      "learning_rate": 0.0001993057475970934,
+      "loss": 1.3198,
+      "step": 922
+    },
+    {
+      "epoch": 0.0665321127369711,
+      "grad_norm": 0.10227217986904835,
+      "learning_rate": 0.00019930299856198017,
+      "loss": 1.3952,
+      "step": 923
+    },
+    {
+      "epoch": 0.066604195199308,
+      "grad_norm": 0.10354224362402509,
+      "learning_rate": 0.00019930024411396512,
+      "loss": 1.3287,
+      "step": 924
+    },
+    {
+      "epoch": 0.06667627766164493,
+      "grad_norm": 0.12867175468991784,
+      "learning_rate": 0.00019929748425319839,
+      "loss": 1.2685,
+      "step": 925
+    },
+    {
+      "epoch": 0.06674836012398183,
+      "grad_norm": 0.16427790468395626,
+      "learning_rate": 0.00019929471897983043,
+      "loss": 1.3272,
+      "step": 926
+    },
+    {
+      "epoch": 0.06682044258631875,
+      "grad_norm": 0.1306385834149794,
+      "learning_rate": 0.00019929194829401196,
+      "loss": 1.5153,
+      "step": 927
+    },
+    {
+      "epoch": 0.06689252504865566,
+      "grad_norm": 0.11712997975225366,
+      "learning_rate": 0.00019928917219589403,
+      "loss": 1.3954,
+      "step": 928
+    },
+    {
+      "epoch": 0.06696460751099258,
+      "grad_norm": 0.10722418300972314,
+      "learning_rate": 0.00019928639068562794,
+      "loss": 1.354,
+      "step": 929
+    },
+    {
+      "epoch": 0.06703668997332948,
+      "grad_norm": 0.11079761885776575,
+      "learning_rate": 0.00019928360376336534,
+      "loss": 1.3205,
+      "step": 930
+    },
+    {
+      "epoch": 0.0671087724356664,
+      "grad_norm": 0.09908277524349775,
+      "learning_rate": 0.00019928081142925807,
+      "loss": 1.3619,
+      "step": 931
+    },
+    {
+      "epoch": 0.06718085489800332,
+      "grad_norm": 0.10341072088166058,
+      "learning_rate": 0.0001992780136834584,
+      "loss": 1.385,
+      "step": 932
+    },
+    {
+      "epoch": 0.06725293736034023,
+      "grad_norm": 0.10755933522820188,
+      "learning_rate": 0.00019927521052611884,
+      "loss": 1.3454,
+      "step": 933
+    },
+    {
+      "epoch": 0.06732501982267715,
+      "grad_norm": 0.11891953909904779,
+      "learning_rate": 0.00019927240195739216,
+      "loss": 1.3395,
+      "step": 934
+    },
+    {
+      "epoch": 0.06739710228501405,
+      "grad_norm": 0.11294385255705809,
+      "learning_rate": 0.00019926958797743144,
+      "loss": 1.2073,
+      "step": 935
+    },
+    {
+      "epoch": 0.06746918474735097,
+      "grad_norm": 0.11760950816682741,
+      "learning_rate": 0.00019926676858639011,
+      "loss": 1.4044,
+      "step": 936
+    },
+    {
+      "epoch": 0.06754126720968788,
+      "grad_norm": 0.12458551079122741,
+      "learning_rate": 0.00019926394378442182,
+      "loss": 1.4339,
+      "step": 937
+    },
+    {
+      "epoch": 0.0676133496720248,
+      "grad_norm": 0.1331539297190386,
+      "learning_rate": 0.00019926111357168055,
+      "loss": 1.3836,
+      "step": 938
+    },
+    {
+      "epoch": 0.0676854321343617,
+      "grad_norm": 0.11280527782300148,
+      "learning_rate": 0.00019925827794832056,
+      "loss": 1.354,
+      "step": 939
+    },
+    {
+      "epoch": 0.06775751459669863,
+      "grad_norm": 0.09892290368215594,
+      "learning_rate": 0.00019925543691449643,
+      "loss": 1.3737,
+      "step": 940
+    },
+    {
+      "epoch": 0.06782959705903553,
+      "grad_norm": 0.137353714315955,
+      "learning_rate": 0.00019925259047036303,
+      "loss": 1.2218,
+      "step": 941
+    },
+    {
+      "epoch": 0.06790167952137245,
+      "grad_norm": 0.13466733557344224,
+      "learning_rate": 0.00019924973861607553,
+      "loss": 1.2785,
+      "step": 942
+    },
+    {
+      "epoch": 0.06797376198370936,
+      "grad_norm": 0.13338350451197586,
+      "learning_rate": 0.00019924688135178936,
+      "loss": 1.3146,
+      "step": 943
+    },
+    {
+      "epoch": 0.06804584444604628,
+      "grad_norm": 0.20768033873313346,
+      "learning_rate": 0.00019924401867766025,
+      "loss": 1.3518,
+      "step": 944
+    },
+    {
+      "epoch": 0.0681179269083832,
+      "grad_norm": 0.10898495951941617,
+      "learning_rate": 0.00019924115059384424,
+      "loss": 1.4312,
+      "step": 945
+    },
+    {
+      "epoch": 0.0681900093707201,
+      "grad_norm": 0.1038049117177923,
+      "learning_rate": 0.00019923827710049773,
+      "loss": 1.3595,
+      "step": 946
+    },
+    {
+      "epoch": 0.06826209183305702,
+      "grad_norm": 0.1325661426364933,
+      "learning_rate": 0.0001992353981977773,
+      "loss": 1.4266,
+      "step": 947
+    },
+    {
+      "epoch": 0.06833417429539393,
+      "grad_norm": 0.27887333471014164,
+      "learning_rate": 0.00019923251388583987,
+      "loss": 1.4847,
+      "step": 948
+    },
+    {
+      "epoch": 0.06840625675773085,
+      "grad_norm": 0.13539534886639484,
+      "learning_rate": 0.00019922962416484268,
+      "loss": 1.3834,
+      "step": 949
+    },
+    {
+      "epoch": 0.06847833922006775,
+      "grad_norm": 0.1069175909380925,
+      "learning_rate": 0.00019922672903494325,
+      "loss": 1.4258,
+      "step": 950
+    },
+    {
+      "epoch": 0.06855042168240467,
+      "grad_norm": 0.12393602678830647,
+      "learning_rate": 0.00019922382849629934,
+      "loss": 1.324,
+      "step": 951
+    },
+    {
+      "epoch": 0.06862250414474158,
+      "grad_norm": 0.2859722960121165,
+      "learning_rate": 0.00019922092254906915,
+      "loss": 1.4462,
+      "step": 952
+    },
+    {
+      "epoch": 0.0686945866070785,
+      "grad_norm": 0.16746564776879083,
+      "learning_rate": 0.000199218011193411,
+      "loss": 1.4348,
+      "step": 953
+    },
+    {
+      "epoch": 0.0687666690694154,
+      "grad_norm": 0.11865823083149263,
+      "learning_rate": 0.00019921509442948358,
+      "loss": 1.3876,
+      "step": 954
+    },
+    {
+      "epoch": 0.06883875153175233,
+      "grad_norm": 0.12354653472152567,
+      "learning_rate": 0.00019921217225744595,
+      "loss": 1.3359,
+      "step": 955
+    },
+    {
+      "epoch": 0.06891083399408923,
+      "grad_norm": 0.11112844659427505,
+      "learning_rate": 0.00019920924467745733,
+      "loss": 1.3587,
+      "step": 956
+    },
+    {
+      "epoch": 0.06898291645642615,
+      "grad_norm": 0.1162733844236941,
+      "learning_rate": 0.00019920631168967733,
+      "loss": 1.3255,
+      "step": 957
+    },
+    {
+      "epoch": 0.06905499891876306,
+      "grad_norm": 0.17299354019893345,
+      "learning_rate": 0.00019920337329426584,
+      "loss": 1.2558,
+      "step": 958
+    },
+    {
+      "epoch": 0.06912708138109998,
+      "grad_norm": 0.12349869446969253,
+      "learning_rate": 0.000199200429491383,
+      "loss": 1.34,
+      "step": 959
+    },
+    {
+      "epoch": 0.0691991638434369,
+      "grad_norm": 0.12063090061668298,
+      "learning_rate": 0.00019919748028118925,
+      "loss": 1.4513,
+      "step": 960
+    },
+    {
+      "epoch": 0.0692712463057738,
+      "grad_norm": 0.1415058211191057,
+      "learning_rate": 0.0001991945256638454,
+      "loss": 1.2938,
+      "step": 961
+    },
+    {
+      "epoch": 0.06934332876811072,
+      "grad_norm": 0.10958346475139882,
+      "learning_rate": 0.00019919156563951248,
+      "loss": 1.3213,
+      "step": 962
+    },
+    {
+      "epoch": 0.06941541123044763,
+      "grad_norm": 0.15092286127126647,
+      "learning_rate": 0.00019918860020835184,
+      "loss": 1.3812,
+      "step": 963
+    },
+    {
+      "epoch": 0.06948749369278455,
+      "grad_norm": 0.10240395707929682,
+      "learning_rate": 0.00019918562937052512,
+      "loss": 1.347,
+      "step": 964
+    },
+    {
+      "epoch": 0.06955957615512146,
+      "grad_norm": 0.11202210127273204,
+      "learning_rate": 0.00019918265312619427,
+      "loss": 1.3384,
+      "step": 965
+    },
+    {
+      "epoch": 0.06963165861745838,
+      "grad_norm": 0.10090191585938066,
+      "learning_rate": 0.0001991796714755215,
+      "loss": 1.3078,
+      "step": 966
+    },
+    {
+      "epoch": 0.06970374107979528,
+      "grad_norm": 0.11920264934641313,
+      "learning_rate": 0.00019917668441866935,
+      "loss": 1.3538,
+      "step": 967
+    },
+    {
+      "epoch": 0.0697758235421322,
+      "grad_norm": 0.11594130749206676,
+      "learning_rate": 0.00019917369195580063,
+      "loss": 1.3098,
+      "step": 968
+    },
+    {
+      "epoch": 0.0698479060044691,
+      "grad_norm": 0.11152032125012604,
+      "learning_rate": 0.00019917069408707848,
+      "loss": 1.4005,
+      "step": 969
+    },
+    {
+      "epoch": 0.06991998846680603,
+      "grad_norm": 0.13082948399281066,
+      "learning_rate": 0.00019916769081266628,
+      "loss": 1.3359,
+      "step": 970
+    },
+    {
+      "epoch": 0.06999207092914293,
+      "grad_norm": 0.1277444332156955,
+      "learning_rate": 0.00019916468213272776,
+      "loss": 1.2213,
+      "step": 971
+    },
+    {
+      "epoch": 0.07006415339147985,
+      "grad_norm": 0.12141865400453074,
+      "learning_rate": 0.00019916166804742692,
+      "loss": 1.468,
+      "step": 972
+    },
+    {
+      "epoch": 0.07013623585381677,
+      "grad_norm": 0.10944635508830258,
+      "learning_rate": 0.00019915864855692803,
+      "loss": 1.3485,
+      "step": 973
+    },
+    {
+      "epoch": 0.07020831831615368,
+      "grad_norm": 0.10626715067799661,
+      "learning_rate": 0.0001991556236613957,
+      "loss": 1.3876,
+      "step": 974
+    },
+    {
+      "epoch": 0.0702804007784906,
+      "grad_norm": 0.11078528019968671,
+      "learning_rate": 0.00019915259336099483,
+      "loss": 1.3018,
+      "step": 975
+    },
+    {
+      "epoch": 0.0703524832408275,
+      "grad_norm": 0.18944079726504698,
+      "learning_rate": 0.00019914955765589057,
+      "loss": 1.3766,
+      "step": 976
+    },
+    {
+      "epoch": 0.07042456570316442,
+      "grad_norm": 0.15060169063765058,
+      "learning_rate": 0.00019914651654624838,
+      "loss": 1.2807,
+      "step": 977
+    },
+    {
+      "epoch": 0.07049664816550133,
+      "grad_norm": 0.1074498571748179,
+      "learning_rate": 0.00019914347003223407,
+      "loss": 1.3274,
+      "step": 978
+    },
+    {
+      "epoch": 0.07056873062783825,
+      "grad_norm": 0.0952516227837784,
+      "learning_rate": 0.00019914041811401365,
+      "loss": 1.2964,
+      "step": 979
+    },
+    {
+      "epoch": 0.07064081309017516,
+      "grad_norm": 0.11660899835046029,
+      "learning_rate": 0.00019913736079175354,
+      "loss": 1.2848,
+      "step": 980
+    },
+    {
+      "epoch": 0.07071289555251208,
+      "grad_norm": 0.17882579582031746,
+      "learning_rate": 0.00019913429806562034,
+      "loss": 1.3419,
+      "step": 981
+    },
+    {
+      "epoch": 0.07078497801484898,
+      "grad_norm": 0.11848752712686335,
+      "learning_rate": 0.00019913122993578104,
+      "loss": 1.2871,
+      "step": 982
+    },
+    {
+      "epoch": 0.0708570604771859,
+      "grad_norm": 0.11482806797589254,
+      "learning_rate": 0.00019912815640240282,
+      "loss": 1.3992,
+      "step": 983
+    },
+    {
+      "epoch": 0.07092914293952281,
+      "grad_norm": 0.12203686266062908,
+      "learning_rate": 0.00019912507746565328,
+      "loss": 1.4123,
+      "step": 984
+    },
+    {
+      "epoch": 0.07100122540185973,
+      "grad_norm": 0.1161965810926321,
+      "learning_rate": 0.0001991219931257002,
+      "loss": 1.2257,
+      "step": 985
+    },
+    {
+      "epoch": 0.07107330786419665,
+      "grad_norm": 0.10341507282026909,
+      "learning_rate": 0.0001991189033827117,
+      "loss": 1.3778,
+      "step": 986
+    },
+    {
+      "epoch": 0.07114539032653355,
+      "grad_norm": 0.09954533608026787,
+      "learning_rate": 0.00019911580823685627,
+      "loss": 1.3554,
+      "step": 987
+    },
+    {
+      "epoch": 0.07121747278887047,
+      "grad_norm": 0.11265275504056788,
+      "learning_rate": 0.00019911270768830256,
+      "loss": 1.3935,
+      "step": 988
+    },
+    {
+      "epoch": 0.07128955525120738,
+      "grad_norm": 0.10280141323360414,
+      "learning_rate": 0.0001991096017372196,
+      "loss": 1.345,
+      "step": 989
+    },
+    {
+      "epoch": 0.0713616377135443,
+      "grad_norm": 0.1227131170766327,
+      "learning_rate": 0.00019910649038377666,
+      "loss": 1.2476,
+      "step": 990
+    },
+    {
+      "epoch": 0.0714337201758812,
+      "grad_norm": 0.11597789130081292,
+      "learning_rate": 0.0001991033736281434,
+      "loss": 1.4177,
+      "step": 991
+    },
+    {
+      "epoch": 0.07150580263821812,
+      "grad_norm": 0.11615630693555959,
+      "learning_rate": 0.00019910025147048963,
+      "loss": 1.4224,
+      "step": 992
+    },
+    {
+      "epoch": 0.07157788510055503,
+      "grad_norm": 0.11924810391249871,
+      "learning_rate": 0.0001990971239109856,
+      "loss": 1.2679,
+      "step": 993
+    },
+    {
+      "epoch": 0.07164996756289195,
+      "grad_norm": 0.10176429834272561,
+      "learning_rate": 0.00019909399094980178,
+      "loss": 1.3891,
+      "step": 994
+    },
+    {
+      "epoch": 0.07172205002522886,
+      "grad_norm": 0.1534765937987868,
+      "learning_rate": 0.00019909085258710892,
+      "loss": 1.3015,
+      "step": 995
+    },
+    {
+      "epoch": 0.07179413248756578,
+      "grad_norm": 0.25110450430598774,
+      "learning_rate": 0.0001990877088230781,
+      "loss": 1.2519,
+      "step": 996
+    },
+    {
+      "epoch": 0.07186621494990268,
+      "grad_norm": 0.1057317573300411,
+      "learning_rate": 0.00019908455965788067,
+      "loss": 1.3477,
+      "step": 997
+    },
+    {
+      "epoch": 0.0719382974122396,
+      "grad_norm": 0.10444138354731436,
+      "learning_rate": 0.00019908140509168834,
+      "loss": 1.3746,
+      "step": 998
+    },
+    {
+      "epoch": 0.07201037987457652,
+      "grad_norm": 0.10147533926707024,
+      "learning_rate": 0.00019907824512467302,
+      "loss": 1.3369,
+      "step": 999
+    },
+    {
+      "epoch": 0.07208246233691343,
+      "grad_norm": 0.13266332420623858,
+      "learning_rate": 0.00019907507975700693,
+      "loss": 1.4029,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07215454479925035,
+      "grad_norm": 0.1681210817377629,
+      "learning_rate": 0.0001990719089888627,
+      "loss": 1.313,
+      "step": 1001
+    },
+    {
+      "epoch": 0.07222662726158725,
+      "grad_norm": 0.10950843252743724,
+      "learning_rate": 0.00019906873282041302,
+      "loss": 1.269,
+      "step": 1002
+    },
+    {
+      "epoch": 0.07229870972392417,
+      "grad_norm": 0.10303570678740476,
+      "learning_rate": 0.00019906555125183116,
+      "loss": 1.4371,
+      "step": 1003
+    },
+    {
+      "epoch": 0.07237079218626108,
+      "grad_norm": 0.1449634675333585,
+      "learning_rate": 0.0001990623642832905,
+      "loss": 1.4196,
+      "step": 1004
+    },
+    {
+      "epoch": 0.072442874648598,
+      "grad_norm": 0.10811436225903194,
+      "learning_rate": 0.00019905917191496473,
+      "loss": 1.2471,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0725149571109349,
+      "grad_norm": 0.13303472593555246,
+      "learning_rate": 0.0001990559741470279,
+      "loss": 1.3055,
+      "step": 1006
+    },
+    {
+      "epoch": 0.07258703957327182,
+      "grad_norm": 0.10089774520138867,
+      "learning_rate": 0.00019905277097965427,
+      "loss": 1.42,
+      "step": 1007
+    },
+    {
+      "epoch": 0.07265912203560873,
+      "grad_norm": 0.11909229121263407,
+      "learning_rate": 0.0001990495624130185,
+      "loss": 1.3768,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07273120449794565,
+      "grad_norm": 0.11016369042923523,
+      "learning_rate": 0.00019904634844729544,
+      "loss": 1.377,
+      "step": 1009
+    },
+    {
+      "epoch": 0.07280328696028256,
+      "grad_norm": 0.10738552282505177,
+      "learning_rate": 0.0001990431290826603,
+      "loss": 1.3949,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07287536942261948,
+      "grad_norm": 0.10474448084613913,
+      "learning_rate": 0.00019903990431928854,
+      "loss": 1.3835,
+      "step": 1011
+    },
+    {
+      "epoch": 0.0729474518849564,
+      "grad_norm": 0.10787643623986339,
+      "learning_rate": 0.00019903667415735598,
+      "loss": 1.3547,
+      "step": 1012
+    },
+    {
+      "epoch": 0.0730195343472933,
+      "grad_norm": 0.12162920071111535,
+      "learning_rate": 0.00019903343859703867,
+      "loss": 1.3931,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07309161680963022,
+      "grad_norm": 0.1081544825457974,
+      "learning_rate": 0.00019903019763851297,
+      "loss": 1.3302,
+      "step": 1014
+    },
+    {
+      "epoch": 0.07316369927196713,
+      "grad_norm": 0.10229997061609322,
+      "learning_rate": 0.00019902695128195554,
+      "loss": 1.3189,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07323578173430405,
+      "grad_norm": 0.17154420916303542,
+      "learning_rate": 0.00019902369952754336,
+      "loss": 1.2797,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07330786419664095,
+      "grad_norm": 0.10172241601364305,
+      "learning_rate": 0.00019902044237545368,
+      "loss": 1.3149,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07337994665897787,
+      "grad_norm": 0.12359977185480543,
+      "learning_rate": 0.000199017179825864,
+      "loss": 1.3282,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07345202912131478,
+      "grad_norm": 0.11396469248142517,
+      "learning_rate": 0.0001990139118789522,
+      "loss": 1.3145,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0735241115836517,
+      "grad_norm": 0.11226840898214181,
+      "learning_rate": 0.0001990106385348964,
+      "loss": 1.3035,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0735961940459886,
+      "grad_norm": 0.1733385082720218,
+      "learning_rate": 0.00019900735979387505,
+      "loss": 1.4088,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07366827650832553,
+      "grad_norm": 0.09829250202466976,
+      "learning_rate": 0.0001990040756560668,
+      "loss": 1.3288,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07374035897066243,
+      "grad_norm": 0.11165028320654474,
+      "learning_rate": 0.00019900078612165075,
+      "loss": 1.2443,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07381244143299935,
+      "grad_norm": 0.2511013199497321,
+      "learning_rate": 0.00019899749119080612,
+      "loss": 1.4254,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07388452389533627,
+      "grad_norm": 0.12110077143343236,
+      "learning_rate": 0.0001989941908637126,
+      "loss": 1.3155,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07395660635767318,
+      "grad_norm": 0.12435578619509384,
+      "learning_rate": 0.00019899088514055004,
+      "loss": 1.3618,
+      "step": 1026
+    },
+    {
+      "epoch": 0.0740286888200101,
+      "grad_norm": 0.09877629360613678,
+      "learning_rate": 0.00019898757402149863,
+      "loss": 1.3144,
+      "step": 1027
+    },
+    {
+      "epoch": 0.074100771282347,
+      "grad_norm": 0.11550759296980759,
+      "learning_rate": 0.0001989842575067389,
+      "loss": 1.3857,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07417285374468392,
+      "grad_norm": 0.11273379200615938,
+      "learning_rate": 0.0001989809355964516,
+      "loss": 1.3682,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07424493620702083,
+      "grad_norm": 0.10138268077396059,
+      "learning_rate": 0.00019897760829081778,
+      "loss": 1.3233,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07431701866935775,
+      "grad_norm": 0.13434173195869106,
+      "learning_rate": 0.0001989742755900188,
+      "loss": 1.2906,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07438910113169465,
+      "grad_norm": 0.3294535978673851,
+      "learning_rate": 0.00019897093749423638,
+      "loss": 1.364,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07446118359403157,
+      "grad_norm": 0.1108406319191207,
+      "learning_rate": 0.00019896759400365248,
+      "loss": 1.3476,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07453326605636848,
+      "grad_norm": 0.11462771306274169,
+      "learning_rate": 0.00019896424511844926,
+      "loss": 1.3862,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0746053485187054,
+      "grad_norm": 0.11826566744952685,
+      "learning_rate": 0.00019896089083880936,
+      "loss": 1.348,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0746774309810423,
+      "grad_norm": 0.12150871926797896,
+      "learning_rate": 0.00019895753116491559,
+      "loss": 1.3344,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07474951344337923,
+      "grad_norm": 0.14380074228172054,
+      "learning_rate": 0.00019895416609695105,
+      "loss": 1.2838,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07482159590571615,
+      "grad_norm": 0.11826688943643754,
+      "learning_rate": 0.00019895079563509916,
+      "loss": 1.3524,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07489367836805305,
+      "grad_norm": 0.10570904845766806,
+      "learning_rate": 0.0001989474197795437,
+      "loss": 1.3427,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07496576083038997,
+      "grad_norm": 0.10862522011385786,
+      "learning_rate": 0.00019894403853046866,
+      "loss": 1.2559,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07503784329272688,
+      "grad_norm": 0.12055170533152396,
+      "learning_rate": 0.00019894065188805833,
+      "loss": 1.2785,
+      "step": 1041
+    },
+    {
+      "epoch": 0.0751099257550638,
+      "grad_norm": 0.10735601365505087,
+      "learning_rate": 0.00019893725985249732,
+      "loss": 1.3264,
+      "step": 1042
+    },
+    {
+      "epoch": 0.0751820082174007,
+      "grad_norm": 0.12022152336973504,
+      "learning_rate": 0.00019893386242397055,
+      "loss": 1.3726,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07525409067973762,
+      "grad_norm": 0.11506777156673405,
+      "learning_rate": 0.00019893045960266316,
+      "loss": 1.3625,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07532617314207453,
+      "grad_norm": 0.1601518242911187,
+      "learning_rate": 0.00019892705138876068,
+      "loss": 1.4626,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07539825560441145,
+      "grad_norm": 0.23133924825788385,
+      "learning_rate": 0.00019892363778244886,
+      "loss": 1.3903,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07547033806674835,
+      "grad_norm": 0.19747172729241216,
+      "learning_rate": 0.00019892021878391384,
+      "loss": 1.3392,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07554242052908527,
+      "grad_norm": 0.12279495462494949,
+      "learning_rate": 0.00019891679439334187,
+      "loss": 1.4284,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07561450299142218,
+      "grad_norm": 0.11204670092275801,
+      "learning_rate": 0.00019891336461091966,
+      "loss": 1.3517,
+      "step": 1049
+    },
+    {
+      "epoch": 0.0756865854537591,
+      "grad_norm": 0.1093863998529526,
+      "learning_rate": 0.0001989099294368342,
+      "loss": 1.3538,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07575866791609602,
+      "grad_norm": 0.10869465194356993,
+      "learning_rate": 0.00019890648887127275,
+      "loss": 1.3534,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07583075037843293,
+      "grad_norm": 0.17249322878712517,
+      "learning_rate": 0.00019890304291442278,
+      "loss": 1.2751,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07590283284076985,
+      "grad_norm": 0.1075603492979219,
+      "learning_rate": 0.00019889959156647215,
+      "loss": 1.3182,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07597491530310675,
+      "grad_norm": 0.2157158432630838,
+      "learning_rate": 0.000198896134827609,
+      "loss": 1.4297,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07604699776544367,
+      "grad_norm": 0.10403272779197031,
+      "learning_rate": 0.00019889267269802176,
+      "loss": 1.2715,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07611908022778058,
+      "grad_norm": 0.1515530058822773,
+      "learning_rate": 0.00019888920517789914,
+      "loss": 1.3595,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0761911626901175,
+      "grad_norm": 0.11132344486241115,
+      "learning_rate": 0.00019888573226743014,
+      "loss": 1.3773,
+      "step": 1057
+    },
+    {
+      "epoch": 0.0762632451524544,
+      "grad_norm": 0.0999282080310766,
+      "learning_rate": 0.0001988822539668041,
+      "loss": 1.2952,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07633532761479132,
+      "grad_norm": 0.11404272658574915,
+      "learning_rate": 0.0001988787702762106,
+      "loss": 1.3702,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07640741007712823,
+      "grad_norm": 0.12821842470971886,
+      "learning_rate": 0.00019887528119583948,
+      "loss": 1.3842,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07647949253946515,
+      "grad_norm": 0.14221521717473906,
+      "learning_rate": 0.000198871786725881,
+      "loss": 1.4369,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07655157500180206,
+      "grad_norm": 0.11277273248054318,
+      "learning_rate": 0.0001988682868665256,
+      "loss": 1.4403,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07662365746413898,
+      "grad_norm": 0.1005188768604402,
+      "learning_rate": 0.00019886478161796408,
+      "loss": 1.3738,
+      "step": 1063
+    },
+    {
+      "epoch": 0.0766957399264759,
+      "grad_norm": 0.11687068153137804,
+      "learning_rate": 0.00019886127098038745,
+      "loss": 1.37,
+      "step": 1064
+    },
+    {
+      "epoch": 0.0767678223888128,
+      "grad_norm": 0.10553634143494821,
+      "learning_rate": 0.00019885775495398714,
+      "loss": 1.4402,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07683990485114972,
+      "grad_norm": 0.11526963521832459,
+      "learning_rate": 0.0001988542335389548,
+      "loss": 1.2622,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07691198731348663,
+      "grad_norm": 0.10436879549612735,
+      "learning_rate": 0.00019885070673548235,
+      "loss": 1.3263,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07698406977582355,
+      "grad_norm": 0.10566549886480656,
+      "learning_rate": 0.000198847174543762,
+      "loss": 1.3663,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07705615223816045,
+      "grad_norm": 0.1009121997272479,
+      "learning_rate": 0.00019884363696398636,
+      "loss": 1.2766,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07712823470049737,
+      "grad_norm": 0.10149862002132408,
+      "learning_rate": 0.0001988400939963482,
+      "loss": 1.3936,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07720031716283428,
+      "grad_norm": 0.11052327876460202,
+      "learning_rate": 0.00019883654564104068,
+      "loss": 1.4071,
+      "step": 1071
+    },
+    {
+      "epoch": 0.0772723996251712,
+      "grad_norm": 0.11675807473637446,
+      "learning_rate": 0.00019883299189825724,
+      "loss": 1.395,
+      "step": 1072
+    },
+    {
+      "epoch": 0.0773444820875081,
+      "grad_norm": 0.1091603544494001,
+      "learning_rate": 0.00019882943276819154,
+      "loss": 1.4424,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07741656454984502,
+      "grad_norm": 0.14116955133004266,
+      "learning_rate": 0.00019882586825103758,
+      "loss": 1.2671,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07748864701218193,
+      "grad_norm": 0.21283629655940745,
+      "learning_rate": 0.00019882229834698967,
+      "loss": 1.2587,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07756072947451885,
+      "grad_norm": 0.10583252058085593,
+      "learning_rate": 0.00019881872305624242,
+      "loss": 1.3191,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07763281193685577,
+      "grad_norm": 0.10278481809142136,
+      "learning_rate": 0.00019881514237899073,
+      "loss": 1.3815,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07770489439919268,
+      "grad_norm": 0.11911034571868741,
+      "learning_rate": 0.00019881155631542975,
+      "loss": 1.2844,
+      "step": 1078
+    },
+    {
+      "epoch": 0.0777769768615296,
+      "grad_norm": 0.11500406186412951,
+      "learning_rate": 0.00019880796486575496,
+      "loss": 1.4616,
+      "step": 1079
+    },
+    {
+      "epoch": 0.0778490593238665,
+      "grad_norm": 0.09783600060433964,
+      "learning_rate": 0.00019880436803016208,
+      "loss": 1.2572,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07792114178620342,
+      "grad_norm": 0.13147732353207853,
+      "learning_rate": 0.00019880076580884722,
+      "loss": 1.4486,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07799322424854033,
+      "grad_norm": 0.10854491415509225,
+      "learning_rate": 0.00019879715820200677,
+      "loss": 1.3878,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07806530671087725,
+      "grad_norm": 0.10700041629317468,
+      "learning_rate": 0.00019879354520983733,
+      "loss": 1.3013,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07813738917321415,
+      "grad_norm": 0.14192581197940674,
+      "learning_rate": 0.00019878992683253582,
+      "loss": 1.3938,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07820947163555107,
+      "grad_norm": 0.09657962802266488,
+      "learning_rate": 0.0001987863030702995,
+      "loss": 1.2965,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07828155409788798,
+      "grad_norm": 0.10601898501790712,
+      "learning_rate": 0.0001987826739233259,
+      "loss": 1.3613,
+      "step": 1086
+    },
+    {
+      "epoch": 0.0783536365602249,
+      "grad_norm": 0.10370529927449099,
+      "learning_rate": 0.00019877903939181284,
+      "loss": 1.2994,
+      "step": 1087
+    },
+    {
+      "epoch": 0.0784257190225618,
+      "grad_norm": 0.10622729545203616,
+      "learning_rate": 0.00019877539947595844,
+      "loss": 1.2886,
+      "step": 1088
+    },
+    {
+      "epoch": 0.07849780148489872,
+      "grad_norm": 0.11471162951175286,
+      "learning_rate": 0.00019877175417596106,
+      "loss": 1.3813,
+      "step": 1089
+    },
+    {
+      "epoch": 0.07856988394723564,
+      "grad_norm": 0.12545931030906257,
+      "learning_rate": 0.00019876810349201947,
+      "loss": 1.4145,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07864196640957255,
+      "grad_norm": 0.11073197785700385,
+      "learning_rate": 0.00019876444742433265,
+      "loss": 1.3031,
+      "step": 1091
+    },
+    {
+      "epoch": 0.07871404887190947,
+      "grad_norm": 0.12399515942152001,
+      "learning_rate": 0.00019876078597309984,
+      "loss": 1.2872,
+      "step": 1092
+    },
+    {
+      "epoch": 0.07878613133424638,
+      "grad_norm": 0.11139005518620088,
+      "learning_rate": 0.00019875711913852067,
+      "loss": 1.4105,
+      "step": 1093
+    },
+    {
+      "epoch": 0.0788582137965833,
+      "grad_norm": 0.2364361083922266,
+      "learning_rate": 0.000198753446920795,
+      "loss": 1.4903,
+      "step": 1094
+    },
+    {
+      "epoch": 0.0789302962589202,
+      "grad_norm": 0.10425415051060798,
+      "learning_rate": 0.000198749769320123,
+      "loss": 1.3678,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07900237872125712,
+      "grad_norm": 0.0942382783588956,
+      "learning_rate": 0.00019874608633670512,
+      "loss": 1.2877,
+      "step": 1096
+    },
+    {
+      "epoch": 0.07907446118359403,
+      "grad_norm": 0.11433980223259481,
+      "learning_rate": 0.00019874239797074212,
+      "loss": 1.4316,
+      "step": 1097
+    },
+    {
+      "epoch": 0.07914654364593095,
+      "grad_norm": 0.13154809941019946,
+      "learning_rate": 0.00019873870422243507,
+      "loss": 1.3911,
+      "step": 1098
+    },
+    {
+      "epoch": 0.07921862610826785,
+      "grad_norm": 0.12203167935267528,
+      "learning_rate": 0.0001987350050919853,
+      "loss": 1.3926,
+      "step": 1099
+    },
+    {
+      "epoch": 0.07929070857060477,
+      "grad_norm": 0.09790084283660713,
+      "learning_rate": 0.00019873130057959445,
+      "loss": 1.415,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07936279103294168,
+      "grad_norm": 0.10635804203298534,
+      "learning_rate": 0.00019872759068546446,
+      "loss": 1.3338,
+      "step": 1101
+    },
+    {
+      "epoch": 0.0794348734952786,
+      "grad_norm": 0.11406733201976885,
+      "learning_rate": 0.00019872387540979748,
+      "loss": 1.3891,
+      "step": 1102
+    },
+    {
+      "epoch": 0.07950695595761552,
+      "grad_norm": 0.09719044324781571,
+      "learning_rate": 0.00019872015475279616,
+      "loss": 1.3501,
+      "step": 1103
+    },
+    {
+      "epoch": 0.07957903841995242,
+      "grad_norm": 0.10594207153983433,
+      "learning_rate": 0.00019871642871466315,
+      "loss": 1.4084,
+      "step": 1104
+    },
+    {
+      "epoch": 0.07965112088228934,
+      "grad_norm": 0.10158842197342967,
+      "learning_rate": 0.00019871269729560166,
+      "loss": 1.3562,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07972320334462625,
+      "grad_norm": 0.12355133302432797,
+      "learning_rate": 0.0001987089604958151,
+      "loss": 1.3649,
+      "step": 1106
+    },
+    {
+      "epoch": 0.07979528580696317,
+      "grad_norm": 0.13800577426337357,
+      "learning_rate": 0.00019870521831550706,
+      "loss": 1.3742,
+      "step": 1107
+    },
+    {
+      "epoch": 0.07986736826930008,
+      "grad_norm": 0.24914769209011042,
+      "learning_rate": 0.0001987014707548816,
+      "loss": 1.2284,
+      "step": 1108
+    },
+    {
+      "epoch": 0.079939450731637,
+      "grad_norm": 0.10611596093592145,
+      "learning_rate": 0.000198697717814143,
+      "loss": 1.2585,
+      "step": 1109
+    },
+    {
+      "epoch": 0.0800115331939739,
+      "grad_norm": 0.10628223352212979,
+      "learning_rate": 0.00019869395949349578,
+      "loss": 1.4109,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08008361565631082,
+      "grad_norm": 0.11643700159539289,
+      "learning_rate": 0.0001986901957931448,
+      "loss": 1.2149,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08015569811864773,
+      "grad_norm": 0.24099546144182404,
+      "learning_rate": 0.0001986864267132953,
+      "loss": 1.461,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08022778058098465,
+      "grad_norm": 0.10734358976061602,
+      "learning_rate": 0.00019868265225415265,
+      "loss": 1.2515,
+      "step": 1113
+    },
+    {
+      "epoch": 0.08029986304332155,
+      "grad_norm": 0.11574490543385953,
+      "learning_rate": 0.00019867887241592258,
+      "loss": 1.3326,
+      "step": 1114
+    },
+    {
+      "epoch": 0.08037194550565847,
+      "grad_norm": 0.11358050526251016,
+      "learning_rate": 0.00019867508719881118,
+      "loss": 1.3346,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0804440279679954,
+      "grad_norm": 0.10632120464387049,
+      "learning_rate": 0.00019867129660302476,
+      "loss": 1.3964,
+      "step": 1116
+    },
+    {
+      "epoch": 0.0805161104303323,
+      "grad_norm": 0.11512211790866692,
+      "learning_rate": 0.00019866750062876994,
+      "loss": 1.2136,
+      "step": 1117
+    },
+    {
+      "epoch": 0.08058819289266922,
+      "grad_norm": 0.1105446776729597,
+      "learning_rate": 0.00019866369927625362,
+      "loss": 1.2953,
+      "step": 1118
+    },
+    {
+      "epoch": 0.08066027535500613,
+      "grad_norm": 0.13380837383400143,
+      "learning_rate": 0.000198659892545683,
+      "loss": 1.3585,
+      "step": 1119
+    },
+    {
+      "epoch": 0.08073235781734305,
+      "grad_norm": 0.21802142388396725,
+      "learning_rate": 0.0001986560804372656,
+      "loss": 1.278,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08080444027967995,
+      "grad_norm": 0.10213923255240653,
+      "learning_rate": 0.00019865226295120926,
+      "loss": 1.2899,
+      "step": 1121
+    },
+    {
+      "epoch": 0.08087652274201687,
+      "grad_norm": 0.1442346574421232,
+      "learning_rate": 0.000198648440087722,
+      "loss": 1.3964,
+      "step": 1122
+    },
+    {
+      "epoch": 0.08094860520435378,
+      "grad_norm": 0.11810454737196545,
+      "learning_rate": 0.0001986446118470122,
+      "loss": 1.3124,
+      "step": 1123
+    },
+    {
+      "epoch": 0.0810206876666907,
+      "grad_norm": 0.09493202164791088,
+      "learning_rate": 0.00019864077822928858,
+      "loss": 1.3408,
+      "step": 1124
+    },
+    {
+      "epoch": 0.0810927701290276,
+      "grad_norm": 0.10546138093479338,
+      "learning_rate": 0.00019863693923476006,
+      "loss": 1.2689,
+      "step": 1125
+    },
+    {
+      "epoch": 0.08116485259136452,
+      "grad_norm": 0.10657436444487868,
+      "learning_rate": 0.00019863309486363592,
+      "loss": 1.2584,
+      "step": 1126
+    },
+    {
+      "epoch": 0.08123693505370143,
+      "grad_norm": 0.1101842485702874,
+      "learning_rate": 0.00019862924511612575,
+      "loss": 1.3073,
+      "step": 1127
+    },
+    {
+      "epoch": 0.08130901751603835,
+      "grad_norm": 0.1257760436345161,
+      "learning_rate": 0.00019862538999243932,
+      "loss": 1.4364,
+      "step": 1128
+    },
+    {
+      "epoch": 0.08138109997837527,
+      "grad_norm": 0.1590206608206496,
+      "learning_rate": 0.0001986215294927868,
+      "loss": 1.4355,
+      "step": 1129
+    },
+    {
+      "epoch": 0.08145318244071217,
+      "grad_norm": 0.1076880568319889,
+      "learning_rate": 0.00019861766361737866,
+      "loss": 1.2774,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0815252649030491,
+      "grad_norm": 0.1967046421502376,
+      "learning_rate": 0.00019861379236642558,
+      "loss": 1.3365,
+      "step": 1131
+    },
+    {
+      "epoch": 0.081597347365386,
+      "grad_norm": 0.11979774017393151,
+      "learning_rate": 0.00019860991574013862,
+      "loss": 1.2705,
+      "step": 1132
+    },
+    {
+      "epoch": 0.08166942982772292,
+      "grad_norm": 0.1172049326725042,
+      "learning_rate": 0.00019860603373872903,
+      "loss": 1.3118,
+      "step": 1133
+    },
+    {
+      "epoch": 0.08174151229005983,
+      "grad_norm": 0.10971364617636005,
+      "learning_rate": 0.00019860214636240845,
+      "loss": 1.3262,
+      "step": 1134
+    },
+    {
+      "epoch": 0.08181359475239675,
+      "grad_norm": 0.1327912755059443,
+      "learning_rate": 0.00019859825361138878,
+      "loss": 1.3414,
+      "step": 1135
+    },
+    {
+      "epoch": 0.08188567721473365,
+      "grad_norm": 0.1017002336321644,
+      "learning_rate": 0.00019859435548588218,
+      "loss": 1.3885,
+      "step": 1136
+    },
+    {
+      "epoch": 0.08195775967707057,
+      "grad_norm": 0.09065006424251026,
+      "learning_rate": 0.00019859045198610117,
+      "loss": 1.2895,
+      "step": 1137
+    },
+    {
+      "epoch": 0.08202984213940748,
+      "grad_norm": 0.11702754679742465,
+      "learning_rate": 0.00019858654311225853,
+      "loss": 1.3154,
+      "step": 1138
+    },
+    {
+      "epoch": 0.0821019246017444,
+      "grad_norm": 0.10481497204617114,
+      "learning_rate": 0.00019858262886456732,
+      "loss": 1.3419,
+      "step": 1139
+    },
+    {
+      "epoch": 0.0821740070640813,
+      "grad_norm": 0.1029837266366933,
+      "learning_rate": 0.00019857870924324086,
+      "loss": 1.3619,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08224608952641822,
+      "grad_norm": 0.10885869595251549,
+      "learning_rate": 0.00019857478424849286,
+      "loss": 1.2864,
+      "step": 1141
+    },
+    {
+      "epoch": 0.08231817198875514,
+      "grad_norm": 0.12333406389965909,
+      "learning_rate": 0.00019857085388053723,
+      "loss": 1.4088,
+      "step": 1142
+    },
+    {
+      "epoch": 0.08239025445109205,
+      "grad_norm": 0.12004396269036041,
+      "learning_rate": 0.0001985669181395882,
+      "loss": 1.3132,
+      "step": 1143
+    },
+    {
+      "epoch": 0.08246233691342897,
+      "grad_norm": 0.17090504232823264,
+      "learning_rate": 0.00019856297702586036,
+      "loss": 1.3737,
+      "step": 1144
+    },
+    {
+      "epoch": 0.08253441937576587,
+      "grad_norm": 0.1067879767412192,
+      "learning_rate": 0.00019855903053956847,
+      "loss": 1.3853,
+      "step": 1145
+    },
+    {
+      "epoch": 0.0826065018381028,
+      "grad_norm": 0.09829709457016718,
+      "learning_rate": 0.0001985550786809277,
+      "loss": 1.3855,
+      "step": 1146
+    },
+    {
+      "epoch": 0.0826785843004397,
+      "grad_norm": 0.15875346964522485,
+      "learning_rate": 0.00019855112145015347,
+      "loss": 1.3385,
+      "step": 1147
+    },
+    {
+      "epoch": 0.08275066676277662,
+      "grad_norm": 0.10167871982126334,
+      "learning_rate": 0.0001985471588474614,
+      "loss": 1.3392,
+      "step": 1148
+    },
+    {
+      "epoch": 0.08282274922511353,
+      "grad_norm": 0.1266641940073995,
+      "learning_rate": 0.0001985431908730676,
+      "loss": 1.2747,
+      "step": 1149
+    },
+    {
+      "epoch": 0.08289483168745045,
+      "grad_norm": 0.18673666205364944,
+      "learning_rate": 0.00019853921752718828,
+      "loss": 1.3838,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08296691414978735,
+      "grad_norm": 0.1105246821243112,
+      "learning_rate": 0.00019853523881004006,
+      "loss": 1.272,
+      "step": 1151
+    },
+    {
+      "epoch": 0.08303899661212427,
+      "grad_norm": 0.41296083827313845,
+      "learning_rate": 0.00019853125472183979,
+      "loss": 1.3848,
+      "step": 1152
+    },
+    {
+      "epoch": 0.08311107907446118,
+      "grad_norm": 0.1221496145846238,
+      "learning_rate": 0.00019852726526280466,
+      "loss": 1.4494,
+      "step": 1153
+    },
+    {
+      "epoch": 0.0831831615367981,
+      "grad_norm": 0.10721333445859951,
+      "learning_rate": 0.00019852327043315212,
+      "loss": 1.3388,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08325524399913502,
+      "grad_norm": 0.2641062526473533,
+      "learning_rate": 0.0001985192702330999,
+      "loss": 1.4096,
+      "step": 1155
+    },
+    {
+      "epoch": 0.08332732646147192,
+      "grad_norm": 0.10791304843322203,
+      "learning_rate": 0.0001985152646628661,
+      "loss": 1.3693,
+      "step": 1156
+    },
+    {
+      "epoch": 0.08339940892380884,
+      "grad_norm": 0.14128043708431962,
+      "learning_rate": 0.00019851125372266902,
+      "loss": 1.3369,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08347149138614575,
+      "grad_norm": 0.11874582049024687,
+      "learning_rate": 0.00019850723741272734,
+      "loss": 1.2901,
+      "step": 1158
+    },
+    {
+      "epoch": 0.08354357384848267,
+      "grad_norm": 0.11479916077899367,
+      "learning_rate": 0.00019850321573325992,
+      "loss": 1.2947,
+      "step": 1159
+    },
+    {
+      "epoch": 0.08361565631081957,
+      "grad_norm": 0.3314884937486859,
+      "learning_rate": 0.00019849918868448602,
+      "loss": 1.2286,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0836877387731565,
+      "grad_norm": 0.11390600220399731,
+      "learning_rate": 0.0001984951562666251,
+      "loss": 1.3596,
+      "step": 1161
+    },
+    {
+      "epoch": 0.0837598212354934,
+      "grad_norm": 0.12617109714222435,
+      "learning_rate": 0.00019849111847989706,
+      "loss": 1.3146,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08383190369783032,
+      "grad_norm": 0.1395568643766966,
+      "learning_rate": 0.00019848707532452191,
+      "loss": 1.3,
+      "step": 1163
+    },
+    {
+      "epoch": 0.08390398616016723,
+      "grad_norm": 0.11920433256181591,
+      "learning_rate": 0.00019848302680072007,
+      "loss": 1.2697,
+      "step": 1164
+    },
+    {
+      "epoch": 0.08397606862250415,
+      "grad_norm": 0.1093008294917657,
+      "learning_rate": 0.0001984789729087122,
+      "loss": 1.215,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08404815108484105,
+      "grad_norm": 0.23183229534085692,
+      "learning_rate": 0.00019847491364871931,
+      "loss": 1.3646,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08412023354717797,
+      "grad_norm": 0.11446706509496703,
+      "learning_rate": 0.00019847084902096263,
+      "loss": 1.3563,
+      "step": 1167
+    },
+    {
+      "epoch": 0.08419231600951488,
+      "grad_norm": 0.21112402127785818,
+      "learning_rate": 0.0001984667790256637,
+      "loss": 1.3719,
+      "step": 1168
+    },
+    {
+      "epoch": 0.0842643984718518,
+      "grad_norm": 0.1176261439476917,
+      "learning_rate": 0.00019846270366304447,
+      "loss": 1.2728,
+      "step": 1169
+    },
+    {
+      "epoch": 0.08433648093418872,
+      "grad_norm": 0.1352747490185362,
+      "learning_rate": 0.00019845862293332698,
+      "loss": 1.4177,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08440856339652562,
+      "grad_norm": 0.11237548094013501,
+      "learning_rate": 0.00019845453683673368,
+      "loss": 1.2826,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08448064585886254,
+      "grad_norm": 0.23822280473138507,
+      "learning_rate": 0.00019845044537348738,
+      "loss": 1.2154,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08455272832119945,
+      "grad_norm": 0.10141842052815203,
+      "learning_rate": 0.00019844634854381102,
+      "loss": 1.2839,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08462481078353637,
+      "grad_norm": 0.11243312308045052,
+      "learning_rate": 0.00019844224634792793,
+      "loss": 1.3139,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08469689324587328,
+      "grad_norm": 0.21844348619268472,
+      "learning_rate": 0.00019843813878606174,
+      "loss": 1.328,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0847689757082102,
+      "grad_norm": 0.13401876526682469,
+      "learning_rate": 0.00019843402585843631,
+      "loss": 1.3165,
+      "step": 1176
+    },
+    {
+      "epoch": 0.0848410581705471,
+      "grad_norm": 0.09946822913272689,
+      "learning_rate": 0.00019842990756527588,
+      "loss": 1.3822,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08491314063288402,
+      "grad_norm": 0.12173066504484917,
+      "learning_rate": 0.00019842578390680488,
+      "loss": 1.3149,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08498522309522093,
+      "grad_norm": 0.12239409751927012,
+      "learning_rate": 0.00019842165488324814,
+      "loss": 1.36,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08505730555755785,
+      "grad_norm": 0.11111860421566971,
+      "learning_rate": 0.0001984175204948307,
+      "loss": 1.3877,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08512938801989475,
+      "grad_norm": 0.1702768502978914,
+      "learning_rate": 0.00019841338074177793,
+      "loss": 1.4089,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08520147048223167,
+      "grad_norm": 0.12249322373599923,
+      "learning_rate": 0.00019840923562431545,
+      "loss": 1.4022,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08527355294456859,
+      "grad_norm": 0.11369535362634275,
+      "learning_rate": 0.00019840508514266928,
+      "loss": 1.2921,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0853456354069055,
+      "grad_norm": 0.18877118786600283,
+      "learning_rate": 0.00019840092929706557,
+      "loss": 1.368,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08541771786924242,
+      "grad_norm": 0.11157454982218756,
+      "learning_rate": 0.0001983967680877309,
+      "loss": 1.4034,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08548980033157932,
+      "grad_norm": 0.1273661360316139,
+      "learning_rate": 0.00019839260151489213,
+      "loss": 1.3037,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08556188279391624,
+      "grad_norm": 0.1106631692030795,
+      "learning_rate": 0.00019838842957877625,
+      "loss": 1.3721,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08563396525625315,
+      "grad_norm": 0.11293196224007901,
+      "learning_rate": 0.00019838425227961082,
+      "loss": 1.2841,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08570604771859007,
+      "grad_norm": 0.10618234281636425,
+      "learning_rate": 0.00019838006961762344,
+      "loss": 1.3267,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08577813018092698,
+      "grad_norm": 0.11677944983134689,
+      "learning_rate": 0.00019837588159304215,
+      "loss": 1.388,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0858502126432639,
+      "grad_norm": 0.1268909284462292,
+      "learning_rate": 0.00019837168820609522,
+      "loss": 1.309,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0859222951056008,
+      "grad_norm": 0.1265603529953798,
+      "learning_rate": 0.0001983674894570112,
+      "loss": 1.3975,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08599437756793772,
+      "grad_norm": 0.11419831243408539,
+      "learning_rate": 0.000198363285346019,
+      "loss": 1.3573,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08606646003027463,
+      "grad_norm": 0.11451655079625994,
+      "learning_rate": 0.00019835907587334775,
+      "loss": 1.3305,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08613854249261155,
+      "grad_norm": 0.11796021257401472,
+      "learning_rate": 0.00019835486103922693,
+      "loss": 1.4146,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08621062495494847,
+      "grad_norm": 0.10557166257550511,
+      "learning_rate": 0.00019835064084388628,
+      "loss": 1.3071,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08628270741728537,
+      "grad_norm": 0.13792340212976528,
+      "learning_rate": 0.00019834641528755582,
+      "loss": 1.4139,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08635478987962229,
+      "grad_norm": 0.13802691914166604,
+      "learning_rate": 0.0001983421843704659,
+      "loss": 1.2758,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0864268723419592,
+      "grad_norm": 0.12214506118401672,
+      "learning_rate": 0.00019833794809284716,
+      "loss": 1.3385,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08649895480429612,
+      "grad_norm": 0.10738954949813878,
+      "learning_rate": 0.00019833370645493047,
+      "loss": 1.3095,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08657103726663302,
+      "grad_norm": 0.1094833292248011,
+      "learning_rate": 0.00019832945945694706,
+      "loss": 1.301,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08664311972896994,
+      "grad_norm": 0.10005399823623792,
+      "learning_rate": 0.00019832520709912844,
+      "loss": 1.3394,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08671520219130685,
+      "grad_norm": 0.1038804096912141,
+      "learning_rate": 0.00019832094938170637,
+      "loss": 1.3278,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08678728465364377,
+      "grad_norm": 0.1393189954233972,
+      "learning_rate": 0.00019831668630491297,
+      "loss": 1.3651,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08685936711598068,
+      "grad_norm": 0.19792194510854164,
+      "learning_rate": 0.00019831241786898062,
+      "loss": 1.396,
+      "step": 1205
+    },
+    {
+      "epoch": 0.0869314495783176,
+      "grad_norm": 0.09966613337423673,
+      "learning_rate": 0.00019830814407414193,
+      "loss": 1.2886,
+      "step": 1206
+    },
+    {
+      "epoch": 0.0870035320406545,
+      "grad_norm": 0.11606619622619384,
+      "learning_rate": 0.00019830386492062992,
+      "loss": 1.3248,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08707561450299142,
+      "grad_norm": 0.12325543190727702,
+      "learning_rate": 0.00019829958040867783,
+      "loss": 1.3244,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08714769696532834,
+      "grad_norm": 0.16515874417288112,
+      "learning_rate": 0.00019829529053851919,
+      "loss": 1.3281,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08721977942766525,
+      "grad_norm": 0.09850563515308847,
+      "learning_rate": 0.00019829099531038782,
+      "loss": 1.3587,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08729186189000217,
+      "grad_norm": 0.09064729537640524,
+      "learning_rate": 0.0001982866947245179,
+      "loss": 1.3448,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08736394435233907,
+      "grad_norm": 0.11748994759997765,
+      "learning_rate": 0.0001982823887811438,
+      "loss": 1.2985,
+      "step": 1212
+    },
+    {
+      "epoch": 0.087436026814676,
+      "grad_norm": 0.12835023171988866,
+      "learning_rate": 0.00019827807748050023,
+      "loss": 1.3988,
+      "step": 1213
+    },
+    {
+      "epoch": 0.0875081092770129,
+      "grad_norm": 0.12398496070158836,
+      "learning_rate": 0.00019827376082282223,
+      "loss": 1.416,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08758019173934982,
+      "grad_norm": 0.10559271030115173,
+      "learning_rate": 0.0001982694388083451,
+      "loss": 1.2872,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08765227420168673,
+      "grad_norm": 0.11289970280716476,
+      "learning_rate": 0.0001982651114373044,
+      "loss": 1.4146,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08772435666402364,
+      "grad_norm": 0.12347906490746927,
+      "learning_rate": 0.000198260778709936,
+      "loss": 1.4108,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08779643912636055,
+      "grad_norm": 0.11500754765932522,
+      "learning_rate": 0.00019825644062647615,
+      "loss": 1.3686,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08786852158869747,
+      "grad_norm": 0.10376594675490998,
+      "learning_rate": 0.00019825209718716123,
+      "loss": 1.3714,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08794060405103438,
+      "grad_norm": 0.10900495914078798,
+      "learning_rate": 0.00019824774839222802,
+      "loss": 1.3638,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0880126865133713,
+      "grad_norm": 0.1602872000799563,
+      "learning_rate": 0.00019824339424191355,
+      "loss": 1.327,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08808476897570822,
+      "grad_norm": 0.11949201292762497,
+      "learning_rate": 0.00019823903473645523,
+      "loss": 1.3549,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08815685143804512,
+      "grad_norm": 0.18169059861198722,
+      "learning_rate": 0.00019823466987609062,
+      "loss": 1.2893,
+      "step": 1223
+    },
+    {
+      "epoch": 0.08822893390038204,
+      "grad_norm": 0.11018524211398839,
+      "learning_rate": 0.0001982302996610577,
+      "loss": 1.3403,
+      "step": 1224
+    },
+    {
+      "epoch": 0.08830101636271895,
+      "grad_norm": 0.12422439507945017,
+      "learning_rate": 0.00019822592409159462,
+      "loss": 1.3388,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08837309882505587,
+      "grad_norm": 0.09734965921032433,
+      "learning_rate": 0.00019822154316793993,
+      "loss": 1.4039,
+      "step": 1226
+    },
+    {
+      "epoch": 0.08844518128739277,
+      "grad_norm": 0.11910911065638599,
+      "learning_rate": 0.00019821715689033242,
+      "loss": 1.3866,
+      "step": 1227
+    },
+    {
+      "epoch": 0.0885172637497297,
+      "grad_norm": 0.2112363960092803,
+      "learning_rate": 0.0001982127652590112,
+      "loss": 1.3072,
+      "step": 1228
+    },
+    {
+      "epoch": 0.0885893462120666,
+      "grad_norm": 0.10456934422697493,
+      "learning_rate": 0.0001982083682742156,
+      "loss": 1.3973,
+      "step": 1229
+    },
+    {
+      "epoch": 0.08866142867440352,
+      "grad_norm": 0.11141278483306366,
+      "learning_rate": 0.00019820396593618535,
+      "loss": 1.3279,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08873351113674043,
+      "grad_norm": 0.10987693161761926,
+      "learning_rate": 0.00019819955824516044,
+      "loss": 1.3266,
+      "step": 1231
+    },
+    {
+      "epoch": 0.08880559359907735,
+      "grad_norm": 0.12497653693949064,
+      "learning_rate": 0.00019819514520138105,
+      "loss": 1.3376,
+      "step": 1232
+    },
+    {
+      "epoch": 0.08887767606141425,
+      "grad_norm": 0.09908645725755544,
+      "learning_rate": 0.00019819072680508776,
+      "loss": 1.3364,
+      "step": 1233
+    },
+    {
+      "epoch": 0.08894975852375117,
+      "grad_norm": 0.11971424907530623,
+      "learning_rate": 0.0001981863030565214,
+      "loss": 1.4668,
+      "step": 1234
+    },
+    {
+      "epoch": 0.08902184098608809,
+      "grad_norm": 0.10197932342968408,
+      "learning_rate": 0.0001981818739559232,
+      "loss": 1.349,
+      "step": 1235
+    },
+    {
+      "epoch": 0.089093923448425,
+      "grad_norm": 0.1050026046659351,
+      "learning_rate": 0.00019817743950353443,
+      "loss": 1.3013,
+      "step": 1236
+    },
+    {
+      "epoch": 0.08916600591076192,
+      "grad_norm": 0.16322275138898898,
+      "learning_rate": 0.00019817299969959693,
+      "loss": 1.313,
+      "step": 1237
+    },
+    {
+      "epoch": 0.08923808837309882,
+      "grad_norm": 0.1129218878666565,
+      "learning_rate": 0.00019816855454435266,
+      "loss": 1.3454,
+      "step": 1238
+    },
+    {
+      "epoch": 0.08931017083543574,
+      "grad_norm": 0.3864966097723567,
+      "learning_rate": 0.00019816410403804394,
+      "loss": 1.3175,
+      "step": 1239
+    },
+    {
+      "epoch": 0.08938225329777265,
+      "grad_norm": 0.11102536484985137,
+      "learning_rate": 0.0001981596481809133,
+      "loss": 1.4267,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08945433576010957,
+      "grad_norm": 0.10702748645629714,
+      "learning_rate": 0.00019815518697320373,
+      "loss": 1.3534,
+      "step": 1241
+    },
+    {
+      "epoch": 0.08952641822244647,
+      "grad_norm": 0.11395046512724354,
+      "learning_rate": 0.0001981507204151583,
+      "loss": 1.3806,
+      "step": 1242
+    },
+    {
+      "epoch": 0.0895985006847834,
+      "grad_norm": 0.10989328842970572,
+      "learning_rate": 0.00019814624850702055,
+      "loss": 1.3139,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0896705831471203,
+      "grad_norm": 0.16238059172157898,
+      "learning_rate": 0.00019814177124903421,
+      "loss": 1.2981,
+      "step": 1244
+    },
+    {
+      "epoch": 0.08974266560945722,
+      "grad_norm": 0.0874461911664559,
+      "learning_rate": 0.00019813728864144335,
+      "loss": 1.3247,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08981474807179413,
+      "grad_norm": 0.1833028733723648,
+      "learning_rate": 0.00019813280068449227,
+      "loss": 1.4188,
+      "step": 1246
+    },
+    {
+      "epoch": 0.08988683053413105,
+      "grad_norm": 0.10526093031112839,
+      "learning_rate": 0.00019812830737842566,
+      "loss": 1.4468,
+      "step": 1247
+    },
+    {
+      "epoch": 0.08995891299646797,
+      "grad_norm": 0.10042687995335693,
+      "learning_rate": 0.0001981238087234884,
+      "loss": 1.3383,
+      "step": 1248
+    },
+    {
+      "epoch": 0.09003099545880487,
+      "grad_norm": 0.1445890648622169,
+      "learning_rate": 0.00019811930471992572,
+      "loss": 1.3585,
+      "step": 1249
+    },
+    {
+      "epoch": 0.09010307792114179,
+      "grad_norm": 0.15386776748436715,
+      "learning_rate": 0.00019811479536798314,
+      "loss": 1.3688,
+      "step": 1250
+    },
+    {
+      "epoch": 0.0901751603834787,
+      "grad_norm": 0.11023448468869673,
+      "learning_rate": 0.00019811028066790644,
+      "loss": 1.3546,
+      "step": 1251
+    },
+    {
+      "epoch": 0.09024724284581562,
+      "grad_norm": 0.11751234883646815,
+      "learning_rate": 0.00019810576061994177,
+      "loss": 1.3744,
+      "step": 1252
+    },
+    {
+      "epoch": 0.09031932530815252,
+      "grad_norm": 0.10849232991520047,
+      "learning_rate": 0.00019810123522433543,
+      "loss": 1.2886,
+      "step": 1253
+    },
+    {
+      "epoch": 0.09039140777048944,
+      "grad_norm": 0.11836087952458178,
+      "learning_rate": 0.00019809670448133412,
+      "loss": 1.2899,
+      "step": 1254
+    },
+    {
+      "epoch": 0.09046349023282635,
+      "grad_norm": 0.13603790392374818,
+      "learning_rate": 0.00019809216839118484,
+      "loss": 1.3533,
+      "step": 1255
+    },
+    {
+      "epoch": 0.09053557269516327,
+      "grad_norm": 0.1289020192345823,
+      "learning_rate": 0.00019808762695413482,
+      "loss": 1.3494,
+      "step": 1256
+    },
+    {
+      "epoch": 0.09060765515750017,
+      "grad_norm": 0.09938489406404703,
+      "learning_rate": 0.0001980830801704316,
+      "loss": 1.3473,
+      "step": 1257
+    },
+    {
+      "epoch": 0.0906797376198371,
+      "grad_norm": 0.12008969602885251,
+      "learning_rate": 0.00019807852804032305,
+      "loss": 1.2695,
+      "step": 1258
+    },
+    {
+      "epoch": 0.090751820082174,
+      "grad_norm": 0.22143316042324457,
+      "learning_rate": 0.0001980739705640573,
+      "loss": 1.461,
+      "step": 1259
+    },
+    {
+      "epoch": 0.09082390254451092,
+      "grad_norm": 0.1191362988534227,
+      "learning_rate": 0.00019806940774188273,
+      "loss": 1.2857,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09089598500684784,
+      "grad_norm": 0.16633614185446274,
+      "learning_rate": 0.00019806483957404808,
+      "loss": 1.358,
+      "step": 1261
+    },
+    {
+      "epoch": 0.09096806746918475,
+      "grad_norm": 0.10714548188461437,
+      "learning_rate": 0.0001980602660608024,
+      "loss": 1.3863,
+      "step": 1262
+    },
+    {
+      "epoch": 0.09104014993152167,
+      "grad_norm": 0.11751889371007838,
+      "learning_rate": 0.0001980556872023949,
+      "loss": 1.2922,
+      "step": 1263
+    },
+    {
+      "epoch": 0.09111223239385857,
+      "grad_norm": 0.10763645455530937,
+      "learning_rate": 0.00019805110299907526,
+      "loss": 1.3603,
+      "step": 1264
+    },
+    {
+      "epoch": 0.09118431485619549,
+      "grad_norm": 0.09592589736480148,
+      "learning_rate": 0.00019804651345109328,
+      "loss": 1.3158,
+      "step": 1265
+    },
+    {
+      "epoch": 0.0912563973185324,
+      "grad_norm": 0.10713126589177038,
+      "learning_rate": 0.0001980419185586992,
+      "loss": 1.3005,
+      "step": 1266
+    },
+    {
+      "epoch": 0.09132847978086932,
+      "grad_norm": 0.11100099039734536,
+      "learning_rate": 0.00019803731832214344,
+      "loss": 1.4237,
+      "step": 1267
+    },
+    {
+      "epoch": 0.09140056224320622,
+      "grad_norm": 0.09399139106906153,
+      "learning_rate": 0.00019803271274167674,
+      "loss": 1.3263,
+      "step": 1268
+    },
+    {
+      "epoch": 0.09147264470554314,
+      "grad_norm": 0.11347816670522043,
+      "learning_rate": 0.0001980281018175502,
+      "loss": 1.2638,
+      "step": 1269
+    },
+    {
+      "epoch": 0.09154472716788005,
+      "grad_norm": 0.10851204754922655,
+      "learning_rate": 0.0001980234855500151,
+      "loss": 1.2596,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09161680963021697,
+      "grad_norm": 0.13519683847946248,
+      "learning_rate": 0.00019801886393932313,
+      "loss": 1.3097,
+      "step": 1271
+    },
+    {
+      "epoch": 0.09168889209255388,
+      "grad_norm": 0.12616714882102648,
+      "learning_rate": 0.00019801423698572617,
+      "loss": 1.4875,
+      "step": 1272
+    },
+    {
+      "epoch": 0.0917609745548908,
+      "grad_norm": 0.0974134612183364,
+      "learning_rate": 0.00019800960468947638,
+      "loss": 1.3247,
+      "step": 1273
+    },
+    {
+      "epoch": 0.09183305701722771,
+      "grad_norm": 0.11531380551495854,
+      "learning_rate": 0.00019800496705082638,
+      "loss": 1.376,
+      "step": 1274
+    },
+    {
+      "epoch": 0.09190513947956462,
+      "grad_norm": 0.1048353091222549,
+      "learning_rate": 0.00019800032407002886,
+      "loss": 1.3666,
+      "step": 1275
+    },
+    {
+      "epoch": 0.09197722194190154,
+      "grad_norm": 0.10784918091122532,
+      "learning_rate": 0.00019799567574733696,
+      "loss": 1.4077,
+      "step": 1276
+    },
+    {
+      "epoch": 0.09204930440423845,
+      "grad_norm": 0.23143690855380206,
+      "learning_rate": 0.00019799102208300403,
+      "loss": 1.3325,
+      "step": 1277
+    },
+    {
+      "epoch": 0.09212138686657537,
+      "grad_norm": 0.10669178005857018,
+      "learning_rate": 0.00019798636307728374,
+      "loss": 1.2863,
+      "step": 1278
+    },
+    {
+      "epoch": 0.09219346932891227,
+      "grad_norm": 0.10272946727561533,
+      "learning_rate": 0.00019798169873043006,
+      "loss": 1.3808,
+      "step": 1279
+    },
+    {
+      "epoch": 0.09226555179124919,
+      "grad_norm": 0.11125924332775655,
+      "learning_rate": 0.0001979770290426972,
+      "loss": 1.2982,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0923376342535861,
+      "grad_norm": 0.23707348246327983,
+      "learning_rate": 0.00019797235401433977,
+      "loss": 1.3921,
+      "step": 1281
+    },
+    {
+      "epoch": 0.09240971671592302,
+      "grad_norm": 0.11385128674139922,
+      "learning_rate": 0.00019796767364561256,
+      "loss": 1.3212,
+      "step": 1282
+    },
+    {
+      "epoch": 0.09248179917825992,
+      "grad_norm": 0.11295294761048996,
+      "learning_rate": 0.00019796298793677068,
+      "loss": 1.324,
+      "step": 1283
+    },
+    {
+      "epoch": 0.09255388164059684,
+      "grad_norm": 0.12825903670243421,
+      "learning_rate": 0.00019795829688806955,
+      "loss": 1.4038,
+      "step": 1284
+    },
+    {
+      "epoch": 0.09262596410293375,
+      "grad_norm": 0.1141580205464222,
+      "learning_rate": 0.00019795360049976486,
+      "loss": 1.4309,
+      "step": 1285
+    },
+    {
+      "epoch": 0.09269804656527067,
+      "grad_norm": 0.15075111453789652,
+      "learning_rate": 0.00019794889877211268,
+      "loss": 1.3679,
+      "step": 1286
+    },
+    {
+      "epoch": 0.09277012902760759,
+      "grad_norm": 0.10593871058678407,
+      "learning_rate": 0.00019794419170536916,
+      "loss": 1.4451,
+      "step": 1287
+    },
+    {
+      "epoch": 0.0928422114899445,
+      "grad_norm": 0.1062691456768403,
+      "learning_rate": 0.00019793947929979105,
+      "loss": 1.3412,
+      "step": 1288
+    },
+    {
+      "epoch": 0.09291429395228142,
+      "grad_norm": 0.11472266582409232,
+      "learning_rate": 0.00019793476155563507,
+      "loss": 1.303,
+      "step": 1289
+    },
+    {
+      "epoch": 0.09298637641461832,
+      "grad_norm": 0.11732551625822192,
+      "learning_rate": 0.00019793003847315844,
+      "loss": 1.3216,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09305845887695524,
+      "grad_norm": 0.104437908597434,
+      "learning_rate": 0.00019792531005261862,
+      "loss": 1.3776,
+      "step": 1291
+    },
+    {
+      "epoch": 0.09313054133929215,
+      "grad_norm": 0.12339520145333115,
+      "learning_rate": 0.00019792057629427335,
+      "loss": 1.3873,
+      "step": 1292
+    },
+    {
+      "epoch": 0.09320262380162907,
+      "grad_norm": 0.10751998906633246,
+      "learning_rate": 0.0001979158371983806,
+      "loss": 1.3448,
+      "step": 1293
+    },
+    {
+      "epoch": 0.09327470626396597,
+      "grad_norm": 0.10302409118440485,
+      "learning_rate": 0.00019791109276519878,
+      "loss": 1.3652,
+      "step": 1294
+    },
+    {
+      "epoch": 0.09334678872630289,
+      "grad_norm": 0.10592904647530967,
+      "learning_rate": 0.00019790634299498647,
+      "loss": 1.2704,
+      "step": 1295
+    },
+    {
+      "epoch": 0.0934188711886398,
+      "grad_norm": 0.10964640144028324,
+      "learning_rate": 0.00019790158788800258,
+      "loss": 1.2917,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09349095365097672,
+      "grad_norm": 0.10262057484354614,
+      "learning_rate": 0.00019789682744450628,
+      "loss": 1.2341,
+      "step": 1297
+    },
+    {
+      "epoch": 0.09356303611331362,
+      "grad_norm": 0.12117075625170862,
+      "learning_rate": 0.00019789206166475706,
+      "loss": 1.3727,
+      "step": 1298
+    },
+    {
+      "epoch": 0.09363511857565054,
+      "grad_norm": 0.13581456229516944,
+      "learning_rate": 0.00019788729054901475,
+      "loss": 1.3203,
+      "step": 1299
+    },
+    {
+      "epoch": 0.09370720103798746,
+      "grad_norm": 0.12422632651862213,
+      "learning_rate": 0.00019788251409753933,
+      "loss": 1.3368,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09377928350032437,
+      "grad_norm": 0.11147898834105953,
+      "learning_rate": 0.00019787773231059124,
+      "loss": 1.3112,
+      "step": 1301
+    },
+    {
+      "epoch": 0.09385136596266129,
+      "grad_norm": 0.3283771946378844,
+      "learning_rate": 0.00019787294518843108,
+      "loss": 1.3323,
+      "step": 1302
+    },
+    {
+      "epoch": 0.0939234484249982,
+      "grad_norm": 0.1210605167909416,
+      "learning_rate": 0.00019786815273131986,
+      "loss": 1.3524,
+      "step": 1303
+    },
+    {
+      "epoch": 0.09399553088733512,
+      "grad_norm": 0.1149564047517315,
+      "learning_rate": 0.00019786335493951872,
+      "loss": 1.3854,
+      "step": 1304
+    },
+    {
+      "epoch": 0.09406761334967202,
+      "grad_norm": 0.10530513450598326,
+      "learning_rate": 0.00019785855181328924,
+      "loss": 1.3487,
+      "step": 1305
+    },
+    {
+      "epoch": 0.09413969581200894,
+      "grad_norm": 0.12813957975108722,
+      "learning_rate": 0.00019785374335289321,
+      "loss": 1.2819,
+      "step": 1306
+    },
+    {
+      "epoch": 0.09421177827434585,
+      "grad_norm": 0.11690066947387513,
+      "learning_rate": 0.00019784892955859276,
+      "loss": 1.3758,
+      "step": 1307
+    },
+    {
+      "epoch": 0.09428386073668277,
+      "grad_norm": 0.10627681052353684,
+      "learning_rate": 0.00019784411043065022,
+      "loss": 1.218,
+      "step": 1308
+    },
+    {
+      "epoch": 0.09435594319901967,
+      "grad_norm": 0.11954925462095166,
+      "learning_rate": 0.00019783928596932838,
+      "loss": 1.4203,
+      "step": 1309
+    },
+    {
+      "epoch": 0.0944280256613566,
+      "grad_norm": 0.19382679657639196,
+      "learning_rate": 0.00019783445617489013,
+      "loss": 1.2206,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0945001081236935,
+      "grad_norm": 0.10776199547091733,
+      "learning_rate": 0.00019782962104759877,
+      "loss": 1.3289,
+      "step": 1311
+    },
+    {
+      "epoch": 0.09457219058603042,
+      "grad_norm": 0.1316246051149742,
+      "learning_rate": 0.00019782478058771784,
+      "loss": 1.3489,
+      "step": 1312
+    },
+    {
+      "epoch": 0.09464427304836734,
+      "grad_norm": 0.10417420427450783,
+      "learning_rate": 0.00019781993479551125,
+      "loss": 1.2392,
+      "step": 1313
+    },
+    {
+      "epoch": 0.09471635551070424,
+      "grad_norm": 0.12936460459070767,
+      "learning_rate": 0.00019781508367124303,
+      "loss": 1.3908,
+      "step": 1314
+    },
+    {
+      "epoch": 0.09478843797304116,
+      "grad_norm": 0.10228901369591116,
+      "learning_rate": 0.00019781022721517772,
+      "loss": 1.2232,
+      "step": 1315
+    },
+    {
+      "epoch": 0.09486052043537807,
+      "grad_norm": 0.12017082858579459,
+      "learning_rate": 0.00019780536542758,
+      "loss": 1.2657,
+      "step": 1316
+    },
+    {
+      "epoch": 0.09493260289771499,
+      "grad_norm": 0.1378085272543346,
+      "learning_rate": 0.00019780049830871483,
+      "loss": 1.3736,
+      "step": 1317
+    },
+    {
+      "epoch": 0.0950046853600519,
+      "grad_norm": 0.1152347138031826,
+      "learning_rate": 0.00019779562585884756,
+      "loss": 1.3583,
+      "step": 1318
+    },
+    {
+      "epoch": 0.09507676782238882,
+      "grad_norm": 0.11482807034592152,
+      "learning_rate": 0.00019779074807824382,
+      "loss": 1.3806,
+      "step": 1319
+    },
+    {
+      "epoch": 0.09514885028472572,
+      "grad_norm": 0.13175449603520112,
+      "learning_rate": 0.0001977858649671694,
+      "loss": 1.3114,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09522093274706264,
+      "grad_norm": 0.14011163839834276,
+      "learning_rate": 0.00019778097652589053,
+      "loss": 1.3494,
+      "step": 1321
+    },
+    {
+      "epoch": 0.09529301520939955,
+      "grad_norm": 0.125576045105913,
+      "learning_rate": 0.00019777608275467365,
+      "loss": 1.3383,
+      "step": 1322
+    },
+    {
+      "epoch": 0.09536509767173647,
+      "grad_norm": 0.12922047548700977,
+      "learning_rate": 0.00019777118365378558,
+      "loss": 1.3775,
+      "step": 1323
+    },
+    {
+      "epoch": 0.09543718013407337,
+      "grad_norm": 0.11018949016043693,
+      "learning_rate": 0.0001977662792234933,
+      "loss": 1.408,
+      "step": 1324
+    },
+    {
+      "epoch": 0.0955092625964103,
+      "grad_norm": 0.12043144722749578,
+      "learning_rate": 0.00019776136946406413,
+      "loss": 1.2634,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09558134505874721,
+      "grad_norm": 0.1335086478790146,
+      "learning_rate": 0.00019775645437576574,
+      "loss": 1.3744,
+      "step": 1326
+    },
+    {
+      "epoch": 0.09565342752108412,
+      "grad_norm": 0.14089765604541338,
+      "learning_rate": 0.00019775153395886603,
+      "loss": 1.3979,
+      "step": 1327
+    },
+    {
+      "epoch": 0.09572550998342104,
+      "grad_norm": 0.1103281225474627,
+      "learning_rate": 0.0001977466082136332,
+      "loss": 1.3379,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09579759244575795,
+      "grad_norm": 0.11612782398103146,
+      "learning_rate": 0.0001977416771403358,
+      "loss": 1.3729,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09586967490809487,
+      "grad_norm": 0.11703881035183099,
+      "learning_rate": 0.00019773674073924252,
+      "loss": 1.3967,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09594175737043177,
+      "grad_norm": 0.11035077814687895,
+      "learning_rate": 0.00019773179901062254,
+      "loss": 1.3857,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09601383983276869,
+      "grad_norm": 0.1100532923888877,
+      "learning_rate": 0.00019772685195474515,
+      "loss": 1.3223,
+      "step": 1332
+    },
+    {
+      "epoch": 0.0960859222951056,
+      "grad_norm": 0.09517193150086126,
+      "learning_rate": 0.00019772189957188005,
+      "loss": 1.3447,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09615800475744252,
+      "grad_norm": 0.12085287596669123,
+      "learning_rate": 0.00019771694186229715,
+      "loss": 1.2424,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09623008721977942,
+      "grad_norm": 0.10150286299385046,
+      "learning_rate": 0.00019771197882626678,
+      "loss": 1.3005,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09630216968211634,
+      "grad_norm": 0.22778359163435197,
+      "learning_rate": 0.00019770701046405934,
+      "loss": 1.3556,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09637425214445325,
+      "grad_norm": 0.11449938202602085,
+      "learning_rate": 0.00019770203677594577,
+      "loss": 1.3429,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09644633460679017,
+      "grad_norm": 0.1121571350394368,
+      "learning_rate": 0.00019769705776219709,
+      "loss": 1.4487,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09651841706912709,
+      "grad_norm": 0.15745292216554033,
+      "learning_rate": 0.00019769207342308477,
+      "loss": 1.3104,
+      "step": 1339
+    },
+    {
+      "epoch": 0.096590499531464,
+      "grad_norm": 0.304663142469487,
+      "learning_rate": 0.00019768708375888047,
+      "loss": 1.3453,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09666258199380091,
+      "grad_norm": 0.11252591827093639,
+      "learning_rate": 0.00019768208876985615,
+      "loss": 1.2547,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09673466445613782,
+      "grad_norm": 0.11555553457723326,
+      "learning_rate": 0.0001976770884562841,
+      "loss": 1.3141,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09680674691847474,
+      "grad_norm": 0.12395050295839263,
+      "learning_rate": 0.0001976720828184369,
+      "loss": 1.3938,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09687882938081165,
+      "grad_norm": 0.10353868317622758,
+      "learning_rate": 0.00019766707185658736,
+      "loss": 1.2889,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09695091184314857,
+      "grad_norm": 0.15968412559799433,
+      "learning_rate": 0.00019766205557100868,
+      "loss": 1.3378,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09702299430548547,
+      "grad_norm": 0.11563697530810951,
+      "learning_rate": 0.00019765703396197425,
+      "loss": 1.3574,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09709507676782239,
+      "grad_norm": 0.11666457144850767,
+      "learning_rate": 0.00019765200702975778,
+      "loss": 1.3926,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0971671592301593,
+      "grad_norm": 0.2202715351584791,
+      "learning_rate": 0.00019764697477463338,
+      "loss": 1.3617,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09723924169249622,
+      "grad_norm": 0.11629888666902971,
+      "learning_rate": 0.00019764193719687523,
+      "loss": 1.3375,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09731132415483312,
+      "grad_norm": 0.1248042125618061,
+      "learning_rate": 0.00019763689429675796,
+      "loss": 1.1911,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09738340661717004,
+      "grad_norm": 0.10594396163975853,
+      "learning_rate": 0.0001976318460745565,
+      "loss": 1.305,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09745548907950696,
+      "grad_norm": 0.10610553544713155,
+      "learning_rate": 0.00019762679253054595,
+      "loss": 1.4362,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09752757154184387,
+      "grad_norm": 0.11293606388508283,
+      "learning_rate": 0.00019762173366500184,
+      "loss": 1.3494,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09759965400418079,
+      "grad_norm": 0.09988355819718517,
+      "learning_rate": 0.0001976166694781999,
+      "loss": 1.2849,
+      "step": 1354
+    },
+    {
+      "epoch": 0.0976717364665177,
+      "grad_norm": 0.09757205003327266,
+      "learning_rate": 0.00019761159997041617,
+      "loss": 1.3258,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09774381892885461,
+      "grad_norm": 0.1193434709374469,
+      "learning_rate": 0.00019760652514192698,
+      "loss": 1.3179,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09781590139119152,
+      "grad_norm": 0.11110452988249932,
+      "learning_rate": 0.00019760144499300893,
+      "loss": 1.2774,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09788798385352844,
+      "grad_norm": 0.10751199138333833,
+      "learning_rate": 0.000197596359523939,
+      "loss": 1.2961,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09796006631586535,
+      "grad_norm": 0.10167385933799637,
+      "learning_rate": 0.00019759126873499433,
+      "loss": 1.3474,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09803214877820227,
+      "grad_norm": 0.10773935545925348,
+      "learning_rate": 0.00019758617262645241,
+      "loss": 1.4217,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09810423124053917,
+      "grad_norm": 0.13207978186086905,
+      "learning_rate": 0.0001975810711985911,
+      "loss": 1.3204,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09817631370287609,
+      "grad_norm": 0.08667309042124798,
+      "learning_rate": 0.00019757596445168843,
+      "loss": 1.3277,
+      "step": 1362
+    },
+    {
+      "epoch": 0.098248396165213,
+      "grad_norm": 0.10704724069696644,
+      "learning_rate": 0.00019757085238602273,
+      "loss": 1.2855,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09832047862754992,
+      "grad_norm": 0.10495652373299548,
+      "learning_rate": 0.0001975657350018727,
+      "loss": 1.3582,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09839256108988684,
+      "grad_norm": 0.10008999440793985,
+      "learning_rate": 0.00019756061229951724,
+      "loss": 1.4074,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09846464355222374,
+      "grad_norm": 0.11773724057254202,
+      "learning_rate": 0.00019755548427923564,
+      "loss": 1.3366,
+      "step": 1366
+    },
+    {
+      "epoch": 0.09853672601456066,
+      "grad_norm": 0.10521833098155361,
+      "learning_rate": 0.00019755035094130735,
+      "loss": 1.2962,
+      "step": 1367
+    },
+    {
+      "epoch": 0.09860880847689757,
+      "grad_norm": 0.1584019061963044,
+      "learning_rate": 0.00019754521228601223,
+      "loss": 1.4916,
+      "step": 1368
+    },
+    {
+      "epoch": 0.09868089093923449,
+      "grad_norm": 0.09653924504147805,
+      "learning_rate": 0.00019754006831363042,
+      "loss": 1.3173,
+      "step": 1369
+    },
+    {
+      "epoch": 0.0987529734015714,
+      "grad_norm": 0.10333396986533515,
+      "learning_rate": 0.0001975349190244422,
+      "loss": 1.4231,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09882505586390831,
+      "grad_norm": 0.1331064487049757,
+      "learning_rate": 0.0001975297644187284,
+      "loss": 1.4117,
+      "step": 1371
+    },
+    {
+      "epoch": 0.09889713832624522,
+      "grad_norm": 0.11309081903261237,
+      "learning_rate": 0.00019752460449676983,
+      "loss": 1.2945,
+      "step": 1372
+    },
+    {
+      "epoch": 0.09896922078858214,
+      "grad_norm": 0.10617110743238095,
+      "learning_rate": 0.0001975194392588479,
+      "loss": 1.3429,
+      "step": 1373
+    },
+    {
+      "epoch": 0.09904130325091905,
+      "grad_norm": 0.10209950843991855,
+      "learning_rate": 0.00019751426870524407,
+      "loss": 1.3572,
+      "step": 1374
+    },
+    {
+      "epoch": 0.09911338571325597,
+      "grad_norm": 0.10529973078341257,
+      "learning_rate": 0.0001975090928362402,
+      "loss": 1.4182,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09918546817559287,
+      "grad_norm": 0.09744027437206969,
+      "learning_rate": 0.00019750391165211846,
+      "loss": 1.2344,
+      "step": 1376
+    },
+    {
+      "epoch": 0.09925755063792979,
+      "grad_norm": 0.08682648464338814,
+      "learning_rate": 0.0001974987251531612,
+      "loss": 1.2814,
+      "step": 1377
+    },
+    {
+      "epoch": 0.0993296331002667,
+      "grad_norm": 0.0957721043205676,
+      "learning_rate": 0.00019749353333965123,
+      "loss": 1.3991,
+      "step": 1378
+    },
+    {
+      "epoch": 0.09940171556260362,
+      "grad_norm": 0.11710332664764127,
+      "learning_rate": 0.00019748833621187144,
+      "loss": 1.2621,
+      "step": 1379
+    },
+    {
+      "epoch": 0.09947379802494054,
+      "grad_norm": 0.11250509727277969,
+      "learning_rate": 0.00019748313377010516,
+      "loss": 1.3132,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09954588048727744,
+      "grad_norm": 0.13384996761480347,
+      "learning_rate": 0.000197477926014636,
+      "loss": 1.274,
+      "step": 1381
+    },
+    {
+      "epoch": 0.09961796294961436,
+      "grad_norm": 0.11931762942130009,
+      "learning_rate": 0.00019747271294574782,
+      "loss": 1.3539,
+      "step": 1382
+    },
+    {
+      "epoch": 0.09969004541195127,
+      "grad_norm": 0.11313374998719887,
+      "learning_rate": 0.00019746749456372474,
+      "loss": 1.2403,
+      "step": 1383
+    },
+    {
+      "epoch": 0.09976212787428819,
+      "grad_norm": 0.11539945119842036,
+      "learning_rate": 0.0001974622708688513,
+      "loss": 1.4778,
+      "step": 1384
+    },
+    {
+      "epoch": 0.0998342103366251,
+      "grad_norm": 0.21493248581035213,
+      "learning_rate": 0.00019745704186141213,
+      "loss": 1.3937,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09990629279896202,
+      "grad_norm": 0.09612295738746636,
+      "learning_rate": 0.00019745180754169228,
+      "loss": 1.2624,
+      "step": 1386
+    },
+    {
+      "epoch": 0.09997837526129892,
+      "grad_norm": 0.12085576628951743,
+      "learning_rate": 0.00019744656790997712,
+      "loss": 1.3339,
+      "step": 1387
+    },
+    {
+      "epoch": 0.10005045772363584,
+      "grad_norm": 0.10578262899407223,
+      "learning_rate": 0.00019744132296655223,
+      "loss": 1.3396,
+      "step": 1388
+    },
+    {
+      "epoch": 0.10012254018597275,
+      "grad_norm": 0.1081317929385026,
+      "learning_rate": 0.0001974360727117035,
+      "loss": 1.432,
+      "step": 1389
+    },
+    {
+      "epoch": 0.10019462264830967,
+      "grad_norm": 0.10151330215098837,
+      "learning_rate": 0.00019743081714571715,
+      "loss": 1.4073,
+      "step": 1390
+    },
+    {
+      "epoch": 0.10026670511064657,
+      "grad_norm": 0.10928525470514835,
+      "learning_rate": 0.00019742555626887955,
+      "loss": 1.4267,
+      "step": 1391
+    },
+    {
+      "epoch": 0.10033878757298349,
+      "grad_norm": 0.1349050530100837,
+      "learning_rate": 0.0001974202900814776,
+      "loss": 1.478,
+      "step": 1392
+    },
+    {
+      "epoch": 0.10041087003532041,
+      "grad_norm": 0.10547290198349046,
+      "learning_rate": 0.00019741501858379828,
+      "loss": 1.3679,
+      "step": 1393
+    },
+    {
+      "epoch": 0.10048295249765732,
+      "grad_norm": 0.18202669345734065,
+      "learning_rate": 0.00019740974177612892,
+      "loss": 1.3959,
+      "step": 1394
+    },
+    {
+      "epoch": 0.10055503495999424,
+      "grad_norm": 0.10841238799338217,
+      "learning_rate": 0.0001974044596587572,
+      "loss": 1.3355,
+      "step": 1395
+    },
+    {
+      "epoch": 0.10062711742233114,
+      "grad_norm": 0.10657755845890768,
+      "learning_rate": 0.00019739917223197105,
+      "loss": 1.4726,
+      "step": 1396
+    },
+    {
+      "epoch": 0.10069919988466806,
+      "grad_norm": 0.13810385871017017,
+      "learning_rate": 0.00019739387949605864,
+      "loss": 1.2608,
+      "step": 1397
+    },
+    {
+      "epoch": 0.10077128234700497,
+      "grad_norm": 0.1700792751846456,
+      "learning_rate": 0.00019738858145130846,
+      "loss": 1.3411,
+      "step": 1398
+    },
+    {
+      "epoch": 0.10084336480934189,
+      "grad_norm": 0.11399236145176064,
+      "learning_rate": 0.00019738327809800933,
+      "loss": 1.4187,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1009154472716788,
+      "grad_norm": 0.09976342049180374,
+      "learning_rate": 0.00019737796943645038,
+      "loss": 1.3376,
+      "step": 1400
+    },
+    {
+      "epoch": 0.10098752973401572,
+      "grad_norm": 0.2579503347417765,
+      "learning_rate": 0.0001973726554669209,
+      "loss": 1.3487,
+      "step": 1401
+    },
+    {
+      "epoch": 0.10105961219635262,
+      "grad_norm": 0.11778888388084852,
+      "learning_rate": 0.00019736733618971056,
+      "loss": 1.3931,
+      "step": 1402
+    },
+    {
+      "epoch": 0.10113169465868954,
+      "grad_norm": 0.14578393608362342,
+      "learning_rate": 0.00019736201160510931,
+      "loss": 1.3409,
+      "step": 1403
+    },
+    {
+      "epoch": 0.10120377712102645,
+      "grad_norm": 0.11091747599344069,
+      "learning_rate": 0.00019735668171340742,
+      "loss": 1.3521,
+      "step": 1404
+    },
+    {
+      "epoch": 0.10127585958336337,
+      "grad_norm": 0.10797044506275666,
+      "learning_rate": 0.0001973513465148954,
+      "loss": 1.3807,
+      "step": 1405
+    },
+    {
+      "epoch": 0.10134794204570029,
+      "grad_norm": 0.12034165018439036,
+      "learning_rate": 0.0001973460060098641,
+      "loss": 1.3326,
+      "step": 1406
+    },
+    {
+      "epoch": 0.1014200245080372,
+      "grad_norm": 0.12987311869511772,
+      "learning_rate": 0.00019734066019860453,
+      "loss": 1.3534,
+      "step": 1407
+    },
+    {
+      "epoch": 0.10149210697037411,
+      "grad_norm": 0.1218471792811035,
+      "learning_rate": 0.0001973353090814082,
+      "loss": 1.3392,
+      "step": 1408
+    },
+    {
+      "epoch": 0.10156418943271102,
+      "grad_norm": 0.10020737939131583,
+      "learning_rate": 0.00019732995265856672,
+      "loss": 1.3005,
+      "step": 1409
+    },
+    {
+      "epoch": 0.10163627189504794,
+      "grad_norm": 0.12088343550677519,
+      "learning_rate": 0.00019732459093037204,
+      "loss": 1.3259,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10170835435738484,
+      "grad_norm": 0.12403348983980976,
+      "learning_rate": 0.00019731922389711653,
+      "loss": 1.2697,
+      "step": 1411
+    },
+    {
+      "epoch": 0.10178043681972176,
+      "grad_norm": 0.2007330328042924,
+      "learning_rate": 0.00019731385155909262,
+      "loss": 1.3366,
+      "step": 1412
+    },
+    {
+      "epoch": 0.10185251928205867,
+      "grad_norm": 0.12008782622837151,
+      "learning_rate": 0.0001973084739165932,
+      "loss": 1.3851,
+      "step": 1413
+    },
+    {
+      "epoch": 0.10192460174439559,
+      "grad_norm": 0.17632440083219098,
+      "learning_rate": 0.00019730309096991147,
+      "loss": 1.3175,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1019966842067325,
+      "grad_norm": 0.11879025942895176,
+      "learning_rate": 0.00019729770271934073,
+      "loss": 1.425,
+      "step": 1415
+    },
+    {
+      "epoch": 0.10206876666906942,
+      "grad_norm": 0.11137450772007146,
+      "learning_rate": 0.00019729230916517478,
+      "loss": 1.4044,
+      "step": 1416
+    },
+    {
+      "epoch": 0.10214084913140632,
+      "grad_norm": 0.09562821939850573,
+      "learning_rate": 0.00019728691030770757,
+      "loss": 1.3232,
+      "step": 1417
+    },
+    {
+      "epoch": 0.10221293159374324,
+      "grad_norm": 0.10734947590001427,
+      "learning_rate": 0.00019728150614723336,
+      "loss": 1.3303,
+      "step": 1418
+    },
+    {
+      "epoch": 0.10228501405608016,
+      "grad_norm": 0.1721322116928204,
+      "learning_rate": 0.00019727609668404678,
+      "loss": 1.4241,
+      "step": 1419
+    },
+    {
+      "epoch": 0.10235709651841707,
+      "grad_norm": 0.1082859341840484,
+      "learning_rate": 0.00019727068191844268,
+      "loss": 1.3937,
+      "step": 1420
+    },
+    {
+      "epoch": 0.10242917898075399,
+      "grad_norm": 0.11028365761527165,
+      "learning_rate": 0.0001972652618507162,
+      "loss": 1.3132,
+      "step": 1421
+    },
+    {
+      "epoch": 0.1025012614430909,
+      "grad_norm": 0.09188925301858525,
+      "learning_rate": 0.00019725983648116277,
+      "loss": 1.4428,
+      "step": 1422
+    },
+    {
+      "epoch": 0.10257334390542781,
+      "grad_norm": 0.11276756338949974,
+      "learning_rate": 0.00019725440581007817,
+      "loss": 1.3956,
+      "step": 1423
+    },
+    {
+      "epoch": 0.10264542636776472,
+      "grad_norm": 0.12569314433289594,
+      "learning_rate": 0.00019724896983775838,
+      "loss": 1.3166,
+      "step": 1424
+    },
+    {
+      "epoch": 0.10271750883010164,
+      "grad_norm": 0.11343364858043092,
+      "learning_rate": 0.00019724352856449975,
+      "loss": 1.331,
+      "step": 1425
+    },
+    {
+      "epoch": 0.10278959129243855,
+      "grad_norm": 0.10195243783228647,
+      "learning_rate": 0.00019723808199059882,
+      "loss": 1.2622,
+      "step": 1426
+    },
+    {
+      "epoch": 0.10286167375477547,
+      "grad_norm": 0.10967672848463475,
+      "learning_rate": 0.0001972326301163525,
+      "loss": 1.3314,
+      "step": 1427
+    },
+    {
+      "epoch": 0.10293375621711237,
+      "grad_norm": 0.09152949955863887,
+      "learning_rate": 0.00019722717294205796,
+      "loss": 1.3483,
+      "step": 1428
+    },
+    {
+      "epoch": 0.10300583867944929,
+      "grad_norm": 0.10247991714274929,
+      "learning_rate": 0.0001972217104680127,
+      "loss": 1.2981,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1030779211417862,
+      "grad_norm": 0.10899032743860972,
+      "learning_rate": 0.00019721624269451445,
+      "loss": 1.4394,
+      "step": 1430
+    },
+    {
+      "epoch": 0.10315000360412312,
+      "grad_norm": 0.11494587224240704,
+      "learning_rate": 0.00019721076962186124,
+      "loss": 1.3681,
+      "step": 1431
+    },
+    {
+      "epoch": 0.10322208606646004,
+      "grad_norm": 0.101600859738019,
+      "learning_rate": 0.0001972052912503514,
+      "loss": 1.3055,
+      "step": 1432
+    },
+    {
+      "epoch": 0.10329416852879694,
+      "grad_norm": 0.20798284009269008,
+      "learning_rate": 0.00019719980758028358,
+      "loss": 1.234,
+      "step": 1433
+    },
+    {
+      "epoch": 0.10336625099113386,
+      "grad_norm": 0.12320331736628648,
+      "learning_rate": 0.00019719431861195668,
+      "loss": 1.288,
+      "step": 1434
+    },
+    {
+      "epoch": 0.10343833345347077,
+      "grad_norm": 0.1017141380887575,
+      "learning_rate": 0.00019718882434566987,
+      "loss": 1.396,
+      "step": 1435
+    },
+    {
+      "epoch": 0.10351041591580769,
+      "grad_norm": 0.11088676780975912,
+      "learning_rate": 0.00019718332478172267,
+      "loss": 1.3749,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1035824983781446,
+      "grad_norm": 0.09773871996746081,
+      "learning_rate": 0.00019717781992041483,
+      "loss": 1.3691,
+      "step": 1437
+    },
+    {
+      "epoch": 0.10365458084048151,
+      "grad_norm": 0.11175231039881411,
+      "learning_rate": 0.00019717230976204646,
+      "loss": 1.253,
+      "step": 1438
+    },
+    {
+      "epoch": 0.10372666330281842,
+      "grad_norm": 0.16839134069545234,
+      "learning_rate": 0.00019716679430691785,
+      "loss": 1.3594,
+      "step": 1439
+    },
+    {
+      "epoch": 0.10379874576515534,
+      "grad_norm": 0.10707373037652895,
+      "learning_rate": 0.0001971612735553297,
+      "loss": 1.4028,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10387082822749225,
+      "grad_norm": 0.09865546800818711,
+      "learning_rate": 0.00019715574750758287,
+      "loss": 1.2948,
+      "step": 1441
+    },
+    {
+      "epoch": 0.10394291068982917,
+      "grad_norm": 0.11429064502729419,
+      "learning_rate": 0.00019715021616397864,
+      "loss": 1.317,
+      "step": 1442
+    },
+    {
+      "epoch": 0.10401499315216607,
+      "grad_norm": 0.12587099007866956,
+      "learning_rate": 0.0001971446795248185,
+      "loss": 1.3526,
+      "step": 1443
+    },
+    {
+      "epoch": 0.10408707561450299,
+      "grad_norm": 0.10438597976497162,
+      "learning_rate": 0.00019713913759040422,
+      "loss": 1.3852,
+      "step": 1444
+    },
+    {
+      "epoch": 0.10415915807683991,
+      "grad_norm": 0.10315477771284136,
+      "learning_rate": 0.00019713359036103797,
+      "loss": 1.2633,
+      "step": 1445
+    },
+    {
+      "epoch": 0.10423124053917682,
+      "grad_norm": 0.11118746160186517,
+      "learning_rate": 0.00019712803783702204,
+      "loss": 1.3031,
+      "step": 1446
+    },
+    {
+      "epoch": 0.10430332300151374,
+      "grad_norm": 0.09941026990203491,
+      "learning_rate": 0.00019712248001865909,
+      "loss": 1.4552,
+      "step": 1447
+    },
+    {
+      "epoch": 0.10437540546385064,
+      "grad_norm": 0.12022438295518564,
+      "learning_rate": 0.00019711691690625212,
+      "loss": 1.3982,
+      "step": 1448
+    },
+    {
+      "epoch": 0.10444748792618756,
+      "grad_norm": 0.10201157312756785,
+      "learning_rate": 0.00019711134850010432,
+      "loss": 1.3244,
+      "step": 1449
+    },
+    {
+      "epoch": 0.10451957038852447,
+      "grad_norm": 0.10347348263625757,
+      "learning_rate": 0.00019710577480051925,
+      "loss": 1.3525,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10459165285086139,
+      "grad_norm": 0.3523417933551231,
+      "learning_rate": 0.00019710019580780074,
+      "loss": 1.345,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1046637353131983,
+      "grad_norm": 0.11192263726455774,
+      "learning_rate": 0.00019709461152225285,
+      "loss": 1.3103,
+      "step": 1452
+    },
+    {
+      "epoch": 0.10473581777553521,
+      "grad_norm": 0.11309001968269335,
+      "learning_rate": 0.00019708902194418002,
+      "loss": 1.3705,
+      "step": 1453
+    },
+    {
+      "epoch": 0.10480790023787212,
+      "grad_norm": 0.1089903910327151,
+      "learning_rate": 0.0001970834270738869,
+      "loss": 1.4561,
+      "step": 1454
+    },
+    {
+      "epoch": 0.10487998270020904,
+      "grad_norm": 0.09798948944748337,
+      "learning_rate": 0.0001970778269116785,
+      "loss": 1.3536,
+      "step": 1455
+    },
+    {
+      "epoch": 0.10495206516254595,
+      "grad_norm": 0.10958950082945487,
+      "learning_rate": 0.00019707222145786,
+      "loss": 1.448,
+      "step": 1456
+    },
+    {
+      "epoch": 0.10502414762488287,
+      "grad_norm": 0.1164884547273835,
+      "learning_rate": 0.00019706661071273703,
+      "loss": 1.362,
+      "step": 1457
+    },
+    {
+      "epoch": 0.10509623008721979,
+      "grad_norm": 0.12083877078899723,
+      "learning_rate": 0.00019706099467661542,
+      "loss": 1.3252,
+      "step": 1458
+    },
+    {
+      "epoch": 0.10516831254955669,
+      "grad_norm": 0.11256072028970605,
+      "learning_rate": 0.0001970553733498012,
+      "loss": 1.3563,
+      "step": 1459
+    },
+    {
+      "epoch": 0.10524039501189361,
+      "grad_norm": 0.16203524774885533,
+      "learning_rate": 0.00019704974673260091,
+      "loss": 1.4425,
+      "step": 1460
+    },
+    {
+      "epoch": 0.10531247747423052,
+      "grad_norm": 0.09702490942120323,
+      "learning_rate": 0.00019704411482532116,
+      "loss": 1.3511,
+      "step": 1461
+    },
+    {
+      "epoch": 0.10538455993656744,
+      "grad_norm": 0.10455097512344974,
+      "learning_rate": 0.00019703847762826896,
+      "loss": 1.341,
+      "step": 1462
+    },
+    {
+      "epoch": 0.10545664239890434,
+      "grad_norm": 0.1788864384843197,
+      "learning_rate": 0.00019703283514175163,
+      "loss": 1.3595,
+      "step": 1463
+    },
+    {
+      "epoch": 0.10552872486124126,
+      "grad_norm": 0.21349658572921254,
+      "learning_rate": 0.00019702718736607668,
+      "loss": 1.3735,
+      "step": 1464
+    },
+    {
+      "epoch": 0.10560080732357817,
+      "grad_norm": 0.12364847172841309,
+      "learning_rate": 0.000197021534301552,
+      "loss": 1.4167,
+      "step": 1465
+    },
+    {
+      "epoch": 0.10567288978591509,
+      "grad_norm": 0.11491987138461296,
+      "learning_rate": 0.0001970158759484857,
+      "loss": 1.2908,
+      "step": 1466
+    },
+    {
+      "epoch": 0.105744972248252,
+      "grad_norm": 0.1202175310101512,
+      "learning_rate": 0.00019701021230718628,
+      "loss": 1.3535,
+      "step": 1467
+    },
+    {
+      "epoch": 0.10581705471058891,
+      "grad_norm": 0.41293573139971257,
+      "learning_rate": 0.00019700454337796236,
+      "loss": 1.422,
+      "step": 1468
+    },
+    {
+      "epoch": 0.10588913717292582,
+      "grad_norm": 0.09872375905238646,
+      "learning_rate": 0.00019699886916112305,
+      "loss": 1.2954,
+      "step": 1469
+    },
+    {
+      "epoch": 0.10596121963526274,
+      "grad_norm": 0.14698537775109888,
+      "learning_rate": 0.00019699318965697756,
+      "loss": 1.4333,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10603330209759966,
+      "grad_norm": 0.11761994770700397,
+      "learning_rate": 0.0001969875048658355,
+      "loss": 1.3594,
+      "step": 1471
+    },
+    {
+      "epoch": 0.10610538455993657,
+      "grad_norm": 0.12420781432038772,
+      "learning_rate": 0.00019698181478800675,
+      "loss": 1.3689,
+      "step": 1472
+    },
+    {
+      "epoch": 0.10617746702227349,
+      "grad_norm": 0.12305775287661418,
+      "learning_rate": 0.0001969761194238015,
+      "loss": 1.432,
+      "step": 1473
+    },
+    {
+      "epoch": 0.10624954948461039,
+      "grad_norm": 0.1355569930050225,
+      "learning_rate": 0.00019697041877353013,
+      "loss": 1.3179,
+      "step": 1474
+    },
+    {
+      "epoch": 0.10632163194694731,
+      "grad_norm": 0.1547117334398643,
+      "learning_rate": 0.00019696471283750343,
+      "loss": 1.37,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10639371440928422,
+      "grad_norm": 0.11890573547387832,
+      "learning_rate": 0.00019695900161603244,
+      "loss": 1.284,
+      "step": 1476
+    },
+    {
+      "epoch": 0.10646579687162114,
+      "grad_norm": 0.11401958300049787,
+      "learning_rate": 0.0001969532851094284,
+      "loss": 1.4117,
+      "step": 1477
+    },
+    {
+      "epoch": 0.10653787933395804,
+      "grad_norm": 0.09875472854463702,
+      "learning_rate": 0.000196947563318003,
+      "loss": 1.3942,
+      "step": 1478
+    },
+    {
+      "epoch": 0.10660996179629496,
+      "grad_norm": 0.10860595278413696,
+      "learning_rate": 0.00019694183624206803,
+      "loss": 1.4121,
+      "step": 1479
+    },
+    {
+      "epoch": 0.10668204425863187,
+      "grad_norm": 0.10112114081896642,
+      "learning_rate": 0.00019693610388193573,
+      "loss": 1.3499,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10675412672096879,
+      "grad_norm": 0.10338271103075636,
+      "learning_rate": 0.00019693036623791857,
+      "loss": 1.3332,
+      "step": 1481
+    },
+    {
+      "epoch": 0.1068262091833057,
+      "grad_norm": 0.15491406095756677,
+      "learning_rate": 0.00019692462331032928,
+      "loss": 1.3327,
+      "step": 1482
+    },
+    {
+      "epoch": 0.10689829164564262,
+      "grad_norm": 0.23205942395261722,
+      "learning_rate": 0.00019691887509948092,
+      "loss": 1.2627,
+      "step": 1483
+    },
+    {
+      "epoch": 0.10697037410797954,
+      "grad_norm": 0.1257312394272933,
+      "learning_rate": 0.0001969131216056868,
+      "loss": 1.3435,
+      "step": 1484
+    },
+    {
+      "epoch": 0.10704245657031644,
+      "grad_norm": 0.10211452323924508,
+      "learning_rate": 0.0001969073628292605,
+      "loss": 1.3467,
+      "step": 1485
+    },
+    {
+      "epoch": 0.10711453903265336,
+      "grad_norm": 0.0915541482915639,
+      "learning_rate": 0.00019690159877051603,
+      "loss": 1.3067,
+      "step": 1486
+    },
+    {
+      "epoch": 0.10718662149499027,
+      "grad_norm": 0.10794568529801646,
+      "learning_rate": 0.00019689582942976748,
+      "loss": 1.3747,
+      "step": 1487
+    },
+    {
+      "epoch": 0.10725870395732719,
+      "grad_norm": 0.10215445670285554,
+      "learning_rate": 0.00019689005480732938,
+      "loss": 1.3886,
+      "step": 1488
+    },
+    {
+      "epoch": 0.10733078641966409,
+      "grad_norm": 0.10763580338046388,
+      "learning_rate": 0.00019688427490351652,
+      "loss": 1.3199,
+      "step": 1489
+    },
+    {
+      "epoch": 0.10740286888200101,
+      "grad_norm": 0.1071315092084587,
+      "learning_rate": 0.00019687848971864389,
+      "loss": 1.2679,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10747495134433792,
+      "grad_norm": 0.13070384931677217,
+      "learning_rate": 0.00019687269925302685,
+      "loss": 1.4114,
+      "step": 1491
+    },
+    {
+      "epoch": 0.10754703380667484,
+      "grad_norm": 0.15632824378002333,
+      "learning_rate": 0.00019686690350698109,
+      "loss": 1.3324,
+      "step": 1492
+    },
+    {
+      "epoch": 0.10761911626901174,
+      "grad_norm": 0.12208464680073998,
+      "learning_rate": 0.0001968611024808225,
+      "loss": 1.2427,
+      "step": 1493
+    },
+    {
+      "epoch": 0.10769119873134866,
+      "grad_norm": 0.13511752642129582,
+      "learning_rate": 0.00019685529617486728,
+      "loss": 1.3134,
+      "step": 1494
+    },
+    {
+      "epoch": 0.10776328119368557,
+      "grad_norm": 0.1036743876270808,
+      "learning_rate": 0.0001968494845894319,
+      "loss": 1.3459,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10783536365602249,
+      "grad_norm": 0.11271915318976201,
+      "learning_rate": 0.00019684366772483317,
+      "loss": 1.3092,
+      "step": 1496
+    },
+    {
+      "epoch": 0.10790744611835941,
+      "grad_norm": 0.11014037652413322,
+      "learning_rate": 0.0001968378455813882,
+      "loss": 1.3222,
+      "step": 1497
+    },
+    {
+      "epoch": 0.10797952858069632,
+      "grad_norm": 0.10092063965320164,
+      "learning_rate": 0.00019683201815941427,
+      "loss": 1.3209,
+      "step": 1498
+    },
+    {
+      "epoch": 0.10805161104303324,
+      "grad_norm": 0.11644519657495422,
+      "learning_rate": 0.00019682618545922907,
+      "loss": 1.2702,
+      "step": 1499
+    },
+    {
+      "epoch": 0.10812369350537014,
+      "grad_norm": 0.10311008738961108,
+      "learning_rate": 0.00019682034748115055,
+      "loss": 1.3439,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10819577596770706,
+      "grad_norm": 0.10050210205325669,
+      "learning_rate": 0.0001968145042254969,
+      "loss": 1.3119,
+      "step": 1501
+    },
+    {
+      "epoch": 0.10826785843004397,
+      "grad_norm": 0.11063414317356322,
+      "learning_rate": 0.00019680865569258663,
+      "loss": 1.3558,
+      "step": 1502
+    },
+    {
+      "epoch": 0.10833994089238089,
+      "grad_norm": 0.10547965128582462,
+      "learning_rate": 0.00019680280188273859,
+      "loss": 1.331,
+      "step": 1503
+    },
+    {
+      "epoch": 0.10841202335471779,
+      "grad_norm": 0.14367915967105216,
+      "learning_rate": 0.00019679694279627177,
+      "loss": 1.2948,
+      "step": 1504
+    },
+    {
+      "epoch": 0.10848410581705471,
+      "grad_norm": 0.10875371224205924,
+      "learning_rate": 0.00019679107843350563,
+      "loss": 1.3566,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10855618827939162,
+      "grad_norm": 0.17489996691702947,
+      "learning_rate": 0.0001967852087947598,
+      "loss": 1.3862,
+      "step": 1506
+    },
+    {
+      "epoch": 0.10862827074172854,
+      "grad_norm": 0.12221506522826761,
+      "learning_rate": 0.0001967793338803542,
+      "loss": 1.386,
+      "step": 1507
+    },
+    {
+      "epoch": 0.10870035320406544,
+      "grad_norm": 0.15185857924503882,
+      "learning_rate": 0.00019677345369060914,
+      "loss": 1.3747,
+      "step": 1508
+    },
+    {
+      "epoch": 0.10877243566640236,
+      "grad_norm": 0.10670889602070265,
+      "learning_rate": 0.00019676756822584505,
+      "loss": 1.3063,
+      "step": 1509
+    },
+    {
+      "epoch": 0.10884451812873928,
+      "grad_norm": 0.10824798259643231,
+      "learning_rate": 0.0001967616774863828,
+      "loss": 1.3175,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10891660059107619,
+      "grad_norm": 0.09917269234410563,
+      "learning_rate": 0.00019675578147254348,
+      "loss": 1.2701,
+      "step": 1511
+    },
+    {
+      "epoch": 0.10898868305341311,
+      "grad_norm": 0.18988380427373217,
+      "learning_rate": 0.00019674988018464847,
+      "loss": 1.3362,
+      "step": 1512
+    },
+    {
+      "epoch": 0.10906076551575002,
+      "grad_norm": 0.117007239697092,
+      "learning_rate": 0.00019674397362301943,
+      "loss": 1.3571,
+      "step": 1513
+    },
+    {
+      "epoch": 0.10913284797808694,
+      "grad_norm": 0.09644526275103973,
+      "learning_rate": 0.00019673806178797834,
+      "loss": 1.2581,
+      "step": 1514
+    },
+    {
+      "epoch": 0.10920493044042384,
+      "grad_norm": 0.10925592651922791,
+      "learning_rate": 0.00019673214467984742,
+      "loss": 1.3677,
+      "step": 1515
+    },
+    {
+      "epoch": 0.10927701290276076,
+      "grad_norm": 0.11271909106686503,
+      "learning_rate": 0.00019672622229894923,
+      "loss": 1.3207,
+      "step": 1516
+    },
+    {
+      "epoch": 0.10934909536509767,
+      "grad_norm": 0.1485898204282482,
+      "learning_rate": 0.0001967202946456066,
+      "loss": 1.3741,
+      "step": 1517
+    },
+    {
+      "epoch": 0.10942117782743459,
+      "grad_norm": 0.1178261101981744,
+      "learning_rate": 0.00019671436172014262,
+      "loss": 1.3378,
+      "step": 1518
+    },
+    {
+      "epoch": 0.1094932602897715,
+      "grad_norm": 0.11162110502877574,
+      "learning_rate": 0.0001967084235228807,
+      "loss": 1.3736,
+      "step": 1519
+    },
+    {
+      "epoch": 0.10956534275210841,
+      "grad_norm": 0.1569221010009317,
+      "learning_rate": 0.0001967024800541445,
+      "loss": 1.3012,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10963742521444532,
+      "grad_norm": 0.10971450378554536,
+      "learning_rate": 0.00019669653131425803,
+      "loss": 1.3542,
+      "step": 1521
+    },
+    {
+      "epoch": 0.10970950767678224,
+      "grad_norm": 0.11221565702475454,
+      "learning_rate": 0.00019669057730354553,
+      "loss": 1.2961,
+      "step": 1522
+    },
+    {
+      "epoch": 0.10978159013911916,
+      "grad_norm": 0.12327368649462837,
+      "learning_rate": 0.00019668461802233153,
+      "loss": 1.3192,
+      "step": 1523
+    },
+    {
+      "epoch": 0.10985367260145606,
+      "grad_norm": 0.10454775306873323,
+      "learning_rate": 0.00019667865347094092,
+      "loss": 1.3036,
+      "step": 1524
+    },
+    {
+      "epoch": 0.10992575506379298,
+      "grad_norm": 0.11757572507629521,
+      "learning_rate": 0.00019667268364969876,
+      "loss": 1.4184,
+      "step": 1525
+    },
+    {
+      "epoch": 0.10999783752612989,
+      "grad_norm": 0.119170032964993,
+      "learning_rate": 0.00019666670855893049,
+      "loss": 1.38,
+      "step": 1526
+    },
+    {
+      "epoch": 0.11006991998846681,
+      "grad_norm": 0.10357997908679441,
+      "learning_rate": 0.00019666072819896177,
+      "loss": 1.3636,
+      "step": 1527
+    },
+    {
+      "epoch": 0.11014200245080372,
+      "grad_norm": 0.10160102204736539,
+      "learning_rate": 0.00019665474257011864,
+      "loss": 1.3179,
+      "step": 1528
+    },
+    {
+      "epoch": 0.11021408491314064,
+      "grad_norm": 0.10846437703930084,
+      "learning_rate": 0.00019664875167272735,
+      "loss": 1.3993,
+      "step": 1529
+    },
+    {
+      "epoch": 0.11028616737547754,
+      "grad_norm": 0.1002212082102922,
+      "learning_rate": 0.0001966427555071144,
+      "loss": 1.3718,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11035824983781446,
+      "grad_norm": 0.12275089489378496,
+      "learning_rate": 0.00019663675407360676,
+      "loss": 1.274,
+      "step": 1531
+    },
+    {
+      "epoch": 0.11043033230015137,
+      "grad_norm": 0.19200133300423877,
+      "learning_rate": 0.00019663074737253142,
+      "loss": 1.282,
+      "step": 1532
+    },
+    {
+      "epoch": 0.11050241476248829,
+      "grad_norm": 0.11137024507355672,
+      "learning_rate": 0.00019662473540421588,
+      "loss": 1.2893,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1105744972248252,
+      "grad_norm": 0.11902208617374718,
+      "learning_rate": 0.00019661871816898784,
+      "loss": 1.3412,
+      "step": 1534
+    },
+    {
+      "epoch": 0.11064657968716211,
+      "grad_norm": 0.10494572205984073,
+      "learning_rate": 0.0001966126956671753,
+      "loss": 1.4553,
+      "step": 1535
+    },
+    {
+      "epoch": 0.11071866214949903,
+      "grad_norm": 0.11588216564172311,
+      "learning_rate": 0.00019660666789910654,
+      "loss": 1.2987,
+      "step": 1536
+    },
+    {
+      "epoch": 0.11079074461183594,
+      "grad_norm": 0.10328846613154583,
+      "learning_rate": 0.00019660063486511006,
+      "loss": 1.3702,
+      "step": 1537
+    },
+    {
+      "epoch": 0.11086282707417286,
+      "grad_norm": 0.09883927243335924,
+      "learning_rate": 0.0001965945965655148,
+      "loss": 1.3268,
+      "step": 1538
+    },
+    {
+      "epoch": 0.11093490953650977,
+      "grad_norm": 0.12632924409901292,
+      "learning_rate": 0.00019658855300064984,
+      "loss": 1.3539,
+      "step": 1539
+    },
+    {
+      "epoch": 0.11100699199884669,
+      "grad_norm": 0.10886649997086308,
+      "learning_rate": 0.00019658250417084466,
+      "loss": 1.3359,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11107907446118359,
+      "grad_norm": 0.09924501577414871,
+      "learning_rate": 0.00019657645007642894,
+      "loss": 1.3674,
+      "step": 1541
+    },
+    {
+      "epoch": 0.11115115692352051,
+      "grad_norm": 0.10300170578564007,
+      "learning_rate": 0.0001965703907177327,
+      "loss": 1.3447,
+      "step": 1542
+    },
+    {
+      "epoch": 0.11122323938585742,
+      "grad_norm": 0.11929648470091296,
+      "learning_rate": 0.0001965643260950862,
+      "loss": 1.2068,
+      "step": 1543
+    },
+    {
+      "epoch": 0.11129532184819434,
+      "grad_norm": 0.11071918699574416,
+      "learning_rate": 0.00019655825620882005,
+      "loss": 1.4481,
+      "step": 1544
+    },
+    {
+      "epoch": 0.11136740431053124,
+      "grad_norm": 0.12783779817844848,
+      "learning_rate": 0.0001965521810592651,
+      "loss": 1.3335,
+      "step": 1545
+    },
+    {
+      "epoch": 0.11143948677286816,
+      "grad_norm": 0.10686670506883138,
+      "learning_rate": 0.00019654610064675251,
+      "loss": 1.3676,
+      "step": 1546
+    },
+    {
+      "epoch": 0.11151156923520507,
+      "grad_norm": 0.10650618978645689,
+      "learning_rate": 0.00019654001497161372,
+      "loss": 1.3336,
+      "step": 1547
+    },
+    {
+      "epoch": 0.11158365169754199,
+      "grad_norm": 0.11961312775289998,
+      "learning_rate": 0.00019653392403418043,
+      "loss": 1.3798,
+      "step": 1548
+    },
+    {
+      "epoch": 0.11165573415987891,
+      "grad_norm": 0.10365403928427926,
+      "learning_rate": 0.00019652782783478466,
+      "loss": 1.317,
+      "step": 1549
+    },
+    {
+      "epoch": 0.11172781662221581,
+      "grad_norm": 0.10637963921795225,
+      "learning_rate": 0.0001965217263737587,
+      "loss": 1.2609,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11179989908455273,
+      "grad_norm": 0.12478691835816068,
+      "learning_rate": 0.0001965156196514352,
+      "loss": 1.2543,
+      "step": 1551
+    },
+    {
+      "epoch": 0.11187198154688964,
+      "grad_norm": 0.11034672012022685,
+      "learning_rate": 0.0001965095076681469,
+      "loss": 1.3358,
+      "step": 1552
+    },
+    {
+      "epoch": 0.11194406400922656,
+      "grad_norm": 0.14784983790053052,
+      "learning_rate": 0.00019650339042422707,
+      "loss": 1.2408,
+      "step": 1553
+    },
+    {
+      "epoch": 0.11201614647156347,
+      "grad_norm": 0.10595543020365346,
+      "learning_rate": 0.00019649726792000913,
+      "loss": 1.3334,
+      "step": 1554
+    },
+    {
+      "epoch": 0.11208822893390039,
+      "grad_norm": 0.29158864943960083,
+      "learning_rate": 0.00019649114015582678,
+      "loss": 1.3945,
+      "step": 1555
+    },
+    {
+      "epoch": 0.11216031139623729,
+      "grad_norm": 0.12753136053742925,
+      "learning_rate": 0.00019648500713201404,
+      "loss": 1.3023,
+      "step": 1556
+    },
+    {
+      "epoch": 0.11223239385857421,
+      "grad_norm": 0.10856918681809231,
+      "learning_rate": 0.00019647886884890527,
+      "loss": 1.4608,
+      "step": 1557
+    },
+    {
+      "epoch": 0.11230447632091112,
+      "grad_norm": 0.10758986289718249,
+      "learning_rate": 0.00019647272530683502,
+      "loss": 1.3693,
+      "step": 1558
+    },
+    {
+      "epoch": 0.11237655878324804,
+      "grad_norm": 0.10077572671635054,
+      "learning_rate": 0.00019646657650613817,
+      "loss": 1.3235,
+      "step": 1559
+    },
+    {
+      "epoch": 0.11244864124558494,
+      "grad_norm": 0.12203965689421141,
+      "learning_rate": 0.00019646042244714985,
+      "loss": 1.361,
+      "step": 1560
+    },
+    {
+      "epoch": 0.11252072370792186,
+      "grad_norm": 0.10792386180346064,
+      "learning_rate": 0.00019645426313020557,
+      "loss": 1.2784,
+      "step": 1561
+    },
+    {
+      "epoch": 0.11259280617025878,
+      "grad_norm": 0.09804400823883119,
+      "learning_rate": 0.00019644809855564102,
+      "loss": 1.3141,
+      "step": 1562
+    },
+    {
+      "epoch": 0.11266488863259569,
+      "grad_norm": 0.10576681120462116,
+      "learning_rate": 0.00019644192872379229,
+      "loss": 1.4159,
+      "step": 1563
+    },
+    {
+      "epoch": 0.11273697109493261,
+      "grad_norm": 0.104641248364246,
+      "learning_rate": 0.00019643575363499562,
+      "loss": 1.3107,
+      "step": 1564
+    },
+    {
+      "epoch": 0.11280905355726951,
+      "grad_norm": 0.11578853059350329,
+      "learning_rate": 0.00019642957328958768,
+      "loss": 1.414,
+      "step": 1565
+    },
+    {
+      "epoch": 0.11288113601960643,
+      "grad_norm": 0.0959176644276553,
+      "learning_rate": 0.00019642338768790526,
+      "loss": 1.3211,
+      "step": 1566
+    },
+    {
+      "epoch": 0.11295321848194334,
+      "grad_norm": 0.09183051504900651,
+      "learning_rate": 0.0001964171968302856,
+      "loss": 1.2699,
+      "step": 1567
+    },
+    {
+      "epoch": 0.11302530094428026,
+      "grad_norm": 0.10290316916450674,
+      "learning_rate": 0.00019641100071706618,
+      "loss": 1.3513,
+      "step": 1568
+    },
+    {
+      "epoch": 0.11309738340661717,
+      "grad_norm": 0.1346203460549771,
+      "learning_rate": 0.00019640479934858465,
+      "loss": 1.3031,
+      "step": 1569
+    },
+    {
+      "epoch": 0.11316946586895409,
+      "grad_norm": 0.100140100823916,
+      "learning_rate": 0.0001963985927251791,
+      "loss": 1.2546,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11324154833129099,
+      "grad_norm": 0.11345479615293445,
+      "learning_rate": 0.00019639238084718787,
+      "loss": 1.3228,
+      "step": 1571
+    },
+    {
+      "epoch": 0.11331363079362791,
+      "grad_norm": 0.1656126760750025,
+      "learning_rate": 0.0001963861637149495,
+      "loss": 1.3382,
+      "step": 1572
+    },
+    {
+      "epoch": 0.11338571325596482,
+      "grad_norm": 0.1255924278215042,
+      "learning_rate": 0.00019637994132880296,
+      "loss": 1.2821,
+      "step": 1573
+    },
+    {
+      "epoch": 0.11345779571830174,
+      "grad_norm": 0.1101964794345219,
+      "learning_rate": 0.00019637371368908736,
+      "loss": 1.2487,
+      "step": 1574
+    },
+    {
+      "epoch": 0.11352987818063866,
+      "grad_norm": 0.09968251015952792,
+      "learning_rate": 0.00019636748079614216,
+      "loss": 1.3422,
+      "step": 1575
+    },
+    {
+      "epoch": 0.11360196064297556,
+      "grad_norm": 0.10369349854866367,
+      "learning_rate": 0.00019636124265030714,
+      "loss": 1.3625,
+      "step": 1576
+    },
+    {
+      "epoch": 0.11367404310531248,
+      "grad_norm": 0.13097198319216086,
+      "learning_rate": 0.0001963549992519223,
+      "loss": 1.4274,
+      "step": 1577
+    },
+    {
+      "epoch": 0.11374612556764939,
+      "grad_norm": 0.11997772558505121,
+      "learning_rate": 0.000196348750601328,
+      "loss": 1.4797,
+      "step": 1578
+    },
+    {
+      "epoch": 0.11381820802998631,
+      "grad_norm": 0.11130454130167192,
+      "learning_rate": 0.00019634249669886483,
+      "loss": 1.3717,
+      "step": 1579
+    },
+    {
+      "epoch": 0.11389029049232322,
+      "grad_norm": 0.11525456069649254,
+      "learning_rate": 0.00019633623754487372,
+      "loss": 1.3386,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11396237295466013,
+      "grad_norm": 0.10703029167233685,
+      "learning_rate": 0.00019632997313969578,
+      "loss": 1.3335,
+      "step": 1581
+    },
+    {
+      "epoch": 0.11403445541699704,
+      "grad_norm": 0.11867149836629247,
+      "learning_rate": 0.0001963237034836725,
+      "loss": 1.3491,
+      "step": 1582
+    },
+    {
+      "epoch": 0.11410653787933396,
+      "grad_norm": 0.09778881972198093,
+      "learning_rate": 0.00019631742857714567,
+      "loss": 1.3636,
+      "step": 1583
+    },
+    {
+      "epoch": 0.11417862034167087,
+      "grad_norm": 0.12816514144070076,
+      "learning_rate": 0.0001963111484204573,
+      "loss": 1.3333,
+      "step": 1584
+    },
+    {
+      "epoch": 0.11425070280400779,
+      "grad_norm": 0.1186193106437118,
+      "learning_rate": 0.0001963048630139497,
+      "loss": 1.3836,
+      "step": 1585
+    },
+    {
+      "epoch": 0.11432278526634469,
+      "grad_norm": 0.10314560352533125,
+      "learning_rate": 0.00019629857235796552,
+      "loss": 1.3353,
+      "step": 1586
+    },
+    {
+      "epoch": 0.11439486772868161,
+      "grad_norm": 0.11069660290257945,
+      "learning_rate": 0.0001962922764528476,
+      "loss": 1.3215,
+      "step": 1587
+    },
+    {
+      "epoch": 0.11446695019101852,
+      "grad_norm": 0.12031472738521892,
+      "learning_rate": 0.0001962859752989392,
+      "loss": 1.3841,
+      "step": 1588
+    },
+    {
+      "epoch": 0.11453903265335544,
+      "grad_norm": 0.10147457073397718,
+      "learning_rate": 0.00019627966889658373,
+      "loss": 1.3357,
+      "step": 1589
+    },
+    {
+      "epoch": 0.11461111511569236,
+      "grad_norm": 0.13277655041671468,
+      "learning_rate": 0.00019627335724612494,
+      "loss": 1.404,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11468319757802926,
+      "grad_norm": 0.11602993672763155,
+      "learning_rate": 0.00019626704034790692,
+      "loss": 1.3122,
+      "step": 1591
+    },
+    {
+      "epoch": 0.11475528004036618,
+      "grad_norm": 0.11417248580294669,
+      "learning_rate": 0.000196260718202274,
+      "loss": 1.2643,
+      "step": 1592
+    },
+    {
+      "epoch": 0.11482736250270309,
+      "grad_norm": 0.16019396424372662,
+      "learning_rate": 0.0001962543908095707,
+      "loss": 1.3634,
+      "step": 1593
+    },
+    {
+      "epoch": 0.11489944496504001,
+      "grad_norm": 0.10727142908758723,
+      "learning_rate": 0.000196248058170142,
+      "loss": 1.329,
+      "step": 1594
+    },
+    {
+      "epoch": 0.11497152742737692,
+      "grad_norm": 0.11341399909732057,
+      "learning_rate": 0.0001962417202843331,
+      "loss": 1.3894,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11504360988971384,
+      "grad_norm": 0.10012844302792027,
+      "learning_rate": 0.00019623537715248942,
+      "loss": 1.3787,
+      "step": 1596
+    },
+    {
+      "epoch": 0.11511569235205074,
+      "grad_norm": 0.11974405792083966,
+      "learning_rate": 0.00019622902877495678,
+      "loss": 1.3095,
+      "step": 1597
+    },
+    {
+      "epoch": 0.11518777481438766,
+      "grad_norm": 0.11242861807146207,
+      "learning_rate": 0.00019622267515208113,
+      "loss": 1.4109,
+      "step": 1598
+    },
+    {
+      "epoch": 0.11525985727672457,
+      "grad_norm": 0.10041793782842194,
+      "learning_rate": 0.00019621631628420887,
+      "loss": 1.3235,
+      "step": 1599
+    },
+    {
+      "epoch": 0.11533193973906149,
+      "grad_norm": 0.20524467013760112,
+      "learning_rate": 0.00019620995217168658,
+      "loss": 1.2801,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11540402220139839,
+      "grad_norm": 0.11196199673156886,
+      "learning_rate": 0.0001962035828148612,
+      "loss": 1.3881,
+      "step": 1601
+    },
+    {
+      "epoch": 0.11547610466373531,
+      "grad_norm": 0.11477798631738663,
+      "learning_rate": 0.0001961972082140799,
+      "loss": 1.2991,
+      "step": 1602
+    },
+    {
+      "epoch": 0.11554818712607223,
+      "grad_norm": 0.09552851556568438,
+      "learning_rate": 0.00019619082836969015,
+      "loss": 1.2927,
+      "step": 1603
+    },
+    {
+      "epoch": 0.11562026958840914,
+      "grad_norm": 0.10196551842692428,
+      "learning_rate": 0.00019618444328203967,
+      "loss": 1.2858,
+      "step": 1604
+    },
+    {
+      "epoch": 0.11569235205074606,
+      "grad_norm": 0.177168620825384,
+      "learning_rate": 0.00019617805295147654,
+      "loss": 1.2724,
+      "step": 1605
+    },
+    {
+      "epoch": 0.11576443451308296,
+      "grad_norm": 0.11904370730165097,
+      "learning_rate": 0.00019617165737834916,
+      "loss": 1.371,
+      "step": 1606
+    },
+    {
+      "epoch": 0.11583651697541988,
+      "grad_norm": 0.1107104364572541,
+      "learning_rate": 0.00019616525656300602,
+      "loss": 1.3928,
+      "step": 1607
+    },
+    {
+      "epoch": 0.11590859943775679,
+      "grad_norm": 0.10533202477023466,
+      "learning_rate": 0.00019615885050579607,
+      "loss": 1.3906,
+      "step": 1608
+    },
+    {
+      "epoch": 0.11598068190009371,
+      "grad_norm": 0.09275691030602375,
+      "learning_rate": 0.00019615243920706853,
+      "loss": 1.2825,
+      "step": 1609
+    },
+    {
+      "epoch": 0.11605276436243062,
+      "grad_norm": 0.1376325521994405,
+      "learning_rate": 0.00019614602266717284,
+      "loss": 1.4401,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11612484682476754,
+      "grad_norm": 0.11354966096153966,
+      "learning_rate": 0.00019613960088645877,
+      "loss": 1.3415,
+      "step": 1611
+    },
+    {
+      "epoch": 0.11619692928710444,
+      "grad_norm": 0.1087127243702572,
+      "learning_rate": 0.00019613317386527637,
+      "loss": 1.3451,
+      "step": 1612
+    },
+    {
+      "epoch": 0.11626901174944136,
+      "grad_norm": 0.12143830551739354,
+      "learning_rate": 0.00019612674160397595,
+      "loss": 1.4232,
+      "step": 1613
+    },
+    {
+      "epoch": 0.11634109421177827,
+      "grad_norm": 0.11743335951748334,
+      "learning_rate": 0.00019612030410290812,
+      "loss": 1.3365,
+      "step": 1614
+    },
+    {
+      "epoch": 0.11641317667411519,
+      "grad_norm": 0.11025802797161534,
+      "learning_rate": 0.00019611386136242386,
+      "loss": 1.3837,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11648525913645211,
+      "grad_norm": 0.11477585998249126,
+      "learning_rate": 0.00019610741338287423,
+      "loss": 1.3625,
+      "step": 1616
+    },
+    {
+      "epoch": 0.11655734159878901,
+      "grad_norm": 0.1308074807695052,
+      "learning_rate": 0.00019610096016461085,
+      "loss": 1.3443,
+      "step": 1617
+    },
+    {
+      "epoch": 0.11662942406112593,
+      "grad_norm": 0.12865098688781315,
+      "learning_rate": 0.00019609450170798533,
+      "loss": 1.3038,
+      "step": 1618
+    },
+    {
+      "epoch": 0.11670150652346284,
+      "grad_norm": 0.11123042994225882,
+      "learning_rate": 0.0001960880380133498,
+      "loss": 1.3277,
+      "step": 1619
+    },
+    {
+      "epoch": 0.11677358898579976,
+      "grad_norm": 0.10355142346422579,
+      "learning_rate": 0.00019608156908105655,
+      "loss": 1.3013,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11684567144813666,
+      "grad_norm": 0.10798221203443373,
+      "learning_rate": 0.00019607509491145823,
+      "loss": 1.4156,
+      "step": 1621
+    },
+    {
+      "epoch": 0.11691775391047358,
+      "grad_norm": 0.1472786430424531,
+      "learning_rate": 0.00019606861550490774,
+      "loss": 1.3112,
+      "step": 1622
+    },
+    {
+      "epoch": 0.11698983637281049,
+      "grad_norm": 0.10956621007088525,
+      "learning_rate": 0.00019606213086175825,
+      "loss": 1.3636,
+      "step": 1623
+    },
+    {
+      "epoch": 0.11706191883514741,
+      "grad_norm": 0.10466040163177605,
+      "learning_rate": 0.0001960556409823632,
+      "loss": 1.2917,
+      "step": 1624
+    },
+    {
+      "epoch": 0.11713400129748432,
+      "grad_norm": 0.10892698456481784,
+      "learning_rate": 0.0001960491458670764,
+      "loss": 1.2249,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11720608375982124,
+      "grad_norm": 0.12648114838916646,
+      "learning_rate": 0.00019604264551625187,
+      "loss": 1.3423,
+      "step": 1626
+    },
+    {
+      "epoch": 0.11727816622215814,
+      "grad_norm": 0.10550499473483733,
+      "learning_rate": 0.00019603613993024395,
+      "loss": 1.3062,
+      "step": 1627
+    },
+    {
+      "epoch": 0.11735024868449506,
+      "grad_norm": 0.09167389462982735,
+      "learning_rate": 0.00019602962910940722,
+      "loss": 1.2908,
+      "step": 1628
+    },
+    {
+      "epoch": 0.11742233114683198,
+      "grad_norm": 0.10917842635471377,
+      "learning_rate": 0.0001960231130540966,
+      "loss": 1.3736,
+      "step": 1629
+    },
+    {
+      "epoch": 0.11749441360916889,
+      "grad_norm": 0.1357063864625358,
+      "learning_rate": 0.00019601659176466727,
+      "loss": 1.2897,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11756649607150581,
+      "grad_norm": 0.11196113935267757,
+      "learning_rate": 0.0001960100652414747,
+      "loss": 1.2416,
+      "step": 1631
+    },
+    {
+      "epoch": 0.11763857853384271,
+      "grad_norm": 0.10914347756774603,
+      "learning_rate": 0.00019600353348487464,
+      "loss": 1.3908,
+      "step": 1632
+    },
+    {
+      "epoch": 0.11771066099617963,
+      "grad_norm": 0.130522271240399,
+      "learning_rate": 0.00019599699649522316,
+      "loss": 1.3469,
+      "step": 1633
+    },
+    {
+      "epoch": 0.11778274345851654,
+      "grad_norm": 0.10729588215101418,
+      "learning_rate": 0.00019599045427287653,
+      "loss": 1.452,
+      "step": 1634
+    },
+    {
+      "epoch": 0.11785482592085346,
+      "grad_norm": 0.10539150066605826,
+      "learning_rate": 0.0001959839068181914,
+      "loss": 1.2796,
+      "step": 1635
+    },
+    {
+      "epoch": 0.11792690838319037,
+      "grad_norm": 0.11416461638247732,
+      "learning_rate": 0.00019597735413152463,
+      "loss": 1.2594,
+      "step": 1636
+    },
+    {
+      "epoch": 0.11799899084552729,
+      "grad_norm": 0.20095995198719294,
+      "learning_rate": 0.00019597079621323346,
+      "loss": 1.3768,
+      "step": 1637
+    },
+    {
+      "epoch": 0.11807107330786419,
+      "grad_norm": 0.10615811551270724,
+      "learning_rate": 0.0001959642330636753,
+      "loss": 1.3436,
+      "step": 1638
+    },
+    {
+      "epoch": 0.11814315577020111,
+      "grad_norm": 0.10708579563950799,
+      "learning_rate": 0.00019595766468320788,
+      "loss": 1.2857,
+      "step": 1639
+    },
+    {
+      "epoch": 0.11821523823253802,
+      "grad_norm": 0.11571384048490545,
+      "learning_rate": 0.00019595109107218932,
+      "loss": 1.3576,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11828732069487494,
+      "grad_norm": 0.1210865188570797,
+      "learning_rate": 0.00019594451223097788,
+      "loss": 1.3733,
+      "step": 1641
+    },
+    {
+      "epoch": 0.11835940315721186,
+      "grad_norm": 0.09841087697366933,
+      "learning_rate": 0.0001959379281599322,
+      "loss": 1.3541,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11843148561954876,
+      "grad_norm": 0.10060538454048959,
+      "learning_rate": 0.00019593133885941112,
+      "loss": 1.4003,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11850356808188568,
+      "grad_norm": 0.10040819803728178,
+      "learning_rate": 0.00019592474432977386,
+      "loss": 1.2525,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11857565054422259,
+      "grad_norm": 0.1798322706019341,
+      "learning_rate": 0.00019591814457137985,
+      "loss": 1.2841,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11864773300655951,
+      "grad_norm": 0.1048665030346019,
+      "learning_rate": 0.00019591153958458884,
+      "loss": 1.4066,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11871981546889641,
+      "grad_norm": 0.10092078081827019,
+      "learning_rate": 0.00019590492936976092,
+      "loss": 1.4159,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11879189793123333,
+      "grad_norm": 0.10449998343254974,
+      "learning_rate": 0.00019589831392725633,
+      "loss": 1.2816,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11886398039357024,
+      "grad_norm": 0.28852339021576817,
+      "learning_rate": 0.00019589169325743574,
+      "loss": 1.3831,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11893606285590716,
+      "grad_norm": 0.1604905328006569,
+      "learning_rate": 0.00019588506736065994,
+      "loss": 1.378,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11900814531824407,
+      "grad_norm": 0.09636496295819116,
+      "learning_rate": 0.00019587843623729016,
+      "loss": 1.3145,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11908022778058099,
+      "grad_norm": 0.10747684497571748,
+      "learning_rate": 0.0001958717998876879,
+      "loss": 1.4124,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11915231024291789,
+      "grad_norm": 0.10458258872804625,
+      "learning_rate": 0.0001958651583122148,
+      "loss": 1.229,
+      "step": 1653
+    },
+    {
+      "epoch": 0.11922439270525481,
+      "grad_norm": 0.09198131134309256,
+      "learning_rate": 0.00019585851151123298,
+      "loss": 1.3003,
+      "step": 1654
+    },
+    {
+      "epoch": 0.11929647516759173,
+      "grad_norm": 0.10999098964573886,
+      "learning_rate": 0.0001958518594851047,
+      "loss": 1.2923,
+      "step": 1655
+    },
+    {
+      "epoch": 0.11936855762992864,
+      "grad_norm": 0.1052615374109633,
+      "learning_rate": 0.00019584520223419255,
+      "loss": 1.4217,
+      "step": 1656
+    },
+    {
+      "epoch": 0.11944064009226556,
+      "grad_norm": 0.11067637131526549,
+      "learning_rate": 0.0001958385397588594,
+      "loss": 1.3866,
+      "step": 1657
+    },
+    {
+      "epoch": 0.11951272255460246,
+      "grad_norm": 0.10222527783044569,
+      "learning_rate": 0.00019583187205946846,
+      "loss": 1.276,
+      "step": 1658
+    },
+    {
+      "epoch": 0.11958480501693938,
+      "grad_norm": 0.10378674891643344,
+      "learning_rate": 0.00019582519913638317,
+      "loss": 1.3754,
+      "step": 1659
+    },
+    {
+      "epoch": 0.11965688747927629,
+      "grad_norm": 0.11542957550642637,
+      "learning_rate": 0.00019581852098996726,
+      "loss": 1.3585,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11972896994161321,
+      "grad_norm": 0.11792002270650495,
+      "learning_rate": 0.0001958118376205847,
+      "loss": 1.3653,
+      "step": 1661
+    },
+    {
+      "epoch": 0.11980105240395011,
+      "grad_norm": 0.1453441826367774,
+      "learning_rate": 0.00019580514902859986,
+      "loss": 1.2907,
+      "step": 1662
+    },
+    {
+      "epoch": 0.11987313486628703,
+      "grad_norm": 0.13453292108584544,
+      "learning_rate": 0.00019579845521437727,
+      "loss": 1.4442,
+      "step": 1663
+    },
+    {
+      "epoch": 0.11994521732862394,
+      "grad_norm": 0.12523361461074953,
+      "learning_rate": 0.00019579175617828187,
+      "loss": 1.3001,
+      "step": 1664
+    },
+    {
+      "epoch": 0.12001729979096086,
+      "grad_norm": 0.10191815729524828,
+      "learning_rate": 0.00019578505192067874,
+      "loss": 1.3107,
+      "step": 1665
+    },
+    {
+      "epoch": 0.12008938225329777,
+      "grad_norm": 0.09819659505422317,
+      "learning_rate": 0.0001957783424419334,
+      "loss": 1.2627,
+      "step": 1666
+    },
+    {
+      "epoch": 0.12016146471563469,
+      "grad_norm": 0.11034904133160763,
+      "learning_rate": 0.00019577162774241154,
+      "loss": 1.3236,
+      "step": 1667
+    },
+    {
+      "epoch": 0.1202335471779716,
+      "grad_norm": 0.11249705633522777,
+      "learning_rate": 0.00019576490782247916,
+      "loss": 1.3945,
+      "step": 1668
+    },
+    {
+      "epoch": 0.12030562964030851,
+      "grad_norm": 0.10574336224328273,
+      "learning_rate": 0.00019575818268250257,
+      "loss": 1.404,
+      "step": 1669
+    },
+    {
+      "epoch": 0.12037771210264543,
+      "grad_norm": 0.20455637498939158,
+      "learning_rate": 0.00019575145232284832,
+      "loss": 1.3977,
+      "step": 1670
+    },
+    {
+      "epoch": 0.12044979456498234,
+      "grad_norm": 0.09798498037059206,
+      "learning_rate": 0.00019574471674388332,
+      "loss": 1.3675,
+      "step": 1671
+    },
+    {
+      "epoch": 0.12052187702731926,
+      "grad_norm": 0.1007193529984958,
+      "learning_rate": 0.00019573797594597472,
+      "loss": 1.3962,
+      "step": 1672
+    },
+    {
+      "epoch": 0.12059395948965616,
+      "grad_norm": 0.10974419797543114,
+      "learning_rate": 0.0001957312299294899,
+      "loss": 1.4302,
+      "step": 1673
+    },
+    {
+      "epoch": 0.12066604195199308,
+      "grad_norm": 0.10614828102527114,
+      "learning_rate": 0.00019572447869479665,
+      "loss": 1.3918,
+      "step": 1674
+    },
+    {
+      "epoch": 0.12073812441432999,
+      "grad_norm": 0.11558695620489634,
+      "learning_rate": 0.00019571772224226291,
+      "loss": 1.5109,
+      "step": 1675
+    },
+    {
+      "epoch": 0.12081020687666691,
+      "grad_norm": 0.10938109781590669,
+      "learning_rate": 0.000195710960572257,
+      "loss": 1.2816,
+      "step": 1676
+    },
+    {
+      "epoch": 0.12088228933900382,
+      "grad_norm": 0.09598339726904935,
+      "learning_rate": 0.0001957041936851475,
+      "loss": 1.2208,
+      "step": 1677
+    },
+    {
+      "epoch": 0.12095437180134073,
+      "grad_norm": 0.12451785421037638,
+      "learning_rate": 0.00019569742158130323,
+      "loss": 1.2653,
+      "step": 1678
+    },
+    {
+      "epoch": 0.12102645426367764,
+      "grad_norm": 0.1072188759271412,
+      "learning_rate": 0.00019569064426109333,
+      "loss": 1.3126,
+      "step": 1679
+    },
+    {
+      "epoch": 0.12109853672601456,
+      "grad_norm": 0.11457938563584978,
+      "learning_rate": 0.0001956838617248873,
+      "loss": 1.312,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12117061918835148,
+      "grad_norm": 0.09767005189189014,
+      "learning_rate": 0.00019567707397305474,
+      "loss": 1.397,
+      "step": 1681
+    },
+    {
+      "epoch": 0.12124270165068839,
+      "grad_norm": 0.12259313455846259,
+      "learning_rate": 0.0001956702810059657,
+      "loss": 1.3225,
+      "step": 1682
+    },
+    {
+      "epoch": 0.1213147841130253,
+      "grad_norm": 0.10465561986157514,
+      "learning_rate": 0.0001956634828239905,
+      "loss": 1.3177,
+      "step": 1683
+    },
+    {
+      "epoch": 0.12138686657536221,
+      "grad_norm": 0.14830137811052937,
+      "learning_rate": 0.0001956566794274996,
+      "loss": 1.3808,
+      "step": 1684
+    },
+    {
+      "epoch": 0.12145894903769913,
+      "grad_norm": 0.12543681020479933,
+      "learning_rate": 0.00019564987081686395,
+      "loss": 1.3902,
+      "step": 1685
+    },
+    {
+      "epoch": 0.12153103150003604,
+      "grad_norm": 0.25895490287219447,
+      "learning_rate": 0.0001956430569924546,
+      "loss": 1.3628,
+      "step": 1686
+    },
+    {
+      "epoch": 0.12160311396237296,
+      "grad_norm": 0.12317573364339358,
+      "learning_rate": 0.00019563623795464303,
+      "loss": 1.324,
+      "step": 1687
+    },
+    {
+      "epoch": 0.12167519642470986,
+      "grad_norm": 0.11501142988377533,
+      "learning_rate": 0.00019562941370380088,
+      "loss": 1.3871,
+      "step": 1688
+    },
+    {
+      "epoch": 0.12174727888704678,
+      "grad_norm": 0.11086033522638643,
+      "learning_rate": 0.00019562258424030016,
+      "loss": 1.3245,
+      "step": 1689
+    },
+    {
+      "epoch": 0.12181936134938369,
+      "grad_norm": 0.09995048337135772,
+      "learning_rate": 0.00019561574956451314,
+      "loss": 1.3991,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12189144381172061,
+      "grad_norm": 0.1573363543108352,
+      "learning_rate": 0.00019560890967681237,
+      "loss": 1.3951,
+      "step": 1691
+    },
+    {
+      "epoch": 0.12196352627405752,
+      "grad_norm": 0.1135180931217498,
+      "learning_rate": 0.00019560206457757066,
+      "loss": 1.3067,
+      "step": 1692
+    },
+    {
+      "epoch": 0.12203560873639444,
+      "grad_norm": 0.2500198005555177,
+      "learning_rate": 0.00019559521426716118,
+      "loss": 1.3366,
+      "step": 1693
+    },
+    {
+      "epoch": 0.12210769119873136,
+      "grad_norm": 0.11023813717144909,
+      "learning_rate": 0.00019558835874595727,
+      "loss": 1.3362,
+      "step": 1694
+    },
+    {
+      "epoch": 0.12217977366106826,
+      "grad_norm": 0.09717964331477484,
+      "learning_rate": 0.00019558149801433266,
+      "loss": 1.3224,
+      "step": 1695
+    },
+    {
+      "epoch": 0.12225185612340518,
+      "grad_norm": 0.10863354663892658,
+      "learning_rate": 0.00019557463207266134,
+      "loss": 1.4659,
+      "step": 1696
+    },
+    {
+      "epoch": 0.12232393858574209,
+      "grad_norm": 0.10282649616549525,
+      "learning_rate": 0.00019556776092131753,
+      "loss": 1.3312,
+      "step": 1697
+    },
+    {
+      "epoch": 0.122396021048079,
+      "grad_norm": 0.09703772470625138,
+      "learning_rate": 0.00019556088456067574,
+      "loss": 1.3002,
+      "step": 1698
+    },
+    {
+      "epoch": 0.12246810351041591,
+      "grad_norm": 0.10736364661899199,
+      "learning_rate": 0.00019555400299111085,
+      "loss": 1.2599,
+      "step": 1699
+    },
+    {
+      "epoch": 0.12254018597275283,
+      "grad_norm": 0.11651751158955734,
+      "learning_rate": 0.00019554711621299795,
+      "loss": 1.3377,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12261226843508974,
+      "grad_norm": 0.11733573340216433,
+      "learning_rate": 0.00019554022422671243,
+      "loss": 1.2654,
+      "step": 1701
+    },
+    {
+      "epoch": 0.12268435089742666,
+      "grad_norm": 0.11136005065881259,
+      "learning_rate": 0.00019553332703262997,
+      "loss": 1.3653,
+      "step": 1702
+    },
+    {
+      "epoch": 0.12275643335976356,
+      "grad_norm": 0.12805629789227874,
+      "learning_rate": 0.0001955264246311265,
+      "loss": 1.3617,
+      "step": 1703
+    },
+    {
+      "epoch": 0.12282851582210048,
+      "grad_norm": 0.117832694371423,
+      "learning_rate": 0.0001955195170225783,
+      "loss": 1.2796,
+      "step": 1704
+    },
+    {
+      "epoch": 0.12290059828443739,
+      "grad_norm": 0.11752554469172953,
+      "learning_rate": 0.00019551260420736189,
+      "loss": 1.4231,
+      "step": 1705
+    },
+    {
+      "epoch": 0.12297268074677431,
+      "grad_norm": 0.11371270844366178,
+      "learning_rate": 0.00019550568618585402,
+      "loss": 1.3892,
+      "step": 1706
+    },
+    {
+      "epoch": 0.12304476320911123,
+      "grad_norm": 0.12124376198828565,
+      "learning_rate": 0.00019549876295843187,
+      "loss": 1.3121,
+      "step": 1707
+    },
+    {
+      "epoch": 0.12311684567144814,
+      "grad_norm": 0.1680418458578921,
+      "learning_rate": 0.0001954918345254728,
+      "loss": 1.2745,
+      "step": 1708
+    },
+    {
+      "epoch": 0.12318892813378506,
+      "grad_norm": 0.11517289708248914,
+      "learning_rate": 0.00019548490088735444,
+      "loss": 1.2883,
+      "step": 1709
+    },
+    {
+      "epoch": 0.12326101059612196,
+      "grad_norm": 0.11593312466603917,
+      "learning_rate": 0.00019547796204445473,
+      "loss": 1.3959,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12333309305845888,
+      "grad_norm": 0.11971086479507916,
+      "learning_rate": 0.00019547101799715196,
+      "loss": 1.3856,
+      "step": 1711
+    },
+    {
+      "epoch": 0.12340517552079579,
+      "grad_norm": 0.10472377986961197,
+      "learning_rate": 0.00019546406874582456,
+      "loss": 1.323,
+      "step": 1712
+    },
+    {
+      "epoch": 0.12347725798313271,
+      "grad_norm": 0.11330133019389182,
+      "learning_rate": 0.00019545711429085137,
+      "loss": 1.3712,
+      "step": 1713
+    },
+    {
+      "epoch": 0.12354934044546961,
+      "grad_norm": 0.12182445248986595,
+      "learning_rate": 0.00019545015463261146,
+      "loss": 1.2725,
+      "step": 1714
+    },
+    {
+      "epoch": 0.12362142290780653,
+      "grad_norm": 0.11186654698190296,
+      "learning_rate": 0.00019544318977148426,
+      "loss": 1.3172,
+      "step": 1715
+    },
+    {
+      "epoch": 0.12369350537014344,
+      "grad_norm": 0.0992362014940531,
+      "learning_rate": 0.00019543621970784928,
+      "loss": 1.3244,
+      "step": 1716
+    },
+    {
+      "epoch": 0.12376558783248036,
+      "grad_norm": 0.11175805189773172,
+      "learning_rate": 0.00019542924444208658,
+      "loss": 1.2184,
+      "step": 1717
+    },
+    {
+      "epoch": 0.12383767029481726,
+      "grad_norm": 0.11356475846119451,
+      "learning_rate": 0.00019542226397457633,
+      "loss": 1.367,
+      "step": 1718
+    },
+    {
+      "epoch": 0.12390975275715418,
+      "grad_norm": 0.10848863599272063,
+      "learning_rate": 0.000195415278305699,
+      "loss": 1.3227,
+      "step": 1719
+    },
+    {
+      "epoch": 0.1239818352194911,
+      "grad_norm": 0.09772562540558995,
+      "learning_rate": 0.00019540828743583538,
+      "loss": 1.3106,
+      "step": 1720
+    },
+    {
+      "epoch": 0.12405391768182801,
+      "grad_norm": 0.10655837924131999,
+      "learning_rate": 0.0001954012913653666,
+      "loss": 1.4082,
+      "step": 1721
+    },
+    {
+      "epoch": 0.12412600014416493,
+      "grad_norm": 0.10536909797078807,
+      "learning_rate": 0.0001953942900946739,
+      "loss": 1.3316,
+      "step": 1722
+    },
+    {
+      "epoch": 0.12419808260650184,
+      "grad_norm": 0.12197990144829565,
+      "learning_rate": 0.00019538728362413895,
+      "loss": 1.2961,
+      "step": 1723
+    },
+    {
+      "epoch": 0.12427016506883876,
+      "grad_norm": 0.12742690808753776,
+      "learning_rate": 0.00019538027195414374,
+      "loss": 1.4227,
+      "step": 1724
+    },
+    {
+      "epoch": 0.12434224753117566,
+      "grad_norm": 0.16541223990998433,
+      "learning_rate": 0.00019537325508507038,
+      "loss": 1.3076,
+      "step": 1725
+    },
+    {
+      "epoch": 0.12441432999351258,
+      "grad_norm": 0.10402156500922177,
+      "learning_rate": 0.0001953662330173014,
+      "loss": 1.2896,
+      "step": 1726
+    },
+    {
+      "epoch": 0.12448641245584949,
+      "grad_norm": 0.17056684546057072,
+      "learning_rate": 0.00019535920575121957,
+      "loss": 1.2107,
+      "step": 1727
+    },
+    {
+      "epoch": 0.12455849491818641,
+      "grad_norm": 0.10218050947837515,
+      "learning_rate": 0.0001953521732872079,
+      "loss": 1.2844,
+      "step": 1728
+    },
+    {
+      "epoch": 0.12463057738052331,
+      "grad_norm": 0.13604082028866546,
+      "learning_rate": 0.00019534513562564972,
+      "loss": 1.394,
+      "step": 1729
+    },
+    {
+      "epoch": 0.12470265984286023,
+      "grad_norm": 0.115048654620511,
+      "learning_rate": 0.0001953380927669287,
+      "loss": 1.2886,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12477474230519714,
+      "grad_norm": 0.10831918829335539,
+      "learning_rate": 0.00019533104471142866,
+      "loss": 1.3449,
+      "step": 1731
+    },
+    {
+      "epoch": 0.12484682476753406,
+      "grad_norm": 0.11315530848044023,
+      "learning_rate": 0.00019532399145953387,
+      "loss": 1.2419,
+      "step": 1732
+    },
+    {
+      "epoch": 0.12491890722987098,
+      "grad_norm": 0.18700527301736455,
+      "learning_rate": 0.00019531693301162874,
+      "loss": 1.3577,
+      "step": 1733
+    },
+    {
+      "epoch": 0.12499098969220789,
+      "grad_norm": 0.11302412747944689,
+      "learning_rate": 0.00019530986936809803,
+      "loss": 1.3823,
+      "step": 1734
+    },
+    {
+      "epoch": 0.1250630721545448,
+      "grad_norm": 0.10905100403307484,
+      "learning_rate": 0.00019530280052932678,
+      "loss": 1.3865,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1251351546168817,
+      "grad_norm": 0.11484573204094155,
+      "learning_rate": 0.00019529572649570027,
+      "loss": 1.3789,
+      "step": 1736
+    },
+    {
+      "epoch": 0.12520723707921863,
+      "grad_norm": 0.1005999189290195,
+      "learning_rate": 0.00019528864726760416,
+      "loss": 1.382,
+      "step": 1737
+    },
+    {
+      "epoch": 0.12527931954155555,
+      "grad_norm": 0.1110646133897432,
+      "learning_rate": 0.00019528156284542427,
+      "loss": 1.3521,
+      "step": 1738
+    },
+    {
+      "epoch": 0.12535140200389244,
+      "grad_norm": 0.10981190932026032,
+      "learning_rate": 0.0001952744732295468,
+      "loss": 1.3382,
+      "step": 1739
+    },
+    {
+      "epoch": 0.12542348446622936,
+      "grad_norm": 0.13820518650766364,
+      "learning_rate": 0.0001952673784203582,
+      "loss": 1.3213,
+      "step": 1740
+    },
+    {
+      "epoch": 0.12549556692856628,
+      "grad_norm": 0.13950194409067962,
+      "learning_rate": 0.00019526027841824522,
+      "loss": 1.2864,
+      "step": 1741
+    },
+    {
+      "epoch": 0.1255676493909032,
+      "grad_norm": 0.16850126030145604,
+      "learning_rate": 0.0001952531732235948,
+      "loss": 1.3208,
+      "step": 1742
+    },
+    {
+      "epoch": 0.1256397318532401,
+      "grad_norm": 0.09826688077585323,
+      "learning_rate": 0.00019524606283679427,
+      "loss": 1.453,
+      "step": 1743
+    },
+    {
+      "epoch": 0.12571181431557701,
+      "grad_norm": 0.11934048410570391,
+      "learning_rate": 0.00019523894725823125,
+      "loss": 1.3589,
+      "step": 1744
+    },
+    {
+      "epoch": 0.12578389677791393,
+      "grad_norm": 0.09974220685831833,
+      "learning_rate": 0.00019523182648829358,
+      "loss": 1.3201,
+      "step": 1745
+    },
+    {
+      "epoch": 0.12585597924025085,
+      "grad_norm": 0.10806735107191845,
+      "learning_rate": 0.00019522470052736935,
+      "loss": 1.364,
+      "step": 1746
+    },
+    {
+      "epoch": 0.12592806170258777,
+      "grad_norm": 0.12129101141781497,
+      "learning_rate": 0.0001952175693758471,
+      "loss": 1.324,
+      "step": 1747
+    },
+    {
+      "epoch": 0.12600014416492467,
+      "grad_norm": 0.16375645269089595,
+      "learning_rate": 0.00019521043303411542,
+      "loss": 1.2375,
+      "step": 1748
+    },
+    {
+      "epoch": 0.12607222662726159,
+      "grad_norm": 0.09434661206130854,
+      "learning_rate": 0.00019520329150256337,
+      "loss": 1.2254,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1261443090895985,
+      "grad_norm": 0.10656951778009666,
+      "learning_rate": 0.00019519614478158027,
+      "loss": 1.3889,
+      "step": 1750
+    },
+    {
+      "epoch": 0.12621639155193543,
+      "grad_norm": 0.1177519625543885,
+      "learning_rate": 0.00019518899287155556,
+      "loss": 1.2898,
+      "step": 1751
+    },
+    {
+      "epoch": 0.12628847401427232,
+      "grad_norm": 0.10746128664860298,
+      "learning_rate": 0.0001951818357728792,
+      "loss": 1.3323,
+      "step": 1752
+    },
+    {
+      "epoch": 0.12636055647660924,
+      "grad_norm": 0.11260340145217203,
+      "learning_rate": 0.00019517467348594125,
+      "loss": 1.3198,
+      "step": 1753
+    },
+    {
+      "epoch": 0.12643263893894616,
+      "grad_norm": 0.12784100137039445,
+      "learning_rate": 0.00019516750601113216,
+      "loss": 1.371,
+      "step": 1754
+    },
+    {
+      "epoch": 0.12650472140128308,
+      "grad_norm": 0.1171384700111091,
+      "learning_rate": 0.00019516033334884258,
+      "loss": 1.3679,
+      "step": 1755
+    },
+    {
+      "epoch": 0.12657680386361997,
+      "grad_norm": 0.10693363516349122,
+      "learning_rate": 0.0001951531554994635,
+      "loss": 1.2715,
+      "step": 1756
+    },
+    {
+      "epoch": 0.1266488863259569,
+      "grad_norm": 0.17809002879551852,
+      "learning_rate": 0.00019514597246338621,
+      "loss": 1.3828,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1267209687882938,
+      "grad_norm": 0.11265727364289103,
+      "learning_rate": 0.0001951387842410022,
+      "loss": 1.3454,
+      "step": 1758
+    },
+    {
+      "epoch": 0.12679305125063073,
+      "grad_norm": 0.09802445748118163,
+      "learning_rate": 0.00019513159083270332,
+      "loss": 1.2952,
+      "step": 1759
+    },
+    {
+      "epoch": 0.12686513371296765,
+      "grad_norm": 0.1130626577860988,
+      "learning_rate": 0.00019512439223888167,
+      "loss": 1.3391,
+      "step": 1760
+    },
+    {
+      "epoch": 0.12693721617530454,
+      "grad_norm": 0.10040575889117825,
+      "learning_rate": 0.00019511718845992962,
+      "loss": 1.3792,
+      "step": 1761
+    },
+    {
+      "epoch": 0.12700929863764146,
+      "grad_norm": 0.10263219521377184,
+      "learning_rate": 0.00019510997949623987,
+      "loss": 1.2754,
+      "step": 1762
+    },
+    {
+      "epoch": 0.12708138109997838,
+      "grad_norm": 0.12247829553467048,
+      "learning_rate": 0.00019510276534820536,
+      "loss": 1.318,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1271534635623153,
+      "grad_norm": 0.13386818290523253,
+      "learning_rate": 0.0001950955460162193,
+      "loss": 1.4488,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1272255460246522,
+      "grad_norm": 0.1105327471797196,
+      "learning_rate": 0.00019508832150067526,
+      "loss": 1.3189,
+      "step": 1765
+    },
+    {
+      "epoch": 0.1272976284869891,
+      "grad_norm": 0.11397938387008737,
+      "learning_rate": 0.000195081091801967,
+      "loss": 1.4608,
+      "step": 1766
+    },
+    {
+      "epoch": 0.12736971094932603,
+      "grad_norm": 0.11171902929606926,
+      "learning_rate": 0.00019507385692048862,
+      "loss": 1.2871,
+      "step": 1767
+    },
+    {
+      "epoch": 0.12744179341166295,
+      "grad_norm": 0.11742275002311368,
+      "learning_rate": 0.0001950666168566345,
+      "loss": 1.2763,
+      "step": 1768
+    },
+    {
+      "epoch": 0.12751387587399984,
+      "grad_norm": 0.10618979096606172,
+      "learning_rate": 0.00019505937161079927,
+      "loss": 1.236,
+      "step": 1769
+    },
+    {
+      "epoch": 0.12758595833633676,
+      "grad_norm": 0.1650432073785699,
+      "learning_rate": 0.00019505212118337784,
+      "loss": 1.3979,
+      "step": 1770
+    },
+    {
+      "epoch": 0.12765804079867368,
+      "grad_norm": 0.11198508524485518,
+      "learning_rate": 0.00019504486557476544,
+      "loss": 1.4197,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1277301232610106,
+      "grad_norm": 0.10906760316117951,
+      "learning_rate": 0.0001950376047853576,
+      "loss": 1.3953,
+      "step": 1772
+    },
+    {
+      "epoch": 0.12780220572334752,
+      "grad_norm": 0.10836408814193048,
+      "learning_rate": 0.00019503033881555006,
+      "loss": 1.2585,
+      "step": 1773
+    },
+    {
+      "epoch": 0.12787428818568441,
+      "grad_norm": 0.10340423201821565,
+      "learning_rate": 0.00019502306766573888,
+      "loss": 1.2693,
+      "step": 1774
+    },
+    {
+      "epoch": 0.12794637064802133,
+      "grad_norm": 0.1129687701886528,
+      "learning_rate": 0.0001950157913363204,
+      "loss": 1.2959,
+      "step": 1775
+    },
+    {
+      "epoch": 0.12801845311035825,
+      "grad_norm": 0.17164798490657557,
+      "learning_rate": 0.00019500850982769128,
+      "loss": 1.283,
+      "step": 1776
+    },
+    {
+      "epoch": 0.12809053557269517,
+      "grad_norm": 0.12118493274788529,
+      "learning_rate": 0.0001950012231402484,
+      "loss": 1.3875,
+      "step": 1777
+    },
+    {
+      "epoch": 0.12816261803503207,
+      "grad_norm": 0.11214354831372285,
+      "learning_rate": 0.00019499393127438893,
+      "loss": 1.3283,
+      "step": 1778
+    },
+    {
+      "epoch": 0.128234700497369,
+      "grad_norm": 0.10437232893998302,
+      "learning_rate": 0.0001949866342305104,
+      "loss": 1.3142,
+      "step": 1779
+    },
+    {
+      "epoch": 0.1283067829597059,
+      "grad_norm": 0.12840310087817988,
+      "learning_rate": 0.0001949793320090105,
+      "loss": 1.3238,
+      "step": 1780
+    },
+    {
+      "epoch": 0.12837886542204283,
+      "grad_norm": 0.10654509246955919,
+      "learning_rate": 0.0001949720246102873,
+      "loss": 1.2732,
+      "step": 1781
+    },
+    {
+      "epoch": 0.12845094788437972,
+      "grad_norm": 0.10801826906185923,
+      "learning_rate": 0.00019496471203473914,
+      "loss": 1.3067,
+      "step": 1782
+    },
+    {
+      "epoch": 0.12852303034671664,
+      "grad_norm": 0.10745415061634206,
+      "learning_rate": 0.00019495739428276458,
+      "loss": 1.3581,
+      "step": 1783
+    },
+    {
+      "epoch": 0.12859511280905356,
+      "grad_norm": 0.11186419157576792,
+      "learning_rate": 0.0001949500713547625,
+      "loss": 1.2852,
+      "step": 1784
+    },
+    {
+      "epoch": 0.12866719527139048,
+      "grad_norm": 0.10469486012776459,
+      "learning_rate": 0.0001949427432511321,
+      "loss": 1.3107,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1287392777337274,
+      "grad_norm": 0.11079024731945109,
+      "learning_rate": 0.00019493540997227285,
+      "loss": 1.3932,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1288113601960643,
+      "grad_norm": 0.12050870744745334,
+      "learning_rate": 0.00019492807151858438,
+      "loss": 1.3617,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1288834426584012,
+      "grad_norm": 0.10559340537326488,
+      "learning_rate": 0.00019492072789046677,
+      "loss": 1.3964,
+      "step": 1788
+    },
+    {
+      "epoch": 0.12895552512073813,
+      "grad_norm": 0.11095941196066271,
+      "learning_rate": 0.00019491337908832031,
+      "loss": 1.3519,
+      "step": 1789
+    },
+    {
+      "epoch": 0.12902760758307505,
+      "grad_norm": 0.10739104334753341,
+      "learning_rate": 0.0001949060251125456,
+      "loss": 1.3816,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12909969004541194,
+      "grad_norm": 0.1075607994445208,
+      "learning_rate": 0.00019489866596354344,
+      "loss": 1.3424,
+      "step": 1791
+    },
+    {
+      "epoch": 0.12917177250774886,
+      "grad_norm": 0.10621814018707365,
+      "learning_rate": 0.000194891301641715,
+      "loss": 1.2596,
+      "step": 1792
+    },
+    {
+      "epoch": 0.12924385497008578,
+      "grad_norm": 0.1344237870318723,
+      "learning_rate": 0.00019488393214746173,
+      "loss": 1.3546,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1293159374324227,
+      "grad_norm": 0.10180361900827359,
+      "learning_rate": 0.00019487655748118526,
+      "loss": 1.4297,
+      "step": 1794
+    },
+    {
+      "epoch": 0.1293880198947596,
+      "grad_norm": 0.10871144719306301,
+      "learning_rate": 0.00019486917764328763,
+      "loss": 1.3103,
+      "step": 1795
+    },
+    {
+      "epoch": 0.1294601023570965,
+      "grad_norm": 0.09819376946711682,
+      "learning_rate": 0.0001948617926341711,
+      "loss": 1.2781,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12953218481943343,
+      "grad_norm": 0.10842805873068197,
+      "learning_rate": 0.0001948544024542382,
+      "loss": 1.3305,
+      "step": 1797
+    },
+    {
+      "epoch": 0.12960426728177035,
+      "grad_norm": 0.0969987666476302,
+      "learning_rate": 0.00019484700710389177,
+      "loss": 1.2583,
+      "step": 1798
+    },
+    {
+      "epoch": 0.12967634974410727,
+      "grad_norm": 0.11262819904106505,
+      "learning_rate": 0.00019483960658353496,
+      "loss": 1.271,
+      "step": 1799
+    },
+    {
+      "epoch": 0.12974843220644416,
+      "grad_norm": 0.10954148146176892,
+      "learning_rate": 0.0001948322008935711,
+      "loss": 1.4002,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12982051466878108,
+      "grad_norm": 0.11808873000605428,
+      "learning_rate": 0.00019482479003440394,
+      "loss": 1.2541,
+      "step": 1801
+    },
+    {
+      "epoch": 0.129892597131118,
+      "grad_norm": 0.14386284562183047,
+      "learning_rate": 0.00019481737400643735,
+      "loss": 1.266,
+      "step": 1802
+    },
+    {
+      "epoch": 0.12996467959345492,
+      "grad_norm": 0.10106534102806249,
+      "learning_rate": 0.00019480995281007565,
+      "loss": 1.4131,
+      "step": 1803
+    },
+    {
+      "epoch": 0.13003676205579182,
+      "grad_norm": 0.10819924740679665,
+      "learning_rate": 0.0001948025264457233,
+      "loss": 1.299,
+      "step": 1804
+    },
+    {
+      "epoch": 0.13010884451812874,
+      "grad_norm": 0.16152468895109193,
+      "learning_rate": 0.00019479509491378516,
+      "loss": 1.3729,
+      "step": 1805
+    },
+    {
+      "epoch": 0.13018092698046566,
+      "grad_norm": 0.11835165606702625,
+      "learning_rate": 0.00019478765821466623,
+      "loss": 1.4374,
+      "step": 1806
+    },
+    {
+      "epoch": 0.13025300944280258,
+      "grad_norm": 0.11770887922849421,
+      "learning_rate": 0.000194780216348772,
+      "loss": 1.4629,
+      "step": 1807
+    },
+    {
+      "epoch": 0.13032509190513947,
+      "grad_norm": 0.11735153006222582,
+      "learning_rate": 0.000194772769316508,
+      "loss": 1.2077,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1303971743674764,
+      "grad_norm": 0.14016352292270928,
+      "learning_rate": 0.00019476531711828027,
+      "loss": 1.4319,
+      "step": 1809
+    },
+    {
+      "epoch": 0.1304692568298133,
+      "grad_norm": 0.12195888433486819,
+      "learning_rate": 0.0001947578597544949,
+      "loss": 1.2816,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13054133929215023,
+      "grad_norm": 0.11173943477032461,
+      "learning_rate": 0.00019475039722555848,
+      "loss": 1.3331,
+      "step": 1811
+    },
+    {
+      "epoch": 0.13061342175448715,
+      "grad_norm": 0.11044227882340962,
+      "learning_rate": 0.00019474292953187774,
+      "loss": 1.3831,
+      "step": 1812
+    },
+    {
+      "epoch": 0.13068550421682404,
+      "grad_norm": 0.1304679856520771,
+      "learning_rate": 0.00019473545667385975,
+      "loss": 1.3368,
+      "step": 1813
+    },
+    {
+      "epoch": 0.13075758667916096,
+      "grad_norm": 0.10807331807110811,
+      "learning_rate": 0.00019472797865191182,
+      "loss": 1.2877,
+      "step": 1814
+    },
+    {
+      "epoch": 0.13082966914149788,
+      "grad_norm": 0.11020659231705268,
+      "learning_rate": 0.00019472049546644162,
+      "loss": 1.3857,
+      "step": 1815
+    },
+    {
+      "epoch": 0.1309017516038348,
+      "grad_norm": 0.1142328793991807,
+      "learning_rate": 0.00019471300711785703,
+      "loss": 1.382,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1309738340661717,
+      "grad_norm": 0.09754090511409326,
+      "learning_rate": 0.00019470551360656623,
+      "loss": 1.4477,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1310459165285086,
+      "grad_norm": 0.1409679853502817,
+      "learning_rate": 0.00019469801493297766,
+      "loss": 1.3745,
+      "step": 1818
+    },
+    {
+      "epoch": 0.13111799899084553,
+      "grad_norm": 0.1841694600716793,
+      "learning_rate": 0.00019469051109750006,
+      "loss": 1.267,
+      "step": 1819
+    },
+    {
+      "epoch": 0.13119008145318245,
+      "grad_norm": 0.11284399320524056,
+      "learning_rate": 0.00019468300210054252,
+      "loss": 1.3527,
+      "step": 1820
+    },
+    {
+      "epoch": 0.13126216391551934,
+      "grad_norm": 0.11642633738039594,
+      "learning_rate": 0.00019467548794251428,
+      "loss": 1.267,
+      "step": 1821
+    },
+    {
+      "epoch": 0.13133424637785626,
+      "grad_norm": 0.10238513260201788,
+      "learning_rate": 0.00019466796862382496,
+      "loss": 1.3015,
+      "step": 1822
+    },
+    {
+      "epoch": 0.13140632884019318,
+      "grad_norm": 0.1234778000252857,
+      "learning_rate": 0.00019466044414488445,
+      "loss": 1.4367,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1314784113025301,
+      "grad_norm": 0.1179658432109079,
+      "learning_rate": 0.00019465291450610286,
+      "loss": 1.3388,
+      "step": 1824
+    },
+    {
+      "epoch": 0.13155049376486702,
+      "grad_norm": 0.112039127796058,
+      "learning_rate": 0.00019464537970789066,
+      "loss": 1.2346,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1316225762272039,
+      "grad_norm": 0.10468765338544479,
+      "learning_rate": 0.0001946378397506585,
+      "loss": 1.3022,
+      "step": 1826
+    },
+    {
+      "epoch": 0.13169465868954083,
+      "grad_norm": 0.2286921684164543,
+      "learning_rate": 0.00019463029463481746,
+      "loss": 1.3289,
+      "step": 1827
+    },
+    {
+      "epoch": 0.13176674115187775,
+      "grad_norm": 0.10603684046280913,
+      "learning_rate": 0.00019462274436077875,
+      "loss": 1.2577,
+      "step": 1828
+    },
+    {
+      "epoch": 0.13183882361421467,
+      "grad_norm": 0.11453837286682396,
+      "learning_rate": 0.00019461518892895395,
+      "loss": 1.4216,
+      "step": 1829
+    },
+    {
+      "epoch": 0.13191090607655157,
+      "grad_norm": 0.11379832503407467,
+      "learning_rate": 0.00019460762833975492,
+      "loss": 1.2987,
+      "step": 1830
+    },
+    {
+      "epoch": 0.13198298853888848,
+      "grad_norm": 0.1102535171421827,
+      "learning_rate": 0.00019460006259359376,
+      "loss": 1.3652,
+      "step": 1831
+    },
+    {
+      "epoch": 0.1320550710012254,
+      "grad_norm": 0.1161824540016957,
+      "learning_rate": 0.00019459249169088288,
+      "loss": 1.3395,
+      "step": 1832
+    },
+    {
+      "epoch": 0.13212715346356232,
+      "grad_norm": 0.11129202272325166,
+      "learning_rate": 0.00019458491563203493,
+      "loss": 1.2461,
+      "step": 1833
+    },
+    {
+      "epoch": 0.13219923592589922,
+      "grad_norm": 0.10899922932424629,
+      "learning_rate": 0.00019457733441746294,
+      "loss": 1.2336,
+      "step": 1834
+    },
+    {
+      "epoch": 0.13227131838823614,
+      "grad_norm": 0.10787341447385047,
+      "learning_rate": 0.00019456974804758006,
+      "loss": 1.3768,
+      "step": 1835
+    },
+    {
+      "epoch": 0.13234340085057306,
+      "grad_norm": 0.11369865432843519,
+      "learning_rate": 0.0001945621565227999,
+      "loss": 1.3324,
+      "step": 1836
+    },
+    {
+      "epoch": 0.13241548331290998,
+      "grad_norm": 0.10757970526514798,
+      "learning_rate": 0.00019455455984353622,
+      "loss": 1.2454,
+      "step": 1837
+    },
+    {
+      "epoch": 0.1324875657752469,
+      "grad_norm": 0.09646505267761679,
+      "learning_rate": 0.0001945469580102031,
+      "loss": 1.3602,
+      "step": 1838
+    },
+    {
+      "epoch": 0.1325596482375838,
+      "grad_norm": 0.12227717173844209,
+      "learning_rate": 0.00019453935102321497,
+      "loss": 1.2998,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1326317306999207,
+      "grad_norm": 0.12282596735591543,
+      "learning_rate": 0.0001945317388829864,
+      "loss": 1.4056,
+      "step": 1840
+    },
+    {
+      "epoch": 0.13270381316225763,
+      "grad_norm": 0.1301067062657616,
+      "learning_rate": 0.00019452412158993236,
+      "loss": 1.3453,
+      "step": 1841
+    },
+    {
+      "epoch": 0.13277589562459455,
+      "grad_norm": 0.1026780099274762,
+      "learning_rate": 0.00019451649914446807,
+      "loss": 1.3415,
+      "step": 1842
+    },
+    {
+      "epoch": 0.13284797808693144,
+      "grad_norm": 0.10962186322313121,
+      "learning_rate": 0.00019450887154700902,
+      "loss": 1.2956,
+      "step": 1843
+    },
+    {
+      "epoch": 0.13292006054926836,
+      "grad_norm": 0.11296831166793551,
+      "learning_rate": 0.00019450123879797095,
+      "loss": 1.3204,
+      "step": 1844
+    },
+    {
+      "epoch": 0.13299214301160528,
+      "grad_norm": 0.10188240908828503,
+      "learning_rate": 0.00019449360089776995,
+      "loss": 1.2364,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1330642254739422,
+      "grad_norm": 0.10363673953450851,
+      "learning_rate": 0.00019448595784682228,
+      "loss": 1.2903,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1331363079362791,
+      "grad_norm": 0.10222360486562598,
+      "learning_rate": 0.00019447830964554468,
+      "loss": 1.433,
+      "step": 1847
+    },
+    {
+      "epoch": 0.133208390398616,
+      "grad_norm": 0.13207306817526585,
+      "learning_rate": 0.00019447065629435396,
+      "loss": 1.3767,
+      "step": 1848
+    },
+    {
+      "epoch": 0.13328047286095293,
+      "grad_norm": 0.10102573177506016,
+      "learning_rate": 0.0001944629977936673,
+      "loss": 1.314,
+      "step": 1849
+    },
+    {
+      "epoch": 0.13335255532328985,
+      "grad_norm": 0.1451555977154735,
+      "learning_rate": 0.00019445533414390214,
+      "loss": 1.3625,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13342463778562677,
+      "grad_norm": 0.1044464750421803,
+      "learning_rate": 0.00019444766534547632,
+      "loss": 1.286,
+      "step": 1851
+    },
+    {
+      "epoch": 0.13349672024796366,
+      "grad_norm": 0.12976454652404554,
+      "learning_rate": 0.00019443999139880775,
+      "loss": 1.3991,
+      "step": 1852
+    },
+    {
+      "epoch": 0.13356880271030058,
+      "grad_norm": 0.1486351066863439,
+      "learning_rate": 0.00019443231230431476,
+      "loss": 1.3151,
+      "step": 1853
+    },
+    {
+      "epoch": 0.1336408851726375,
+      "grad_norm": 0.10228527373718951,
+      "learning_rate": 0.00019442462806241594,
+      "loss": 1.251,
+      "step": 1854
+    },
+    {
+      "epoch": 0.13371296763497442,
+      "grad_norm": 0.12895732895791032,
+      "learning_rate": 0.00019441693867353012,
+      "loss": 1.4374,
+      "step": 1855
+    },
+    {
+      "epoch": 0.13378505009731131,
+      "grad_norm": 0.20733786897691567,
+      "learning_rate": 0.0001944092441380765,
+      "loss": 1.3177,
+      "step": 1856
+    },
+    {
+      "epoch": 0.13385713255964823,
+      "grad_norm": 0.11328043803269543,
+      "learning_rate": 0.00019440154445647445,
+      "loss": 1.2951,
+      "step": 1857
+    },
+    {
+      "epoch": 0.13392921502198515,
+      "grad_norm": 0.21825418339246325,
+      "learning_rate": 0.00019439383962914367,
+      "loss": 1.3109,
+      "step": 1858
+    },
+    {
+      "epoch": 0.13400129748432207,
+      "grad_norm": 0.10929147223304937,
+      "learning_rate": 0.0001943861296565042,
+      "loss": 1.2906,
+      "step": 1859
+    },
+    {
+      "epoch": 0.13407337994665897,
+      "grad_norm": 0.18048956653252785,
+      "learning_rate": 0.00019437841453897624,
+      "loss": 1.3647,
+      "step": 1860
+    },
+    {
+      "epoch": 0.13414546240899589,
+      "grad_norm": 0.1102277099258072,
+      "learning_rate": 0.00019437069427698033,
+      "loss": 1.3378,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1342175448713328,
+      "grad_norm": 0.10705361584015215,
+      "learning_rate": 0.00019436296887093737,
+      "loss": 1.278,
+      "step": 1862
+    },
+    {
+      "epoch": 0.13428962733366973,
+      "grad_norm": 0.12578285060585234,
+      "learning_rate": 0.00019435523832126835,
+      "loss": 1.2876,
+      "step": 1863
+    },
+    {
+      "epoch": 0.13436170979600665,
+      "grad_norm": 0.11491347829271599,
+      "learning_rate": 0.00019434750262839476,
+      "loss": 1.3587,
+      "step": 1864
+    },
+    {
+      "epoch": 0.13443379225834354,
+      "grad_norm": 0.09967837066587161,
+      "learning_rate": 0.0001943397617927382,
+      "loss": 1.3152,
+      "step": 1865
+    },
+    {
+      "epoch": 0.13450587472068046,
+      "grad_norm": 0.13326420608593637,
+      "learning_rate": 0.00019433201581472065,
+      "loss": 1.3123,
+      "step": 1866
+    },
+    {
+      "epoch": 0.13457795718301738,
+      "grad_norm": 0.11126383174063818,
+      "learning_rate": 0.0001943242646947643,
+      "loss": 1.3721,
+      "step": 1867
+    },
+    {
+      "epoch": 0.1346500396453543,
+      "grad_norm": 0.1572912515445738,
+      "learning_rate": 0.00019431650843329169,
+      "loss": 1.2854,
+      "step": 1868
+    },
+    {
+      "epoch": 0.1347221221076912,
+      "grad_norm": 0.11164606828332127,
+      "learning_rate": 0.00019430874703072558,
+      "loss": 1.3336,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1347942045700281,
+      "grad_norm": 0.10483989584528879,
+      "learning_rate": 0.00019430098048748906,
+      "loss": 1.3579,
+      "step": 1870
+    },
+    {
+      "epoch": 0.13486628703236503,
+      "grad_norm": 0.10509209144497367,
+      "learning_rate": 0.00019429320880400545,
+      "loss": 1.3702,
+      "step": 1871
+    },
+    {
+      "epoch": 0.13493836949470195,
+      "grad_norm": 0.10611788503886341,
+      "learning_rate": 0.0001942854319806984,
+      "loss": 1.3563,
+      "step": 1872
+    },
+    {
+      "epoch": 0.13501045195703884,
+      "grad_norm": 0.10390410019547112,
+      "learning_rate": 0.0001942776500179918,
+      "loss": 1.316,
+      "step": 1873
+    },
+    {
+      "epoch": 0.13508253441937576,
+      "grad_norm": 0.10510445530315385,
+      "learning_rate": 0.00019426986291630984,
+      "loss": 1.2927,
+      "step": 1874
+    },
+    {
+      "epoch": 0.13515461688171268,
+      "grad_norm": 0.11564369787090018,
+      "learning_rate": 0.00019426207067607697,
+      "loss": 1.3487,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1352266993440496,
+      "grad_norm": 0.10732837990704194,
+      "learning_rate": 0.000194254273297718,
+      "loss": 1.3773,
+      "step": 1876
+    },
+    {
+      "epoch": 0.13529878180638652,
+      "grad_norm": 0.22039279260879008,
+      "learning_rate": 0.0001942464707816579,
+      "loss": 1.3094,
+      "step": 1877
+    },
+    {
+      "epoch": 0.1353708642687234,
+      "grad_norm": 0.18584726782047845,
+      "learning_rate": 0.00019423866312832195,
+      "loss": 1.1937,
+      "step": 1878
+    },
+    {
+      "epoch": 0.13544294673106033,
+      "grad_norm": 0.11841274410340044,
+      "learning_rate": 0.0001942308503381358,
+      "loss": 1.3812,
+      "step": 1879
+    },
+    {
+      "epoch": 0.13551502919339725,
+      "grad_norm": 0.11205799420869467,
+      "learning_rate": 0.00019422303241152533,
+      "loss": 1.2347,
+      "step": 1880
+    },
+    {
+      "epoch": 0.13558711165573417,
+      "grad_norm": 0.10871173184593645,
+      "learning_rate": 0.0001942152093489166,
+      "loss": 1.3048,
+      "step": 1881
+    },
+    {
+      "epoch": 0.13565919411807106,
+      "grad_norm": 0.10738305474241192,
+      "learning_rate": 0.00019420738115073611,
+      "loss": 1.3648,
+      "step": 1882
+    },
+    {
+      "epoch": 0.13573127658040798,
+      "grad_norm": 0.1212548038011961,
+      "learning_rate": 0.00019419954781741054,
+      "loss": 1.3438,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1358033590427449,
+      "grad_norm": 0.12310941377603063,
+      "learning_rate": 0.00019419170934936687,
+      "loss": 1.348,
+      "step": 1884
+    },
+    {
+      "epoch": 0.13587544150508182,
+      "grad_norm": 0.09782858370159653,
+      "learning_rate": 0.0001941838657470324,
+      "loss": 1.3605,
+      "step": 1885
+    },
+    {
+      "epoch": 0.13594752396741872,
+      "grad_norm": 0.10756090646535298,
+      "learning_rate": 0.00019417601701083464,
+      "loss": 1.2808,
+      "step": 1886
+    },
+    {
+      "epoch": 0.13601960642975564,
+      "grad_norm": 0.11828506885319998,
+      "learning_rate": 0.00019416816314120144,
+      "loss": 1.3315,
+      "step": 1887
+    },
+    {
+      "epoch": 0.13609168889209255,
+      "grad_norm": 0.11619080724532203,
+      "learning_rate": 0.00019416030413856088,
+      "loss": 1.2653,
+      "step": 1888
+    },
+    {
+      "epoch": 0.13616377135442947,
+      "grad_norm": 0.11598043219043239,
+      "learning_rate": 0.0001941524400033414,
+      "loss": 1.1713,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1362358538167664,
+      "grad_norm": 0.23947548702986868,
+      "learning_rate": 0.00019414457073597158,
+      "loss": 1.4311,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1363079362791033,
+      "grad_norm": 0.10026341330504983,
+      "learning_rate": 0.00019413669633688044,
+      "loss": 1.2614,
+      "step": 1891
+    },
+    {
+      "epoch": 0.1363800187414402,
+      "grad_norm": 0.11958635137529869,
+      "learning_rate": 0.00019412881680649715,
+      "loss": 1.3151,
+      "step": 1892
+    },
+    {
+      "epoch": 0.13645210120377713,
+      "grad_norm": 0.20286438630389428,
+      "learning_rate": 0.00019412093214525128,
+      "loss": 1.3122,
+      "step": 1893
+    },
+    {
+      "epoch": 0.13652418366611405,
+      "grad_norm": 0.1115322069355837,
+      "learning_rate": 0.00019411304235357254,
+      "loss": 1.4001,
+      "step": 1894
+    },
+    {
+      "epoch": 0.13659626612845094,
+      "grad_norm": 0.1285465365408839,
+      "learning_rate": 0.00019410514743189108,
+      "loss": 1.304,
+      "step": 1895
+    },
+    {
+      "epoch": 0.13666834859078786,
+      "grad_norm": 0.11725060097424538,
+      "learning_rate": 0.00019409724738063714,
+      "loss": 1.3679,
+      "step": 1896
+    },
+    {
+      "epoch": 0.13674043105312478,
+      "grad_norm": 0.12147335311337944,
+      "learning_rate": 0.00019408934220024144,
+      "loss": 1.3533,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1368125135154617,
+      "grad_norm": 0.10417712684334854,
+      "learning_rate": 0.00019408143189113484,
+      "loss": 1.3269,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1368845959777986,
+      "grad_norm": 0.10683720207667145,
+      "learning_rate": 0.0001940735164537485,
+      "loss": 1.3388,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1369566784401355,
+      "grad_norm": 0.12348676146227135,
+      "learning_rate": 0.00019406559588851395,
+      "loss": 1.3266,
+      "step": 1900
+    },
+    {
+      "epoch": 0.13702876090247243,
+      "grad_norm": 0.13102819145466904,
+      "learning_rate": 0.00019405767019586284,
+      "loss": 1.3946,
+      "step": 1901
+    },
+    {
+      "epoch": 0.13710084336480935,
+      "grad_norm": 0.11160825461946527,
+      "learning_rate": 0.00019404973937622727,
+      "loss": 1.3702,
+      "step": 1902
+    },
+    {
+      "epoch": 0.13717292582714624,
+      "grad_norm": 0.11474922365624378,
+      "learning_rate": 0.00019404180343003953,
+      "loss": 1.283,
+      "step": 1903
+    },
+    {
+      "epoch": 0.13724500828948316,
+      "grad_norm": 0.10402887597021855,
+      "learning_rate": 0.00019403386235773216,
+      "loss": 1.3619,
+      "step": 1904
+    },
+    {
+      "epoch": 0.13731709075182008,
+      "grad_norm": 0.1234078008777589,
+      "learning_rate": 0.000194025916159738,
+      "loss": 1.262,
+      "step": 1905
+    },
+    {
+      "epoch": 0.137389173214157,
+      "grad_norm": 0.10249042352161258,
+      "learning_rate": 0.0001940179648364903,
+      "loss": 1.2662,
+      "step": 1906
+    },
+    {
+      "epoch": 0.13746125567649392,
+      "grad_norm": 0.10497965801428433,
+      "learning_rate": 0.00019401000838842238,
+      "loss": 1.3041,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1375333381388308,
+      "grad_norm": 0.1301719276609821,
+      "learning_rate": 0.00019400204681596796,
+      "loss": 1.3024,
+      "step": 1908
+    },
+    {
+      "epoch": 0.13760542060116773,
+      "grad_norm": 0.22768803721094757,
+      "learning_rate": 0.00019399408011956098,
+      "loss": 1.3506,
+      "step": 1909
+    },
+    {
+      "epoch": 0.13767750306350465,
+      "grad_norm": 0.0959211924018738,
+      "learning_rate": 0.00019398610829963583,
+      "loss": 1.3274,
+      "step": 1910
+    },
+    {
+      "epoch": 0.13774958552584157,
+      "grad_norm": 0.10011587183905456,
+      "learning_rate": 0.00019397813135662692,
+      "loss": 1.3739,
+      "step": 1911
+    },
+    {
+      "epoch": 0.13782166798817846,
+      "grad_norm": 0.11082587660470405,
+      "learning_rate": 0.00019397014929096903,
+      "loss": 1.3156,
+      "step": 1912
+    },
+    {
+      "epoch": 0.13789375045051538,
+      "grad_norm": 0.09580607691722823,
+      "learning_rate": 0.0001939621621030974,
+      "loss": 1.3671,
+      "step": 1913
+    },
+    {
+      "epoch": 0.1379658329128523,
+      "grad_norm": 0.11341307586975286,
+      "learning_rate": 0.00019395416979344732,
+      "loss": 1.2754,
+      "step": 1914
+    },
+    {
+      "epoch": 0.13803791537518922,
+      "grad_norm": 0.1169442352058445,
+      "learning_rate": 0.0001939461723624544,
+      "loss": 1.4326,
+      "step": 1915
+    },
+    {
+      "epoch": 0.13810999783752612,
+      "grad_norm": 0.10834086773885362,
+      "learning_rate": 0.00019393816981055467,
+      "loss": 1.2174,
+      "step": 1916
+    },
+    {
+      "epoch": 0.13818208029986304,
+      "grad_norm": 0.10809782900962062,
+      "learning_rate": 0.00019393016213818423,
+      "loss": 1.273,
+      "step": 1917
+    },
+    {
+      "epoch": 0.13825416276219996,
+      "grad_norm": 0.11411124712610557,
+      "learning_rate": 0.0001939221493457797,
+      "loss": 1.3113,
+      "step": 1918
+    },
+    {
+      "epoch": 0.13832624522453688,
+      "grad_norm": 0.11808493440807503,
+      "learning_rate": 0.00019391413143377773,
+      "loss": 1.3396,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1383983276868738,
+      "grad_norm": 0.10461370769524361,
+      "learning_rate": 0.00019390610840261548,
+      "loss": 1.3686,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1384704101492107,
+      "grad_norm": 0.11326892919157289,
+      "learning_rate": 0.00019389808025273015,
+      "loss": 1.3889,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1385424926115476,
+      "grad_norm": 0.09656182018985583,
+      "learning_rate": 0.00019389004698455945,
+      "loss": 1.299,
+      "step": 1922
+    },
+    {
+      "epoch": 0.13861457507388453,
+      "grad_norm": 0.15378373978153656,
+      "learning_rate": 0.0001938820085985412,
+      "loss": 1.2582,
+      "step": 1923
+    },
+    {
+      "epoch": 0.13868665753622145,
+      "grad_norm": 0.1056665786550205,
+      "learning_rate": 0.00019387396509511359,
+      "loss": 1.2896,
+      "step": 1924
+    },
+    {
+      "epoch": 0.13875873999855834,
+      "grad_norm": 0.09221678889090956,
+      "learning_rate": 0.00019386591647471506,
+      "loss": 1.2444,
+      "step": 1925
+    },
+    {
+      "epoch": 0.13883082246089526,
+      "grad_norm": 0.11511890051019838,
+      "learning_rate": 0.00019385786273778434,
+      "loss": 1.3112,
+      "step": 1926
+    },
+    {
+      "epoch": 0.13890290492323218,
+      "grad_norm": 0.12589671225680293,
+      "learning_rate": 0.00019384980388476044,
+      "loss": 1.3012,
+      "step": 1927
+    },
+    {
+      "epoch": 0.1389749873855691,
+      "grad_norm": 0.11434280756238428,
+      "learning_rate": 0.0001938417399160826,
+      "loss": 1.3242,
+      "step": 1928
+    },
+    {
+      "epoch": 0.139047069847906,
+      "grad_norm": 0.14861001253698103,
+      "learning_rate": 0.0001938336708321904,
+      "loss": 1.3028,
+      "step": 1929
+    },
+    {
+      "epoch": 0.1391191523102429,
+      "grad_norm": 0.11375308264052833,
+      "learning_rate": 0.0001938255966335237,
+      "loss": 1.2552,
+      "step": 1930
+    },
+    {
+      "epoch": 0.13919123477257983,
+      "grad_norm": 0.12407428434004242,
+      "learning_rate": 0.0001938175173205226,
+      "loss": 1.3244,
+      "step": 1931
+    },
+    {
+      "epoch": 0.13926331723491675,
+      "grad_norm": 0.11083988050948718,
+      "learning_rate": 0.00019380943289362747,
+      "loss": 1.4292,
+      "step": 1932
+    },
+    {
+      "epoch": 0.13933539969725367,
+      "grad_norm": 0.1560601862644791,
+      "learning_rate": 0.000193801343353279,
+      "loss": 1.2932,
+      "step": 1933
+    },
+    {
+      "epoch": 0.13940748215959056,
+      "grad_norm": 0.24381051277046972,
+      "learning_rate": 0.00019379324869991816,
+      "loss": 1.3165,
+      "step": 1934
+    },
+    {
+      "epoch": 0.13947956462192748,
+      "grad_norm": 0.213472511770603,
+      "learning_rate": 0.00019378514893398616,
+      "loss": 1.3851,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1395516470842644,
+      "grad_norm": 0.1101562197139619,
+      "learning_rate": 0.00019377704405592453,
+      "loss": 1.2904,
+      "step": 1936
+    },
+    {
+      "epoch": 0.13962372954660132,
+      "grad_norm": 0.11198971040890927,
+      "learning_rate": 0.00019376893406617503,
+      "loss": 1.3005,
+      "step": 1937
+    },
+    {
+      "epoch": 0.1396958120089382,
+      "grad_norm": 0.12890288801674296,
+      "learning_rate": 0.00019376081896517976,
+      "loss": 1.3036,
+      "step": 1938
+    },
+    {
+      "epoch": 0.13976789447127513,
+      "grad_norm": 0.10538529700763284,
+      "learning_rate": 0.00019375269875338105,
+      "loss": 1.3154,
+      "step": 1939
+    },
+    {
+      "epoch": 0.13983997693361205,
+      "grad_norm": 0.13063883890080666,
+      "learning_rate": 0.00019374457343122149,
+      "loss": 1.3112,
+      "step": 1940
+    },
+    {
+      "epoch": 0.13991205939594897,
+      "grad_norm": 0.09974258227403905,
+      "learning_rate": 0.00019373644299914403,
+      "loss": 1.3688,
+      "step": 1941
+    },
+    {
+      "epoch": 0.13998414185828587,
+      "grad_norm": 0.14192615298546565,
+      "learning_rate": 0.00019372830745759186,
+      "loss": 1.3683,
+      "step": 1942
+    },
+    {
+      "epoch": 0.14005622432062279,
+      "grad_norm": 0.11154573073710136,
+      "learning_rate": 0.00019372016680700837,
+      "loss": 1.3217,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1401283067829597,
+      "grad_norm": 0.1152038513535374,
+      "learning_rate": 0.0001937120210478374,
+      "loss": 1.2269,
+      "step": 1944
+    },
+    {
+      "epoch": 0.14020038924529662,
+      "grad_norm": 0.10728201153821974,
+      "learning_rate": 0.0001937038701805229,
+      "loss": 1.3105,
+      "step": 1945
+    },
+    {
+      "epoch": 0.14027247170763354,
+      "grad_norm": 0.11186691670578686,
+      "learning_rate": 0.00019369571420550913,
+      "loss": 1.3613,
+      "step": 1946
+    },
+    {
+      "epoch": 0.14034455416997044,
+      "grad_norm": 0.14522856567316386,
+      "learning_rate": 0.00019368755312324076,
+      "loss": 1.4012,
+      "step": 1947
+    },
+    {
+      "epoch": 0.14041663663230736,
+      "grad_norm": 0.18069224338279785,
+      "learning_rate": 0.00019367938693416257,
+      "loss": 1.1935,
+      "step": 1948
+    },
+    {
+      "epoch": 0.14048871909464428,
+      "grad_norm": 0.11378158482250438,
+      "learning_rate": 0.0001936712156387197,
+      "loss": 1.3488,
+      "step": 1949
+    },
+    {
+      "epoch": 0.1405608015569812,
+      "grad_norm": 0.11170357420624336,
+      "learning_rate": 0.0001936630392373576,
+      "loss": 1.2966,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1406328840193181,
+      "grad_norm": 0.09751933395045435,
+      "learning_rate": 0.0001936548577305219,
+      "loss": 1.3731,
+      "step": 1951
+    },
+    {
+      "epoch": 0.140704966481655,
+      "grad_norm": 0.1111543994232747,
+      "learning_rate": 0.00019364667111865862,
+      "loss": 1.3337,
+      "step": 1952
+    },
+    {
+      "epoch": 0.14077704894399193,
+      "grad_norm": 0.10632684570771034,
+      "learning_rate": 0.00019363847940221396,
+      "loss": 1.3619,
+      "step": 1953
+    },
+    {
+      "epoch": 0.14084913140632885,
+      "grad_norm": 0.10466364114754108,
+      "learning_rate": 0.00019363028258163447,
+      "loss": 1.3279,
+      "step": 1954
+    },
+    {
+      "epoch": 0.14092121386866574,
+      "grad_norm": 0.12106526167292192,
+      "learning_rate": 0.00019362208065736695,
+      "loss": 1.3219,
+      "step": 1955
+    },
+    {
+      "epoch": 0.14099329633100266,
+      "grad_norm": 0.09675104101825237,
+      "learning_rate": 0.0001936138736298584,
+      "loss": 1.4217,
+      "step": 1956
+    },
+    {
+      "epoch": 0.14106537879333958,
+      "grad_norm": 0.11880420181192265,
+      "learning_rate": 0.00019360566149955632,
+      "loss": 1.4052,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1411374612556765,
+      "grad_norm": 0.12058503937496338,
+      "learning_rate": 0.00019359744426690822,
+      "loss": 1.3022,
+      "step": 1958
+    },
+    {
+      "epoch": 0.14120954371801342,
+      "grad_norm": 0.10397917550772211,
+      "learning_rate": 0.00019358922193236212,
+      "loss": 1.2848,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1412816261803503,
+      "grad_norm": 0.1031864329010215,
+      "learning_rate": 0.00019358099449636612,
+      "loss": 1.3545,
+      "step": 1960
+    },
+    {
+      "epoch": 0.14135370864268723,
+      "grad_norm": 0.11721267409861319,
+      "learning_rate": 0.00019357276195936868,
+      "loss": 1.3932,
+      "step": 1961
+    },
+    {
+      "epoch": 0.14142579110502415,
+      "grad_norm": 0.11183397720789832,
+      "learning_rate": 0.00019356452432181862,
+      "loss": 1.2408,
+      "step": 1962
+    },
+    {
+      "epoch": 0.14149787356736107,
+      "grad_norm": 0.1833039671052304,
+      "learning_rate": 0.00019355628158416495,
+      "loss": 1.2981,
+      "step": 1963
+    },
+    {
+      "epoch": 0.14156995602969796,
+      "grad_norm": 0.10980635663420496,
+      "learning_rate": 0.00019354803374685694,
+      "loss": 1.3267,
+      "step": 1964
+    },
+    {
+      "epoch": 0.14164203849203488,
+      "grad_norm": 0.10161428305486261,
+      "learning_rate": 0.00019353978081034417,
+      "loss": 1.3773,
+      "step": 1965
+    },
+    {
+      "epoch": 0.1417141209543718,
+      "grad_norm": 0.11450097072162077,
+      "learning_rate": 0.00019353152277507652,
+      "loss": 1.3472,
+      "step": 1966
+    },
+    {
+      "epoch": 0.14178620341670872,
+      "grad_norm": 0.10729522493385209,
+      "learning_rate": 0.00019352325964150415,
+      "loss": 1.3249,
+      "step": 1967
+    },
+    {
+      "epoch": 0.14185828587904561,
+      "grad_norm": 0.11687158260473467,
+      "learning_rate": 0.00019351499141007742,
+      "loss": 1.4096,
+      "step": 1968
+    },
+    {
+      "epoch": 0.14193036834138253,
+      "grad_norm": 0.13119185332406585,
+      "learning_rate": 0.00019350671808124705,
+      "loss": 1.3702,
+      "step": 1969
+    },
+    {
+      "epoch": 0.14200245080371945,
+      "grad_norm": 0.10416626053101415,
+      "learning_rate": 0.00019349843965546404,
+      "loss": 1.3951,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14207453326605637,
+      "grad_norm": 0.09504957263949101,
+      "learning_rate": 0.00019349015613317957,
+      "loss": 1.3336,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1421466157283933,
+      "grad_norm": 0.1660732864140865,
+      "learning_rate": 0.00019348186751484524,
+      "loss": 1.3413,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1422186981907302,
+      "grad_norm": 0.13654934378471265,
+      "learning_rate": 0.00019347357380091279,
+      "loss": 1.3083,
+      "step": 1973
+    },
+    {
+      "epoch": 0.1422907806530671,
+      "grad_norm": 0.11986160869056624,
+      "learning_rate": 0.00019346527499183432,
+      "loss": 1.4014,
+      "step": 1974
+    },
+    {
+      "epoch": 0.14236286311540403,
+      "grad_norm": 0.10485102678392541,
+      "learning_rate": 0.00019345697108806223,
+      "loss": 1.3858,
+      "step": 1975
+    },
+    {
+      "epoch": 0.14243494557774095,
+      "grad_norm": 0.314057366373724,
+      "learning_rate": 0.00019344866209004908,
+      "loss": 1.2991,
+      "step": 1976
+    },
+    {
+      "epoch": 0.14250702804007784,
+      "grad_norm": 0.11931500121728725,
+      "learning_rate": 0.00019344034799824789,
+      "loss": 1.3484,
+      "step": 1977
+    },
+    {
+      "epoch": 0.14257911050241476,
+      "grad_norm": 0.10435593445042969,
+      "learning_rate": 0.00019343202881311173,
+      "loss": 1.3263,
+      "step": 1978
+    },
+    {
+      "epoch": 0.14265119296475168,
+      "grad_norm": 0.12319088024795223,
+      "learning_rate": 0.00019342370453509419,
+      "loss": 1.3037,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1427232754270886,
+      "grad_norm": 0.11528485219161254,
+      "learning_rate": 0.00019341537516464894,
+      "loss": 1.3314,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1427953578894255,
+      "grad_norm": 0.12635939513368633,
+      "learning_rate": 0.00019340704070223,
+      "loss": 1.4026,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1428674403517624,
+      "grad_norm": 0.10243810476379994,
+      "learning_rate": 0.00019339870114829175,
+      "loss": 1.273,
+      "step": 1982
+    },
+    {
+      "epoch": 0.14293952281409933,
+      "grad_norm": 0.10243040542415624,
+      "learning_rate": 0.00019339035650328869,
+      "loss": 1.4091,
+      "step": 1983
+    },
+    {
+      "epoch": 0.14301160527643625,
+      "grad_norm": 0.1384950534283273,
+      "learning_rate": 0.0001933820067676757,
+      "loss": 1.2805,
+      "step": 1984
+    },
+    {
+      "epoch": 0.14308368773877317,
+      "grad_norm": 0.10623576437996451,
+      "learning_rate": 0.0001933736519419079,
+      "loss": 1.2598,
+      "step": 1985
+    },
+    {
+      "epoch": 0.14315577020111006,
+      "grad_norm": 0.12454533579176895,
+      "learning_rate": 0.0001933652920264408,
+      "loss": 1.3999,
+      "step": 1986
+    },
+    {
+      "epoch": 0.14322785266344698,
+      "grad_norm": 0.11286461535616962,
+      "learning_rate": 0.00019335692702172994,
+      "loss": 1.3156,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1432999351257839,
+      "grad_norm": 0.11220436303123618,
+      "learning_rate": 0.0001933485569282314,
+      "loss": 1.2708,
+      "step": 1988
+    },
+    {
+      "epoch": 0.14337201758812082,
+      "grad_norm": 0.1171501462468315,
+      "learning_rate": 0.0001933401817464014,
+      "loss": 1.3457,
+      "step": 1989
+    },
+    {
+      "epoch": 0.1434441000504577,
+      "grad_norm": 0.19727765983066747,
+      "learning_rate": 0.0001933318014766964,
+      "loss": 1.3755,
+      "step": 1990
+    },
+    {
+      "epoch": 0.14351618251279463,
+      "grad_norm": 0.15597299244160895,
+      "learning_rate": 0.00019332341611957325,
+      "loss": 1.3935,
+      "step": 1991
+    },
+    {
+      "epoch": 0.14358826497513155,
+      "grad_norm": 0.13676765939281196,
+      "learning_rate": 0.00019331502567548908,
+      "loss": 1.3062,
+      "step": 1992
+    },
+    {
+      "epoch": 0.14366034743746847,
+      "grad_norm": 0.20870801438569503,
+      "learning_rate": 0.00019330663014490115,
+      "loss": 1.3054,
+      "step": 1993
+    },
+    {
+      "epoch": 0.14373242989980536,
+      "grad_norm": 0.1162908311799796,
+      "learning_rate": 0.00019329822952826714,
+      "loss": 1.3097,
+      "step": 1994
+    },
+    {
+      "epoch": 0.14380451236214228,
+      "grad_norm": 0.11678269007696329,
+      "learning_rate": 0.00019328982382604498,
+      "loss": 1.3125,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1438765948244792,
+      "grad_norm": 0.09927825080551517,
+      "learning_rate": 0.0001932814130386928,
+      "loss": 1.3044,
+      "step": 1996
+    },
+    {
+      "epoch": 0.14394867728681612,
+      "grad_norm": 0.1000336838179997,
+      "learning_rate": 0.0001932729971666691,
+      "loss": 1.3275,
+      "step": 1997
+    },
+    {
+      "epoch": 0.14402075974915304,
+      "grad_norm": 0.12129392100864808,
+      "learning_rate": 0.0001932645762104326,
+      "loss": 1.3841,
+      "step": 1998
+    },
+    {
+      "epoch": 0.14409284221148994,
+      "grad_norm": 0.10359419984654059,
+      "learning_rate": 0.00019325615017044232,
+      "loss": 1.4604,
+      "step": 1999
+    },
+    {
+      "epoch": 0.14416492467382686,
+      "grad_norm": 0.10079256494477368,
+      "learning_rate": 0.00019324771904715758,
+      "loss": 1.3166,
+      "step": 2000
+    },
+    {
+      "epoch": 0.14423700713616378,
+      "grad_norm": 0.10345617735166733,
+      "learning_rate": 0.00019323928284103791,
+      "loss": 1.2451,
+      "step": 2001
+    },
+    {
+      "epoch": 0.1443090895985007,
+      "grad_norm": 0.11158909759623442,
+      "learning_rate": 0.00019323084155254322,
+      "loss": 1.2239,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1443811720608376,
+      "grad_norm": 0.0967139431924691,
+      "learning_rate": 0.00019322239518213357,
+      "loss": 1.248,
+      "step": 2003
+    },
+    {
+      "epoch": 0.1444532545231745,
+      "grad_norm": 0.132858525016854,
+      "learning_rate": 0.0001932139437302694,
+      "loss": 1.4293,
+      "step": 2004
+    },
+    {
+      "epoch": 0.14452533698551143,
+      "grad_norm": 0.11870190015924464,
+      "learning_rate": 0.00019320548719741138,
+      "loss": 1.2271,
+      "step": 2005
+    },
+    {
+      "epoch": 0.14459741944784835,
+      "grad_norm": 0.17241224737682445,
+      "learning_rate": 0.00019319702558402046,
+      "loss": 1.3106,
+      "step": 2006
+    },
+    {
+      "epoch": 0.14466950191018524,
+      "grad_norm": 0.11326046555200694,
+      "learning_rate": 0.0001931885588905579,
+      "loss": 1.3139,
+      "step": 2007
+    },
+    {
+      "epoch": 0.14474158437252216,
+      "grad_norm": 0.11506639404487759,
+      "learning_rate": 0.0001931800871174852,
+      "loss": 1.3043,
+      "step": 2008
+    },
+    {
+      "epoch": 0.14481366683485908,
+      "grad_norm": 0.16851674444618311,
+      "learning_rate": 0.0001931716102652641,
+      "loss": 1.3129,
+      "step": 2009
+    },
+    {
+      "epoch": 0.144885749297196,
+      "grad_norm": 0.13753098131319,
+      "learning_rate": 0.00019316312833435673,
+      "loss": 1.3303,
+      "step": 2010
+    },
+    {
+      "epoch": 0.14495783175953292,
+      "grad_norm": 0.1092427736488733,
+      "learning_rate": 0.0001931546413252254,
+      "loss": 1.3224,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1450299142218698,
+      "grad_norm": 0.10887217008123402,
+      "learning_rate": 0.0001931461492383327,
+      "loss": 1.256,
+      "step": 2012
+    },
+    {
+      "epoch": 0.14510199668420673,
+      "grad_norm": 0.10984887340055492,
+      "learning_rate": 0.0001931376520741416,
+      "loss": 1.2689,
+      "step": 2013
+    },
+    {
+      "epoch": 0.14517407914654365,
+      "grad_norm": 0.08923994522006325,
+      "learning_rate": 0.00019312914983311523,
+      "loss": 1.2791,
+      "step": 2014
+    },
+    {
+      "epoch": 0.14524616160888057,
+      "grad_norm": 0.11193426161990387,
+      "learning_rate": 0.00019312064251571703,
+      "loss": 1.3091,
+      "step": 2015
+    },
+    {
+      "epoch": 0.14531824407121746,
+      "grad_norm": 0.11167184874143742,
+      "learning_rate": 0.0001931121301224107,
+      "loss": 1.4116,
+      "step": 2016
+    },
+    {
+      "epoch": 0.14539032653355438,
+      "grad_norm": 0.10641965697886001,
+      "learning_rate": 0.00019310361265366033,
+      "loss": 1.2948,
+      "step": 2017
+    },
+    {
+      "epoch": 0.1454624089958913,
+      "grad_norm": 0.10489241287734676,
+      "learning_rate": 0.0001930950901099301,
+      "loss": 1.3325,
+      "step": 2018
+    },
+    {
+      "epoch": 0.14553449145822822,
+      "grad_norm": 0.09961206865190754,
+      "learning_rate": 0.0001930865624916846,
+      "loss": 1.2794,
+      "step": 2019
+    },
+    {
+      "epoch": 0.1456065739205651,
+      "grad_norm": 0.11278010161353298,
+      "learning_rate": 0.0001930780297993887,
+      "loss": 1.3178,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14567865638290203,
+      "grad_norm": 0.11065301807720239,
+      "learning_rate": 0.00019306949203350747,
+      "loss": 1.3582,
+      "step": 2021
+    },
+    {
+      "epoch": 0.14575073884523895,
+      "grad_norm": 0.10894674296058939,
+      "learning_rate": 0.0001930609491945063,
+      "loss": 1.3718,
+      "step": 2022
+    },
+    {
+      "epoch": 0.14582282130757587,
+      "grad_norm": 0.10280038054276996,
+      "learning_rate": 0.00019305240128285081,
+      "loss": 1.3572,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1458949037699128,
+      "grad_norm": 0.10696202979217424,
+      "learning_rate": 0.00019304384829900702,
+      "loss": 1.3112,
+      "step": 2024
+    },
+    {
+      "epoch": 0.14596698623224968,
+      "grad_norm": 0.11891247412162402,
+      "learning_rate": 0.0001930352902434411,
+      "loss": 1.3543,
+      "step": 2025
+    },
+    {
+      "epoch": 0.1460390686945866,
+      "grad_norm": 0.12154692853333697,
+      "learning_rate": 0.00019302672711661955,
+      "loss": 1.2777,
+      "step": 2026
+    },
+    {
+      "epoch": 0.14611115115692352,
+      "grad_norm": 0.11228144677376341,
+      "learning_rate": 0.00019301815891900912,
+      "loss": 1.2748,
+      "step": 2027
+    },
+    {
+      "epoch": 0.14618323361926044,
+      "grad_norm": 0.15063312086537817,
+      "learning_rate": 0.00019300958565107688,
+      "loss": 1.3344,
+      "step": 2028
+    },
+    {
+      "epoch": 0.14625531608159734,
+      "grad_norm": 0.12043938295182408,
+      "learning_rate": 0.00019300100731329013,
+      "loss": 1.285,
+      "step": 2029
+    },
+    {
+      "epoch": 0.14632739854393426,
+      "grad_norm": 0.09956419077059729,
+      "learning_rate": 0.00019299242390611644,
+      "loss": 1.2368,
+      "step": 2030
+    },
+    {
+      "epoch": 0.14639948100627118,
+      "grad_norm": 0.11770482834940005,
+      "learning_rate": 0.00019298383543002375,
+      "loss": 1.3704,
+      "step": 2031
+    },
+    {
+      "epoch": 0.1464715634686081,
+      "grad_norm": 0.1199334702140474,
+      "learning_rate": 0.00019297524188548017,
+      "loss": 1.3664,
+      "step": 2032
+    },
+    {
+      "epoch": 0.146543645930945,
+      "grad_norm": 0.12533608669935586,
+      "learning_rate": 0.00019296664327295412,
+      "loss": 1.371,
+      "step": 2033
+    },
+    {
+      "epoch": 0.1466157283932819,
+      "grad_norm": 0.11557957240083229,
+      "learning_rate": 0.0001929580395929143,
+      "loss": 1.271,
+      "step": 2034
+    },
+    {
+      "epoch": 0.14668781085561883,
+      "grad_norm": 0.11285590316247034,
+      "learning_rate": 0.00019294943084582967,
+      "loss": 1.2927,
+      "step": 2035
+    },
+    {
+      "epoch": 0.14675989331795575,
+      "grad_norm": 0.11524306892141223,
+      "learning_rate": 0.00019294081703216955,
+      "loss": 1.3799,
+      "step": 2036
+    },
+    {
+      "epoch": 0.14683197578029267,
+      "grad_norm": 0.12492457694502337,
+      "learning_rate": 0.00019293219815240344,
+      "loss": 1.3963,
+      "step": 2037
+    },
+    {
+      "epoch": 0.14690405824262956,
+      "grad_norm": 0.1037280574373411,
+      "learning_rate": 0.0001929235742070011,
+      "loss": 1.2253,
+      "step": 2038
+    },
+    {
+      "epoch": 0.14697614070496648,
+      "grad_norm": 0.11552612387914207,
+      "learning_rate": 0.00019291494519643268,
+      "loss": 1.393,
+      "step": 2039
+    },
+    {
+      "epoch": 0.1470482231673034,
+      "grad_norm": 0.1756524316953672,
+      "learning_rate": 0.00019290631112116845,
+      "loss": 1.3685,
+      "step": 2040
+    },
+    {
+      "epoch": 0.14712030562964032,
+      "grad_norm": 0.1136669271395032,
+      "learning_rate": 0.00019289767198167916,
+      "loss": 1.3054,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1471923880919772,
+      "grad_norm": 0.10644803145341503,
+      "learning_rate": 0.00019288902777843563,
+      "loss": 1.3442,
+      "step": 2042
+    },
+    {
+      "epoch": 0.14726447055431413,
+      "grad_norm": 0.10128502798867915,
+      "learning_rate": 0.0001928803785119091,
+      "loss": 1.3368,
+      "step": 2043
+    },
+    {
+      "epoch": 0.14733655301665105,
+      "grad_norm": 0.11396129886894654,
+      "learning_rate": 0.00019287172418257098,
+      "loss": 1.2524,
+      "step": 2044
+    },
+    {
+      "epoch": 0.14740863547898797,
+      "grad_norm": 0.12853548453070718,
+      "learning_rate": 0.00019286306479089307,
+      "loss": 1.3555,
+      "step": 2045
+    },
+    {
+      "epoch": 0.14748071794132486,
+      "grad_norm": 0.1059631125272775,
+      "learning_rate": 0.00019285440033734732,
+      "loss": 1.4655,
+      "step": 2046
+    },
+    {
+      "epoch": 0.14755280040366178,
+      "grad_norm": 0.1631263828821402,
+      "learning_rate": 0.00019284573082240607,
+      "loss": 1.3741,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1476248828659987,
+      "grad_norm": 0.1029617846330115,
+      "learning_rate": 0.00019283705624654186,
+      "loss": 1.3137,
+      "step": 2048
+    },
+    {
+      "epoch": 0.14769696532833562,
+      "grad_norm": 0.09626753791578865,
+      "learning_rate": 0.00019282837661022755,
+      "loss": 1.3151,
+      "step": 2049
+    },
+    {
+      "epoch": 0.14776904779067254,
+      "grad_norm": 0.11243406010332255,
+      "learning_rate": 0.00019281969191393622,
+      "loss": 1.3057,
+      "step": 2050
+    },
+    {
+      "epoch": 0.14784113025300943,
+      "grad_norm": 0.11284392390589539,
+      "learning_rate": 0.00019281100215814135,
+      "loss": 1.3491,
+      "step": 2051
+    },
+    {
+      "epoch": 0.14791321271534635,
+      "grad_norm": 0.1063220338234662,
+      "learning_rate": 0.0001928023073433165,
+      "loss": 1.4154,
+      "step": 2052
+    },
+    {
+      "epoch": 0.14798529517768327,
+      "grad_norm": 0.10568350105639163,
+      "learning_rate": 0.00019279360746993568,
+      "loss": 1.3375,
+      "step": 2053
+    },
+    {
+      "epoch": 0.1480573776400202,
+      "grad_norm": 0.1042098168842622,
+      "learning_rate": 0.0001927849025384731,
+      "loss": 1.312,
+      "step": 2054
+    },
+    {
+      "epoch": 0.14812946010235709,
+      "grad_norm": 0.11304066636973084,
+      "learning_rate": 0.00019277619254940326,
+      "loss": 1.37,
+      "step": 2055
+    },
+    {
+      "epoch": 0.148201542564694,
+      "grad_norm": 0.10004860323954122,
+      "learning_rate": 0.00019276747750320095,
+      "loss": 1.2987,
+      "step": 2056
+    },
+    {
+      "epoch": 0.14827362502703093,
+      "grad_norm": 0.11988775737520413,
+      "learning_rate": 0.00019275875740034116,
+      "loss": 1.3524,
+      "step": 2057
+    },
+    {
+      "epoch": 0.14834570748936785,
+      "grad_norm": 0.10351376401898708,
+      "learning_rate": 0.00019275003224129925,
+      "loss": 1.3429,
+      "step": 2058
+    },
+    {
+      "epoch": 0.14841778995170474,
+      "grad_norm": 0.17803510950556875,
+      "learning_rate": 0.0001927413020265508,
+      "loss": 1.3369,
+      "step": 2059
+    },
+    {
+      "epoch": 0.14848987241404166,
+      "grad_norm": 0.09866770304777706,
+      "learning_rate": 0.00019273256675657175,
+      "loss": 1.3036,
+      "step": 2060
+    },
+    {
+      "epoch": 0.14856195487637858,
+      "grad_norm": 0.11217688385698536,
+      "learning_rate": 0.00019272382643183817,
+      "loss": 1.3949,
+      "step": 2061
+    },
+    {
+      "epoch": 0.1486340373387155,
+      "grad_norm": 0.16060567070042675,
+      "learning_rate": 0.00019271508105282652,
+      "loss": 1.3988,
+      "step": 2062
+    },
+    {
+      "epoch": 0.14870611980105242,
+      "grad_norm": 0.0999704708605788,
+      "learning_rate": 0.0001927063306200135,
+      "loss": 1.2223,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1487782022633893,
+      "grad_norm": 0.10140412440505321,
+      "learning_rate": 0.00019269757513387613,
+      "loss": 1.3466,
+      "step": 2064
+    },
+    {
+      "epoch": 0.14885028472572623,
+      "grad_norm": 0.10592044328709539,
+      "learning_rate": 0.00019268881459489156,
+      "loss": 1.2976,
+      "step": 2065
+    },
+    {
+      "epoch": 0.14892236718806315,
+      "grad_norm": 0.10297204517676213,
+      "learning_rate": 0.0001926800490035374,
+      "loss": 1.3209,
+      "step": 2066
+    },
+    {
+      "epoch": 0.14899444965040007,
+      "grad_norm": 0.09343005674267385,
+      "learning_rate": 0.0001926712783602914,
+      "loss": 1.2285,
+      "step": 2067
+    },
+    {
+      "epoch": 0.14906653211273696,
+      "grad_norm": 0.2005287667866991,
+      "learning_rate": 0.00019266250266563174,
+      "loss": 1.3106,
+      "step": 2068
+    },
+    {
+      "epoch": 0.14913861457507388,
+      "grad_norm": 0.08879225802134103,
+      "learning_rate": 0.00019265372192003665,
+      "loss": 1.2856,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1492106970374108,
+      "grad_norm": 0.1401257518294958,
+      "learning_rate": 0.00019264493612398481,
+      "loss": 1.4255,
+      "step": 2070
+    },
+    {
+      "epoch": 0.14928277949974772,
+      "grad_norm": 0.11787873984303185,
+      "learning_rate": 0.00019263614527795513,
+      "loss": 1.2675,
+      "step": 2071
+    },
+    {
+      "epoch": 0.1493548619620846,
+      "grad_norm": 0.11897614912976384,
+      "learning_rate": 0.00019262734938242678,
+      "loss": 1.2774,
+      "step": 2072
+    },
+    {
+      "epoch": 0.14942694442442153,
+      "grad_norm": 0.132404137468897,
+      "learning_rate": 0.00019261854843787926,
+      "loss": 1.2102,
+      "step": 2073
+    },
+    {
+      "epoch": 0.14949902688675845,
+      "grad_norm": 0.11022094578475379,
+      "learning_rate": 0.00019260974244479222,
+      "loss": 1.3139,
+      "step": 2074
+    },
+    {
+      "epoch": 0.14957110934909537,
+      "grad_norm": 0.1262133196581483,
+      "learning_rate": 0.00019260093140364572,
+      "loss": 1.396,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1496431918114323,
+      "grad_norm": 0.10811839485508853,
+      "learning_rate": 0.00019259211531492002,
+      "loss": 1.2993,
+      "step": 2076
+    },
+    {
+      "epoch": 0.14971527427376918,
+      "grad_norm": 0.23244173108589003,
+      "learning_rate": 0.0001925832941790957,
+      "loss": 1.3356,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1497873567361061,
+      "grad_norm": 0.259426310050802,
+      "learning_rate": 0.00019257446799665355,
+      "loss": 1.3601,
+      "step": 2078
+    },
+    {
+      "epoch": 0.14985943919844302,
+      "grad_norm": 0.11436213643077207,
+      "learning_rate": 0.00019256563676807472,
+      "loss": 1.2671,
+      "step": 2079
+    },
+    {
+      "epoch": 0.14993152166077994,
+      "grad_norm": 0.10177969036162934,
+      "learning_rate": 0.00019255680049384054,
+      "loss": 1.3269,
+      "step": 2080
+    },
+    {
+      "epoch": 0.15000360412311683,
+      "grad_norm": 0.12695039753772255,
+      "learning_rate": 0.00019254795917443275,
+      "loss": 1.2848,
+      "step": 2081
+    },
+    {
+      "epoch": 0.15007568658545375,
+      "grad_norm": 0.11132454604097017,
+      "learning_rate": 0.00019253911281033319,
+      "loss": 1.3573,
+      "step": 2082
+    },
+    {
+      "epoch": 0.15014776904779067,
+      "grad_norm": 0.10593483590193367,
+      "learning_rate": 0.0001925302614020241,
+      "loss": 1.3688,
+      "step": 2083
+    },
+    {
+      "epoch": 0.1502198515101276,
+      "grad_norm": 0.09656530434722724,
+      "learning_rate": 0.00019252140494998798,
+      "loss": 1.2852,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1502919339724645,
+      "grad_norm": 0.09400243028934344,
+      "learning_rate": 0.00019251254345470758,
+      "loss": 1.277,
+      "step": 2085
+    },
+    {
+      "epoch": 0.1503640164348014,
+      "grad_norm": 0.10168934010143701,
+      "learning_rate": 0.00019250367691666588,
+      "loss": 1.3941,
+      "step": 2086
+    },
+    {
+      "epoch": 0.15043609889713833,
+      "grad_norm": 0.11402727015876934,
+      "learning_rate": 0.00019249480533634628,
+      "loss": 1.3298,
+      "step": 2087
+    },
+    {
+      "epoch": 0.15050818135947525,
+      "grad_norm": 0.12172503565856856,
+      "learning_rate": 0.00019248592871423225,
+      "loss": 1.2807,
+      "step": 2088
+    },
+    {
+      "epoch": 0.15058026382181217,
+      "grad_norm": 0.10625119748381029,
+      "learning_rate": 0.00019247704705080773,
+      "loss": 1.4077,
+      "step": 2089
+    },
+    {
+      "epoch": 0.15065234628414906,
+      "grad_norm": 0.10272888445637253,
+      "learning_rate": 0.00019246816034655683,
+      "loss": 1.2119,
+      "step": 2090
+    },
+    {
+      "epoch": 0.15072442874648598,
+      "grad_norm": 0.10921648330803436,
+      "learning_rate": 0.00019245926860196396,
+      "loss": 1.348,
+      "step": 2091
+    },
+    {
+      "epoch": 0.1507965112088229,
+      "grad_norm": 0.15013966529779407,
+      "learning_rate": 0.00019245037181751375,
+      "loss": 1.1758,
+      "step": 2092
+    },
+    {
+      "epoch": 0.15086859367115982,
+      "grad_norm": 0.11154195629844676,
+      "learning_rate": 0.00019244146999369122,
+      "loss": 1.2868,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1509406761334967,
+      "grad_norm": 0.10110654487243825,
+      "learning_rate": 0.00019243256313098155,
+      "loss": 1.423,
+      "step": 2094
+    },
+    {
+      "epoch": 0.15101275859583363,
+      "grad_norm": 0.0983854202606571,
+      "learning_rate": 0.00019242365122987027,
+      "loss": 1.3277,
+      "step": 2095
+    },
+    {
+      "epoch": 0.15108484105817055,
+      "grad_norm": 0.11111128215192462,
+      "learning_rate": 0.00019241473429084315,
+      "loss": 1.2696,
+      "step": 2096
+    },
+    {
+      "epoch": 0.15115692352050747,
+      "grad_norm": 0.11175613806217162,
+      "learning_rate": 0.00019240581231438626,
+      "loss": 1.4676,
+      "step": 2097
+    },
+    {
+      "epoch": 0.15122900598284436,
+      "grad_norm": 0.10733997219422398,
+      "learning_rate": 0.0001923968853009859,
+      "loss": 1.3635,
+      "step": 2098
+    },
+    {
+      "epoch": 0.15130108844518128,
+      "grad_norm": 0.12875362489784875,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 1.2722,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1513731709075182,
+      "grad_norm": 0.10810575469034092,
+      "learning_rate": 0.0001923790161653015,
+      "loss": 1.3525,
+      "step": 2100
+    },
+    {
+      "epoch": 0.15144525336985512,
+      "grad_norm": 0.10830343033882826,
+      "learning_rate": 0.00019237007404399148,
+      "loss": 1.2794,
+      "step": 2101
+    },
+    {
+      "epoch": 0.15151733583219204,
+      "grad_norm": 0.10700466433484372,
+      "learning_rate": 0.00019236112688768606,
+      "loss": 1.3322,
+      "step": 2102
+    },
+    {
+      "epoch": 0.15158941829452893,
+      "grad_norm": 0.09910120140531703,
+      "learning_rate": 0.00019235217469687293,
+      "loss": 1.3011,
+      "step": 2103
+    },
+    {
+      "epoch": 0.15166150075686585,
+      "grad_norm": 0.09510672556284329,
+      "learning_rate": 0.0001923432174720401,
+      "loss": 1.3741,
+      "step": 2104
+    },
+    {
+      "epoch": 0.15173358321920277,
+      "grad_norm": 0.09784522243899106,
+      "learning_rate": 0.00019233425521367577,
+      "loss": 1.3069,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1518056656815397,
+      "grad_norm": 0.10260171490926658,
+      "learning_rate": 0.00019232528792226846,
+      "loss": 1.2594,
+      "step": 2106
+    },
+    {
+      "epoch": 0.15187774814387658,
+      "grad_norm": 0.10637773575853125,
+      "learning_rate": 0.00019231631559830703,
+      "loss": 1.3092,
+      "step": 2107
+    },
+    {
+      "epoch": 0.1519498306062135,
+      "grad_norm": 0.11536223919238943,
+      "learning_rate": 0.0001923073382422805,
+      "loss": 1.3452,
+      "step": 2108
+    },
+    {
+      "epoch": 0.15202191306855042,
+      "grad_norm": 0.2434113023108464,
+      "learning_rate": 0.00019229835585467825,
+      "loss": 1.3777,
+      "step": 2109
+    },
+    {
+      "epoch": 0.15209399553088734,
+      "grad_norm": 0.10256642843790192,
+      "learning_rate": 0.0001922893684359899,
+      "loss": 1.2966,
+      "step": 2110
+    },
+    {
+      "epoch": 0.15216607799322424,
+      "grad_norm": 0.10436610546299521,
+      "learning_rate": 0.00019228037598670525,
+      "loss": 1.2722,
+      "step": 2111
+    },
+    {
+      "epoch": 0.15223816045556116,
+      "grad_norm": 0.18929247150707942,
+      "learning_rate": 0.0001922713785073146,
+      "loss": 1.3688,
+      "step": 2112
+    },
+    {
+      "epoch": 0.15231024291789808,
+      "grad_norm": 0.11017059315721593,
+      "learning_rate": 0.00019226237599830834,
+      "loss": 1.3562,
+      "step": 2113
+    },
+    {
+      "epoch": 0.152382325380235,
+      "grad_norm": 0.10509095066579499,
+      "learning_rate": 0.00019225336846017717,
+      "loss": 1.3128,
+      "step": 2114
+    },
+    {
+      "epoch": 0.15245440784257192,
+      "grad_norm": 0.11390677523274344,
+      "learning_rate": 0.00019224435589341207,
+      "loss": 1.4154,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1525264903049088,
+      "grad_norm": 0.11013725626623753,
+      "learning_rate": 0.0001922353382985044,
+      "loss": 1.3207,
+      "step": 2116
+    },
+    {
+      "epoch": 0.15259857276724573,
+      "grad_norm": 0.097589091057085,
+      "learning_rate": 0.00019222631567594557,
+      "loss": 1.2946,
+      "step": 2117
+    },
+    {
+      "epoch": 0.15267065522958265,
+      "grad_norm": 0.1085344125305648,
+      "learning_rate": 0.0001922172880262275,
+      "loss": 1.2314,
+      "step": 2118
+    },
+    {
+      "epoch": 0.15274273769191957,
+      "grad_norm": 0.12226674420910556,
+      "learning_rate": 0.00019220825534984218,
+      "loss": 1.3158,
+      "step": 2119
+    },
+    {
+      "epoch": 0.15281482015425646,
+      "grad_norm": 0.11764242357422479,
+      "learning_rate": 0.00019219921764728208,
+      "loss": 1.2191,
+      "step": 2120
+    },
+    {
+      "epoch": 0.15288690261659338,
+      "grad_norm": 0.10003310936849859,
+      "learning_rate": 0.00019219017491903974,
+      "loss": 1.3165,
+      "step": 2121
+    },
+    {
+      "epoch": 0.1529589850789303,
+      "grad_norm": 0.1097664865575588,
+      "learning_rate": 0.00019218112716560813,
+      "loss": 1.3743,
+      "step": 2122
+    },
+    {
+      "epoch": 0.15303106754126722,
+      "grad_norm": 0.13610785386091312,
+      "learning_rate": 0.00019217207438748042,
+      "loss": 1.4281,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1531031500036041,
+      "grad_norm": 0.1203547933566413,
+      "learning_rate": 0.00019216301658515005,
+      "loss": 1.325,
+      "step": 2124
+    },
+    {
+      "epoch": 0.15317523246594103,
+      "grad_norm": 0.1346226622215571,
+      "learning_rate": 0.00019215395375911073,
+      "loss": 1.375,
+      "step": 2125
+    },
+    {
+      "epoch": 0.15324731492827795,
+      "grad_norm": 0.1165944390116016,
+      "learning_rate": 0.00019214488590985656,
+      "loss": 1.3598,
+      "step": 2126
+    },
+    {
+      "epoch": 0.15331939739061487,
+      "grad_norm": 0.11230065137113092,
+      "learning_rate": 0.00019213581303788172,
+      "loss": 1.3684,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1533914798529518,
+      "grad_norm": 0.10736882087111398,
+      "learning_rate": 0.0001921267351436808,
+      "loss": 1.3436,
+      "step": 2128
+    },
+    {
+      "epoch": 0.15346356231528868,
+      "grad_norm": 0.12347019912533579,
+      "learning_rate": 0.00019211765222774865,
+      "loss": 1.3012,
+      "step": 2129
+    },
+    {
+      "epoch": 0.1535356447776256,
+      "grad_norm": 0.10510324371067038,
+      "learning_rate": 0.0001921085642905803,
+      "loss": 1.2938,
+      "step": 2130
+    },
+    {
+      "epoch": 0.15360772723996252,
+      "grad_norm": 0.11184005837718965,
+      "learning_rate": 0.0001920994713326712,
+      "loss": 1.3408,
+      "step": 2131
+    },
+    {
+      "epoch": 0.15367980970229944,
+      "grad_norm": 0.10992387951523255,
+      "learning_rate": 0.00019209037335451697,
+      "loss": 1.1678,
+      "step": 2132
+    },
+    {
+      "epoch": 0.15375189216463633,
+      "grad_norm": 0.10079507559845882,
+      "learning_rate": 0.00019208127035661354,
+      "loss": 1.3193,
+      "step": 2133
+    },
+    {
+      "epoch": 0.15382397462697325,
+      "grad_norm": 0.10524646895590918,
+      "learning_rate": 0.00019207216233945706,
+      "loss": 1.3425,
+      "step": 2134
+    },
+    {
+      "epoch": 0.15389605708931017,
+      "grad_norm": 0.10009118480459606,
+      "learning_rate": 0.00019206304930354404,
+      "loss": 1.378,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1539681395516471,
+      "grad_norm": 0.10515281621889566,
+      "learning_rate": 0.0001920539312493712,
+      "loss": 1.3549,
+      "step": 2136
+    },
+    {
+      "epoch": 0.15404022201398399,
+      "grad_norm": 0.1114573318841526,
+      "learning_rate": 0.0001920448081774356,
+      "loss": 1.3207,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1541123044763209,
+      "grad_norm": 0.09834817060441375,
+      "learning_rate": 0.0001920356800882345,
+      "loss": 1.4047,
+      "step": 2138
+    },
+    {
+      "epoch": 0.15418438693865782,
+      "grad_norm": 0.10202272008343091,
+      "learning_rate": 0.00019202654698226542,
+      "loss": 1.2932,
+      "step": 2139
+    },
+    {
+      "epoch": 0.15425646940099474,
+      "grad_norm": 0.10059663874560106,
+      "learning_rate": 0.0001920174088600263,
+      "loss": 1.2548,
+      "step": 2140
+    },
+    {
+      "epoch": 0.15432855186333166,
+      "grad_norm": 0.10288813410228807,
+      "learning_rate": 0.00019200826572201513,
+      "loss": 1.3265,
+      "step": 2141
+    },
+    {
+      "epoch": 0.15440063432566856,
+      "grad_norm": 0.1147719794233791,
+      "learning_rate": 0.00019199911756873038,
+      "loss": 1.3729,
+      "step": 2142
+    },
+    {
+      "epoch": 0.15447271678800548,
+      "grad_norm": 0.10204778733245673,
+      "learning_rate": 0.00019198996440067064,
+      "loss": 1.3263,
+      "step": 2143
+    },
+    {
+      "epoch": 0.1545447992503424,
+      "grad_norm": 0.17023443898715387,
+      "learning_rate": 0.0001919808062183349,
+      "loss": 1.3134,
+      "step": 2144
+    },
+    {
+      "epoch": 0.15461688171267932,
+      "grad_norm": 0.0937239067972613,
+      "learning_rate": 0.00019197164302222234,
+      "loss": 1.2994,
+      "step": 2145
+    },
+    {
+      "epoch": 0.1546889641750162,
+      "grad_norm": 0.10079242544155934,
+      "learning_rate": 0.00019196247481283242,
+      "loss": 1.1943,
+      "step": 2146
+    },
+    {
+      "epoch": 0.15476104663735313,
+      "grad_norm": 0.12652106875765923,
+      "learning_rate": 0.00019195330159066493,
+      "loss": 1.3038,
+      "step": 2147
+    },
+    {
+      "epoch": 0.15483312909969005,
+      "grad_norm": 0.12057428828787001,
+      "learning_rate": 0.00019194412335621984,
+      "loss": 1.3091,
+      "step": 2148
+    },
+    {
+      "epoch": 0.15490521156202697,
+      "grad_norm": 0.10970247138889297,
+      "learning_rate": 0.00019193494010999747,
+      "loss": 1.354,
+      "step": 2149
+    },
+    {
+      "epoch": 0.15497729402436386,
+      "grad_norm": 0.10097054597664604,
+      "learning_rate": 0.00019192575185249838,
+      "loss": 1.3095,
+      "step": 2150
+    },
+    {
+      "epoch": 0.15504937648670078,
+      "grad_norm": 0.10356350124727631,
+      "learning_rate": 0.00019191655858422346,
+      "loss": 1.3627,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1551214589490377,
+      "grad_norm": 0.10316962475163066,
+      "learning_rate": 0.00019190736030567376,
+      "loss": 1.3953,
+      "step": 2152
+    },
+    {
+      "epoch": 0.15519354141137462,
+      "grad_norm": 0.12213080284380809,
+      "learning_rate": 0.00019189815701735073,
+      "loss": 1.2751,
+      "step": 2153
+    },
+    {
+      "epoch": 0.15526562387371154,
+      "grad_norm": 0.1144657796615708,
+      "learning_rate": 0.00019188894871975594,
+      "loss": 1.306,
+      "step": 2154
+    },
+    {
+      "epoch": 0.15533770633604843,
+      "grad_norm": 0.12780430389301387,
+      "learning_rate": 0.00019187973541339145,
+      "loss": 1.2849,
+      "step": 2155
+    },
+    {
+      "epoch": 0.15540978879838535,
+      "grad_norm": 0.10437376715201496,
+      "learning_rate": 0.00019187051709875936,
+      "loss": 1.4029,
+      "step": 2156
+    },
+    {
+      "epoch": 0.15548187126072227,
+      "grad_norm": 0.10395192615959614,
+      "learning_rate": 0.0001918612937763622,
+      "loss": 1.2794,
+      "step": 2157
+    },
+    {
+      "epoch": 0.1555539537230592,
+      "grad_norm": 0.11599985744151001,
+      "learning_rate": 0.0001918520654467027,
+      "loss": 1.2862,
+      "step": 2158
+    },
+    {
+      "epoch": 0.15562603618539608,
+      "grad_norm": 0.11974541345077482,
+      "learning_rate": 0.0001918428321102839,
+      "loss": 1.3333,
+      "step": 2159
+    },
+    {
+      "epoch": 0.155698118647733,
+      "grad_norm": 0.10825914272617576,
+      "learning_rate": 0.00019183359376760911,
+      "loss": 1.3782,
+      "step": 2160
+    },
+    {
+      "epoch": 0.15577020111006992,
+      "grad_norm": 0.10826779402311715,
+      "learning_rate": 0.0001918243504191819,
+      "loss": 1.3106,
+      "step": 2161
+    },
+    {
+      "epoch": 0.15584228357240684,
+      "grad_norm": 0.10585556939259415,
+      "learning_rate": 0.00019181510206550613,
+      "loss": 1.3714,
+      "step": 2162
+    },
+    {
+      "epoch": 0.15591436603474373,
+      "grad_norm": 0.1032645073184311,
+      "learning_rate": 0.00019180584870708586,
+      "loss": 1.295,
+      "step": 2163
+    },
+    {
+      "epoch": 0.15598644849708065,
+      "grad_norm": 0.10315755012493905,
+      "learning_rate": 0.00019179659034442552,
+      "loss": 1.3515,
+      "step": 2164
+    },
+    {
+      "epoch": 0.15605853095941757,
+      "grad_norm": 0.11330997874883912,
+      "learning_rate": 0.0001917873269780298,
+      "loss": 1.3567,
+      "step": 2165
+    },
+    {
+      "epoch": 0.1561306134217545,
+      "grad_norm": 0.16957139606422586,
+      "learning_rate": 0.00019177805860840356,
+      "loss": 1.2749,
+      "step": 2166
+    },
+    {
+      "epoch": 0.1562026958840914,
+      "grad_norm": 0.11405589528515178,
+      "learning_rate": 0.00019176878523605208,
+      "loss": 1.3545,
+      "step": 2167
+    },
+    {
+      "epoch": 0.1562747783464283,
+      "grad_norm": 0.10457585371807458,
+      "learning_rate": 0.0001917595068614808,
+      "loss": 1.3597,
+      "step": 2168
+    },
+    {
+      "epoch": 0.15634686080876523,
+      "grad_norm": 0.10294964239813235,
+      "learning_rate": 0.00019175022348519554,
+      "loss": 1.2866,
+      "step": 2169
+    },
+    {
+      "epoch": 0.15641894327110215,
+      "grad_norm": 0.1072944192072695,
+      "learning_rate": 0.00019174093510770224,
+      "loss": 1.4265,
+      "step": 2170
+    },
+    {
+      "epoch": 0.15649102573343907,
+      "grad_norm": 0.10979975145790957,
+      "learning_rate": 0.00019173164172950723,
+      "loss": 1.3822,
+      "step": 2171
+    },
+    {
+      "epoch": 0.15656310819577596,
+      "grad_norm": 0.11191484901855026,
+      "learning_rate": 0.00019172234335111714,
+      "loss": 1.3477,
+      "step": 2172
+    },
+    {
+      "epoch": 0.15663519065811288,
+      "grad_norm": 0.11147550169544897,
+      "learning_rate": 0.00019171303997303873,
+      "loss": 1.396,
+      "step": 2173
+    },
+    {
+      "epoch": 0.1567072731204498,
+      "grad_norm": 0.10484256829388491,
+      "learning_rate": 0.00019170373159577917,
+      "loss": 1.3219,
+      "step": 2174
+    },
+    {
+      "epoch": 0.15677935558278672,
+      "grad_norm": 0.11235182087088168,
+      "learning_rate": 0.0001916944182198458,
+      "loss": 1.3459,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1568514380451236,
+      "grad_norm": 0.10888821680429765,
+      "learning_rate": 0.00019168509984574634,
+      "loss": 1.2995,
+      "step": 2176
+    },
+    {
+      "epoch": 0.15692352050746053,
+      "grad_norm": 0.10862817303549757,
+      "learning_rate": 0.00019167577647398871,
+      "loss": 1.2404,
+      "step": 2177
+    },
+    {
+      "epoch": 0.15699560296979745,
+      "grad_norm": 0.10181657697584369,
+      "learning_rate": 0.00019166644810508107,
+      "loss": 1.3574,
+      "step": 2178
+    },
+    {
+      "epoch": 0.15706768543213437,
+      "grad_norm": 0.1374940116377615,
+      "learning_rate": 0.00019165711473953197,
+      "loss": 1.3685,
+      "step": 2179
+    },
+    {
+      "epoch": 0.1571397678944713,
+      "grad_norm": 0.10101785179579358,
+      "learning_rate": 0.00019164777637785011,
+      "loss": 1.3183,
+      "step": 2180
+    },
+    {
+      "epoch": 0.15721185035680818,
+      "grad_norm": 0.10759803073579002,
+      "learning_rate": 0.00019163843302054455,
+      "loss": 1.3462,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1572839328191451,
+      "grad_norm": 0.1505233497085046,
+      "learning_rate": 0.00019162908466812458,
+      "loss": 1.3292,
+      "step": 2182
+    },
+    {
+      "epoch": 0.15735601528148202,
+      "grad_norm": 0.10821556989380152,
+      "learning_rate": 0.0001916197313210997,
+      "loss": 1.2631,
+      "step": 2183
+    },
+    {
+      "epoch": 0.15742809774381894,
+      "grad_norm": 0.1678707449062026,
+      "learning_rate": 0.00019161037297997987,
+      "loss": 1.3424,
+      "step": 2184
+    },
+    {
+      "epoch": 0.15750018020615583,
+      "grad_norm": 0.10012246005412843,
+      "learning_rate": 0.00019160100964527512,
+      "loss": 1.3363,
+      "step": 2185
+    },
+    {
+      "epoch": 0.15757226266849275,
+      "grad_norm": 0.1624516338204757,
+      "learning_rate": 0.00019159164131749587,
+      "loss": 1.3378,
+      "step": 2186
+    },
+    {
+      "epoch": 0.15764434513082967,
+      "grad_norm": 0.11704206713635948,
+      "learning_rate": 0.00019158226799715275,
+      "loss": 1.3497,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1577164275931666,
+      "grad_norm": 0.11941842525388312,
+      "learning_rate": 0.0001915728896847567,
+      "loss": 1.3762,
+      "step": 2188
+    },
+    {
+      "epoch": 0.15778851005550348,
+      "grad_norm": 0.10340789590632986,
+      "learning_rate": 0.00019156350638081896,
+      "loss": 1.2573,
+      "step": 2189
+    },
+    {
+      "epoch": 0.1578605925178404,
+      "grad_norm": 0.11615204942275353,
+      "learning_rate": 0.00019155411808585094,
+      "loss": 1.3261,
+      "step": 2190
+    },
+    {
+      "epoch": 0.15793267498017732,
+      "grad_norm": 0.105715396254105,
+      "learning_rate": 0.00019154472480036443,
+      "loss": 1.3658,
+      "step": 2191
+    },
+    {
+      "epoch": 0.15800475744251424,
+      "grad_norm": 0.10561272702210447,
+      "learning_rate": 0.00019153532652487143,
+      "loss": 1.3266,
+      "step": 2192
+    },
+    {
+      "epoch": 0.15807683990485116,
+      "grad_norm": 0.11479793949947353,
+      "learning_rate": 0.00019152592325988424,
+      "loss": 1.3568,
+      "step": 2193
+    },
+    {
+      "epoch": 0.15814892236718806,
+      "grad_norm": 0.09686010563586812,
+      "learning_rate": 0.00019151651500591544,
+      "loss": 1.3469,
+      "step": 2194
+    },
+    {
+      "epoch": 0.15822100482952497,
+      "grad_norm": 0.11301870190302898,
+      "learning_rate": 0.00019150710176347782,
+      "loss": 1.3527,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1582930872918619,
+      "grad_norm": 0.10789045736474277,
+      "learning_rate": 0.00019149768353308452,
+      "loss": 1.3126,
+      "step": 2196
+    },
+    {
+      "epoch": 0.15836516975419881,
+      "grad_norm": 0.11356891666743736,
+      "learning_rate": 0.0001914882603152489,
+      "loss": 1.3676,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1584372522165357,
+      "grad_norm": 0.10070134872295955,
+      "learning_rate": 0.00019147883211048457,
+      "loss": 1.2849,
+      "step": 2198
+    },
+    {
+      "epoch": 0.15850933467887263,
+      "grad_norm": 0.10756024855378739,
+      "learning_rate": 0.00019146939891930557,
+      "loss": 1.3795,
+      "step": 2199
+    },
+    {
+      "epoch": 0.15858141714120955,
+      "grad_norm": 0.11336402432663144,
+      "learning_rate": 0.00019145996074222595,
+      "loss": 1.3664,
+      "step": 2200
+    },
+    {
+      "epoch": 0.15865349960354647,
+      "grad_norm": 0.13851071317686525,
+      "learning_rate": 0.0001914505175797603,
+      "loss": 1.4244,
+      "step": 2201
+    },
+    {
+      "epoch": 0.15872558206588336,
+      "grad_norm": 0.2241412561084089,
+      "learning_rate": 0.0001914410694324233,
+      "loss": 1.4202,
+      "step": 2202
+    },
+    {
+      "epoch": 0.15879766452822028,
+      "grad_norm": 0.10688838150793659,
+      "learning_rate": 0.00019143161630072995,
+      "loss": 1.3489,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1588697469905572,
+      "grad_norm": 0.09846978392727197,
+      "learning_rate": 0.00019142215818519553,
+      "loss": 1.2821,
+      "step": 2204
+    },
+    {
+      "epoch": 0.15894182945289412,
+      "grad_norm": 0.10464542215534901,
+      "learning_rate": 0.00019141269508633562,
+      "loss": 1.3894,
+      "step": 2205
+    },
+    {
+      "epoch": 0.15901391191523104,
+      "grad_norm": 0.10329998678953056,
+      "learning_rate": 0.000191403227004666,
+      "loss": 1.342,
+      "step": 2206
+    },
+    {
+      "epoch": 0.15908599437756793,
+      "grad_norm": 0.134645955967905,
+      "learning_rate": 0.00019139375394070283,
+      "loss": 1.2463,
+      "step": 2207
+    },
+    {
+      "epoch": 0.15915807683990485,
+      "grad_norm": 0.11369527803026862,
+      "learning_rate": 0.0001913842758949624,
+      "loss": 1.3187,
+      "step": 2208
+    },
+    {
+      "epoch": 0.15923015930224177,
+      "grad_norm": 0.11254986244335917,
+      "learning_rate": 0.0001913747928679614,
+      "loss": 1.3665,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1593022417645787,
+      "grad_norm": 0.1056197705101916,
+      "learning_rate": 0.00019136530486021672,
+      "loss": 1.4218,
+      "step": 2210
+    },
+    {
+      "epoch": 0.15937432422691558,
+      "grad_norm": 0.12499502464754604,
+      "learning_rate": 0.00019135581187224555,
+      "loss": 1.275,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1594464066892525,
+      "grad_norm": 0.11603275312399985,
+      "learning_rate": 0.00019134631390456535,
+      "loss": 1.3377,
+      "step": 2212
+    },
+    {
+      "epoch": 0.15951848915158942,
+      "grad_norm": 0.135823853754083,
+      "learning_rate": 0.00019133681095769386,
+      "loss": 1.3302,
+      "step": 2213
+    },
+    {
+      "epoch": 0.15959057161392634,
+      "grad_norm": 0.14669396848313243,
+      "learning_rate": 0.000191327303032149,
+      "loss": 1.3809,
+      "step": 2214
+    },
+    {
+      "epoch": 0.15966265407626323,
+      "grad_norm": 0.13976091369292068,
+      "learning_rate": 0.00019131779012844912,
+      "loss": 1.2958,
+      "step": 2215
+    },
+    {
+      "epoch": 0.15973473653860015,
+      "grad_norm": 0.11987421871609037,
+      "learning_rate": 0.00019130827224711272,
+      "loss": 1.3285,
+      "step": 2216
+    },
+    {
+      "epoch": 0.15980681900093707,
+      "grad_norm": 0.09681008810304959,
+      "learning_rate": 0.00019129874938865861,
+      "loss": 1.2272,
+      "step": 2217
+    },
+    {
+      "epoch": 0.159878901463274,
+      "grad_norm": 0.10177841811906295,
+      "learning_rate": 0.00019128922155360587,
+      "loss": 1.2593,
+      "step": 2218
+    },
+    {
+      "epoch": 0.1599509839256109,
+      "grad_norm": 0.10581800633787997,
+      "learning_rate": 0.00019127968874247387,
+      "loss": 1.2983,
+      "step": 2219
+    },
+    {
+      "epoch": 0.1600230663879478,
+      "grad_norm": 0.10953897353163625,
+      "learning_rate": 0.00019127015095578222,
+      "loss": 1.374,
+      "step": 2220
+    },
+    {
+      "epoch": 0.16009514885028472,
+      "grad_norm": 0.10261162670916803,
+      "learning_rate": 0.0001912606081940508,
+      "loss": 1.2375,
+      "step": 2221
+    },
+    {
+      "epoch": 0.16016723131262164,
+      "grad_norm": 0.1059264750655961,
+      "learning_rate": 0.00019125106045779982,
+      "loss": 1.2006,
+      "step": 2222
+    },
+    {
+      "epoch": 0.16023931377495856,
+      "grad_norm": 0.11503891742838848,
+      "learning_rate": 0.0001912415077475497,
+      "loss": 1.3163,
+      "step": 2223
+    },
+    {
+      "epoch": 0.16031139623729546,
+      "grad_norm": 0.11446582918275526,
+      "learning_rate": 0.0001912319500638211,
+      "loss": 1.3275,
+      "step": 2224
+    },
+    {
+      "epoch": 0.16038347869963238,
+      "grad_norm": 0.10634616225041466,
+      "learning_rate": 0.00019122238740713505,
+      "loss": 1.2484,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1604555611619693,
+      "grad_norm": 0.10271148833409295,
+      "learning_rate": 0.0001912128197780128,
+      "loss": 1.3367,
+      "step": 2226
+    },
+    {
+      "epoch": 0.16052764362430622,
+      "grad_norm": 0.11653130728716073,
+      "learning_rate": 0.00019120324717697583,
+      "loss": 1.4421,
+      "step": 2227
+    },
+    {
+      "epoch": 0.1605997260866431,
+      "grad_norm": 0.09960550555470613,
+      "learning_rate": 0.00019119366960454598,
+      "loss": 1.2946,
+      "step": 2228
+    },
+    {
+      "epoch": 0.16067180854898003,
+      "grad_norm": 0.10205586952112515,
+      "learning_rate": 0.00019118408706124532,
+      "loss": 1.3181,
+      "step": 2229
+    },
+    {
+      "epoch": 0.16074389101131695,
+      "grad_norm": 0.10743549872145934,
+      "learning_rate": 0.00019117449954759612,
+      "loss": 1.321,
+      "step": 2230
+    },
+    {
+      "epoch": 0.16081597347365387,
+      "grad_norm": 0.10282994383582045,
+      "learning_rate": 0.00019116490706412104,
+      "loss": 1.3373,
+      "step": 2231
+    },
+    {
+      "epoch": 0.1608880559359908,
+      "grad_norm": 0.2382177172927358,
+      "learning_rate": 0.00019115530961134295,
+      "loss": 1.3707,
+      "step": 2232
+    },
+    {
+      "epoch": 0.16096013839832768,
+      "grad_norm": 0.09993543400824026,
+      "learning_rate": 0.00019114570718978496,
+      "loss": 1.2884,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1610322208606646,
+      "grad_norm": 0.11860789969319925,
+      "learning_rate": 0.00019113609979997054,
+      "loss": 1.3296,
+      "step": 2234
+    },
+    {
+      "epoch": 0.16110430332300152,
+      "grad_norm": 0.13324217850549738,
+      "learning_rate": 0.00019112648744242335,
+      "loss": 1.3518,
+      "step": 2235
+    },
+    {
+      "epoch": 0.16117638578533844,
+      "grad_norm": 0.29159405673763655,
+      "learning_rate": 0.00019111687011766733,
+      "loss": 1.3224,
+      "step": 2236
+    },
+    {
+      "epoch": 0.16124846824767533,
+      "grad_norm": 0.11582378723581523,
+      "learning_rate": 0.00019110724782622676,
+      "loss": 1.267,
+      "step": 2237
+    },
+    {
+      "epoch": 0.16132055071001225,
+      "grad_norm": 0.1112716473731516,
+      "learning_rate": 0.00019109762056862613,
+      "loss": 1.2925,
+      "step": 2238
+    },
+    {
+      "epoch": 0.16139263317234917,
+      "grad_norm": 0.10511517793310368,
+      "learning_rate": 0.00019108798834539017,
+      "loss": 1.2494,
+      "step": 2239
+    },
+    {
+      "epoch": 0.1614647156346861,
+      "grad_norm": 0.10673145097942988,
+      "learning_rate": 0.00019107835115704394,
+      "loss": 1.3738,
+      "step": 2240
+    },
+    {
+      "epoch": 0.16153679809702298,
+      "grad_norm": 0.11920481937789576,
+      "learning_rate": 0.00019106870900411277,
+      "loss": 1.3299,
+      "step": 2241
+    },
+    {
+      "epoch": 0.1616088805593599,
+      "grad_norm": 0.11167234702920953,
+      "learning_rate": 0.00019105906188712223,
+      "loss": 1.344,
+      "step": 2242
+    },
+    {
+      "epoch": 0.16168096302169682,
+      "grad_norm": 0.11373060295628355,
+      "learning_rate": 0.0001910494098065982,
+      "loss": 1.3047,
+      "step": 2243
+    },
+    {
+      "epoch": 0.16175304548403374,
+      "grad_norm": 0.11625019241069151,
+      "learning_rate": 0.00019103975276306678,
+      "loss": 1.3498,
+      "step": 2244
+    },
+    {
+      "epoch": 0.16182512794637066,
+      "grad_norm": 0.10589678435902754,
+      "learning_rate": 0.00019103009075705437,
+      "loss": 1.3104,
+      "step": 2245
+    },
+    {
+      "epoch": 0.16189721040870755,
+      "grad_norm": 0.10901096423075675,
+      "learning_rate": 0.00019102042378908761,
+      "loss": 1.3276,
+      "step": 2246
+    },
+    {
+      "epoch": 0.16196929287104447,
+      "grad_norm": 0.21888497422185932,
+      "learning_rate": 0.00019101075185969348,
+      "loss": 1.3016,
+      "step": 2247
+    },
+    {
+      "epoch": 0.1620413753333814,
+      "grad_norm": 0.10675809806659864,
+      "learning_rate": 0.00019100107496939917,
+      "loss": 1.3655,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1621134577957183,
+      "grad_norm": 0.10151705540228659,
+      "learning_rate": 0.00019099139311873215,
+      "loss": 1.3806,
+      "step": 2249
+    },
+    {
+      "epoch": 0.1621855402580552,
+      "grad_norm": 0.11237129415541836,
+      "learning_rate": 0.0001909817063082202,
+      "loss": 1.4248,
+      "step": 2250
+    },
+    {
+      "epoch": 0.16225762272039213,
+      "grad_norm": 0.11843320615332897,
+      "learning_rate": 0.00019097201453839127,
+      "loss": 1.2767,
+      "step": 2251
+    },
+    {
+      "epoch": 0.16232970518272904,
+      "grad_norm": 0.11914630069914534,
+      "learning_rate": 0.0001909623178097737,
+      "loss": 1.2868,
+      "step": 2252
+    },
+    {
+      "epoch": 0.16240178764506596,
+      "grad_norm": 0.12212492549885845,
+      "learning_rate": 0.00019095261612289607,
+      "loss": 1.3096,
+      "step": 2253
+    },
+    {
+      "epoch": 0.16247387010740286,
+      "grad_norm": 0.15235748095841933,
+      "learning_rate": 0.00019094290947828713,
+      "loss": 1.366,
+      "step": 2254
+    },
+    {
+      "epoch": 0.16254595256973978,
+      "grad_norm": 0.10569914494448528,
+      "learning_rate": 0.00019093319787647604,
+      "loss": 1.2143,
+      "step": 2255
+    },
+    {
+      "epoch": 0.1626180350320767,
+      "grad_norm": 0.12182601267543537,
+      "learning_rate": 0.00019092348131799215,
+      "loss": 1.4434,
+      "step": 2256
+    },
+    {
+      "epoch": 0.16269011749441362,
+      "grad_norm": 0.11425855190241686,
+      "learning_rate": 0.00019091375980336513,
+      "loss": 1.3431,
+      "step": 2257
+    },
+    {
+      "epoch": 0.16276219995675054,
+      "grad_norm": 0.09776555583376448,
+      "learning_rate": 0.00019090403333312485,
+      "loss": 1.4102,
+      "step": 2258
+    },
+    {
+      "epoch": 0.16283428241908743,
+      "grad_norm": 0.12073525286429716,
+      "learning_rate": 0.00019089430190780149,
+      "loss": 1.3746,
+      "step": 2259
+    },
+    {
+      "epoch": 0.16290636488142435,
+      "grad_norm": 0.1190830716620678,
+      "learning_rate": 0.00019088456552792554,
+      "loss": 1.2766,
+      "step": 2260
+    },
+    {
+      "epoch": 0.16297844734376127,
+      "grad_norm": 0.11449590367542838,
+      "learning_rate": 0.00019087482419402764,
+      "loss": 1.2559,
+      "step": 2261
+    },
+    {
+      "epoch": 0.1630505298060982,
+      "grad_norm": 0.097537174445678,
+      "learning_rate": 0.00019086507790663885,
+      "loss": 1.3518,
+      "step": 2262
+    },
+    {
+      "epoch": 0.16312261226843508,
+      "grad_norm": 0.11264461419039211,
+      "learning_rate": 0.00019085532666629045,
+      "loss": 1.2341,
+      "step": 2263
+    },
+    {
+      "epoch": 0.163194694730772,
+      "grad_norm": 0.10910390493844033,
+      "learning_rate": 0.0001908455704735139,
+      "loss": 1.2941,
+      "step": 2264
+    },
+    {
+      "epoch": 0.16326677719310892,
+      "grad_norm": 0.11757817996175396,
+      "learning_rate": 0.00019083580932884104,
+      "loss": 1.3094,
+      "step": 2265
+    },
+    {
+      "epoch": 0.16333885965544584,
+      "grad_norm": 0.10504302893207033,
+      "learning_rate": 0.00019082604323280393,
+      "loss": 1.3281,
+      "step": 2266
+    },
+    {
+      "epoch": 0.16341094211778273,
+      "grad_norm": 0.09453054604012608,
+      "learning_rate": 0.0001908162721859349,
+      "loss": 1.3382,
+      "step": 2267
+    },
+    {
+      "epoch": 0.16348302458011965,
+      "grad_norm": 0.11300487690449207,
+      "learning_rate": 0.0001908064961887666,
+      "loss": 1.3995,
+      "step": 2268
+    },
+    {
+      "epoch": 0.16355510704245657,
+      "grad_norm": 0.14707519312355108,
+      "learning_rate": 0.00019079671524183185,
+      "loss": 1.3934,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1636271895047935,
+      "grad_norm": 0.11214487471502041,
+      "learning_rate": 0.00019078692934566385,
+      "loss": 1.3548,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1636992719671304,
+      "grad_norm": 0.10797933001137629,
+      "learning_rate": 0.00019077713850079598,
+      "loss": 1.2645,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1637713544294673,
+      "grad_norm": 0.11811328014320788,
+      "learning_rate": 0.00019076734270776196,
+      "loss": 1.2918,
+      "step": 2272
+    },
+    {
+      "epoch": 0.16384343689180422,
+      "grad_norm": 0.11266657368281144,
+      "learning_rate": 0.00019075754196709572,
+      "loss": 1.2551,
+      "step": 2273
+    },
+    {
+      "epoch": 0.16391551935414114,
+      "grad_norm": 0.10524943213298055,
+      "learning_rate": 0.0001907477362793315,
+      "loss": 1.2786,
+      "step": 2274
+    },
+    {
+      "epoch": 0.16398760181647806,
+      "grad_norm": 0.1602762259947097,
+      "learning_rate": 0.0001907379256450038,
+      "loss": 1.2342,
+      "step": 2275
+    },
+    {
+      "epoch": 0.16405968427881495,
+      "grad_norm": 0.09834850941403127,
+      "learning_rate": 0.00019072811006464743,
+      "loss": 1.3173,
+      "step": 2276
+    },
+    {
+      "epoch": 0.16413176674115187,
+      "grad_norm": 0.10173171059724835,
+      "learning_rate": 0.00019071828953879733,
+      "loss": 1.3124,
+      "step": 2277
+    },
+    {
+      "epoch": 0.1642038492034888,
+      "grad_norm": 0.11022970607565434,
+      "learning_rate": 0.0001907084640679889,
+      "loss": 1.3406,
+      "step": 2278
+    },
+    {
+      "epoch": 0.16427593166582571,
+      "grad_norm": 0.10745925499068965,
+      "learning_rate": 0.00019069863365275765,
+      "loss": 1.3539,
+      "step": 2279
+    },
+    {
+      "epoch": 0.1643480141281626,
+      "grad_norm": 0.0993640059084939,
+      "learning_rate": 0.00019068879829363947,
+      "loss": 1.4145,
+      "step": 2280
+    },
+    {
+      "epoch": 0.16442009659049953,
+      "grad_norm": 0.10873972933351436,
+      "learning_rate": 0.00019067895799117045,
+      "loss": 1.2885,
+      "step": 2281
+    },
+    {
+      "epoch": 0.16449217905283645,
+      "grad_norm": 0.11021350112832706,
+      "learning_rate": 0.00019066911274588698,
+      "loss": 1.4032,
+      "step": 2282
+    },
+    {
+      "epoch": 0.16456426151517337,
+      "grad_norm": 0.12636127787116547,
+      "learning_rate": 0.00019065926255832573,
+      "loss": 1.3965,
+      "step": 2283
+    },
+    {
+      "epoch": 0.16463634397751029,
+      "grad_norm": 0.11183619944374593,
+      "learning_rate": 0.00019064940742902358,
+      "loss": 1.3952,
+      "step": 2284
+    },
+    {
+      "epoch": 0.16470842643984718,
+      "grad_norm": 0.10610529121759578,
+      "learning_rate": 0.00019063954735851777,
+      "loss": 1.3626,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1647805089021841,
+      "grad_norm": 0.11867025344395259,
+      "learning_rate": 0.00019062968234734577,
+      "loss": 1.4053,
+      "step": 2286
+    },
+    {
+      "epoch": 0.16485259136452102,
+      "grad_norm": 0.19173081165106665,
+      "learning_rate": 0.00019061981239604524,
+      "loss": 1.347,
+      "step": 2287
+    },
+    {
+      "epoch": 0.16492467382685794,
+      "grad_norm": 0.11030595486817149,
+      "learning_rate": 0.00019060993750515427,
+      "loss": 1.3099,
+      "step": 2288
+    },
+    {
+      "epoch": 0.16499675628919483,
+      "grad_norm": 0.10952906462086891,
+      "learning_rate": 0.00019060005767521104,
+      "loss": 1.3635,
+      "step": 2289
+    },
+    {
+      "epoch": 0.16506883875153175,
+      "grad_norm": 0.09923218594276231,
+      "learning_rate": 0.00019059017290675416,
+      "loss": 1.2881,
+      "step": 2290
+    },
+    {
+      "epoch": 0.16514092121386867,
+      "grad_norm": 0.1423993225892486,
+      "learning_rate": 0.0001905802832003224,
+      "loss": 1.2536,
+      "step": 2291
+    },
+    {
+      "epoch": 0.1652130036762056,
+      "grad_norm": 0.1266937381014742,
+      "learning_rate": 0.00019057038855645487,
+      "loss": 1.2767,
+      "step": 2292
+    },
+    {
+      "epoch": 0.16528508613854248,
+      "grad_norm": 0.10537723413946708,
+      "learning_rate": 0.00019056048897569093,
+      "loss": 1.2728,
+      "step": 2293
+    },
+    {
+      "epoch": 0.1653571686008794,
+      "grad_norm": 0.17832411603140957,
+      "learning_rate": 0.00019055058445857013,
+      "loss": 1.3527,
+      "step": 2294
+    },
+    {
+      "epoch": 0.16542925106321632,
+      "grad_norm": 0.10378249185523127,
+      "learning_rate": 0.00019054067500563236,
+      "loss": 1.3486,
+      "step": 2295
+    },
+    {
+      "epoch": 0.16550133352555324,
+      "grad_norm": 0.10323398689921269,
+      "learning_rate": 0.00019053076061741783,
+      "loss": 1.3598,
+      "step": 2296
+    },
+    {
+      "epoch": 0.16557341598789016,
+      "grad_norm": 0.10592501431515637,
+      "learning_rate": 0.00019052084129446694,
+      "loss": 1.2818,
+      "step": 2297
+    },
+    {
+      "epoch": 0.16564549845022705,
+      "grad_norm": 0.10606867152665175,
+      "learning_rate": 0.00019051091703732037,
+      "loss": 1.3712,
+      "step": 2298
+    },
+    {
+      "epoch": 0.16571758091256397,
+      "grad_norm": 0.11147109856549034,
+      "learning_rate": 0.0001905009878465191,
+      "loss": 1.3015,
+      "step": 2299
+    },
+    {
+      "epoch": 0.1657896633749009,
+      "grad_norm": 0.1103198020614005,
+      "learning_rate": 0.00019049105372260432,
+      "loss": 1.3278,
+      "step": 2300
+    },
+    {
+      "epoch": 0.1658617458372378,
+      "grad_norm": 0.11685687145788297,
+      "learning_rate": 0.00019048111466611758,
+      "loss": 1.374,
+      "step": 2301
+    },
+    {
+      "epoch": 0.1659338282995747,
+      "grad_norm": 0.10146062696953212,
+      "learning_rate": 0.0001904711706776006,
+      "loss": 1.2629,
+      "step": 2302
+    },
+    {
+      "epoch": 0.16600591076191162,
+      "grad_norm": 0.1034616147775385,
+      "learning_rate": 0.00019046122175759547,
+      "loss": 1.299,
+      "step": 2303
+    },
+    {
+      "epoch": 0.16607799322424854,
+      "grad_norm": 0.14034312418910336,
+      "learning_rate": 0.00019045126790664444,
+      "loss": 1.3735,
+      "step": 2304
+    },
+    {
+      "epoch": 0.16615007568658546,
+      "grad_norm": 0.11450457233443563,
+      "learning_rate": 0.00019044130912529013,
+      "loss": 1.401,
+      "step": 2305
+    },
+    {
+      "epoch": 0.16622215814892236,
+      "grad_norm": 0.09523606997504766,
+      "learning_rate": 0.00019043134541407532,
+      "loss": 1.339,
+      "step": 2306
+    },
+    {
+      "epoch": 0.16629424061125928,
+      "grad_norm": 0.11323990392459625,
+      "learning_rate": 0.00019042137677354322,
+      "loss": 1.3985,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1663663230735962,
+      "grad_norm": 0.10355373293023457,
+      "learning_rate": 0.00019041140320423712,
+      "loss": 1.407,
+      "step": 2308
+    },
+    {
+      "epoch": 0.16643840553593311,
+      "grad_norm": 0.10979753282552066,
+      "learning_rate": 0.0001904014247067007,
+      "loss": 1.354,
+      "step": 2309
+    },
+    {
+      "epoch": 0.16651048799827003,
+      "grad_norm": 0.10007091037493397,
+      "learning_rate": 0.00019039144128147787,
+      "loss": 1.4112,
+      "step": 2310
+    },
+    {
+      "epoch": 0.16658257046060693,
+      "grad_norm": 0.10733273591583792,
+      "learning_rate": 0.00019038145292911285,
+      "loss": 1.3232,
+      "step": 2311
+    },
+    {
+      "epoch": 0.16665465292294385,
+      "grad_norm": 0.11065642500234328,
+      "learning_rate": 0.00019037145965015005,
+      "loss": 1.3934,
+      "step": 2312
+    },
+    {
+      "epoch": 0.16672673538528077,
+      "grad_norm": 0.11646794476523854,
+      "learning_rate": 0.00019036146144513425,
+      "loss": 1.3901,
+      "step": 2313
+    },
+    {
+      "epoch": 0.1667988178476177,
+      "grad_norm": 0.09737071723594527,
+      "learning_rate": 0.0001903514583146104,
+      "loss": 1.3888,
+      "step": 2314
+    },
+    {
+      "epoch": 0.16687090030995458,
+      "grad_norm": 0.09754916995525041,
+      "learning_rate": 0.0001903414502591237,
+      "loss": 1.3699,
+      "step": 2315
+    },
+    {
+      "epoch": 0.1669429827722915,
+      "grad_norm": 0.10300558494986359,
+      "learning_rate": 0.0001903314372792198,
+      "loss": 1.3565,
+      "step": 2316
+    },
+    {
+      "epoch": 0.16701506523462842,
+      "grad_norm": 0.10767291284292646,
+      "learning_rate": 0.00019032141937544444,
+      "loss": 1.1652,
+      "step": 2317
+    },
+    {
+      "epoch": 0.16708714769696534,
+      "grad_norm": 0.11083005918768549,
+      "learning_rate": 0.00019031139654834367,
+      "loss": 1.3385,
+      "step": 2318
+    },
+    {
+      "epoch": 0.16715923015930223,
+      "grad_norm": 0.11647202869802076,
+      "learning_rate": 0.00019030136879846383,
+      "loss": 1.3807,
+      "step": 2319
+    },
+    {
+      "epoch": 0.16723131262163915,
+      "grad_norm": 0.10886171911117307,
+      "learning_rate": 0.00019029133612635154,
+      "loss": 1.3703,
+      "step": 2320
+    },
+    {
+      "epoch": 0.16730339508397607,
+      "grad_norm": 0.10462985688118806,
+      "learning_rate": 0.00019028129853255367,
+      "loss": 1.3183,
+      "step": 2321
+    },
+    {
+      "epoch": 0.167375477546313,
+      "grad_norm": 0.10485715690015564,
+      "learning_rate": 0.00019027125601761734,
+      "loss": 1.2759,
+      "step": 2322
+    },
+    {
+      "epoch": 0.16744756000864988,
+      "grad_norm": 0.11672031445876088,
+      "learning_rate": 0.00019026120858208997,
+      "loss": 1.4149,
+      "step": 2323
+    },
+    {
+      "epoch": 0.1675196424709868,
+      "grad_norm": 0.10880660968051044,
+      "learning_rate": 0.00019025115622651922,
+      "loss": 1.3403,
+      "step": 2324
+    },
+    {
+      "epoch": 0.16759172493332372,
+      "grad_norm": 0.11525855926499598,
+      "learning_rate": 0.00019024109895145308,
+      "loss": 1.2851,
+      "step": 2325
+    },
+    {
+      "epoch": 0.16766380739566064,
+      "grad_norm": 0.11847438322249719,
+      "learning_rate": 0.00019023103675743972,
+      "loss": 1.3204,
+      "step": 2326
+    },
+    {
+      "epoch": 0.16773588985799756,
+      "grad_norm": 0.1001653422552871,
+      "learning_rate": 0.0001902209696450276,
+      "loss": 1.3156,
+      "step": 2327
+    },
+    {
+      "epoch": 0.16780797232033445,
+      "grad_norm": 0.11298648720233494,
+      "learning_rate": 0.0001902108976147655,
+      "loss": 1.3098,
+      "step": 2328
+    },
+    {
+      "epoch": 0.16788005478267137,
+      "grad_norm": 0.15750518210036232,
+      "learning_rate": 0.00019020082066720245,
+      "loss": 1.2904,
+      "step": 2329
+    },
+    {
+      "epoch": 0.1679521372450083,
+      "grad_norm": 0.11936665929545469,
+      "learning_rate": 0.0001901907388028877,
+      "loss": 1.3749,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1680242197073452,
+      "grad_norm": 0.4227746485592971,
+      "learning_rate": 0.00019018065202237083,
+      "loss": 1.3315,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1680963021696821,
+      "grad_norm": 0.10185936878671235,
+      "learning_rate": 0.00019017056032620162,
+      "loss": 1.2578,
+      "step": 2332
+    },
+    {
+      "epoch": 0.16816838463201902,
+      "grad_norm": 0.10619870694037074,
+      "learning_rate": 0.0001901604637149302,
+      "loss": 1.3748,
+      "step": 2333
+    },
+    {
+      "epoch": 0.16824046709435594,
+      "grad_norm": 0.1175244746445815,
+      "learning_rate": 0.0001901503621891069,
+      "loss": 1.2851,
+      "step": 2334
+    },
+    {
+      "epoch": 0.16831254955669286,
+      "grad_norm": 0.11372838450555306,
+      "learning_rate": 0.00019014025574928237,
+      "loss": 1.3904,
+      "step": 2335
+    },
+    {
+      "epoch": 0.16838463201902976,
+      "grad_norm": 0.10070903801452265,
+      "learning_rate": 0.00019013014439600746,
+      "loss": 1.3727,
+      "step": 2336
+    },
+    {
+      "epoch": 0.16845671448136668,
+      "grad_norm": 0.09959220399886252,
+      "learning_rate": 0.00019012002812983337,
+      "loss": 1.3191,
+      "step": 2337
+    },
+    {
+      "epoch": 0.1685287969437036,
+      "grad_norm": 0.1060800504052883,
+      "learning_rate": 0.00019010990695131152,
+      "loss": 1.3179,
+      "step": 2338
+    },
+    {
+      "epoch": 0.16860087940604052,
+      "grad_norm": 0.11630994722342106,
+      "learning_rate": 0.0001900997808609936,
+      "loss": 1.3485,
+      "step": 2339
+    },
+    {
+      "epoch": 0.16867296186837744,
+      "grad_norm": 0.10812462492756587,
+      "learning_rate": 0.00019008964985943151,
+      "loss": 1.2625,
+      "step": 2340
+    },
+    {
+      "epoch": 0.16874504433071433,
+      "grad_norm": 0.1294356576413497,
+      "learning_rate": 0.0001900795139471776,
+      "loss": 1.3445,
+      "step": 2341
+    },
+    {
+      "epoch": 0.16881712679305125,
+      "grad_norm": 0.17484783610554397,
+      "learning_rate": 0.00019006937312478428,
+      "loss": 1.3346,
+      "step": 2342
+    },
+    {
+      "epoch": 0.16888920925538817,
+      "grad_norm": 0.11592221908413837,
+      "learning_rate": 0.00019005922739280434,
+      "loss": 1.2707,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1689612917177251,
+      "grad_norm": 0.11106275047832805,
+      "learning_rate": 0.00019004907675179082,
+      "loss": 1.3559,
+      "step": 2344
+    },
+    {
+      "epoch": 0.16903337418006198,
+      "grad_norm": 0.11122511187301187,
+      "learning_rate": 0.000190038921202297,
+      "loss": 1.3836,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1691054566423989,
+      "grad_norm": 0.1131774657584591,
+      "learning_rate": 0.00019002876074487648,
+      "loss": 1.2307,
+      "step": 2346
+    },
+    {
+      "epoch": 0.16917753910473582,
+      "grad_norm": 0.1072025867516343,
+      "learning_rate": 0.0001900185953800831,
+      "loss": 1.3486,
+      "step": 2347
+    },
+    {
+      "epoch": 0.16924962156707274,
+      "grad_norm": 0.104782442507731,
+      "learning_rate": 0.0001900084251084709,
+      "loss": 1.356,
+      "step": 2348
+    },
+    {
+      "epoch": 0.16932170402940963,
+      "grad_norm": 0.13230807235341116,
+      "learning_rate": 0.0001899982499305943,
+      "loss": 1.3601,
+      "step": 2349
+    },
+    {
+      "epoch": 0.16939378649174655,
+      "grad_norm": 0.10114662188870224,
+      "learning_rate": 0.00018998806984700796,
+      "loss": 1.3237,
+      "step": 2350
+    },
+    {
+      "epoch": 0.16946586895408347,
+      "grad_norm": 0.10065753053345457,
+      "learning_rate": 0.00018997788485826674,
+      "loss": 1.184,
+      "step": 2351
+    },
+    {
+      "epoch": 0.1695379514164204,
+      "grad_norm": 0.11637794252820242,
+      "learning_rate": 0.0001899676949649258,
+      "loss": 1.3775,
+      "step": 2352
+    },
+    {
+      "epoch": 0.1696100338787573,
+      "grad_norm": 0.10605417750145486,
+      "learning_rate": 0.00018995750016754067,
+      "loss": 1.2989,
+      "step": 2353
+    },
+    {
+      "epoch": 0.1696821163410942,
+      "grad_norm": 0.12196465771367535,
+      "learning_rate": 0.00018994730046666694,
+      "loss": 1.2852,
+      "step": 2354
+    },
+    {
+      "epoch": 0.16975419880343112,
+      "grad_norm": 0.10516002114908952,
+      "learning_rate": 0.0001899370958628607,
+      "loss": 1.2573,
+      "step": 2355
+    },
+    {
+      "epoch": 0.16982628126576804,
+      "grad_norm": 0.11256020992864664,
+      "learning_rate": 0.00018992688635667811,
+      "loss": 1.364,
+      "step": 2356
+    },
+    {
+      "epoch": 0.16989836372810496,
+      "grad_norm": 0.10806376892483979,
+      "learning_rate": 0.00018991667194867567,
+      "loss": 1.3062,
+      "step": 2357
+    },
+    {
+      "epoch": 0.16997044619044185,
+      "grad_norm": 0.1100134898334168,
+      "learning_rate": 0.00018990645263941021,
+      "loss": 1.3233,
+      "step": 2358
+    },
+    {
+      "epoch": 0.17004252865277877,
+      "grad_norm": 0.12669471405153027,
+      "learning_rate": 0.0001898962284294388,
+      "loss": 1.3596,
+      "step": 2359
+    },
+    {
+      "epoch": 0.1701146111151157,
+      "grad_norm": 0.11579392129044734,
+      "learning_rate": 0.00018988599931931866,
+      "loss": 1.2557,
+      "step": 2360
+    },
+    {
+      "epoch": 0.1701866935774526,
+      "grad_norm": 0.30098888794094547,
+      "learning_rate": 0.00018987576530960743,
+      "loss": 1.3916,
+      "step": 2361
+    },
+    {
+      "epoch": 0.1702587760397895,
+      "grad_norm": 0.10709760620121467,
+      "learning_rate": 0.00018986552640086295,
+      "loss": 1.328,
+      "step": 2362
+    },
+    {
+      "epoch": 0.17033085850212643,
+      "grad_norm": 0.11064614297960002,
+      "learning_rate": 0.0001898552825936433,
+      "loss": 1.2561,
+      "step": 2363
+    },
+    {
+      "epoch": 0.17040294096446335,
+      "grad_norm": 0.1000080734804425,
+      "learning_rate": 0.00018984503388850686,
+      "loss": 1.3669,
+      "step": 2364
+    },
+    {
+      "epoch": 0.17047502342680027,
+      "grad_norm": 0.11580471606888766,
+      "learning_rate": 0.00018983478028601234,
+      "loss": 1.3788,
+      "step": 2365
+    },
+    {
+      "epoch": 0.17054710588913718,
+      "grad_norm": 0.1199690253908704,
+      "learning_rate": 0.0001898245217867186,
+      "loss": 1.3451,
+      "step": 2366
+    },
+    {
+      "epoch": 0.17061918835147408,
+      "grad_norm": 0.14144218779656886,
+      "learning_rate": 0.00018981425839118484,
+      "loss": 1.2157,
+      "step": 2367
+    },
+    {
+      "epoch": 0.170691270813811,
+      "grad_norm": 0.11471974764254243,
+      "learning_rate": 0.00018980399009997046,
+      "loss": 1.4217,
+      "step": 2368
+    },
+    {
+      "epoch": 0.17076335327614792,
+      "grad_norm": 0.11299013826588457,
+      "learning_rate": 0.00018979371691363523,
+      "loss": 1.2637,
+      "step": 2369
+    },
+    {
+      "epoch": 0.17083543573848484,
+      "grad_norm": 0.10810610147996418,
+      "learning_rate": 0.0001897834388327391,
+      "loss": 1.3725,
+      "step": 2370
+    },
+    {
+      "epoch": 0.17090751820082173,
+      "grad_norm": 0.1000309482414558,
+      "learning_rate": 0.00018977315585784235,
+      "loss": 1.2906,
+      "step": 2371
+    },
+    {
+      "epoch": 0.17097960066315865,
+      "grad_norm": 0.13696764291784647,
+      "learning_rate": 0.00018976286798950547,
+      "loss": 1.2726,
+      "step": 2372
+    },
+    {
+      "epoch": 0.17105168312549557,
+      "grad_norm": 0.10685483340902513,
+      "learning_rate": 0.0001897525752282892,
+      "loss": 1.1816,
+      "step": 2373
+    },
+    {
+      "epoch": 0.1711237655878325,
+      "grad_norm": 0.12987410968474505,
+      "learning_rate": 0.0001897422775747547,
+      "loss": 1.2996,
+      "step": 2374
+    },
+    {
+      "epoch": 0.17119584805016938,
+      "grad_norm": 0.13064294216417674,
+      "learning_rate": 0.00018973197502946315,
+      "loss": 1.3012,
+      "step": 2375
+    },
+    {
+      "epoch": 0.1712679305125063,
+      "grad_norm": 0.11165163292911214,
+      "learning_rate": 0.00018972166759297622,
+      "loss": 1.2417,
+      "step": 2376
+    },
+    {
+      "epoch": 0.17134001297484322,
+      "grad_norm": 0.13274445917402272,
+      "learning_rate": 0.00018971135526585573,
+      "loss": 1.399,
+      "step": 2377
+    },
+    {
+      "epoch": 0.17141209543718014,
+      "grad_norm": 0.10880313521458199,
+      "learning_rate": 0.00018970103804866382,
+      "loss": 1.3002,
+      "step": 2378
+    },
+    {
+      "epoch": 0.17148417789951706,
+      "grad_norm": 0.12270359230211916,
+      "learning_rate": 0.00018969071594196284,
+      "loss": 1.2202,
+      "step": 2379
+    },
+    {
+      "epoch": 0.17155626036185395,
+      "grad_norm": 0.13027838982480866,
+      "learning_rate": 0.00018968038894631544,
+      "loss": 1.2673,
+      "step": 2380
+    },
+    {
+      "epoch": 0.17162834282419087,
+      "grad_norm": 0.10411068580452171,
+      "learning_rate": 0.00018967005706228452,
+      "loss": 1.3672,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1717004252865278,
+      "grad_norm": 0.09551160227645628,
+      "learning_rate": 0.00018965972029043332,
+      "loss": 1.3813,
+      "step": 2382
+    },
+    {
+      "epoch": 0.1717725077488647,
+      "grad_norm": 0.11094427963539386,
+      "learning_rate": 0.0001896493786313252,
+      "loss": 1.3225,
+      "step": 2383
+    },
+    {
+      "epoch": 0.1718445902112016,
+      "grad_norm": 0.10547135642904562,
+      "learning_rate": 0.00018963903208552393,
+      "loss": 1.279,
+      "step": 2384
+    },
+    {
+      "epoch": 0.17191667267353852,
+      "grad_norm": 0.10459366911446097,
+      "learning_rate": 0.0001896286806535935,
+      "loss": 1.3028,
+      "step": 2385
+    },
+    {
+      "epoch": 0.17198875513587544,
+      "grad_norm": 0.10659034830348292,
+      "learning_rate": 0.0001896183243360981,
+      "loss": 1.3569,
+      "step": 2386
+    },
+    {
+      "epoch": 0.17206083759821236,
+      "grad_norm": 0.11195537819224922,
+      "learning_rate": 0.0001896079631336023,
+      "loss": 1.2908,
+      "step": 2387
+    },
+    {
+      "epoch": 0.17213292006054925,
+      "grad_norm": 0.10721511258132803,
+      "learning_rate": 0.00018959759704667087,
+      "loss": 1.267,
+      "step": 2388
+    },
+    {
+      "epoch": 0.17220500252288617,
+      "grad_norm": 0.11867117423487444,
+      "learning_rate": 0.0001895872260758688,
+      "loss": 1.3314,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1722770849852231,
+      "grad_norm": 0.11417482113110625,
+      "learning_rate": 0.00018957685022176146,
+      "loss": 1.2828,
+      "step": 2390
+    },
+    {
+      "epoch": 0.17234916744756001,
+      "grad_norm": 0.10886196052797992,
+      "learning_rate": 0.0001895664694849144,
+      "loss": 1.3972,
+      "step": 2391
+    },
+    {
+      "epoch": 0.17242124990989693,
+      "grad_norm": 0.10931467096950315,
+      "learning_rate": 0.00018955608386589346,
+      "loss": 1.3786,
+      "step": 2392
+    },
+    {
+      "epoch": 0.17249333237223383,
+      "grad_norm": 0.10467663867309672,
+      "learning_rate": 0.00018954569336526478,
+      "loss": 1.2795,
+      "step": 2393
+    },
+    {
+      "epoch": 0.17256541483457075,
+      "grad_norm": 0.15713229182053,
+      "learning_rate": 0.0001895352979835947,
+      "loss": 1.37,
+      "step": 2394
+    },
+    {
+      "epoch": 0.17263749729690767,
+      "grad_norm": 0.12916011045175332,
+      "learning_rate": 0.00018952489772144988,
+      "loss": 1.3384,
+      "step": 2395
+    },
+    {
+      "epoch": 0.17270957975924459,
+      "grad_norm": 0.11732109004816138,
+      "learning_rate": 0.0001895144925793972,
+      "loss": 1.3721,
+      "step": 2396
+    },
+    {
+      "epoch": 0.17278166222158148,
+      "grad_norm": 0.1030431759096412,
+      "learning_rate": 0.00018950408255800387,
+      "loss": 1.3564,
+      "step": 2397
+    },
+    {
+      "epoch": 0.1728537446839184,
+      "grad_norm": 0.1009264420213361,
+      "learning_rate": 0.00018949366765783732,
+      "loss": 1.3124,
+      "step": 2398
+    },
+    {
+      "epoch": 0.17292582714625532,
+      "grad_norm": 0.11304517822240677,
+      "learning_rate": 0.00018948324787946525,
+      "loss": 1.4139,
+      "step": 2399
+    },
+    {
+      "epoch": 0.17299790960859224,
+      "grad_norm": 0.10666880440107447,
+      "learning_rate": 0.00018947282322345562,
+      "loss": 1.2799,
+      "step": 2400
+    },
+    {
+      "epoch": 0.17306999207092913,
+      "grad_norm": 0.10496156778282678,
+      "learning_rate": 0.00018946239369037668,
+      "loss": 1.3078,
+      "step": 2401
+    },
+    {
+      "epoch": 0.17314207453326605,
+      "grad_norm": 0.10802005514659688,
+      "learning_rate": 0.00018945195928079693,
+      "loss": 1.2936,
+      "step": 2402
+    },
+    {
+      "epoch": 0.17321415699560297,
+      "grad_norm": 0.10131607351403857,
+      "learning_rate": 0.00018944151999528516,
+      "loss": 1.2716,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1732862394579399,
+      "grad_norm": 0.09572074757607409,
+      "learning_rate": 0.00018943107583441036,
+      "loss": 1.3251,
+      "step": 2404
+    },
+    {
+      "epoch": 0.1733583219202768,
+      "grad_norm": 0.10214927399635566,
+      "learning_rate": 0.00018942062679874186,
+      "loss": 1.3504,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1734304043826137,
+      "grad_norm": 0.11771339729517416,
+      "learning_rate": 0.0001894101728888492,
+      "loss": 1.3455,
+      "step": 2406
+    },
+    {
+      "epoch": 0.17350248684495062,
+      "grad_norm": 0.1137384657915831,
+      "learning_rate": 0.00018939971410530224,
+      "loss": 1.3453,
+      "step": 2407
+    },
+    {
+      "epoch": 0.17357456930728754,
+      "grad_norm": 0.11206234754146123,
+      "learning_rate": 0.0001893892504486711,
+      "loss": 1.4029,
+      "step": 2408
+    },
+    {
+      "epoch": 0.17364665176962446,
+      "grad_norm": 0.10917520074048025,
+      "learning_rate": 0.00018937878191952606,
+      "loss": 1.259,
+      "step": 2409
+    },
+    {
+      "epoch": 0.17371873423196135,
+      "grad_norm": 0.11810741480903009,
+      "learning_rate": 0.00018936830851843785,
+      "loss": 1.3961,
+      "step": 2410
+    },
+    {
+      "epoch": 0.17379081669429827,
+      "grad_norm": 0.10449794976282009,
+      "learning_rate": 0.0001893578302459773,
+      "loss": 1.2892,
+      "step": 2411
+    },
+    {
+      "epoch": 0.1738628991566352,
+      "grad_norm": 0.1719735601348386,
+      "learning_rate": 0.00018934734710271558,
+      "loss": 1.2736,
+      "step": 2412
+    },
+    {
+      "epoch": 0.1739349816189721,
+      "grad_norm": 0.38634950783894223,
+      "learning_rate": 0.00018933685908922412,
+      "loss": 1.2696,
+      "step": 2413
+    },
+    {
+      "epoch": 0.174007064081309,
+      "grad_norm": 0.09599723227561507,
+      "learning_rate": 0.0001893263662060746,
+      "loss": 1.3283,
+      "step": 2414
+    },
+    {
+      "epoch": 0.17407914654364592,
+      "grad_norm": 0.10415218139644784,
+      "learning_rate": 0.00018931586845383897,
+      "loss": 1.2246,
+      "step": 2415
+    },
+    {
+      "epoch": 0.17415122900598284,
+      "grad_norm": 0.0957479292601988,
+      "learning_rate": 0.0001893053658330895,
+      "loss": 1.2587,
+      "step": 2416
+    },
+    {
+      "epoch": 0.17422331146831976,
+      "grad_norm": 0.10223836230145628,
+      "learning_rate": 0.00018929485834439863,
+      "loss": 1.3334,
+      "step": 2417
+    },
+    {
+      "epoch": 0.17429539393065668,
+      "grad_norm": 0.10761822394528074,
+      "learning_rate": 0.00018928434598833912,
+      "loss": 1.4339,
+      "step": 2418
+    },
+    {
+      "epoch": 0.17436747639299358,
+      "grad_norm": 0.16929916208509724,
+      "learning_rate": 0.000189273828765484,
+      "loss": 1.4306,
+      "step": 2419
+    },
+    {
+      "epoch": 0.1744395588553305,
+      "grad_norm": 0.10739904568840619,
+      "learning_rate": 0.00018926330667640655,
+      "loss": 1.2831,
+      "step": 2420
+    },
+    {
+      "epoch": 0.17451164131766742,
+      "grad_norm": 0.09991501855475501,
+      "learning_rate": 0.0001892527797216803,
+      "loss": 1.3725,
+      "step": 2421
+    },
+    {
+      "epoch": 0.17458372378000434,
+      "grad_norm": 0.09658827833164386,
+      "learning_rate": 0.0001892422479018791,
+      "loss": 1.2787,
+      "step": 2422
+    },
+    {
+      "epoch": 0.17465580624234123,
+      "grad_norm": 0.09937548565350882,
+      "learning_rate": 0.00018923171121757698,
+      "loss": 1.3216,
+      "step": 2423
+    },
+    {
+      "epoch": 0.17472788870467815,
+      "grad_norm": 0.11423052561723335,
+      "learning_rate": 0.00018922116966934834,
+      "loss": 1.3876,
+      "step": 2424
+    },
+    {
+      "epoch": 0.17479997116701507,
+      "grad_norm": 0.09973503800272224,
+      "learning_rate": 0.00018921062325776773,
+      "loss": 1.1873,
+      "step": 2425
+    },
+    {
+      "epoch": 0.174872053629352,
+      "grad_norm": 0.11292735248527326,
+      "learning_rate": 0.00018920007198341008,
+      "loss": 1.2953,
+      "step": 2426
+    },
+    {
+      "epoch": 0.17494413609168888,
+      "grad_norm": 0.11415928822194911,
+      "learning_rate": 0.00018918951584685046,
+      "loss": 1.2331,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1750162185540258,
+      "grad_norm": 0.10431104955841336,
+      "learning_rate": 0.00018917895484866435,
+      "loss": 1.3555,
+      "step": 2428
+    },
+    {
+      "epoch": 0.17508830101636272,
+      "grad_norm": 0.11721870576172148,
+      "learning_rate": 0.00018916838898942734,
+      "loss": 1.2966,
+      "step": 2429
+    },
+    {
+      "epoch": 0.17516038347869964,
+      "grad_norm": 0.13702134812224412,
+      "learning_rate": 0.00018915781826971543,
+      "loss": 1.2593,
+      "step": 2430
+    },
+    {
+      "epoch": 0.17523246594103656,
+      "grad_norm": 0.10555003120211448,
+      "learning_rate": 0.00018914724269010483,
+      "loss": 1.3451,
+      "step": 2431
+    },
+    {
+      "epoch": 0.17530454840337345,
+      "grad_norm": 0.10527917638878491,
+      "learning_rate": 0.00018913666225117194,
+      "loss": 1.3055,
+      "step": 2432
+    },
+    {
+      "epoch": 0.17537663086571037,
+      "grad_norm": 0.14568984795401166,
+      "learning_rate": 0.00018912607695349348,
+      "loss": 1.1304,
+      "step": 2433
+    },
+    {
+      "epoch": 0.1754487133280473,
+      "grad_norm": 0.11838605331324555,
+      "learning_rate": 0.00018911548679764653,
+      "loss": 1.3724,
+      "step": 2434
+    },
+    {
+      "epoch": 0.1755207957903842,
+      "grad_norm": 0.09733559145709914,
+      "learning_rate": 0.0001891048917842083,
+      "loss": 1.2739,
+      "step": 2435
+    },
+    {
+      "epoch": 0.1755928782527211,
+      "grad_norm": 0.11561016905432972,
+      "learning_rate": 0.0001890942919137563,
+      "loss": 1.3802,
+      "step": 2436
+    },
+    {
+      "epoch": 0.17566496071505802,
+      "grad_norm": 0.10806982530531752,
+      "learning_rate": 0.00018908368718686833,
+      "loss": 1.3128,
+      "step": 2437
+    },
+    {
+      "epoch": 0.17573704317739494,
+      "grad_norm": 0.12610309608752307,
+      "learning_rate": 0.00018907307760412243,
+      "loss": 1.2833,
+      "step": 2438
+    },
+    {
+      "epoch": 0.17580912563973186,
+      "grad_norm": 0.11824623518480805,
+      "learning_rate": 0.00018906246316609696,
+      "loss": 1.3356,
+      "step": 2439
+    },
+    {
+      "epoch": 0.17588120810206875,
+      "grad_norm": 0.10515734180016846,
+      "learning_rate": 0.00018905184387337044,
+      "loss": 1.2645,
+      "step": 2440
+    },
+    {
+      "epoch": 0.17595329056440567,
+      "grad_norm": 0.09989841236899286,
+      "learning_rate": 0.00018904121972652176,
+      "loss": 1.2992,
+      "step": 2441
+    },
+    {
+      "epoch": 0.1760253730267426,
+      "grad_norm": 0.10417592871262958,
+      "learning_rate": 0.00018903059072613004,
+      "loss": 1.3218,
+      "step": 2442
+    },
+    {
+      "epoch": 0.1760974554890795,
+      "grad_norm": 0.10191324383569109,
+      "learning_rate": 0.0001890199568727746,
+      "loss": 1.2985,
+      "step": 2443
+    },
+    {
+      "epoch": 0.17616953795141643,
+      "grad_norm": 0.10638851528072835,
+      "learning_rate": 0.00018900931816703516,
+      "loss": 1.3195,
+      "step": 2444
+    },
+    {
+      "epoch": 0.17624162041375332,
+      "grad_norm": 0.11357055633514247,
+      "learning_rate": 0.00018899867460949154,
+      "loss": 1.3835,
+      "step": 2445
+    },
+    {
+      "epoch": 0.17631370287609024,
+      "grad_norm": 0.09435851178669827,
+      "learning_rate": 0.00018898802620072398,
+      "loss": 1.4235,
+      "step": 2446
+    },
+    {
+      "epoch": 0.17638578533842716,
+      "grad_norm": 0.10677987488359683,
+      "learning_rate": 0.00018897737294131284,
+      "loss": 1.3032,
+      "step": 2447
+    },
+    {
+      "epoch": 0.17645786780076408,
+      "grad_norm": 0.11762574825860027,
+      "learning_rate": 0.0001889667148318389,
+      "loss": 1.3073,
+      "step": 2448
+    },
+    {
+      "epoch": 0.17652995026310098,
+      "grad_norm": 0.33665092789384626,
+      "learning_rate": 0.00018895605187288305,
+      "loss": 1.2992,
+      "step": 2449
+    },
+    {
+      "epoch": 0.1766020327254379,
+      "grad_norm": 0.09671481641341262,
+      "learning_rate": 0.00018894538406502657,
+      "loss": 1.3886,
+      "step": 2450
+    },
+    {
+      "epoch": 0.17667411518777482,
+      "grad_norm": 0.10732571037047337,
+      "learning_rate": 0.00018893471140885094,
+      "loss": 1.3766,
+      "step": 2451
+    },
+    {
+      "epoch": 0.17674619765011174,
+      "grad_norm": 0.1253696042524264,
+      "learning_rate": 0.0001889240339049379,
+      "loss": 1.3905,
+      "step": 2452
+    },
+    {
+      "epoch": 0.17681828011244863,
+      "grad_norm": 0.11608522016673191,
+      "learning_rate": 0.00018891335155386946,
+      "loss": 1.268,
+      "step": 2453
+    },
+    {
+      "epoch": 0.17689036257478555,
+      "grad_norm": 0.10945568040880384,
+      "learning_rate": 0.00018890266435622792,
+      "loss": 1.2813,
+      "step": 2454
+    },
+    {
+      "epoch": 0.17696244503712247,
+      "grad_norm": 0.10745386073105494,
+      "learning_rate": 0.00018889197231259584,
+      "loss": 1.3324,
+      "step": 2455
+    },
+    {
+      "epoch": 0.1770345274994594,
+      "grad_norm": 0.10730094368960749,
+      "learning_rate": 0.00018888127542355602,
+      "loss": 1.2523,
+      "step": 2456
+    },
+    {
+      "epoch": 0.1771066099617963,
+      "grad_norm": 0.10729111017660214,
+      "learning_rate": 0.0001888705736896915,
+      "loss": 1.1957,
+      "step": 2457
+    },
+    {
+      "epoch": 0.1771786924241332,
+      "grad_norm": 0.10418125114839806,
+      "learning_rate": 0.0001888598671115857,
+      "loss": 1.3197,
+      "step": 2458
+    },
+    {
+      "epoch": 0.17725077488647012,
+      "grad_norm": 0.11116747237995848,
+      "learning_rate": 0.00018884915568982217,
+      "loss": 1.3053,
+      "step": 2459
+    },
+    {
+      "epoch": 0.17732285734880704,
+      "grad_norm": 0.16033427364850117,
+      "learning_rate": 0.00018883843942498475,
+      "loss": 1.2682,
+      "step": 2460
+    },
+    {
+      "epoch": 0.17739493981114396,
+      "grad_norm": 0.1118485928447318,
+      "learning_rate": 0.00018882771831765766,
+      "loss": 1.3413,
+      "step": 2461
+    },
+    {
+      "epoch": 0.17746702227348085,
+      "grad_norm": 0.11481420378072288,
+      "learning_rate": 0.00018881699236842522,
+      "loss": 1.4174,
+      "step": 2462
+    },
+    {
+      "epoch": 0.17753910473581777,
+      "grad_norm": 0.24186295408888533,
+      "learning_rate": 0.00018880626157787212,
+      "loss": 1.3117,
+      "step": 2463
+    },
+    {
+      "epoch": 0.1776111871981547,
+      "grad_norm": 0.10420841600471048,
+      "learning_rate": 0.00018879552594658328,
+      "loss": 1.409,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1776832696604916,
+      "grad_norm": 0.11462085436653227,
+      "learning_rate": 0.00018878478547514389,
+      "loss": 1.3587,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1777553521228285,
+      "grad_norm": 0.16249519578450836,
+      "learning_rate": 0.0001887740401641394,
+      "loss": 1.3736,
+      "step": 2466
+    },
+    {
+      "epoch": 0.17782743458516542,
+      "grad_norm": 0.14358564717812727,
+      "learning_rate": 0.0001887632900141555,
+      "loss": 1.327,
+      "step": 2467
+    },
+    {
+      "epoch": 0.17789951704750234,
+      "grad_norm": 0.10750154549088661,
+      "learning_rate": 0.00018875253502577825,
+      "loss": 1.2669,
+      "step": 2468
+    },
+    {
+      "epoch": 0.17797159950983926,
+      "grad_norm": 0.10478415547603259,
+      "learning_rate": 0.0001887417751995938,
+      "loss": 1.391,
+      "step": 2469
+    },
+    {
+      "epoch": 0.17804368197217618,
+      "grad_norm": 0.12413001465816717,
+      "learning_rate": 0.0001887310105361887,
+      "loss": 1.3129,
+      "step": 2470
+    },
+    {
+      "epoch": 0.17811576443451307,
+      "grad_norm": 0.10788523985363327,
+      "learning_rate": 0.00018872024103614972,
+      "loss": 1.3537,
+      "step": 2471
+    },
+    {
+      "epoch": 0.17818784689685,
+      "grad_norm": 0.11594696410809632,
+      "learning_rate": 0.00018870946670006392,
+      "loss": 1.3089,
+      "step": 2472
+    },
+    {
+      "epoch": 0.1782599293591869,
+      "grad_norm": 0.19491941358621923,
+      "learning_rate": 0.00018869868752851852,
+      "loss": 1.3632,
+      "step": 2473
+    },
+    {
+      "epoch": 0.17833201182152383,
+      "grad_norm": 0.11616895209069801,
+      "learning_rate": 0.00018868790352210115,
+      "loss": 1.1894,
+      "step": 2474
+    },
+    {
+      "epoch": 0.17840409428386073,
+      "grad_norm": 0.11451312068499483,
+      "learning_rate": 0.00018867711468139963,
+      "loss": 1.3757,
+      "step": 2475
+    },
+    {
+      "epoch": 0.17847617674619765,
+      "grad_norm": 0.1025674298943867,
+      "learning_rate": 0.00018866632100700197,
+      "loss": 1.3988,
+      "step": 2476
+    },
+    {
+      "epoch": 0.17854825920853457,
+      "grad_norm": 0.10751468599904018,
+      "learning_rate": 0.00018865552249949663,
+      "loss": 1.3735,
+      "step": 2477
+    },
+    {
+      "epoch": 0.17862034167087149,
+      "grad_norm": 0.09779504116511331,
+      "learning_rate": 0.0001886447191594722,
+      "loss": 1.3143,
+      "step": 2478
+    },
+    {
+      "epoch": 0.17869242413320838,
+      "grad_norm": 0.11360387449676865,
+      "learning_rate": 0.00018863391098751747,
+      "loss": 1.2154,
+      "step": 2479
+    },
+    {
+      "epoch": 0.1787645065955453,
+      "grad_norm": 0.10187232482618071,
+      "learning_rate": 0.00018862309798422172,
+      "loss": 1.316,
+      "step": 2480
+    },
+    {
+      "epoch": 0.17883658905788222,
+      "grad_norm": 0.0981148952347619,
+      "learning_rate": 0.00018861228015017425,
+      "loss": 1.2131,
+      "step": 2481
+    },
+    {
+      "epoch": 0.17890867152021914,
+      "grad_norm": 0.10930082353548526,
+      "learning_rate": 0.00018860145748596477,
+      "loss": 1.353,
+      "step": 2482
+    },
+    {
+      "epoch": 0.17898075398255606,
+      "grad_norm": 0.10785418628887276,
+      "learning_rate": 0.00018859062999218319,
+      "loss": 1.323,
+      "step": 2483
+    },
+    {
+      "epoch": 0.17905283644489295,
+      "grad_norm": 0.09899134161749217,
+      "learning_rate": 0.00018857979766941975,
+      "loss": 1.2962,
+      "step": 2484
+    },
+    {
+      "epoch": 0.17912491890722987,
+      "grad_norm": 0.166565343864564,
+      "learning_rate": 0.00018856896051826487,
+      "loss": 1.2685,
+      "step": 2485
+    },
+    {
+      "epoch": 0.1791970013695668,
+      "grad_norm": 0.10359554291616108,
+      "learning_rate": 0.00018855811853930927,
+      "loss": 1.1472,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1792690838319037,
+      "grad_norm": 0.11031978732478848,
+      "learning_rate": 0.00018854727173314394,
+      "loss": 1.2722,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1793411662942406,
+      "grad_norm": 0.09764436861372536,
+      "learning_rate": 0.00018853642010036015,
+      "loss": 1.3347,
+      "step": 2488
+    },
+    {
+      "epoch": 0.17941324875657752,
+      "grad_norm": 0.1081955688760174,
+      "learning_rate": 0.0001885255636415494,
+      "loss": 1.3329,
+      "step": 2489
+    },
+    {
+      "epoch": 0.17948533121891444,
+      "grad_norm": 0.1225195734078159,
+      "learning_rate": 0.00018851470235730345,
+      "loss": 1.3104,
+      "step": 2490
+    },
+    {
+      "epoch": 0.17955741368125136,
+      "grad_norm": 0.09805287457651865,
+      "learning_rate": 0.00018850383624821433,
+      "loss": 1.2216,
+      "step": 2491
+    },
+    {
+      "epoch": 0.17962949614358825,
+      "grad_norm": 0.10310748487502187,
+      "learning_rate": 0.00018849296531487438,
+      "loss": 1.3857,
+      "step": 2492
+    },
+    {
+      "epoch": 0.17970157860592517,
+      "grad_norm": 0.11285873373577858,
+      "learning_rate": 0.00018848208955787613,
+      "loss": 1.3984,
+      "step": 2493
+    },
+    {
+      "epoch": 0.1797736610682621,
+      "grad_norm": 0.26695467428793357,
+      "learning_rate": 0.0001884712089778124,
+      "loss": 1.3587,
+      "step": 2494
+    },
+    {
+      "epoch": 0.179845743530599,
+      "grad_norm": 0.10822125703964759,
+      "learning_rate": 0.00018846032357527635,
+      "loss": 1.3514,
+      "step": 2495
+    },
+    {
+      "epoch": 0.17991782599293593,
+      "grad_norm": 0.09871163718992543,
+      "learning_rate": 0.00018844943335086123,
+      "loss": 1.3421,
+      "step": 2496
+    },
+    {
+      "epoch": 0.17998990845527282,
+      "grad_norm": 0.11327547753198218,
+      "learning_rate": 0.0001884385383051607,
+      "loss": 1.3093,
+      "step": 2497
+    },
+    {
+      "epoch": 0.18006199091760974,
+      "grad_norm": 0.12070521099574674,
+      "learning_rate": 0.00018842763843876864,
+      "loss": 1.2989,
+      "step": 2498
+    },
+    {
+      "epoch": 0.18013407337994666,
+      "grad_norm": 0.0978740137890369,
+      "learning_rate": 0.0001884167337522792,
+      "loss": 1.2743,
+      "step": 2499
+    },
+    {
+      "epoch": 0.18020615584228358,
+      "grad_norm": 0.17216681020321198,
+      "learning_rate": 0.00018840582424628678,
+      "loss": 1.3143,
+      "step": 2500
+    },
+    {
+      "epoch": 0.18027823830462048,
+      "grad_norm": 0.12053889418950581,
+      "learning_rate": 0.00018839490992138603,
+      "loss": 1.3107,
+      "step": 2501
+    },
+    {
+      "epoch": 0.1803503207669574,
+      "grad_norm": 0.11116971200673358,
+      "learning_rate": 0.00018838399077817188,
+      "loss": 1.3015,
+      "step": 2502
+    },
+    {
+      "epoch": 0.18042240322929431,
+      "grad_norm": 0.1088827934448522,
+      "learning_rate": 0.00018837306681723952,
+      "loss": 1.2983,
+      "step": 2503
+    },
+    {
+      "epoch": 0.18049448569163123,
+      "grad_norm": 0.11172954789723219,
+      "learning_rate": 0.00018836213803918444,
+      "loss": 1.3246,
+      "step": 2504
+    },
+    {
+      "epoch": 0.18056656815396813,
+      "grad_norm": 0.19458341768697587,
+      "learning_rate": 0.0001883512044446023,
+      "loss": 1.3004,
+      "step": 2505
+    },
+    {
+      "epoch": 0.18063865061630505,
+      "grad_norm": 0.1898742247730242,
+      "learning_rate": 0.00018834026603408913,
+      "loss": 1.3552,
+      "step": 2506
+    },
+    {
+      "epoch": 0.18071073307864197,
+      "grad_norm": 0.09661320942028445,
+      "learning_rate": 0.00018832932280824112,
+      "loss": 1.3232,
+      "step": 2507
+    },
+    {
+      "epoch": 0.1807828155409789,
+      "grad_norm": 0.10841872571238925,
+      "learning_rate": 0.00018831837476765482,
+      "loss": 1.2419,
+      "step": 2508
+    },
+    {
+      "epoch": 0.1808548980033158,
+      "grad_norm": 0.14039292494364342,
+      "learning_rate": 0.00018830742191292696,
+      "loss": 1.3023,
+      "step": 2509
+    },
+    {
+      "epoch": 0.1809269804656527,
+      "grad_norm": 0.11824581046107896,
+      "learning_rate": 0.00018829646424465456,
+      "loss": 1.2599,
+      "step": 2510
+    },
+    {
+      "epoch": 0.18099906292798962,
+      "grad_norm": 0.11068531327886433,
+      "learning_rate": 0.00018828550176343498,
+      "loss": 1.302,
+      "step": 2511
+    },
+    {
+      "epoch": 0.18107114539032654,
+      "grad_norm": 0.10744076957373347,
+      "learning_rate": 0.0001882745344698657,
+      "loss": 1.3632,
+      "step": 2512
+    },
+    {
+      "epoch": 0.18114322785266346,
+      "grad_norm": 0.10457423775789897,
+      "learning_rate": 0.00018826356236454462,
+      "loss": 1.3093,
+      "step": 2513
+    },
+    {
+      "epoch": 0.18121531031500035,
+      "grad_norm": 0.12432140490934494,
+      "learning_rate": 0.00018825258544806972,
+      "loss": 1.2749,
+      "step": 2514
+    },
+    {
+      "epoch": 0.18128739277733727,
+      "grad_norm": 0.09950069143389026,
+      "learning_rate": 0.00018824160372103935,
+      "loss": 1.3391,
+      "step": 2515
+    },
+    {
+      "epoch": 0.1813594752396742,
+      "grad_norm": 0.11620898091169664,
+      "learning_rate": 0.00018823061718405222,
+      "loss": 1.189,
+      "step": 2516
+    },
+    {
+      "epoch": 0.1814315577020111,
+      "grad_norm": 0.11066816211872359,
+      "learning_rate": 0.00018821962583770704,
+      "loss": 1.3938,
+      "step": 2517
+    },
+    {
+      "epoch": 0.181503640164348,
+      "grad_norm": 0.12106661567521382,
+      "learning_rate": 0.00018820862968260305,
+      "loss": 1.3104,
+      "step": 2518
+    },
+    {
+      "epoch": 0.18157572262668492,
+      "grad_norm": 0.10636710899253006,
+      "learning_rate": 0.00018819762871933967,
+      "loss": 1.3264,
+      "step": 2519
+    },
+    {
+      "epoch": 0.18164780508902184,
+      "grad_norm": 0.11276983851377465,
+      "learning_rate": 0.0001881866229485164,
+      "loss": 1.3387,
+      "step": 2520
+    },
+    {
+      "epoch": 0.18171988755135876,
+      "grad_norm": 0.10742768142014865,
+      "learning_rate": 0.00018817561237073326,
+      "loss": 1.2626,
+      "step": 2521
+    },
+    {
+      "epoch": 0.18179197001369568,
+      "grad_norm": 0.1204837651247006,
+      "learning_rate": 0.00018816459698659045,
+      "loss": 1.2541,
+      "step": 2522
+    },
+    {
+      "epoch": 0.18186405247603257,
+      "grad_norm": 0.10147141306642653,
+      "learning_rate": 0.0001881535767966883,
+      "loss": 1.4037,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1819361349383695,
+      "grad_norm": 0.1083401029344176,
+      "learning_rate": 0.00018814255180162759,
+      "loss": 1.3176,
+      "step": 2524
+    },
+    {
+      "epoch": 0.1820082174007064,
+      "grad_norm": 0.11961113264438719,
+      "learning_rate": 0.00018813152200200926,
+      "loss": 1.3254,
+      "step": 2525
+    },
+    {
+      "epoch": 0.18208029986304333,
+      "grad_norm": 0.11059810588020645,
+      "learning_rate": 0.00018812048739843455,
+      "loss": 1.3196,
+      "step": 2526
+    },
+    {
+      "epoch": 0.18215238232538022,
+      "grad_norm": 0.10360079052442586,
+      "learning_rate": 0.00018810944799150492,
+      "loss": 1.3356,
+      "step": 2527
+    },
+    {
+      "epoch": 0.18222446478771714,
+      "grad_norm": 0.10167651908579739,
+      "learning_rate": 0.00018809840378182212,
+      "loss": 1.3292,
+      "step": 2528
+    },
+    {
+      "epoch": 0.18229654725005406,
+      "grad_norm": 0.10645323297567434,
+      "learning_rate": 0.00018808735476998818,
+      "loss": 1.2796,
+      "step": 2529
+    },
+    {
+      "epoch": 0.18236862971239098,
+      "grad_norm": 0.11229000197380866,
+      "learning_rate": 0.00018807630095660533,
+      "loss": 1.3591,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18244071217472788,
+      "grad_norm": 0.11247106018285627,
+      "learning_rate": 0.00018806524234227614,
+      "loss": 1.314,
+      "step": 2531
+    },
+    {
+      "epoch": 0.1825127946370648,
+      "grad_norm": 0.11275668441530223,
+      "learning_rate": 0.00018805417892760336,
+      "loss": 1.3176,
+      "step": 2532
+    },
+    {
+      "epoch": 0.18258487709940172,
+      "grad_norm": 0.11839184864607358,
+      "learning_rate": 0.0001880431107131901,
+      "loss": 1.2327,
+      "step": 2533
+    },
+    {
+      "epoch": 0.18265695956173864,
+      "grad_norm": 0.11713160934418654,
+      "learning_rate": 0.00018803203769963967,
+      "loss": 1.3627,
+      "step": 2534
+    },
+    {
+      "epoch": 0.18272904202407556,
+      "grad_norm": 0.10501536538573199,
+      "learning_rate": 0.00018802095988755557,
+      "loss": 1.2699,
+      "step": 2535
+    },
+    {
+      "epoch": 0.18280112448641245,
+      "grad_norm": 0.22496305716538945,
+      "learning_rate": 0.00018800987727754173,
+      "loss": 1.2637,
+      "step": 2536
+    },
+    {
+      "epoch": 0.18287320694874937,
+      "grad_norm": 0.14687985580243645,
+      "learning_rate": 0.00018799878987020224,
+      "loss": 1.3825,
+      "step": 2537
+    },
+    {
+      "epoch": 0.1829452894110863,
+      "grad_norm": 0.1252478474578962,
+      "learning_rate": 0.0001879876976661414,
+      "loss": 1.3279,
+      "step": 2538
+    },
+    {
+      "epoch": 0.1830173718734232,
+      "grad_norm": 0.10911763185523793,
+      "learning_rate": 0.00018797660066596393,
+      "loss": 1.2968,
+      "step": 2539
+    },
+    {
+      "epoch": 0.1830894543357601,
+      "grad_norm": 0.09856934158902861,
+      "learning_rate": 0.0001879654988702746,
+      "loss": 1.3767,
+      "step": 2540
+    },
+    {
+      "epoch": 0.18316153679809702,
+      "grad_norm": 0.10816213442906598,
+      "learning_rate": 0.00018795439227967865,
+      "loss": 1.4076,
+      "step": 2541
+    },
+    {
+      "epoch": 0.18323361926043394,
+      "grad_norm": 0.11724520336597734,
+      "learning_rate": 0.0001879432808947815,
+      "loss": 1.2595,
+      "step": 2542
+    },
+    {
+      "epoch": 0.18330570172277086,
+      "grad_norm": 0.10043835609378642,
+      "learning_rate": 0.00018793216471618874,
+      "loss": 1.378,
+      "step": 2543
+    },
+    {
+      "epoch": 0.18337778418510775,
+      "grad_norm": 0.11753550450145342,
+      "learning_rate": 0.0001879210437445063,
+      "loss": 1.3539,
+      "step": 2544
+    },
+    {
+      "epoch": 0.18344986664744467,
+      "grad_norm": 0.13217682356730626,
+      "learning_rate": 0.00018790991798034047,
+      "loss": 1.4129,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1835219491097816,
+      "grad_norm": 0.1428029375110193,
+      "learning_rate": 0.00018789878742429762,
+      "loss": 1.3019,
+      "step": 2546
+    },
+    {
+      "epoch": 0.1835940315721185,
+      "grad_norm": 0.0980674770847328,
+      "learning_rate": 0.00018788765207698448,
+      "loss": 1.185,
+      "step": 2547
+    },
+    {
+      "epoch": 0.18366611403445543,
+      "grad_norm": 0.11383209011574885,
+      "learning_rate": 0.00018787651193900805,
+      "loss": 1.3525,
+      "step": 2548
+    },
+    {
+      "epoch": 0.18373819649679232,
+      "grad_norm": 0.1255757189855905,
+      "learning_rate": 0.00018786536701097557,
+      "loss": 1.306,
+      "step": 2549
+    },
+    {
+      "epoch": 0.18381027895912924,
+      "grad_norm": 0.11645653015138252,
+      "learning_rate": 0.0001878542172934945,
+      "loss": 1.324,
+      "step": 2550
+    },
+    {
+      "epoch": 0.18388236142146616,
+      "grad_norm": 0.108705670307555,
+      "learning_rate": 0.00018784306278717263,
+      "loss": 1.3649,
+      "step": 2551
+    },
+    {
+      "epoch": 0.18395444388380308,
+      "grad_norm": 0.10751584705784464,
+      "learning_rate": 0.00018783190349261797,
+      "loss": 1.3764,
+      "step": 2552
+    },
+    {
+      "epoch": 0.18402652634613997,
+      "grad_norm": 0.15424103173266165,
+      "learning_rate": 0.0001878207394104388,
+      "loss": 1.282,
+      "step": 2553
+    },
+    {
+      "epoch": 0.1840986088084769,
+      "grad_norm": 0.10178376979475787,
+      "learning_rate": 0.00018780957054124366,
+      "loss": 1.3564,
+      "step": 2554
+    },
+    {
+      "epoch": 0.1841706912708138,
+      "grad_norm": 0.10067399908138576,
+      "learning_rate": 0.00018779839688564138,
+      "loss": 1.3276,
+      "step": 2555
+    },
+    {
+      "epoch": 0.18424277373315073,
+      "grad_norm": 0.11597740155223166,
+      "learning_rate": 0.000187787218444241,
+      "loss": 1.3574,
+      "step": 2556
+    },
+    {
+      "epoch": 0.18431485619548763,
+      "grad_norm": 0.11258907702156301,
+      "learning_rate": 0.00018777603521765183,
+      "loss": 1.3546,
+      "step": 2557
+    },
+    {
+      "epoch": 0.18438693865782455,
+      "grad_norm": 0.11281501526509219,
+      "learning_rate": 0.00018776484720648347,
+      "loss": 1.3484,
+      "step": 2558
+    },
+    {
+      "epoch": 0.18445902112016146,
+      "grad_norm": 0.11689787429289623,
+      "learning_rate": 0.0001877536544113458,
+      "loss": 1.3684,
+      "step": 2559
+    },
+    {
+      "epoch": 0.18453110358249838,
+      "grad_norm": 0.11111801831283726,
+      "learning_rate": 0.00018774245683284889,
+      "loss": 1.3061,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1846031860448353,
+      "grad_norm": 0.11868314815713356,
+      "learning_rate": 0.0001877312544716031,
+      "loss": 1.3848,
+      "step": 2561
+    },
+    {
+      "epoch": 0.1846752685071722,
+      "grad_norm": 0.09985720391641105,
+      "learning_rate": 0.00018772004732821906,
+      "loss": 1.3422,
+      "step": 2562
+    },
+    {
+      "epoch": 0.18474735096950912,
+      "grad_norm": 0.10554201329816641,
+      "learning_rate": 0.0001877088354033077,
+      "loss": 1.2905,
+      "step": 2563
+    },
+    {
+      "epoch": 0.18481943343184604,
+      "grad_norm": 0.10635419566081855,
+      "learning_rate": 0.00018769761869748015,
+      "loss": 1.3187,
+      "step": 2564
+    },
+    {
+      "epoch": 0.18489151589418296,
+      "grad_norm": 0.11174805532495129,
+      "learning_rate": 0.0001876863972113478,
+      "loss": 1.3439,
+      "step": 2565
+    },
+    {
+      "epoch": 0.18496359835651985,
+      "grad_norm": 0.11380559474352361,
+      "learning_rate": 0.00018767517094552235,
+      "loss": 1.2888,
+      "step": 2566
+    },
+    {
+      "epoch": 0.18503568081885677,
+      "grad_norm": 0.1319201120024553,
+      "learning_rate": 0.0001876639399006157,
+      "loss": 1.3538,
+      "step": 2567
+    },
+    {
+      "epoch": 0.1851077632811937,
+      "grad_norm": 0.16172198003014232,
+      "learning_rate": 0.0001876527040772401,
+      "loss": 1.3318,
+      "step": 2568
+    },
+    {
+      "epoch": 0.1851798457435306,
+      "grad_norm": 0.10889469888337276,
+      "learning_rate": 0.00018764146347600793,
+      "loss": 1.3716,
+      "step": 2569
+    },
+    {
+      "epoch": 0.1852519282058675,
+      "grad_norm": 0.10253719305545672,
+      "learning_rate": 0.00018763021809753194,
+      "loss": 1.2354,
+      "step": 2570
+    },
+    {
+      "epoch": 0.18532401066820442,
+      "grad_norm": 0.10792302381216719,
+      "learning_rate": 0.0001876189679424251,
+      "loss": 1.302,
+      "step": 2571
+    },
+    {
+      "epoch": 0.18539609313054134,
+      "grad_norm": 0.1078631481345351,
+      "learning_rate": 0.00018760771301130067,
+      "loss": 1.2951,
+      "step": 2572
+    },
+    {
+      "epoch": 0.18546817559287826,
+      "grad_norm": 0.1873760691270662,
+      "learning_rate": 0.00018759645330477208,
+      "loss": 1.1888,
+      "step": 2573
+    },
+    {
+      "epoch": 0.18554025805521518,
+      "grad_norm": 0.10756296827677109,
+      "learning_rate": 0.00018758518882345314,
+      "loss": 1.3824,
+      "step": 2574
+    },
+    {
+      "epoch": 0.18561234051755207,
+      "grad_norm": 0.13172401971075556,
+      "learning_rate": 0.00018757391956795787,
+      "loss": 1.3329,
+      "step": 2575
+    },
+    {
+      "epoch": 0.185684422979889,
+      "grad_norm": 0.12158640720401624,
+      "learning_rate": 0.00018756264553890052,
+      "loss": 1.3512,
+      "step": 2576
+    },
+    {
+      "epoch": 0.1857565054422259,
+      "grad_norm": 0.11060658644183607,
+      "learning_rate": 0.00018755136673689562,
+      "loss": 1.4053,
+      "step": 2577
+    },
+    {
+      "epoch": 0.18582858790456283,
+      "grad_norm": 0.11263983046691389,
+      "learning_rate": 0.00018754008316255799,
+      "loss": 1.3114,
+      "step": 2578
+    },
+    {
+      "epoch": 0.18590067036689972,
+      "grad_norm": 0.1039095459057427,
+      "learning_rate": 0.00018752879481650264,
+      "loss": 1.402,
+      "step": 2579
+    },
+    {
+      "epoch": 0.18597275282923664,
+      "grad_norm": 0.11949353796202829,
+      "learning_rate": 0.00018751750169934494,
+      "loss": 1.3206,
+      "step": 2580
+    },
+    {
+      "epoch": 0.18604483529157356,
+      "grad_norm": 0.12374336647365484,
+      "learning_rate": 0.00018750620381170044,
+      "loss": 1.4561,
+      "step": 2581
+    },
+    {
+      "epoch": 0.18611691775391048,
+      "grad_norm": 0.11882439575323475,
+      "learning_rate": 0.00018749490115418498,
+      "loss": 1.3556,
+      "step": 2582
+    },
+    {
+      "epoch": 0.18618900021624737,
+      "grad_norm": 0.12107279659222799,
+      "learning_rate": 0.00018748359372741465,
+      "loss": 1.3469,
+      "step": 2583
+    },
+    {
+      "epoch": 0.1862610826785843,
+      "grad_norm": 0.10270044259472223,
+      "learning_rate": 0.00018747228153200582,
+      "loss": 1.3563,
+      "step": 2584
+    },
+    {
+      "epoch": 0.18633316514092121,
+      "grad_norm": 0.11053840318767086,
+      "learning_rate": 0.0001874609645685751,
+      "loss": 1.3688,
+      "step": 2585
+    },
+    {
+      "epoch": 0.18640524760325813,
+      "grad_norm": 0.09468945046540443,
+      "learning_rate": 0.00018744964283773935,
+      "loss": 1.3349,
+      "step": 2586
+    },
+    {
+      "epoch": 0.18647733006559505,
+      "grad_norm": 0.11926618792335138,
+      "learning_rate": 0.00018743831634011573,
+      "loss": 1.2134,
+      "step": 2587
+    },
+    {
+      "epoch": 0.18654941252793195,
+      "grad_norm": 0.11079531516580998,
+      "learning_rate": 0.00018742698507632162,
+      "loss": 1.1844,
+      "step": 2588
+    },
+    {
+      "epoch": 0.18662149499026887,
+      "grad_norm": 0.09957397483262906,
+      "learning_rate": 0.0001874156490469747,
+      "loss": 1.3758,
+      "step": 2589
+    },
+    {
+      "epoch": 0.18669357745260579,
+      "grad_norm": 0.12200347716585064,
+      "learning_rate": 0.0001874043082526928,
+      "loss": 1.3218,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1867656599149427,
+      "grad_norm": 0.1001345986594159,
+      "learning_rate": 0.00018739296269409424,
+      "loss": 1.3237,
+      "step": 2591
+    },
+    {
+      "epoch": 0.1868377423772796,
+      "grad_norm": 0.1117935162004999,
+      "learning_rate": 0.0001873816123717973,
+      "loss": 1.3226,
+      "step": 2592
+    },
+    {
+      "epoch": 0.18690982483961652,
+      "grad_norm": 0.11200845364195945,
+      "learning_rate": 0.00018737025728642078,
+      "loss": 1.4735,
+      "step": 2593
+    },
+    {
+      "epoch": 0.18698190730195344,
+      "grad_norm": 0.10873657205929131,
+      "learning_rate": 0.00018735889743858362,
+      "loss": 1.2693,
+      "step": 2594
+    },
+    {
+      "epoch": 0.18705398976429036,
+      "grad_norm": 0.10907023938675428,
+      "learning_rate": 0.00018734753282890497,
+      "loss": 1.3921,
+      "step": 2595
+    },
+    {
+      "epoch": 0.18712607222662725,
+      "grad_norm": 0.10389879361899163,
+      "learning_rate": 0.0001873361634580044,
+      "loss": 1.401,
+      "step": 2596
+    },
+    {
+      "epoch": 0.18719815468896417,
+      "grad_norm": 0.1082381083040205,
+      "learning_rate": 0.00018732478932650154,
+      "loss": 1.3434,
+      "step": 2597
+    },
+    {
+      "epoch": 0.1872702371513011,
+      "grad_norm": 0.10810197987817119,
+      "learning_rate": 0.0001873134104350165,
+      "loss": 1.2892,
+      "step": 2598
+    },
+    {
+      "epoch": 0.187342319613638,
+      "grad_norm": 0.19379961714698066,
+      "learning_rate": 0.0001873020267841694,
+      "loss": 1.3241,
+      "step": 2599
+    },
+    {
+      "epoch": 0.18741440207597493,
+      "grad_norm": 0.1153914773938588,
+      "learning_rate": 0.00018729063837458082,
+      "loss": 1.2417,
+      "step": 2600
+    },
+    {
+      "epoch": 0.18748648453831182,
+      "grad_norm": 0.11193592454745321,
+      "learning_rate": 0.00018727924520687154,
+      "loss": 1.3133,
+      "step": 2601
+    },
+    {
+      "epoch": 0.18755856700064874,
+      "grad_norm": 0.1483284918886877,
+      "learning_rate": 0.00018726784728166258,
+      "loss": 1.3213,
+      "step": 2602
+    },
+    {
+      "epoch": 0.18763064946298566,
+      "grad_norm": 0.11649661835088679,
+      "learning_rate": 0.0001872564445995752,
+      "loss": 1.2176,
+      "step": 2603
+    },
+    {
+      "epoch": 0.18770273192532258,
+      "grad_norm": 0.11807224150738628,
+      "learning_rate": 0.00018724503716123102,
+      "loss": 1.345,
+      "step": 2604
+    },
+    {
+      "epoch": 0.18777481438765947,
+      "grad_norm": 0.10088525845306652,
+      "learning_rate": 0.00018723362496725175,
+      "loss": 1.3519,
+      "step": 2605
+    },
+    {
+      "epoch": 0.1878468968499964,
+      "grad_norm": 0.13021291248220962,
+      "learning_rate": 0.0001872222080182595,
+      "loss": 1.4378,
+      "step": 2606
+    },
+    {
+      "epoch": 0.1879189793123333,
+      "grad_norm": 0.1423176532627825,
+      "learning_rate": 0.00018721078631487665,
+      "loss": 1.441,
+      "step": 2607
+    },
+    {
+      "epoch": 0.18799106177467023,
+      "grad_norm": 0.10974384747808304,
+      "learning_rate": 0.0001871993598577257,
+      "loss": 1.3389,
+      "step": 2608
+    },
+    {
+      "epoch": 0.18806314423700712,
+      "grad_norm": 0.11144033908852921,
+      "learning_rate": 0.00018718792864742953,
+      "loss": 1.2938,
+      "step": 2609
+    },
+    {
+      "epoch": 0.18813522669934404,
+      "grad_norm": 0.10726089392307792,
+      "learning_rate": 0.00018717649268461126,
+      "loss": 1.2593,
+      "step": 2610
+    },
+    {
+      "epoch": 0.18820730916168096,
+      "grad_norm": 0.10558861319545298,
+      "learning_rate": 0.0001871650519698942,
+      "loss": 1.4089,
+      "step": 2611
+    },
+    {
+      "epoch": 0.18827939162401788,
+      "grad_norm": 0.0961946300115577,
+      "learning_rate": 0.00018715360650390205,
+      "loss": 1.2134,
+      "step": 2612
+    },
+    {
+      "epoch": 0.1883514740863548,
+      "grad_norm": 0.10949401919441835,
+      "learning_rate": 0.0001871421562872586,
+      "loss": 1.2666,
+      "step": 2613
+    },
+    {
+      "epoch": 0.1884235565486917,
+      "grad_norm": 0.10062788681278442,
+      "learning_rate": 0.00018713070132058806,
+      "loss": 1.3366,
+      "step": 2614
+    },
+    {
+      "epoch": 0.18849563901102862,
+      "grad_norm": 0.1126036798061021,
+      "learning_rate": 0.00018711924160451485,
+      "loss": 1.3838,
+      "step": 2615
+    },
+    {
+      "epoch": 0.18856772147336553,
+      "grad_norm": 0.12264044866394923,
+      "learning_rate": 0.00018710777713966352,
+      "loss": 1.2692,
+      "step": 2616
+    },
+    {
+      "epoch": 0.18863980393570245,
+      "grad_norm": 0.13483001679851958,
+      "learning_rate": 0.00018709630792665906,
+      "loss": 1.2959,
+      "step": 2617
+    },
+    {
+      "epoch": 0.18871188639803935,
+      "grad_norm": 0.2393087073156164,
+      "learning_rate": 0.00018708483396612664,
+      "loss": 1.3798,
+      "step": 2618
+    },
+    {
+      "epoch": 0.18878396886037627,
+      "grad_norm": 0.1024867269784781,
+      "learning_rate": 0.0001870733552586917,
+      "loss": 1.3561,
+      "step": 2619
+    },
+    {
+      "epoch": 0.1888560513227132,
+      "grad_norm": 0.11724201622588178,
+      "learning_rate": 0.0001870618718049799,
+      "loss": 1.3335,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1889281337850501,
+      "grad_norm": 0.11285849792169442,
+      "learning_rate": 0.0001870503836056172,
+      "loss": 1.2933,
+      "step": 2621
+    },
+    {
+      "epoch": 0.189000216247387,
+      "grad_norm": 0.10511515331408026,
+      "learning_rate": 0.00018703889066122985,
+      "loss": 1.3868,
+      "step": 2622
+    },
+    {
+      "epoch": 0.18907229870972392,
+      "grad_norm": 0.12755440567550821,
+      "learning_rate": 0.00018702739297244425,
+      "loss": 1.2519,
+      "step": 2623
+    },
+    {
+      "epoch": 0.18914438117206084,
+      "grad_norm": 0.11072982855094526,
+      "learning_rate": 0.0001870158905398872,
+      "loss": 1.3563,
+      "step": 2624
+    },
+    {
+      "epoch": 0.18921646363439776,
+      "grad_norm": 0.11235819805931234,
+      "learning_rate": 0.0001870043833641856,
+      "loss": 1.2995,
+      "step": 2625
+    },
+    {
+      "epoch": 0.18928854609673468,
+      "grad_norm": 0.10342247052847192,
+      "learning_rate": 0.0001869928714459668,
+      "loss": 1.344,
+      "step": 2626
+    },
+    {
+      "epoch": 0.18936062855907157,
+      "grad_norm": 0.09882929250178799,
+      "learning_rate": 0.00018698135478585823,
+      "loss": 1.3885,
+      "step": 2627
+    },
+    {
+      "epoch": 0.1894327110214085,
+      "grad_norm": 0.12050012967869432,
+      "learning_rate": 0.00018696983338448763,
+      "loss": 1.3036,
+      "step": 2628
+    },
+    {
+      "epoch": 0.1895047934837454,
+      "grad_norm": 0.10266316401042962,
+      "learning_rate": 0.00018695830724248312,
+      "loss": 1.3213,
+      "step": 2629
+    },
+    {
+      "epoch": 0.18957687594608233,
+      "grad_norm": 0.10521053494763251,
+      "learning_rate": 0.00018694677636047285,
+      "loss": 1.395,
+      "step": 2630
+    },
+    {
+      "epoch": 0.18964895840841922,
+      "grad_norm": 0.10036584300007852,
+      "learning_rate": 0.00018693524073908546,
+      "loss": 1.2245,
+      "step": 2631
+    },
+    {
+      "epoch": 0.18972104087075614,
+      "grad_norm": 0.10959103605988148,
+      "learning_rate": 0.0001869237003789497,
+      "loss": 1.2628,
+      "step": 2632
+    },
+    {
+      "epoch": 0.18979312333309306,
+      "grad_norm": 0.11130677377389395,
+      "learning_rate": 0.0001869121552806946,
+      "loss": 1.2443,
+      "step": 2633
+    },
+    {
+      "epoch": 0.18986520579542998,
+      "grad_norm": 0.10448461312352848,
+      "learning_rate": 0.00018690060544494954,
+      "loss": 1.2362,
+      "step": 2634
+    },
+    {
+      "epoch": 0.18993728825776687,
+      "grad_norm": 0.1418000171442362,
+      "learning_rate": 0.000186889050872344,
+      "loss": 1.3941,
+      "step": 2635
+    },
+    {
+      "epoch": 0.1900093707201038,
+      "grad_norm": 0.18205500794763696,
+      "learning_rate": 0.0001868774915635079,
+      "loss": 1.2909,
+      "step": 2636
+    },
+    {
+      "epoch": 0.1900814531824407,
+      "grad_norm": 0.10295825724013652,
+      "learning_rate": 0.00018686592751907127,
+      "loss": 1.232,
+      "step": 2637
+    },
+    {
+      "epoch": 0.19015353564477763,
+      "grad_norm": 0.10221836626202371,
+      "learning_rate": 0.00018685435873966446,
+      "loss": 1.3266,
+      "step": 2638
+    },
+    {
+      "epoch": 0.19022561810711455,
+      "grad_norm": 0.11539624383759994,
+      "learning_rate": 0.00018684278522591806,
+      "loss": 1.2326,
+      "step": 2639
+    },
+    {
+      "epoch": 0.19029770056945144,
+      "grad_norm": 0.10777088279706921,
+      "learning_rate": 0.00018683120697846297,
+      "loss": 1.2374,
+      "step": 2640
+    },
+    {
+      "epoch": 0.19036978303178836,
+      "grad_norm": 0.1298624640986684,
+      "learning_rate": 0.0001868196239979303,
+      "loss": 1.316,
+      "step": 2641
+    },
+    {
+      "epoch": 0.19044186549412528,
+      "grad_norm": 0.11025985973637988,
+      "learning_rate": 0.00018680803628495137,
+      "loss": 1.3594,
+      "step": 2642
+    },
+    {
+      "epoch": 0.1905139479564622,
+      "grad_norm": 0.12176806188101659,
+      "learning_rate": 0.00018679644384015788,
+      "loss": 1.4091,
+      "step": 2643
+    },
+    {
+      "epoch": 0.1905860304187991,
+      "grad_norm": 0.13213720593972528,
+      "learning_rate": 0.00018678484666418167,
+      "loss": 1.385,
+      "step": 2644
+    },
+    {
+      "epoch": 0.19065811288113602,
+      "grad_norm": 0.10375576866575481,
+      "learning_rate": 0.00018677324475765498,
+      "loss": 1.2887,
+      "step": 2645
+    },
+    {
+      "epoch": 0.19073019534347294,
+      "grad_norm": 0.11911173735190753,
+      "learning_rate": 0.0001867616381212101,
+      "loss": 1.2604,
+      "step": 2646
+    },
+    {
+      "epoch": 0.19080227780580986,
+      "grad_norm": 0.11069984879997545,
+      "learning_rate": 0.00018675002675547976,
+      "loss": 1.3833,
+      "step": 2647
+    },
+    {
+      "epoch": 0.19087436026814675,
+      "grad_norm": 0.1094125909709141,
+      "learning_rate": 0.00018673841066109688,
+      "loss": 1.2485,
+      "step": 2648
+    },
+    {
+      "epoch": 0.19094644273048367,
+      "grad_norm": 0.12166592806704658,
+      "learning_rate": 0.00018672678983869463,
+      "loss": 1.3694,
+      "step": 2649
+    },
+    {
+      "epoch": 0.1910185251928206,
+      "grad_norm": 0.12436484425458363,
+      "learning_rate": 0.00018671516428890648,
+      "loss": 1.3927,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1910906076551575,
+      "grad_norm": 0.10911095874348238,
+      "learning_rate": 0.0001867035340123661,
+      "loss": 1.3205,
+      "step": 2651
+    },
+    {
+      "epoch": 0.19116269011749443,
+      "grad_norm": 0.10673076003157286,
+      "learning_rate": 0.0001866918990097074,
+      "loss": 1.3062,
+      "step": 2652
+    },
+    {
+      "epoch": 0.19123477257983132,
+      "grad_norm": 0.11359934157935941,
+      "learning_rate": 0.00018668025928156466,
+      "loss": 1.3307,
+      "step": 2653
+    },
+    {
+      "epoch": 0.19130685504216824,
+      "grad_norm": 0.10600511585673583,
+      "learning_rate": 0.00018666861482857235,
+      "loss": 1.2904,
+      "step": 2654
+    },
+    {
+      "epoch": 0.19137893750450516,
+      "grad_norm": 0.11520611137367195,
+      "learning_rate": 0.00018665696565136513,
+      "loss": 1.2661,
+      "step": 2655
+    },
+    {
+      "epoch": 0.19145101996684208,
+      "grad_norm": 0.10744273025399073,
+      "learning_rate": 0.00018664531175057803,
+      "loss": 1.3883,
+      "step": 2656
+    },
+    {
+      "epoch": 0.19152310242917897,
+      "grad_norm": 0.12041407577921236,
+      "learning_rate": 0.00018663365312684632,
+      "loss": 1.3466,
+      "step": 2657
+    },
+    {
+      "epoch": 0.1915951848915159,
+      "grad_norm": 0.11651498068224711,
+      "learning_rate": 0.00018662198978080543,
+      "loss": 1.4303,
+      "step": 2658
+    },
+    {
+      "epoch": 0.1916672673538528,
+      "grad_norm": 0.11044462838406637,
+      "learning_rate": 0.00018661032171309118,
+      "loss": 1.2964,
+      "step": 2659
+    },
+    {
+      "epoch": 0.19173934981618973,
+      "grad_norm": 0.11402282551750417,
+      "learning_rate": 0.00018659864892433956,
+      "loss": 1.2836,
+      "step": 2660
+    },
+    {
+      "epoch": 0.19181143227852662,
+      "grad_norm": 0.10634318272959867,
+      "learning_rate": 0.00018658697141518685,
+      "loss": 1.3025,
+      "step": 2661
+    },
+    {
+      "epoch": 0.19188351474086354,
+      "grad_norm": 0.10589849486111479,
+      "learning_rate": 0.00018657528918626954,
+      "loss": 1.3626,
+      "step": 2662
+    },
+    {
+      "epoch": 0.19195559720320046,
+      "grad_norm": 0.13879466688737202,
+      "learning_rate": 0.00018656360223822445,
+      "loss": 1.2991,
+      "step": 2663
+    },
+    {
+      "epoch": 0.19202767966553738,
+      "grad_norm": 0.10023817853055339,
+      "learning_rate": 0.00018655191057168863,
+      "loss": 1.2862,
+      "step": 2664
+    },
+    {
+      "epoch": 0.1920997621278743,
+      "grad_norm": 0.12607556296331765,
+      "learning_rate": 0.00018654021418729937,
+      "loss": 1.382,
+      "step": 2665
+    },
+    {
+      "epoch": 0.1921718445902112,
+      "grad_norm": 0.10424105552010655,
+      "learning_rate": 0.00018652851308569423,
+      "loss": 1.4756,
+      "step": 2666
+    },
+    {
+      "epoch": 0.1922439270525481,
+      "grad_norm": 0.11202021599828287,
+      "learning_rate": 0.000186516807267511,
+      "loss": 1.3362,
+      "step": 2667
+    },
+    {
+      "epoch": 0.19231600951488503,
+      "grad_norm": 0.1183005587526956,
+      "learning_rate": 0.0001865050967333878,
+      "loss": 1.355,
+      "step": 2668
+    },
+    {
+      "epoch": 0.19238809197722195,
+      "grad_norm": 0.10793067116902307,
+      "learning_rate": 0.00018649338148396289,
+      "loss": 1.2996,
+      "step": 2669
+    },
+    {
+      "epoch": 0.19246017443955885,
+      "grad_norm": 0.09621837697113128,
+      "learning_rate": 0.0001864816615198749,
+      "loss": 1.3705,
+      "step": 2670
+    },
+    {
+      "epoch": 0.19253225690189577,
+      "grad_norm": 0.12964901223437558,
+      "learning_rate": 0.0001864699368417627,
+      "loss": 1.3072,
+      "step": 2671
+    },
+    {
+      "epoch": 0.19260433936423269,
+      "grad_norm": 0.11106409222867417,
+      "learning_rate": 0.00018645820745026538,
+      "loss": 1.2681,
+      "step": 2672
+    },
+    {
+      "epoch": 0.1926764218265696,
+      "grad_norm": 0.11314998360835916,
+      "learning_rate": 0.00018644647334602223,
+      "loss": 1.3739,
+      "step": 2673
+    },
+    {
+      "epoch": 0.1927485042889065,
+      "grad_norm": 0.09674128990282567,
+      "learning_rate": 0.00018643473452967295,
+      "loss": 1.3521,
+      "step": 2674
+    },
+    {
+      "epoch": 0.19282058675124342,
+      "grad_norm": 0.10929308357345353,
+      "learning_rate": 0.00018642299100185734,
+      "loss": 1.3417,
+      "step": 2675
+    },
+    {
+      "epoch": 0.19289266921358034,
+      "grad_norm": 0.10814403577069007,
+      "learning_rate": 0.00018641124276321555,
+      "loss": 1.3929,
+      "step": 2676
+    },
+    {
+      "epoch": 0.19296475167591726,
+      "grad_norm": 0.1067218909588352,
+      "learning_rate": 0.00018639948981438799,
+      "loss": 1.3956,
+      "step": 2677
+    },
+    {
+      "epoch": 0.19303683413825418,
+      "grad_norm": 0.14450630243076767,
+      "learning_rate": 0.00018638773215601527,
+      "loss": 1.3484,
+      "step": 2678
+    },
+    {
+      "epoch": 0.19310891660059107,
+      "grad_norm": 0.1222637077072012,
+      "learning_rate": 0.00018637596978873835,
+      "loss": 1.3174,
+      "step": 2679
+    },
+    {
+      "epoch": 0.193180999062928,
+      "grad_norm": 0.15777165369068205,
+      "learning_rate": 0.0001863642027131983,
+      "loss": 1.2575,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1932530815252649,
+      "grad_norm": 0.11357274004755265,
+      "learning_rate": 0.00018635243093003654,
+      "loss": 1.3313,
+      "step": 2681
+    },
+    {
+      "epoch": 0.19332516398760183,
+      "grad_norm": 0.11735497943958857,
+      "learning_rate": 0.0001863406544398948,
+      "loss": 1.3134,
+      "step": 2682
+    },
+    {
+      "epoch": 0.19339724644993872,
+      "grad_norm": 0.11759327841404157,
+      "learning_rate": 0.00018632887324341498,
+      "loss": 1.2939,
+      "step": 2683
+    },
+    {
+      "epoch": 0.19346932891227564,
+      "grad_norm": 0.11320266188220333,
+      "learning_rate": 0.0001863170873412392,
+      "loss": 1.4271,
+      "step": 2684
+    },
+    {
+      "epoch": 0.19354141137461256,
+      "grad_norm": 0.10515132385706596,
+      "learning_rate": 0.00018630529673400998,
+      "loss": 1.2513,
+      "step": 2685
+    },
+    {
+      "epoch": 0.19361349383694948,
+      "grad_norm": 0.11418814947864693,
+      "learning_rate": 0.00018629350142236993,
+      "loss": 1.2773,
+      "step": 2686
+    },
+    {
+      "epoch": 0.19368557629928637,
+      "grad_norm": 0.13009015080844016,
+      "learning_rate": 0.00018628170140696213,
+      "loss": 1.3036,
+      "step": 2687
+    },
+    {
+      "epoch": 0.1937576587616233,
+      "grad_norm": 0.10995308885380538,
+      "learning_rate": 0.00018626989668842965,
+      "loss": 1.3131,
+      "step": 2688
+    },
+    {
+      "epoch": 0.1938297412239602,
+      "grad_norm": 0.11662976849349517,
+      "learning_rate": 0.000186258087267416,
+      "loss": 1.2976,
+      "step": 2689
+    },
+    {
+      "epoch": 0.19390182368629713,
+      "grad_norm": 0.1316057029458674,
+      "learning_rate": 0.00018624627314456493,
+      "loss": 1.253,
+      "step": 2690
+    },
+    {
+      "epoch": 0.19397390614863405,
+      "grad_norm": 0.18904458049338232,
+      "learning_rate": 0.0001862344543205204,
+      "loss": 1.3361,
+      "step": 2691
+    },
+    {
+      "epoch": 0.19404598861097094,
+      "grad_norm": 0.13997835693906005,
+      "learning_rate": 0.00018622263079592664,
+      "loss": 1.3,
+      "step": 2692
+    },
+    {
+      "epoch": 0.19411807107330786,
+      "grad_norm": 0.0996994244583873,
+      "learning_rate": 0.00018621080257142812,
+      "loss": 1.3485,
+      "step": 2693
+    },
+    {
+      "epoch": 0.19419015353564478,
+      "grad_norm": 0.10707185836908459,
+      "learning_rate": 0.00018619896964766955,
+      "loss": 1.4232,
+      "step": 2694
+    },
+    {
+      "epoch": 0.1942622359979817,
+      "grad_norm": 0.1173661314819115,
+      "learning_rate": 0.00018618713202529606,
+      "loss": 1.3037,
+      "step": 2695
+    },
+    {
+      "epoch": 0.1943343184603186,
+      "grad_norm": 0.1070145802819054,
+      "learning_rate": 0.00018617528970495278,
+      "loss": 1.2525,
+      "step": 2696
+    },
+    {
+      "epoch": 0.19440640092265551,
+      "grad_norm": 0.10085632820717455,
+      "learning_rate": 0.00018616344268728523,
+      "loss": 1.374,
+      "step": 2697
+    },
+    {
+      "epoch": 0.19447848338499243,
+      "grad_norm": 0.10554080403114412,
+      "learning_rate": 0.00018615159097293927,
+      "loss": 1.3056,
+      "step": 2698
+    },
+    {
+      "epoch": 0.19455056584732935,
+      "grad_norm": 0.10292607750544494,
+      "learning_rate": 0.00018613973456256083,
+      "loss": 1.2132,
+      "step": 2699
+    },
+    {
+      "epoch": 0.19462264830966625,
+      "grad_norm": 0.10487257616995346,
+      "learning_rate": 0.00018612787345679625,
+      "loss": 1.3413,
+      "step": 2700
+    },
+    {
+      "epoch": 0.19469473077200317,
+      "grad_norm": 0.10845469369617795,
+      "learning_rate": 0.00018611600765629205,
+      "loss": 1.3464,
+      "step": 2701
+    },
+    {
+      "epoch": 0.19476681323434009,
+      "grad_norm": 0.10928432088391081,
+      "learning_rate": 0.00018610413716169499,
+      "loss": 1.3024,
+      "step": 2702
+    },
+    {
+      "epoch": 0.194838895696677,
+      "grad_norm": 0.14901138518987023,
+      "learning_rate": 0.00018609226197365217,
+      "loss": 1.4165,
+      "step": 2703
+    },
+    {
+      "epoch": 0.19491097815901393,
+      "grad_norm": 0.3385264492928032,
+      "learning_rate": 0.00018608038209281085,
+      "loss": 1.305,
+      "step": 2704
+    },
+    {
+      "epoch": 0.19498306062135082,
+      "grad_norm": 0.2435363356410392,
+      "learning_rate": 0.00018606849751981862,
+      "loss": 1.2978,
+      "step": 2705
+    },
+    {
+      "epoch": 0.19505514308368774,
+      "grad_norm": 0.1159503862346053,
+      "learning_rate": 0.0001860566082553233,
+      "loss": 1.334,
+      "step": 2706
+    },
+    {
+      "epoch": 0.19512722554602466,
+      "grad_norm": 0.11290489677885875,
+      "learning_rate": 0.0001860447142999729,
+      "loss": 1.3712,
+      "step": 2707
+    },
+    {
+      "epoch": 0.19519930800836158,
+      "grad_norm": 0.10850060585026043,
+      "learning_rate": 0.00018603281565441585,
+      "loss": 1.3968,
+      "step": 2708
+    },
+    {
+      "epoch": 0.19527139047069847,
+      "grad_norm": 0.12772396571330386,
+      "learning_rate": 0.00018602091231930063,
+      "loss": 1.1753,
+      "step": 2709
+    },
+    {
+      "epoch": 0.1953434729330354,
+      "grad_norm": 0.11493734079067759,
+      "learning_rate": 0.00018600900429527616,
+      "loss": 1.3261,
+      "step": 2710
+    },
+    {
+      "epoch": 0.1954155553953723,
+      "grad_norm": 0.12766240205502102,
+      "learning_rate": 0.00018599709158299146,
+      "loss": 1.3034,
+      "step": 2711
+    },
+    {
+      "epoch": 0.19548763785770923,
+      "grad_norm": 0.11653051620356034,
+      "learning_rate": 0.00018598517418309593,
+      "loss": 1.2566,
+      "step": 2712
+    },
+    {
+      "epoch": 0.19555972032004612,
+      "grad_norm": 0.11136516842334589,
+      "learning_rate": 0.00018597325209623918,
+      "loss": 1.3526,
+      "step": 2713
+    },
+    {
+      "epoch": 0.19563180278238304,
+      "grad_norm": 0.11551857884424839,
+      "learning_rate": 0.00018596132532307103,
+      "loss": 1.3245,
+      "step": 2714
+    },
+    {
+      "epoch": 0.19570388524471996,
+      "grad_norm": 0.10336283613806718,
+      "learning_rate": 0.00018594939386424163,
+      "loss": 1.2455,
+      "step": 2715
+    },
+    {
+      "epoch": 0.19577596770705688,
+      "grad_norm": 0.11508517460683368,
+      "learning_rate": 0.00018593745772040135,
+      "loss": 1.3091,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1958480501693938,
+      "grad_norm": 0.1325504848082111,
+      "learning_rate": 0.00018592551689220076,
+      "loss": 1.3611,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1959201326317307,
+      "grad_norm": 0.10099290914984563,
+      "learning_rate": 0.00018591357138029083,
+      "loss": 1.261,
+      "step": 2718
+    },
+    {
+      "epoch": 0.1959922150940676,
+      "grad_norm": 0.10300403735798985,
+      "learning_rate": 0.00018590162118532267,
+      "loss": 1.3153,
+      "step": 2719
+    },
+    {
+      "epoch": 0.19606429755640453,
+      "grad_norm": 0.12732550786483005,
+      "learning_rate": 0.0001858896663079476,
+      "loss": 1.3103,
+      "step": 2720
+    },
+    {
+      "epoch": 0.19613638001874145,
+      "grad_norm": 0.11339540244954983,
+      "learning_rate": 0.00018587770674881738,
+      "loss": 1.2929,
+      "step": 2721
+    },
+    {
+      "epoch": 0.19620846248107834,
+      "grad_norm": 0.2750965760899001,
+      "learning_rate": 0.00018586574250858382,
+      "loss": 1.2165,
+      "step": 2722
+    },
+    {
+      "epoch": 0.19628054494341526,
+      "grad_norm": 0.14389798980023596,
+      "learning_rate": 0.00018585377358789913,
+      "loss": 1.3207,
+      "step": 2723
+    },
+    {
+      "epoch": 0.19635262740575218,
+      "grad_norm": 0.13977972285377654,
+      "learning_rate": 0.0001858417999874157,
+      "loss": 1.4036,
+      "step": 2724
+    },
+    {
+      "epoch": 0.1964247098680891,
+      "grad_norm": 0.11239961915407062,
+      "learning_rate": 0.00018582982170778622,
+      "loss": 1.3274,
+      "step": 2725
+    },
+    {
+      "epoch": 0.196496792330426,
+      "grad_norm": 0.1092485533400403,
+      "learning_rate": 0.00018581783874966362,
+      "loss": 1.4071,
+      "step": 2726
+    },
+    {
+      "epoch": 0.19656887479276292,
+      "grad_norm": 0.2425216874566072,
+      "learning_rate": 0.00018580585111370104,
+      "loss": 1.2714,
+      "step": 2727
+    },
+    {
+      "epoch": 0.19664095725509984,
+      "grad_norm": 0.1029475091789455,
+      "learning_rate": 0.00018579385880055193,
+      "loss": 1.3261,
+      "step": 2728
+    },
+    {
+      "epoch": 0.19671303971743676,
+      "grad_norm": 0.12455793384386535,
+      "learning_rate": 0.00018578186181086997,
+      "loss": 1.2737,
+      "step": 2729
+    },
+    {
+      "epoch": 0.19678512217977367,
+      "grad_norm": 0.11019739725522497,
+      "learning_rate": 0.00018576986014530916,
+      "loss": 1.3347,
+      "step": 2730
+    },
+    {
+      "epoch": 0.19685720464211057,
+      "grad_norm": 0.11221567309358796,
+      "learning_rate": 0.0001857578538045236,
+      "loss": 1.3295,
+      "step": 2731
+    },
+    {
+      "epoch": 0.1969292871044475,
+      "grad_norm": 0.11017011662634776,
+      "learning_rate": 0.0001857458427891678,
+      "loss": 1.2759,
+      "step": 2732
+    },
+    {
+      "epoch": 0.1970013695667844,
+      "grad_norm": 0.20295611341369454,
+      "learning_rate": 0.0001857338270998965,
+      "loss": 1.355,
+      "step": 2733
+    },
+    {
+      "epoch": 0.19707345202912133,
+      "grad_norm": 0.1119882075602629,
+      "learning_rate": 0.0001857218067373646,
+      "loss": 1.4007,
+      "step": 2734
+    },
+    {
+      "epoch": 0.19714553449145822,
+      "grad_norm": 0.11913190605734839,
+      "learning_rate": 0.00018570978170222736,
+      "loss": 1.3203,
+      "step": 2735
+    },
+    {
+      "epoch": 0.19721761695379514,
+      "grad_norm": 0.13396829644458,
+      "learning_rate": 0.00018569775199514022,
+      "loss": 1.3106,
+      "step": 2736
+    },
+    {
+      "epoch": 0.19728969941613206,
+      "grad_norm": 0.11531320553633635,
+      "learning_rate": 0.00018568571761675893,
+      "loss": 1.2274,
+      "step": 2737
+    },
+    {
+      "epoch": 0.19736178187846898,
+      "grad_norm": 0.11169605434673852,
+      "learning_rate": 0.00018567367856773948,
+      "loss": 1.3213,
+      "step": 2738
+    },
+    {
+      "epoch": 0.19743386434080587,
+      "grad_norm": 0.10150409898883075,
+      "learning_rate": 0.00018566163484873806,
+      "loss": 1.2025,
+      "step": 2739
+    },
+    {
+      "epoch": 0.1975059468031428,
+      "grad_norm": 0.11895338293810163,
+      "learning_rate": 0.0001856495864604112,
+      "loss": 1.3217,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1975780292654797,
+      "grad_norm": 0.22749874687164712,
+      "learning_rate": 0.00018563753340341562,
+      "loss": 1.2708,
+      "step": 2741
+    },
+    {
+      "epoch": 0.19765011172781663,
+      "grad_norm": 0.19037929918372498,
+      "learning_rate": 0.0001856254756784084,
+      "loss": 1.3116,
+      "step": 2742
+    },
+    {
+      "epoch": 0.19772219419015352,
+      "grad_norm": 0.11411789824005306,
+      "learning_rate": 0.00018561341328604666,
+      "loss": 1.3611,
+      "step": 2743
+    },
+    {
+      "epoch": 0.19779427665249044,
+      "grad_norm": 0.1375207470860487,
+      "learning_rate": 0.000185601346226988,
+      "loss": 1.3616,
+      "step": 2744
+    },
+    {
+      "epoch": 0.19786635911482736,
+      "grad_norm": 0.20654158725943764,
+      "learning_rate": 0.00018558927450189013,
+      "loss": 1.3849,
+      "step": 2745
+    },
+    {
+      "epoch": 0.19793844157716428,
+      "grad_norm": 0.1156112611741174,
+      "learning_rate": 0.00018557719811141115,
+      "loss": 1.2867,
+      "step": 2746
+    },
+    {
+      "epoch": 0.1980105240395012,
+      "grad_norm": 0.21187222709071968,
+      "learning_rate": 0.00018556511705620924,
+      "loss": 1.199,
+      "step": 2747
+    },
+    {
+      "epoch": 0.1980826065018381,
+      "grad_norm": 0.09526706574773992,
+      "learning_rate": 0.00018555303133694297,
+      "loss": 1.2993,
+      "step": 2748
+    },
+    {
+      "epoch": 0.198154688964175,
+      "grad_norm": 0.1189136249162116,
+      "learning_rate": 0.0001855409409542711,
+      "loss": 1.3397,
+      "step": 2749
+    },
+    {
+      "epoch": 0.19822677142651193,
+      "grad_norm": 0.10460871186604805,
+      "learning_rate": 0.00018552884590885272,
+      "loss": 1.2709,
+      "step": 2750
+    },
+    {
+      "epoch": 0.19829885388884885,
+      "grad_norm": 0.1070672646818605,
+      "learning_rate": 0.00018551674620134704,
+      "loss": 1.2912,
+      "step": 2751
+    },
+    {
+      "epoch": 0.19837093635118574,
+      "grad_norm": 0.11607113671841893,
+      "learning_rate": 0.00018550464183241365,
+      "loss": 1.3113,
+      "step": 2752
+    },
+    {
+      "epoch": 0.19844301881352266,
+      "grad_norm": 0.10611431525404313,
+      "learning_rate": 0.00018549253280271231,
+      "loss": 1.3681,
+      "step": 2753
+    },
+    {
+      "epoch": 0.19851510127585958,
+      "grad_norm": 0.10904671537470381,
+      "learning_rate": 0.0001854804191129031,
+      "loss": 1.2152,
+      "step": 2754
+    },
+    {
+      "epoch": 0.1985871837381965,
+      "grad_norm": 0.1065478246243059,
+      "learning_rate": 0.00018546830076364635,
+      "loss": 1.2562,
+      "step": 2755
+    },
+    {
+      "epoch": 0.1986592662005334,
+      "grad_norm": 0.10623976400268047,
+      "learning_rate": 0.00018545617775560258,
+      "loss": 1.2539,
+      "step": 2756
+    },
+    {
+      "epoch": 0.19873134866287032,
+      "grad_norm": 0.14190666201563862,
+      "learning_rate": 0.0001854440500894326,
+      "loss": 1.3354,
+      "step": 2757
+    },
+    {
+      "epoch": 0.19880343112520724,
+      "grad_norm": 0.11190448296817572,
+      "learning_rate": 0.00018543191776579745,
+      "loss": 1.2367,
+      "step": 2758
+    },
+    {
+      "epoch": 0.19887551358754416,
+      "grad_norm": 0.10437348109115062,
+      "learning_rate": 0.00018541978078535852,
+      "loss": 1.3651,
+      "step": 2759
+    },
+    {
+      "epoch": 0.19894759604988108,
+      "grad_norm": 0.11154012277268922,
+      "learning_rate": 0.00018540763914877737,
+      "loss": 1.301,
+      "step": 2760
+    },
+    {
+      "epoch": 0.19901967851221797,
+      "grad_norm": 0.10685146176557493,
+      "learning_rate": 0.0001853954928567158,
+      "loss": 1.3603,
+      "step": 2761
+    },
+    {
+      "epoch": 0.1990917609745549,
+      "grad_norm": 0.10406116091869198,
+      "learning_rate": 0.00018538334190983585,
+      "loss": 1.2742,
+      "step": 2762
+    },
+    {
+      "epoch": 0.1991638434368918,
+      "grad_norm": 0.14082145810454982,
+      "learning_rate": 0.00018537118630879994,
+      "loss": 1.2655,
+      "step": 2763
+    },
+    {
+      "epoch": 0.19923592589922873,
+      "grad_norm": 0.10277855776637235,
+      "learning_rate": 0.00018535902605427065,
+      "loss": 1.2769,
+      "step": 2764
+    },
+    {
+      "epoch": 0.19930800836156562,
+      "grad_norm": 0.11549827248364981,
+      "learning_rate": 0.00018534686114691078,
+      "loss": 1.3332,
+      "step": 2765
+    },
+    {
+      "epoch": 0.19938009082390254,
+      "grad_norm": 0.10052062516628686,
+      "learning_rate": 0.00018533469158738344,
+      "loss": 1.3373,
+      "step": 2766
+    },
+    {
+      "epoch": 0.19945217328623946,
+      "grad_norm": 0.1071489306128591,
+      "learning_rate": 0.00018532251737635196,
+      "loss": 1.285,
+      "step": 2767
+    },
+    {
+      "epoch": 0.19952425574857638,
+      "grad_norm": 0.11879382400024399,
+      "learning_rate": 0.00018531033851447996,
+      "loss": 1.4189,
+      "step": 2768
+    },
+    {
+      "epoch": 0.19959633821091327,
+      "grad_norm": 0.09749588186673916,
+      "learning_rate": 0.00018529815500243135,
+      "loss": 1.3035,
+      "step": 2769
+    },
+    {
+      "epoch": 0.1996684206732502,
+      "grad_norm": 0.11712850028553065,
+      "learning_rate": 0.00018528596684087015,
+      "loss": 1.2185,
+      "step": 2770
+    },
+    {
+      "epoch": 0.1997405031355871,
+      "grad_norm": 0.10373302386354757,
+      "learning_rate": 0.0001852737740304608,
+      "loss": 1.3481,
+      "step": 2771
+    },
+    {
+      "epoch": 0.19981258559792403,
+      "grad_norm": 0.1687099088840803,
+      "learning_rate": 0.00018526157657186787,
+      "loss": 1.3267,
+      "step": 2772
+    },
+    {
+      "epoch": 0.19988466806026095,
+      "grad_norm": 0.10245102175438502,
+      "learning_rate": 0.00018524937446575624,
+      "loss": 1.2709,
+      "step": 2773
+    },
+    {
+      "epoch": 0.19995675052259784,
+      "grad_norm": 0.16058226869878064,
+      "learning_rate": 0.00018523716771279104,
+      "loss": 1.3383,
+      "step": 2774
+    },
+    {
+      "epoch": 0.20002883298493476,
+      "grad_norm": 0.1146406143297301,
+      "learning_rate": 0.00018522495631363768,
+      "loss": 1.3253,
+      "step": 2775
+    },
+    {
+      "epoch": 0.20010091544727168,
+      "grad_norm": 0.11221147291514806,
+      "learning_rate": 0.00018521274026896172,
+      "loss": 1.3402,
+      "step": 2776
+    },
+    {
+      "epoch": 0.2001729979096086,
+      "grad_norm": 0.10983859609195337,
+      "learning_rate": 0.00018520051957942907,
+      "loss": 1.3205,
+      "step": 2777
+    },
+    {
+      "epoch": 0.2002450803719455,
+      "grad_norm": 0.21620799398699087,
+      "learning_rate": 0.0001851882942457059,
+      "loss": 1.2728,
+      "step": 2778
+    },
+    {
+      "epoch": 0.20031716283428241,
+      "grad_norm": 0.11454134359640464,
+      "learning_rate": 0.00018517606426845856,
+      "loss": 1.3655,
+      "step": 2779
+    },
+    {
+      "epoch": 0.20038924529661933,
+      "grad_norm": 0.12255340113676798,
+      "learning_rate": 0.00018516382964835372,
+      "loss": 1.3876,
+      "step": 2780
+    },
+    {
+      "epoch": 0.20046132775895625,
+      "grad_norm": 0.11286836345963715,
+      "learning_rate": 0.00018515159038605828,
+      "loss": 1.2657,
+      "step": 2781
+    },
+    {
+      "epoch": 0.20053341022129315,
+      "grad_norm": 0.10779943903047458,
+      "learning_rate": 0.00018513934648223937,
+      "loss": 1.4166,
+      "step": 2782
+    },
+    {
+      "epoch": 0.20060549268363007,
+      "grad_norm": 0.1109955423417722,
+      "learning_rate": 0.0001851270979375644,
+      "loss": 1.3148,
+      "step": 2783
+    },
+    {
+      "epoch": 0.20067757514596699,
+      "grad_norm": 0.14702049952212703,
+      "learning_rate": 0.000185114844752701,
+      "loss": 1.4367,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2007496576083039,
+      "grad_norm": 0.11361864717176202,
+      "learning_rate": 0.00018510258692831712,
+      "loss": 1.2648,
+      "step": 2785
+    },
+    {
+      "epoch": 0.20082174007064083,
+      "grad_norm": 0.10976166845446969,
+      "learning_rate": 0.00018509032446508092,
+      "loss": 1.2342,
+      "step": 2786
+    },
+    {
+      "epoch": 0.20089382253297772,
+      "grad_norm": 0.10580705132154075,
+      "learning_rate": 0.00018507805736366078,
+      "loss": 1.2623,
+      "step": 2787
+    },
+    {
+      "epoch": 0.20096590499531464,
+      "grad_norm": 0.11773535411456516,
+      "learning_rate": 0.00018506578562472537,
+      "loss": 1.2295,
+      "step": 2788
+    },
+    {
+      "epoch": 0.20103798745765156,
+      "grad_norm": 0.13803611927154627,
+      "learning_rate": 0.00018505350924894365,
+      "loss": 1.373,
+      "step": 2789
+    },
+    {
+      "epoch": 0.20111006991998848,
+      "grad_norm": 0.118931505556613,
+      "learning_rate": 0.00018504122823698472,
+      "loss": 1.3144,
+      "step": 2790
+    },
+    {
+      "epoch": 0.20118215238232537,
+      "grad_norm": 0.11547659731269874,
+      "learning_rate": 0.0001850289425895181,
+      "loss": 1.2983,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2012542348446623,
+      "grad_norm": 0.10723776752528674,
+      "learning_rate": 0.00018501665230721338,
+      "loss": 1.3715,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2013263173069992,
+      "grad_norm": 0.10360789769528812,
+      "learning_rate": 0.00018500435739074057,
+      "loss": 1.3354,
+      "step": 2793
+    },
+    {
+      "epoch": 0.20139839976933613,
+      "grad_norm": 0.12396590721127246,
+      "learning_rate": 0.00018499205784076978,
+      "loss": 1.3298,
+      "step": 2794
+    },
+    {
+      "epoch": 0.20147048223167302,
+      "grad_norm": 0.18927488205593182,
+      "learning_rate": 0.0001849797536579715,
+      "loss": 1.2984,
+      "step": 2795
+    },
+    {
+      "epoch": 0.20154256469400994,
+      "grad_norm": 0.10740368521346086,
+      "learning_rate": 0.0001849674448430164,
+      "loss": 1.2527,
+      "step": 2796
+    },
+    {
+      "epoch": 0.20161464715634686,
+      "grad_norm": 0.10426673426639496,
+      "learning_rate": 0.0001849551313965754,
+      "loss": 1.3131,
+      "step": 2797
+    },
+    {
+      "epoch": 0.20168672961868378,
+      "grad_norm": 0.11910753264415776,
+      "learning_rate": 0.0001849428133193197,
+      "loss": 1.2857,
+      "step": 2798
+    },
+    {
+      "epoch": 0.2017588120810207,
+      "grad_norm": 0.09825448026419982,
+      "learning_rate": 0.00018493049061192072,
+      "loss": 1.2956,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2018308945433576,
+      "grad_norm": 0.10476141197385888,
+      "learning_rate": 0.00018491816327505025,
+      "loss": 1.3732,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2019029770056945,
+      "grad_norm": 0.10488633574867579,
+      "learning_rate": 0.00018490583130938016,
+      "loss": 1.2814,
+      "step": 2801
+    },
+    {
+      "epoch": 0.20197505946803143,
+      "grad_norm": 0.10684685811048673,
+      "learning_rate": 0.00018489349471558267,
+      "loss": 1.282,
+      "step": 2802
+    },
+    {
+      "epoch": 0.20204714193036835,
+      "grad_norm": 0.1019217320249586,
+      "learning_rate": 0.00018488115349433025,
+      "loss": 1.3028,
+      "step": 2803
+    },
+    {
+      "epoch": 0.20211922439270524,
+      "grad_norm": 0.10369114778123927,
+      "learning_rate": 0.0001848688076462956,
+      "loss": 1.3651,
+      "step": 2804
+    },
+    {
+      "epoch": 0.20219130685504216,
+      "grad_norm": 0.10324958777436573,
+      "learning_rate": 0.00018485645717215168,
+      "loss": 1.3252,
+      "step": 2805
+    },
+    {
+      "epoch": 0.20226338931737908,
+      "grad_norm": 0.11592871073608467,
+      "learning_rate": 0.00018484410207257168,
+      "loss": 1.3306,
+      "step": 2806
+    },
+    {
+      "epoch": 0.202335471779716,
+      "grad_norm": 0.0979076060101401,
+      "learning_rate": 0.00018483174234822908,
+      "loss": 1.2473,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2024075542420529,
+      "grad_norm": 0.12235150448009541,
+      "learning_rate": 0.0001848193779997976,
+      "loss": 1.2304,
+      "step": 2808
+    },
+    {
+      "epoch": 0.20247963670438981,
+      "grad_norm": 0.10506848328359981,
+      "learning_rate": 0.0001848070090279512,
+      "loss": 1.2969,
+      "step": 2809
+    },
+    {
+      "epoch": 0.20255171916672673,
+      "grad_norm": 0.11570173445476992,
+      "learning_rate": 0.0001847946354333641,
+      "loss": 1.3657,
+      "step": 2810
+    },
+    {
+      "epoch": 0.20262380162906365,
+      "grad_norm": 0.09239261765227307,
+      "learning_rate": 0.00018478225721671077,
+      "loss": 1.3313,
+      "step": 2811
+    },
+    {
+      "epoch": 0.20269588409140057,
+      "grad_norm": 0.10809065890131056,
+      "learning_rate": 0.00018476987437866598,
+      "loss": 1.1899,
+      "step": 2812
+    },
+    {
+      "epoch": 0.20276796655373747,
+      "grad_norm": 0.1063722827935771,
+      "learning_rate": 0.00018475748691990462,
+      "loss": 1.3285,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2028400490160744,
+      "grad_norm": 0.11950018464234657,
+      "learning_rate": 0.00018474509484110198,
+      "loss": 1.3776,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2029121314784113,
+      "grad_norm": 0.11510412829558807,
+      "learning_rate": 0.0001847326981429335,
+      "loss": 1.2161,
+      "step": 2815
+    },
+    {
+      "epoch": 0.20298421394074823,
+      "grad_norm": 0.11530214610694184,
+      "learning_rate": 0.00018472029682607496,
+      "loss": 1.2905,
+      "step": 2816
+    },
+    {
+      "epoch": 0.20305629640308512,
+      "grad_norm": 0.1153960877129043,
+      "learning_rate": 0.00018470789089120227,
+      "loss": 1.239,
+      "step": 2817
+    },
+    {
+      "epoch": 0.20312837886542204,
+      "grad_norm": 0.11337613690524827,
+      "learning_rate": 0.00018469548033899175,
+      "loss": 1.2297,
+      "step": 2818
+    },
+    {
+      "epoch": 0.20320046132775896,
+      "grad_norm": 0.1059621864677661,
+      "learning_rate": 0.00018468306517011984,
+      "loss": 1.3224,
+      "step": 2819
+    },
+    {
+      "epoch": 0.20327254379009588,
+      "grad_norm": 0.10323834937485403,
+      "learning_rate": 0.00018467064538526328,
+      "loss": 1.3866,
+      "step": 2820
+    },
+    {
+      "epoch": 0.20334462625243277,
+      "grad_norm": 0.12238196591164108,
+      "learning_rate": 0.00018465822098509904,
+      "loss": 1.3471,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2034167087147697,
+      "grad_norm": 0.13102352011004717,
+      "learning_rate": 0.0001846457919703044,
+      "loss": 1.3414,
+      "step": 2822
+    },
+    {
+      "epoch": 0.2034887911771066,
+      "grad_norm": 0.11676821490134887,
+      "learning_rate": 0.00018463335834155683,
+      "loss": 1.1925,
+      "step": 2823
+    },
+    {
+      "epoch": 0.20356087363944353,
+      "grad_norm": 0.10816217613764108,
+      "learning_rate": 0.00018462092009953408,
+      "loss": 1.2885,
+      "step": 2824
+    },
+    {
+      "epoch": 0.20363295610178045,
+      "grad_norm": 0.10362318767020519,
+      "learning_rate": 0.00018460847724491414,
+      "loss": 1.1972,
+      "step": 2825
+    },
+    {
+      "epoch": 0.20370503856411734,
+      "grad_norm": 0.11554007440355028,
+      "learning_rate": 0.00018459602977837527,
+      "loss": 1.264,
+      "step": 2826
+    },
+    {
+      "epoch": 0.20377712102645426,
+      "grad_norm": 0.12173701286551655,
+      "learning_rate": 0.00018458357770059593,
+      "loss": 1.3824,
+      "step": 2827
+    },
+    {
+      "epoch": 0.20384920348879118,
+      "grad_norm": 0.10013595838206123,
+      "learning_rate": 0.00018457112101225492,
+      "loss": 1.2923,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2039212859511281,
+      "grad_norm": 0.09414175079065339,
+      "learning_rate": 0.00018455865971403124,
+      "loss": 1.3429,
+      "step": 2829
+    },
+    {
+      "epoch": 0.203993368413465,
+      "grad_norm": 0.10772457375711016,
+      "learning_rate": 0.0001845461938066041,
+      "loss": 1.3251,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2040654508758019,
+      "grad_norm": 0.11924318211579443,
+      "learning_rate": 0.00018453372329065303,
+      "loss": 1.2643,
+      "step": 2831
+    },
+    {
+      "epoch": 0.20413753333813883,
+      "grad_norm": 0.11177921031488607,
+      "learning_rate": 0.00018452124816685777,
+      "loss": 1.2806,
+      "step": 2832
+    },
+    {
+      "epoch": 0.20420961580047575,
+      "grad_norm": 0.09790716455054219,
+      "learning_rate": 0.00018450876843589836,
+      "loss": 1.239,
+      "step": 2833
+    },
+    {
+      "epoch": 0.20428169826281264,
+      "grad_norm": 0.10438564046107696,
+      "learning_rate": 0.000184496284098455,
+      "loss": 1.3239,
+      "step": 2834
+    },
+    {
+      "epoch": 0.20435378072514956,
+      "grad_norm": 0.10040669258509173,
+      "learning_rate": 0.00018448379515520827,
+      "loss": 1.2446,
+      "step": 2835
+    },
+    {
+      "epoch": 0.20442586318748648,
+      "grad_norm": 0.10685879767006032,
+      "learning_rate": 0.00018447130160683886,
+      "loss": 1.3973,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2044979456498234,
+      "grad_norm": 0.11015384041356077,
+      "learning_rate": 0.0001844588034540278,
+      "loss": 1.3347,
+      "step": 2837
+    },
+    {
+      "epoch": 0.20457002811216032,
+      "grad_norm": 0.10683586133845292,
+      "learning_rate": 0.0001844463006974564,
+      "loss": 1.2647,
+      "step": 2838
+    },
+    {
+      "epoch": 0.20464211057449722,
+      "grad_norm": 0.11585483560194591,
+      "learning_rate": 0.00018443379333780613,
+      "loss": 1.2838,
+      "step": 2839
+    },
+    {
+      "epoch": 0.20471419303683414,
+      "grad_norm": 0.11485178905280694,
+      "learning_rate": 0.00018442128137575873,
+      "loss": 1.318,
+      "step": 2840
+    },
+    {
+      "epoch": 0.20478627549917106,
+      "grad_norm": 0.09317572701549919,
+      "learning_rate": 0.00018440876481199626,
+      "loss": 1.3968,
+      "step": 2841
+    },
+    {
+      "epoch": 0.20485835796150798,
+      "grad_norm": 0.10770971066852093,
+      "learning_rate": 0.00018439624364720097,
+      "loss": 1.2587,
+      "step": 2842
+    },
+    {
+      "epoch": 0.20493044042384487,
+      "grad_norm": 0.16777811118129754,
+      "learning_rate": 0.00018438371788205536,
+      "loss": 1.3107,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2050025228861818,
+      "grad_norm": 0.11839743239177417,
+      "learning_rate": 0.0001843711875172422,
+      "loss": 1.2744,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2050746053485187,
+      "grad_norm": 0.1035189030092311,
+      "learning_rate": 0.00018435865255344452,
+      "loss": 1.3131,
+      "step": 2845
+    },
+    {
+      "epoch": 0.20514668781085563,
+      "grad_norm": 0.11081364095244457,
+      "learning_rate": 0.00018434611299134558,
+      "loss": 1.3526,
+      "step": 2846
+    },
+    {
+      "epoch": 0.20521877027319252,
+      "grad_norm": 0.11190378718525713,
+      "learning_rate": 0.00018433356883162892,
+      "loss": 1.389,
+      "step": 2847
+    },
+    {
+      "epoch": 0.20529085273552944,
+      "grad_norm": 0.10358589310901753,
+      "learning_rate": 0.00018432102007497826,
+      "loss": 1.2193,
+      "step": 2848
+    },
+    {
+      "epoch": 0.20536293519786636,
+      "grad_norm": 0.11734613110790706,
+      "learning_rate": 0.00018430846672207768,
+      "loss": 1.3347,
+      "step": 2849
+    },
+    {
+      "epoch": 0.20543501766020328,
+      "grad_norm": 0.10418702756863514,
+      "learning_rate": 0.0001842959087736114,
+      "loss": 1.3633,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2055071001225402,
+      "grad_norm": 0.12208004128359663,
+      "learning_rate": 0.00018428334623026395,
+      "loss": 1.2894,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2055791825848771,
+      "grad_norm": 0.0976075544132287,
+      "learning_rate": 0.0001842707790927201,
+      "loss": 1.2439,
+      "step": 2852
+    },
+    {
+      "epoch": 0.205651265047214,
+      "grad_norm": 0.10226951510510644,
+      "learning_rate": 0.0001842582073616649,
+      "loss": 1.2415,
+      "step": 2853
+    },
+    {
+      "epoch": 0.20572334750955093,
+      "grad_norm": 0.10770958004924541,
+      "learning_rate": 0.00018424563103778364,
+      "loss": 1.4082,
+      "step": 2854
+    },
+    {
+      "epoch": 0.20579542997188785,
+      "grad_norm": 0.1328943634996466,
+      "learning_rate": 0.00018423305012176176,
+      "loss": 1.2741,
+      "step": 2855
+    },
+    {
+      "epoch": 0.20586751243422474,
+      "grad_norm": 0.10588282628876056,
+      "learning_rate": 0.00018422046461428508,
+      "loss": 1.3248,
+      "step": 2856
+    },
+    {
+      "epoch": 0.20593959489656166,
+      "grad_norm": 0.11276538063162425,
+      "learning_rate": 0.00018420787451603961,
+      "loss": 1.3566,
+      "step": 2857
+    },
+    {
+      "epoch": 0.20601167735889858,
+      "grad_norm": 0.10240673092804352,
+      "learning_rate": 0.00018419527982771166,
+      "loss": 1.2929,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2060837598212355,
+      "grad_norm": 0.106881102527288,
+      "learning_rate": 0.00018418268054998772,
+      "loss": 1.3255,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2061558422835724,
+      "grad_norm": 0.33501270813117084,
+      "learning_rate": 0.00018417007668355457,
+      "loss": 1.2849,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2062279247459093,
+      "grad_norm": 0.1156994679268259,
+      "learning_rate": 0.00018415746822909922,
+      "loss": 1.2751,
+      "step": 2861
+    },
+    {
+      "epoch": 0.20630000720824623,
+      "grad_norm": 0.10312113842026259,
+      "learning_rate": 0.000184144855187309,
+      "loss": 1.3887,
+      "step": 2862
+    },
+    {
+      "epoch": 0.20637208967058315,
+      "grad_norm": 0.11459701308966717,
+      "learning_rate": 0.00018413223755887136,
+      "loss": 1.4198,
+      "step": 2863
+    },
+    {
+      "epoch": 0.20644417213292007,
+      "grad_norm": 0.1017597090861847,
+      "learning_rate": 0.0001841196153444741,
+      "loss": 1.3426,
+      "step": 2864
+    },
+    {
+      "epoch": 0.20651625459525697,
+      "grad_norm": 0.24739276123861695,
+      "learning_rate": 0.00018410698854480525,
+      "loss": 1.3944,
+      "step": 2865
+    },
+    {
+      "epoch": 0.20658833705759388,
+      "grad_norm": 0.11074603342155227,
+      "learning_rate": 0.00018409435716055308,
+      "loss": 1.2487,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2066604195199308,
+      "grad_norm": 0.1017087418124017,
+      "learning_rate": 0.00018408172119240614,
+      "loss": 1.3537,
+      "step": 2867
+    },
+    {
+      "epoch": 0.20673250198226772,
+      "grad_norm": 0.11319361388142203,
+      "learning_rate": 0.00018406908064105314,
+      "loss": 1.4601,
+      "step": 2868
+    },
+    {
+      "epoch": 0.20680458444460462,
+      "grad_norm": 0.10549117351101438,
+      "learning_rate": 0.00018405643550718317,
+      "loss": 1.2684,
+      "step": 2869
+    },
+    {
+      "epoch": 0.20687666690694154,
+      "grad_norm": 0.10894771216993364,
+      "learning_rate": 0.00018404378579148546,
+      "loss": 1.3068,
+      "step": 2870
+    },
+    {
+      "epoch": 0.20694874936927846,
+      "grad_norm": 0.10691284594296273,
+      "learning_rate": 0.00018403113149464953,
+      "loss": 1.3602,
+      "step": 2871
+    },
+    {
+      "epoch": 0.20702083183161538,
+      "grad_norm": 0.10833855609293848,
+      "learning_rate": 0.00018401847261736523,
+      "loss": 1.3386,
+      "step": 2872
+    },
+    {
+      "epoch": 0.20709291429395227,
+      "grad_norm": 0.12389311906817642,
+      "learning_rate": 0.00018400580916032246,
+      "loss": 1.2979,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2071649967562892,
+      "grad_norm": 0.11009378999846406,
+      "learning_rate": 0.0001839931411242116,
+      "loss": 1.3861,
+      "step": 2874
+    },
+    {
+      "epoch": 0.2072370792186261,
+      "grad_norm": 0.10835272754726583,
+      "learning_rate": 0.0001839804685097231,
+      "loss": 1.3824,
+      "step": 2875
+    },
+    {
+      "epoch": 0.20730916168096303,
+      "grad_norm": 0.10649743444024459,
+      "learning_rate": 0.00018396779131754775,
+      "loss": 1.3564,
+      "step": 2876
+    },
+    {
+      "epoch": 0.20738124414329995,
+      "grad_norm": 0.15367743605882997,
+      "learning_rate": 0.00018395510954837658,
+      "loss": 1.222,
+      "step": 2877
+    },
+    {
+      "epoch": 0.20745332660563684,
+      "grad_norm": 0.11327441570716332,
+      "learning_rate": 0.00018394242320290087,
+      "loss": 1.3086,
+      "step": 2878
+    },
+    {
+      "epoch": 0.20752540906797376,
+      "grad_norm": 0.11090849928823791,
+      "learning_rate": 0.00018392973228181209,
+      "loss": 1.3228,
+      "step": 2879
+    },
+    {
+      "epoch": 0.20759749153031068,
+      "grad_norm": 0.11181260119067024,
+      "learning_rate": 0.00018391703678580207,
+      "loss": 1.3037,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2076695739926476,
+      "grad_norm": 0.11445317658463812,
+      "learning_rate": 0.0001839043367155628,
+      "loss": 1.3138,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2077416564549845,
+      "grad_norm": 0.11220383035032211,
+      "learning_rate": 0.00018389163207178656,
+      "loss": 1.3515,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2078137389173214,
+      "grad_norm": 0.10770536316697617,
+      "learning_rate": 0.00018387892285516582,
+      "loss": 1.2803,
+      "step": 2883
+    },
+    {
+      "epoch": 0.20788582137965833,
+      "grad_norm": 0.10745948759577786,
+      "learning_rate": 0.00018386620906639343,
+      "loss": 1.2181,
+      "step": 2884
+    },
+    {
+      "epoch": 0.20795790384199525,
+      "grad_norm": 0.10600362247540665,
+      "learning_rate": 0.00018385349070616233,
+      "loss": 1.4117,
+      "step": 2885
+    },
+    {
+      "epoch": 0.20802998630433214,
+      "grad_norm": 0.10916435880823787,
+      "learning_rate": 0.00018384076777516582,
+      "loss": 1.3831,
+      "step": 2886
+    },
+    {
+      "epoch": 0.20810206876666906,
+      "grad_norm": 0.13193981543877684,
+      "learning_rate": 0.0001838280402740974,
+      "loss": 1.3227,
+      "step": 2887
+    },
+    {
+      "epoch": 0.20817415122900598,
+      "grad_norm": 0.10912596699942263,
+      "learning_rate": 0.00018381530820365085,
+      "loss": 1.3262,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2082462336913429,
+      "grad_norm": 0.28957908176878605,
+      "learning_rate": 0.00018380257156452013,
+      "loss": 1.361,
+      "step": 2889
+    },
+    {
+      "epoch": 0.20831831615367982,
+      "grad_norm": 0.20080591499343287,
+      "learning_rate": 0.00018378983035739958,
+      "loss": 1.3374,
+      "step": 2890
+    },
+    {
+      "epoch": 0.20839039861601671,
+      "grad_norm": 0.13059788340338765,
+      "learning_rate": 0.0001837770845829837,
+      "loss": 1.2765,
+      "step": 2891
+    },
+    {
+      "epoch": 0.20846248107835363,
+      "grad_norm": 0.10247462991786946,
+      "learning_rate": 0.00018376433424196716,
+      "loss": 1.27,
+      "step": 2892
+    },
+    {
+      "epoch": 0.20853456354069055,
+      "grad_norm": 0.11601391574707544,
+      "learning_rate": 0.00018375157933504503,
+      "loss": 1.2521,
+      "step": 2893
+    },
+    {
+      "epoch": 0.20860664600302747,
+      "grad_norm": 0.12606265613691336,
+      "learning_rate": 0.00018373881986291258,
+      "loss": 1.2485,
+      "step": 2894
+    },
+    {
+      "epoch": 0.20867872846536437,
+      "grad_norm": 0.10926631523421948,
+      "learning_rate": 0.0001837260558262653,
+      "loss": 1.3197,
+      "step": 2895
+    },
+    {
+      "epoch": 0.20875081092770129,
+      "grad_norm": 0.10330718630474725,
+      "learning_rate": 0.00018371328722579895,
+      "loss": 1.1831,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2088228933900382,
+      "grad_norm": 0.12893422259656542,
+      "learning_rate": 0.0001837005140622095,
+      "loss": 1.2636,
+      "step": 2897
+    },
+    {
+      "epoch": 0.20889497585237513,
+      "grad_norm": 0.10857966027797793,
+      "learning_rate": 0.00018368773633619325,
+      "loss": 1.3659,
+      "step": 2898
+    },
+    {
+      "epoch": 0.20896705831471202,
+      "grad_norm": 0.11374572789097037,
+      "learning_rate": 0.00018367495404844668,
+      "loss": 1.3137,
+      "step": 2899
+    },
+    {
+      "epoch": 0.20903914077704894,
+      "grad_norm": 0.10230001642787044,
+      "learning_rate": 0.0001836621671996665,
+      "loss": 1.3902,
+      "step": 2900
+    },
+    {
+      "epoch": 0.20911122323938586,
+      "grad_norm": 0.10537329897005103,
+      "learning_rate": 0.0001836493757905498,
+      "loss": 1.4103,
+      "step": 2901
+    },
+    {
+      "epoch": 0.20918330570172278,
+      "grad_norm": 0.10012106051323638,
+      "learning_rate": 0.00018363657982179373,
+      "loss": 1.3125,
+      "step": 2902
+    },
+    {
+      "epoch": 0.2092553881640597,
+      "grad_norm": 0.1088939155516287,
+      "learning_rate": 0.00018362377929409583,
+      "loss": 1.3038,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2093274706263966,
+      "grad_norm": 0.19663184890777707,
+      "learning_rate": 0.00018361097420815386,
+      "loss": 1.3392,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2093995530887335,
+      "grad_norm": 0.10642636460025953,
+      "learning_rate": 0.00018359816456466575,
+      "loss": 1.3427,
+      "step": 2905
+    },
+    {
+      "epoch": 0.20947163555107043,
+      "grad_norm": 0.10361947594282399,
+      "learning_rate": 0.00018358535036432983,
+      "loss": 1.3479,
+      "step": 2906
+    },
+    {
+      "epoch": 0.20954371801340735,
+      "grad_norm": 0.1005901913578703,
+      "learning_rate": 0.0001835725316078445,
+      "loss": 1.2836,
+      "step": 2907
+    },
+    {
+      "epoch": 0.20961580047574424,
+      "grad_norm": 0.0996716389207294,
+      "learning_rate": 0.00018355970829590857,
+      "loss": 1.3377,
+      "step": 2908
+    },
+    {
+      "epoch": 0.20968788293808116,
+      "grad_norm": 0.10859810256022127,
+      "learning_rate": 0.00018354688042922096,
+      "loss": 1.2502,
+      "step": 2909
+    },
+    {
+      "epoch": 0.20975996540041808,
+      "grad_norm": 0.2640719562487404,
+      "learning_rate": 0.00018353404800848096,
+      "loss": 1.3468,
+      "step": 2910
+    },
+    {
+      "epoch": 0.209832047862755,
+      "grad_norm": 0.12376676256770425,
+      "learning_rate": 0.000183521211034388,
+      "loss": 1.3049,
+      "step": 2911
+    },
+    {
+      "epoch": 0.2099041303250919,
+      "grad_norm": 0.10852950346049739,
+      "learning_rate": 0.00018350836950764188,
+      "loss": 1.3424,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2099762127874288,
+      "grad_norm": 0.09973943279523449,
+      "learning_rate": 0.0001834955234289425,
+      "loss": 1.3359,
+      "step": 2913
+    },
+    {
+      "epoch": 0.21004829524976573,
+      "grad_norm": 0.1712843585874355,
+      "learning_rate": 0.00018348267279899013,
+      "loss": 1.2517,
+      "step": 2914
+    },
+    {
+      "epoch": 0.21012037771210265,
+      "grad_norm": 0.11038416595892352,
+      "learning_rate": 0.00018346981761848522,
+      "loss": 1.3449,
+      "step": 2915
+    },
+    {
+      "epoch": 0.21019246017443957,
+      "grad_norm": 0.1166089419659546,
+      "learning_rate": 0.00018345695788812854,
+      "loss": 1.3223,
+      "step": 2916
+    },
+    {
+      "epoch": 0.21026454263677646,
+      "grad_norm": 0.38682197078645675,
+      "learning_rate": 0.000183444093608621,
+      "loss": 1.3888,
+      "step": 2917
+    },
+    {
+      "epoch": 0.21033662509911338,
+      "grad_norm": 0.11906365915000346,
+      "learning_rate": 0.00018343122478066385,
+      "loss": 1.2067,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2104087075614503,
+      "grad_norm": 0.11325714183952337,
+      "learning_rate": 0.00018341835140495854,
+      "loss": 1.2025,
+      "step": 2919
+    },
+    {
+      "epoch": 0.21048079002378722,
+      "grad_norm": 0.11418199259589887,
+      "learning_rate": 0.00018340547348220682,
+      "loss": 1.3199,
+      "step": 2920
+    },
+    {
+      "epoch": 0.21055287248612412,
+      "grad_norm": 0.10443120096485858,
+      "learning_rate": 0.00018339259101311063,
+      "loss": 1.2833,
+      "step": 2921
+    },
+    {
+      "epoch": 0.21062495494846104,
+      "grad_norm": 0.1425359366914154,
+      "learning_rate": 0.0001833797039983722,
+      "loss": 1.2671,
+      "step": 2922
+    },
+    {
+      "epoch": 0.21069703741079795,
+      "grad_norm": 0.11139002845642028,
+      "learning_rate": 0.00018336681243869392,
+      "loss": 1.2763,
+      "step": 2923
+    },
+    {
+      "epoch": 0.21076911987313487,
+      "grad_norm": 0.11050342744584897,
+      "learning_rate": 0.00018335391633477857,
+      "loss": 1.2948,
+      "step": 2924
+    },
+    {
+      "epoch": 0.21084120233547177,
+      "grad_norm": 0.10821163222840768,
+      "learning_rate": 0.00018334101568732905,
+      "loss": 1.3597,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2109132847978087,
+      "grad_norm": 0.09743517436962537,
+      "learning_rate": 0.00018332811049704862,
+      "loss": 1.263,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2109853672601456,
+      "grad_norm": 0.11055531307809988,
+      "learning_rate": 0.00018331520076464065,
+      "loss": 1.3459,
+      "step": 2927
+    },
+    {
+      "epoch": 0.21105744972248253,
+      "grad_norm": 0.10530247237307198,
+      "learning_rate": 0.0001833022864908089,
+      "loss": 1.377,
+      "step": 2928
+    },
+    {
+      "epoch": 0.21112953218481945,
+      "grad_norm": 0.11007081041056148,
+      "learning_rate": 0.00018328936767625728,
+      "loss": 1.3031,
+      "step": 2929
+    },
+    {
+      "epoch": 0.21120161464715634,
+      "grad_norm": 0.1304672045438983,
+      "learning_rate": 0.00018327644432169002,
+      "loss": 1.2933,
+      "step": 2930
+    },
+    {
+      "epoch": 0.21127369710949326,
+      "grad_norm": 0.30691645935251627,
+      "learning_rate": 0.0001832635164278115,
+      "loss": 1.3305,
+      "step": 2931
+    },
+    {
+      "epoch": 0.21134577957183018,
+      "grad_norm": 0.10273571289641709,
+      "learning_rate": 0.00018325058399532647,
+      "loss": 1.3033,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2114178620341671,
+      "grad_norm": 0.10991266579810913,
+      "learning_rate": 0.0001832376470249398,
+      "loss": 1.2527,
+      "step": 2933
+    },
+    {
+      "epoch": 0.211489944496504,
+      "grad_norm": 0.1172721479897722,
+      "learning_rate": 0.00018322470551735673,
+      "loss": 1.1214,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2115620269588409,
+      "grad_norm": 0.1437095314427446,
+      "learning_rate": 0.00018321175947328268,
+      "loss": 1.2867,
+      "step": 2935
+    },
+    {
+      "epoch": 0.21163410942117783,
+      "grad_norm": 0.19001305112730837,
+      "learning_rate": 0.00018319880889342325,
+      "loss": 1.3486,
+      "step": 2936
+    },
+    {
+      "epoch": 0.21170619188351475,
+      "grad_norm": 0.11689020144326964,
+      "learning_rate": 0.00018318585377848442,
+      "loss": 1.3266,
+      "step": 2937
+    },
+    {
+      "epoch": 0.21177827434585164,
+      "grad_norm": 0.1052606755677612,
+      "learning_rate": 0.0001831728941291724,
+      "loss": 1.3113,
+      "step": 2938
+    },
+    {
+      "epoch": 0.21185035680818856,
+      "grad_norm": 0.11772992562248757,
+      "learning_rate": 0.00018315992994619355,
+      "loss": 1.3926,
+      "step": 2939
+    },
+    {
+      "epoch": 0.21192243927052548,
+      "grad_norm": 0.10810481170824751,
+      "learning_rate": 0.00018314696123025454,
+      "loss": 1.2336,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2119945217328624,
+      "grad_norm": 0.11447415673775276,
+      "learning_rate": 0.0001831339879820623,
+      "loss": 1.2983,
+      "step": 2941
+    },
+    {
+      "epoch": 0.21206660419519932,
+      "grad_norm": 0.13058759241945006,
+      "learning_rate": 0.00018312101020232394,
+      "loss": 1.281,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2121386866575362,
+      "grad_norm": 0.10052309779214075,
+      "learning_rate": 0.00018310802789174693,
+      "loss": 1.3238,
+      "step": 2943
+    },
+    {
+      "epoch": 0.21221076911987313,
+      "grad_norm": 0.12753164194776165,
+      "learning_rate": 0.0001830950410510389,
+      "loss": 1.2734,
+      "step": 2944
+    },
+    {
+      "epoch": 0.21228285158221005,
+      "grad_norm": 0.10893009945284945,
+      "learning_rate": 0.00018308204968090774,
+      "loss": 1.2915,
+      "step": 2945
+    },
+    {
+      "epoch": 0.21235493404454697,
+      "grad_norm": 0.13025901817032323,
+      "learning_rate": 0.00018306905378206157,
+      "loss": 1.4238,
+      "step": 2946
+    },
+    {
+      "epoch": 0.21242701650688386,
+      "grad_norm": 0.11576440686868168,
+      "learning_rate": 0.00018305605335520886,
+      "loss": 1.2799,
+      "step": 2947
+    },
+    {
+      "epoch": 0.21249909896922078,
+      "grad_norm": 0.11359705873433934,
+      "learning_rate": 0.00018304304840105817,
+      "loss": 1.3596,
+      "step": 2948
+    },
+    {
+      "epoch": 0.2125711814315577,
+      "grad_norm": 0.11241494044730936,
+      "learning_rate": 0.00018303003892031843,
+      "loss": 1.3331,
+      "step": 2949
+    },
+    {
+      "epoch": 0.21264326389389462,
+      "grad_norm": 0.09422837112307615,
+      "learning_rate": 0.00018301702491369876,
+      "loss": 1.262,
+      "step": 2950
+    },
+    {
+      "epoch": 0.21271534635623152,
+      "grad_norm": 0.11196696326009258,
+      "learning_rate": 0.00018300400638190855,
+      "loss": 1.3349,
+      "step": 2951
+    },
+    {
+      "epoch": 0.21278742881856844,
+      "grad_norm": 0.12345318721872842,
+      "learning_rate": 0.0001829909833256574,
+      "loss": 1.2484,
+      "step": 2952
+    },
+    {
+      "epoch": 0.21285951128090536,
+      "grad_norm": 0.10385629982726023,
+      "learning_rate": 0.0001829779557456552,
+      "loss": 1.2676,
+      "step": 2953
+    },
+    {
+      "epoch": 0.21293159374324228,
+      "grad_norm": 0.0972649326193182,
+      "learning_rate": 0.00018296492364261208,
+      "loss": 1.3239,
+      "step": 2954
+    },
+    {
+      "epoch": 0.2130036762055792,
+      "grad_norm": 0.10318672232640037,
+      "learning_rate": 0.00018295188701723838,
+      "loss": 1.2536,
+      "step": 2955
+    },
+    {
+      "epoch": 0.2130757586679161,
+      "grad_norm": 0.10640542648925579,
+      "learning_rate": 0.00018293884587024475,
+      "loss": 1.2988,
+      "step": 2956
+    },
+    {
+      "epoch": 0.213147841130253,
+      "grad_norm": 0.10370918997525051,
+      "learning_rate": 0.000182925800202342,
+      "loss": 1.187,
+      "step": 2957
+    },
+    {
+      "epoch": 0.21321992359258993,
+      "grad_norm": 0.11535001739649325,
+      "learning_rate": 0.0001829127500142413,
+      "loss": 1.4029,
+      "step": 2958
+    },
+    {
+      "epoch": 0.21329200605492685,
+      "grad_norm": 0.12663843896151875,
+      "learning_rate": 0.00018289969530665395,
+      "loss": 1.3221,
+      "step": 2959
+    },
+    {
+      "epoch": 0.21336408851726374,
+      "grad_norm": 0.10382391048553562,
+      "learning_rate": 0.00018288663608029157,
+      "loss": 1.3155,
+      "step": 2960
+    },
+    {
+      "epoch": 0.21343617097960066,
+      "grad_norm": 0.1360134971085304,
+      "learning_rate": 0.00018287357233586597,
+      "loss": 1.3932,
+      "step": 2961
+    },
+    {
+      "epoch": 0.21350825344193758,
+      "grad_norm": 0.6206570316399912,
+      "learning_rate": 0.0001828605040740893,
+      "loss": 1.3597,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2135803359042745,
+      "grad_norm": 0.10702775716578203,
+      "learning_rate": 0.00018284743129567388,
+      "loss": 1.2578,
+      "step": 2963
+    },
+    {
+      "epoch": 0.2136524183666114,
+      "grad_norm": 0.10179561631230319,
+      "learning_rate": 0.00018283435400133225,
+      "loss": 1.3391,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2137245008289483,
+      "grad_norm": 0.11524472540674714,
+      "learning_rate": 0.00018282127219177726,
+      "loss": 1.2286,
+      "step": 2965
+    },
+    {
+      "epoch": 0.21379658329128523,
+      "grad_norm": 0.10387264286664143,
+      "learning_rate": 0.00018280818586772203,
+      "loss": 1.1895,
+      "step": 2966
+    },
+    {
+      "epoch": 0.21386866575362215,
+      "grad_norm": 0.11131172076484347,
+      "learning_rate": 0.00018279509502987983,
+      "loss": 1.3059,
+      "step": 2967
+    },
+    {
+      "epoch": 0.21394074821595907,
+      "grad_norm": 0.10321645201181079,
+      "learning_rate": 0.00018278199967896423,
+      "loss": 1.2726,
+      "step": 2968
+    },
+    {
+      "epoch": 0.21401283067829596,
+      "grad_norm": 0.23874599268333307,
+      "learning_rate": 0.00018276889981568906,
+      "loss": 1.4176,
+      "step": 2969
+    },
+    {
+      "epoch": 0.21408491314063288,
+      "grad_norm": 0.12709143668927203,
+      "learning_rate": 0.0001827557954407684,
+      "loss": 1.2738,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2141569956029698,
+      "grad_norm": 0.11455187269709922,
+      "learning_rate": 0.00018274268655491652,
+      "loss": 1.4184,
+      "step": 2971
+    },
+    {
+      "epoch": 0.21422907806530672,
+      "grad_norm": 0.1170764455974419,
+      "learning_rate": 0.000182729573158848,
+      "loss": 1.3216,
+      "step": 2972
+    },
+    {
+      "epoch": 0.2143011605276436,
+      "grad_norm": 0.11912428300176055,
+      "learning_rate": 0.0001827164552532776,
+      "loss": 1.2621,
+      "step": 2973
+    },
+    {
+      "epoch": 0.21437324298998053,
+      "grad_norm": 0.12120905157405285,
+      "learning_rate": 0.0001827033328389204,
+      "loss": 1.3383,
+      "step": 2974
+    },
+    {
+      "epoch": 0.21444532545231745,
+      "grad_norm": 0.13186096863409041,
+      "learning_rate": 0.00018269020591649168,
+      "loss": 1.3404,
+      "step": 2975
+    },
+    {
+      "epoch": 0.21451740791465437,
+      "grad_norm": 0.13829291458567902,
+      "learning_rate": 0.00018267707448670696,
+      "loss": 1.3385,
+      "step": 2976
+    },
+    {
+      "epoch": 0.21458949037699127,
+      "grad_norm": 0.10978263291540363,
+      "learning_rate": 0.00018266393855028204,
+      "loss": 1.2998,
+      "step": 2977
+    },
+    {
+      "epoch": 0.21466157283932819,
+      "grad_norm": 0.1089669246517979,
+      "learning_rate": 0.00018265079810793293,
+      "loss": 1.2666,
+      "step": 2978
+    },
+    {
+      "epoch": 0.2147336553016651,
+      "grad_norm": 0.10399701080879417,
+      "learning_rate": 0.00018263765316037592,
+      "loss": 1.373,
+      "step": 2979
+    },
+    {
+      "epoch": 0.21480573776400202,
+      "grad_norm": 0.12470306428587362,
+      "learning_rate": 0.00018262450370832748,
+      "loss": 1.2422,
+      "step": 2980
+    },
+    {
+      "epoch": 0.21487782022633894,
+      "grad_norm": 0.11860736764582185,
+      "learning_rate": 0.00018261134975250441,
+      "loss": 1.2804,
+      "step": 2981
+    },
+    {
+      "epoch": 0.21494990268867584,
+      "grad_norm": 0.10621258642882886,
+      "learning_rate": 0.00018259819129362375,
+      "loss": 1.2943,
+      "step": 2982
+    },
+    {
+      "epoch": 0.21502198515101276,
+      "grad_norm": 0.10648876270257994,
+      "learning_rate": 0.00018258502833240272,
+      "loss": 1.3457,
+      "step": 2983
+    },
+    {
+      "epoch": 0.21509406761334968,
+      "grad_norm": 0.10417844963378595,
+      "learning_rate": 0.0001825718608695588,
+      "loss": 1.2831,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2151661500756866,
+      "grad_norm": 0.10540890666359566,
+      "learning_rate": 0.00018255868890580977,
+      "loss": 1.2551,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2152382325380235,
+      "grad_norm": 0.10201938858813153,
+      "learning_rate": 0.00018254551244187357,
+      "loss": 1.2948,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2153103150003604,
+      "grad_norm": 0.2940981205440516,
+      "learning_rate": 0.00018253233147846851,
+      "loss": 1.2993,
+      "step": 2987
+    },
+    {
+      "epoch": 0.21538239746269733,
+      "grad_norm": 0.10430689739335933,
+      "learning_rate": 0.00018251914601631302,
+      "loss": 1.3141,
+      "step": 2988
+    },
+    {
+      "epoch": 0.21545447992503425,
+      "grad_norm": 0.12252652189032276,
+      "learning_rate": 0.0001825059560561258,
+      "loss": 1.246,
+      "step": 2989
+    },
+    {
+      "epoch": 0.21552656238737114,
+      "grad_norm": 0.1062913673580392,
+      "learning_rate": 0.0001824927615986259,
+      "loss": 1.339,
+      "step": 2990
+    },
+    {
+      "epoch": 0.21559864484970806,
+      "grad_norm": 0.11218394448481744,
+      "learning_rate": 0.00018247956264453249,
+      "loss": 1.3744,
+      "step": 2991
+    },
+    {
+      "epoch": 0.21567072731204498,
+      "grad_norm": 0.11627331795330403,
+      "learning_rate": 0.000182466359194565,
+      "loss": 1.3296,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2157428097743819,
+      "grad_norm": 0.10646063866234125,
+      "learning_rate": 0.00018245315124944318,
+      "loss": 1.3353,
+      "step": 2993
+    },
+    {
+      "epoch": 0.21581489223671882,
+      "grad_norm": 0.11833464329374321,
+      "learning_rate": 0.00018243993880988696,
+      "loss": 1.2708,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2158869746990557,
+      "grad_norm": 0.1138773912062882,
+      "learning_rate": 0.00018242672187661658,
+      "loss": 1.2765,
+      "step": 2995
+    },
+    {
+      "epoch": 0.21595905716139263,
+      "grad_norm": 0.11542687266402635,
+      "learning_rate": 0.00018241350045035244,
+      "loss": 1.3527,
+      "step": 2996
+    },
+    {
+      "epoch": 0.21603113962372955,
+      "grad_norm": 0.10770493347483268,
+      "learning_rate": 0.0001824002745318152,
+      "loss": 1.2602,
+      "step": 2997
+    },
+    {
+      "epoch": 0.21610322208606647,
+      "grad_norm": 0.10008041109977217,
+      "learning_rate": 0.00018238704412172586,
+      "loss": 1.2896,
+      "step": 2998
+    },
+    {
+      "epoch": 0.21617530454840336,
+      "grad_norm": 0.1055728580480517,
+      "learning_rate": 0.00018237380922080554,
+      "loss": 1.2018,
+      "step": 2999
+    },
+    {
+      "epoch": 0.21624738701074028,
+      "grad_norm": 0.11323432254214702,
+      "learning_rate": 0.0001823605698297757,
+      "loss": 1.4249,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2163194694730772,
+      "grad_norm": 0.20754647683713057,
+      "learning_rate": 0.00018234732594935794,
+      "loss": 1.3514,
+      "step": 3001
+    },
+    {
+      "epoch": 0.21639155193541412,
+      "grad_norm": 0.1038401896892828,
+      "learning_rate": 0.00018233407758027424,
+      "loss": 1.2406,
+      "step": 3002
+    },
+    {
+      "epoch": 0.21646363439775101,
+      "grad_norm": 0.11890242288941659,
+      "learning_rate": 0.00018232082472324676,
+      "loss": 1.3996,
+      "step": 3003
+    },
+    {
+      "epoch": 0.21653571686008793,
+      "grad_norm": 0.10785699520836059,
+      "learning_rate": 0.00018230756737899783,
+      "loss": 1.2527,
+      "step": 3004
+    },
+    {
+      "epoch": 0.21660779932242485,
+      "grad_norm": 0.09977266257082122,
+      "learning_rate": 0.00018229430554825017,
+      "loss": 1.233,
+      "step": 3005
+    },
+    {
+      "epoch": 0.21667988178476177,
+      "grad_norm": 0.11987556042000932,
+      "learning_rate": 0.0001822810392317266,
+      "loss": 1.2259,
+      "step": 3006
+    },
+    {
+      "epoch": 0.2167519642470987,
+      "grad_norm": 0.11745807561453847,
+      "learning_rate": 0.0001822677684301503,
+      "loss": 1.3114,
+      "step": 3007
+    },
+    {
+      "epoch": 0.21682404670943559,
+      "grad_norm": 0.11192645470347873,
+      "learning_rate": 0.00018225449314424462,
+      "loss": 1.2412,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2168961291717725,
+      "grad_norm": 0.11469234578523646,
+      "learning_rate": 0.0001822412133747332,
+      "loss": 1.3407,
+      "step": 3009
+    },
+    {
+      "epoch": 0.21696821163410943,
+      "grad_norm": 0.10198896387303936,
+      "learning_rate": 0.0001822279291223399,
+      "loss": 1.243,
+      "step": 3010
+    },
+    {
+      "epoch": 0.21704029409644635,
+      "grad_norm": 0.11149775995693742,
+      "learning_rate": 0.00018221464038778885,
+      "loss": 1.3366,
+      "step": 3011
+    },
+    {
+      "epoch": 0.21711237655878324,
+      "grad_norm": 0.11548531501911133,
+      "learning_rate": 0.00018220134717180435,
+      "loss": 1.3451,
+      "step": 3012
+    },
+    {
+      "epoch": 0.21718445902112016,
+      "grad_norm": 0.11785693647435656,
+      "learning_rate": 0.00018218804947511106,
+      "loss": 1.4077,
+      "step": 3013
+    },
+    {
+      "epoch": 0.21725654148345708,
+      "grad_norm": 0.18159529076570433,
+      "learning_rate": 0.00018217474729843378,
+      "loss": 1.315,
+      "step": 3014
+    },
+    {
+      "epoch": 0.217328623945794,
+      "grad_norm": 0.12086531802296546,
+      "learning_rate": 0.00018216144064249763,
+      "loss": 1.4144,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2174007064081309,
+      "grad_norm": 0.11283706843975529,
+      "learning_rate": 0.00018214812950802792,
+      "loss": 1.3634,
+      "step": 3016
+    },
+    {
+      "epoch": 0.2174727888704678,
+      "grad_norm": 0.12549642876939224,
+      "learning_rate": 0.00018213481389575025,
+      "loss": 1.2601,
+      "step": 3017
+    },
+    {
+      "epoch": 0.21754487133280473,
+      "grad_norm": 0.11737243825321247,
+      "learning_rate": 0.0001821214938063904,
+      "loss": 1.2605,
+      "step": 3018
+    },
+    {
+      "epoch": 0.21761695379514165,
+      "grad_norm": 0.1163893736283323,
+      "learning_rate": 0.00018210816924067448,
+      "loss": 1.2238,
+      "step": 3019
+    },
+    {
+      "epoch": 0.21768903625747857,
+      "grad_norm": 0.10605716192633302,
+      "learning_rate": 0.00018209484019932875,
+      "loss": 1.3543,
+      "step": 3020
+    },
+    {
+      "epoch": 0.21776111871981546,
+      "grad_norm": 0.17259032555133053,
+      "learning_rate": 0.0001820815066830798,
+      "loss": 1.3583,
+      "step": 3021
+    },
+    {
+      "epoch": 0.21783320118215238,
+      "grad_norm": 0.13197002659710905,
+      "learning_rate": 0.00018206816869265443,
+      "loss": 1.3645,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2179052836444893,
+      "grad_norm": 0.10766512491631414,
+      "learning_rate": 0.00018205482622877966,
+      "loss": 1.3687,
+      "step": 3023
+    },
+    {
+      "epoch": 0.21797736610682622,
+      "grad_norm": 0.18915585863613701,
+      "learning_rate": 0.0001820414792921828,
+      "loss": 1.3175,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2180494485691631,
+      "grad_norm": 0.09674250791549045,
+      "learning_rate": 0.0001820281278835913,
+      "loss": 1.2945,
+      "step": 3025
+    },
+    {
+      "epoch": 0.21812153103150003,
+      "grad_norm": 0.20209069999897578,
+      "learning_rate": 0.000182014772003733,
+      "loss": 1.2242,
+      "step": 3026
+    },
+    {
+      "epoch": 0.21819361349383695,
+      "grad_norm": 0.10103549781204466,
+      "learning_rate": 0.0001820014116533359,
+      "loss": 1.3245,
+      "step": 3027
+    },
+    {
+      "epoch": 0.21826569595617387,
+      "grad_norm": 0.10829348038044724,
+      "learning_rate": 0.00018198804683312826,
+      "loss": 1.3037,
+      "step": 3028
+    },
+    {
+      "epoch": 0.21833777841851076,
+      "grad_norm": 0.11072523627789026,
+      "learning_rate": 0.00018197467754383858,
+      "loss": 1.3866,
+      "step": 3029
+    },
+    {
+      "epoch": 0.21840986088084768,
+      "grad_norm": 0.10994249234599769,
+      "learning_rate": 0.00018196130378619562,
+      "loss": 1.2496,
+      "step": 3030
+    },
+    {
+      "epoch": 0.2184819433431846,
+      "grad_norm": 0.11052812956476665,
+      "learning_rate": 0.0001819479255609284,
+      "loss": 1.3377,
+      "step": 3031
+    },
+    {
+      "epoch": 0.21855402580552152,
+      "grad_norm": 0.09577600285741338,
+      "learning_rate": 0.00018193454286876602,
+      "loss": 1.2995,
+      "step": 3032
+    },
+    {
+      "epoch": 0.21862610826785844,
+      "grad_norm": 0.10748348698040705,
+      "learning_rate": 0.0001819211557104381,
+      "loss": 1.3639,
+      "step": 3033
+    },
+    {
+      "epoch": 0.21869819073019534,
+      "grad_norm": 0.10439878805436867,
+      "learning_rate": 0.00018190776408667428,
+      "loss": 1.2321,
+      "step": 3034
+    },
+    {
+      "epoch": 0.21877027319253226,
+      "grad_norm": 0.10605884822015131,
+      "learning_rate": 0.00018189436799820457,
+      "loss": 1.2863,
+      "step": 3035
+    },
+    {
+      "epoch": 0.21884235565486918,
+      "grad_norm": 0.11341152055147037,
+      "learning_rate": 0.00018188096744575918,
+      "loss": 1.3726,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2189144381172061,
+      "grad_norm": 0.31444065504382507,
+      "learning_rate": 0.0001818675624300685,
+      "loss": 1.3275,
+      "step": 3037
+    },
+    {
+      "epoch": 0.218986520579543,
+      "grad_norm": 0.10961592352740679,
+      "learning_rate": 0.0001818541529518633,
+      "loss": 1.3762,
+      "step": 3038
+    },
+    {
+      "epoch": 0.2190586030418799,
+      "grad_norm": 0.11802365744337366,
+      "learning_rate": 0.0001818407390118744,
+      "loss": 1.3337,
+      "step": 3039
+    },
+    {
+      "epoch": 0.21913068550421683,
+      "grad_norm": 0.10816237331679555,
+      "learning_rate": 0.0001818273206108331,
+      "loss": 1.3129,
+      "step": 3040
+    },
+    {
+      "epoch": 0.21920276796655375,
+      "grad_norm": 0.1171754857297476,
+      "learning_rate": 0.00018181389774947078,
+      "loss": 1.285,
+      "step": 3041
+    },
+    {
+      "epoch": 0.21927485042889064,
+      "grad_norm": 0.10827539480811328,
+      "learning_rate": 0.0001818004704285191,
+      "loss": 1.3695,
+      "step": 3042
+    },
+    {
+      "epoch": 0.21934693289122756,
+      "grad_norm": 0.13763790304565826,
+      "learning_rate": 0.00018178703864870998,
+      "loss": 1.2253,
+      "step": 3043
+    },
+    {
+      "epoch": 0.21941901535356448,
+      "grad_norm": 0.10602504314582689,
+      "learning_rate": 0.00018177360241077557,
+      "loss": 1.2914,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2194910978159014,
+      "grad_norm": 0.11578847992597126,
+      "learning_rate": 0.00018176016171544826,
+      "loss": 1.3882,
+      "step": 3045
+    },
+    {
+      "epoch": 0.21956318027823832,
+      "grad_norm": 0.1474020395827384,
+      "learning_rate": 0.00018174671656346069,
+      "loss": 1.2517,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2196352627405752,
+      "grad_norm": 0.11329703621044492,
+      "learning_rate": 0.0001817332669555457,
+      "loss": 1.28,
+      "step": 3047
+    },
+    {
+      "epoch": 0.21970734520291213,
+      "grad_norm": 0.10524054333075546,
+      "learning_rate": 0.0001817198128924365,
+      "loss": 1.2488,
+      "step": 3048
+    },
+    {
+      "epoch": 0.21977942766524905,
+      "grad_norm": 0.10492780408784828,
+      "learning_rate": 0.0001817063543748664,
+      "loss": 1.2975,
+      "step": 3049
+    },
+    {
+      "epoch": 0.21985151012758597,
+      "grad_norm": 0.11261555903772505,
+      "learning_rate": 0.00018169289140356902,
+      "loss": 1.3635,
+      "step": 3050
+    },
+    {
+      "epoch": 0.21992359258992286,
+      "grad_norm": 0.12682133359044165,
+      "learning_rate": 0.0001816794239792782,
+      "loss": 1.3326,
+      "step": 3051
+    },
+    {
+      "epoch": 0.21999567505225978,
+      "grad_norm": 0.09912365710090865,
+      "learning_rate": 0.00018166595210272805,
+      "loss": 1.3296,
+      "step": 3052
+    },
+    {
+      "epoch": 0.2200677575145967,
+      "grad_norm": 0.11227030964240167,
+      "learning_rate": 0.00018165247577465294,
+      "loss": 1.3132,
+      "step": 3053
+    },
+    {
+      "epoch": 0.22013983997693362,
+      "grad_norm": 0.10653638097705087,
+      "learning_rate": 0.0001816389949957874,
+      "loss": 1.3644,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2202119224392705,
+      "grad_norm": 0.10330223187399985,
+      "learning_rate": 0.00018162550976686626,
+      "loss": 1.3425,
+      "step": 3055
+    },
+    {
+      "epoch": 0.22028400490160743,
+      "grad_norm": 0.09574575648852174,
+      "learning_rate": 0.00018161202008862458,
+      "loss": 1.2857,
+      "step": 3056
+    },
+    {
+      "epoch": 0.22035608736394435,
+      "grad_norm": 0.11027767266281165,
+      "learning_rate": 0.0001815985259617977,
+      "loss": 1.3475,
+      "step": 3057
+    },
+    {
+      "epoch": 0.22042816982628127,
+      "grad_norm": 0.13700115923013173,
+      "learning_rate": 0.00018158502738712114,
+      "loss": 1.3511,
+      "step": 3058
+    },
+    {
+      "epoch": 0.2205002522886182,
+      "grad_norm": 0.11505643809284287,
+      "learning_rate": 0.00018157152436533072,
+      "loss": 1.3485,
+      "step": 3059
+    },
+    {
+      "epoch": 0.22057233475095508,
+      "grad_norm": 0.11484534876781946,
+      "learning_rate": 0.00018155801689716245,
+      "loss": 1.2754,
+      "step": 3060
+    },
+    {
+      "epoch": 0.220644417213292,
+      "grad_norm": 0.1567677271937715,
+      "learning_rate": 0.00018154450498335264,
+      "loss": 1.2754,
+      "step": 3061
+    },
+    {
+      "epoch": 0.22071649967562892,
+      "grad_norm": 0.1208434181911948,
+      "learning_rate": 0.00018153098862463778,
+      "loss": 1.3233,
+      "step": 3062
+    },
+    {
+      "epoch": 0.22078858213796584,
+      "grad_norm": 0.09997167594961617,
+      "learning_rate": 0.00018151746782175466,
+      "loss": 1.324,
+      "step": 3063
+    },
+    {
+      "epoch": 0.22086066460030274,
+      "grad_norm": 0.10556073794832152,
+      "learning_rate": 0.00018150394257544027,
+      "loss": 1.3836,
+      "step": 3064
+    },
+    {
+      "epoch": 0.22093274706263966,
+      "grad_norm": 0.11287394967646747,
+      "learning_rate": 0.00018149041288643183,
+      "loss": 1.3624,
+      "step": 3065
+    },
+    {
+      "epoch": 0.22100482952497658,
+      "grad_norm": 0.10814814846857682,
+      "learning_rate": 0.00018147687875546685,
+      "loss": 1.3083,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2210769119873135,
+      "grad_norm": 0.10338027512240922,
+      "learning_rate": 0.00018146334018328308,
+      "loss": 1.3704,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2211489944496504,
+      "grad_norm": 0.11019525107327693,
+      "learning_rate": 0.00018144979717061847,
+      "loss": 1.3341,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2212210769119873,
+      "grad_norm": 0.1219120784172822,
+      "learning_rate": 0.00018143624971821124,
+      "loss": 1.3207,
+      "step": 3069
+    },
+    {
+      "epoch": 0.22129315937432423,
+      "grad_norm": 0.11014781718730776,
+      "learning_rate": 0.00018142269782679988,
+      "loss": 1.2952,
+      "step": 3070
+    },
+    {
+      "epoch": 0.22136524183666115,
+      "grad_norm": 0.11033700614465825,
+      "learning_rate": 0.00018140914149712302,
+      "loss": 1.2862,
+      "step": 3071
+    },
+    {
+      "epoch": 0.22143732429899807,
+      "grad_norm": 0.10572903021213782,
+      "learning_rate": 0.00018139558072991966,
+      "loss": 1.4223,
+      "step": 3072
+    },
+    {
+      "epoch": 0.22150940676133496,
+      "grad_norm": 0.10042832343155109,
+      "learning_rate": 0.00018138201552592896,
+      "loss": 1.3364,
+      "step": 3073
+    },
+    {
+      "epoch": 0.22158148922367188,
+      "grad_norm": 0.11145063017585696,
+      "learning_rate": 0.00018136844588589036,
+      "loss": 1.2835,
+      "step": 3074
+    },
+    {
+      "epoch": 0.2216535716860088,
+      "grad_norm": 0.1335396087496954,
+      "learning_rate": 0.00018135487181054351,
+      "loss": 1.1434,
+      "step": 3075
+    },
+    {
+      "epoch": 0.22172565414834572,
+      "grad_norm": 0.10061995436603793,
+      "learning_rate": 0.00018134129330062834,
+      "loss": 1.3057,
+      "step": 3076
+    },
+    {
+      "epoch": 0.2217977366106826,
+      "grad_norm": 0.10889980491287295,
+      "learning_rate": 0.00018132771035688498,
+      "loss": 1.3855,
+      "step": 3077
+    },
+    {
+      "epoch": 0.22186981907301953,
+      "grad_norm": 0.1062471172687284,
+      "learning_rate": 0.0001813141229800538,
+      "loss": 1.3512,
+      "step": 3078
+    },
+    {
+      "epoch": 0.22194190153535645,
+      "grad_norm": 0.09769577377671121,
+      "learning_rate": 0.00018130053117087547,
+      "loss": 1.2722,
+      "step": 3079
+    },
+    {
+      "epoch": 0.22201398399769337,
+      "grad_norm": 0.10138913632389443,
+      "learning_rate": 0.00018128693493009088,
+      "loss": 1.3169,
+      "step": 3080
+    },
+    {
+      "epoch": 0.22208606646003026,
+      "grad_norm": 0.11377160814489125,
+      "learning_rate": 0.00018127333425844107,
+      "loss": 1.3121,
+      "step": 3081
+    },
+    {
+      "epoch": 0.22215814892236718,
+      "grad_norm": 0.11126002706659939,
+      "learning_rate": 0.00018125972915666752,
+      "loss": 1.4229,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2222302313847041,
+      "grad_norm": 0.12304683989983321,
+      "learning_rate": 0.00018124611962551173,
+      "loss": 1.1995,
+      "step": 3083
+    },
+    {
+      "epoch": 0.22230231384704102,
+      "grad_norm": 0.11934420141191199,
+      "learning_rate": 0.00018123250566571557,
+      "loss": 1.3458,
+      "step": 3084
+    },
+    {
+      "epoch": 0.22237439630937794,
+      "grad_norm": 0.10847695744424585,
+      "learning_rate": 0.00018121888727802113,
+      "loss": 1.4315,
+      "step": 3085
+    },
+    {
+      "epoch": 0.22244647877171483,
+      "grad_norm": 0.11862372784501661,
+      "learning_rate": 0.00018120526446317072,
+      "loss": 1.3257,
+      "step": 3086
+    },
+    {
+      "epoch": 0.22251856123405175,
+      "grad_norm": 0.11655863305124113,
+      "learning_rate": 0.0001811916372219069,
+      "loss": 1.2752,
+      "step": 3087
+    },
+    {
+      "epoch": 0.22259064369638867,
+      "grad_norm": 0.14220526949247503,
+      "learning_rate": 0.00018117800555497252,
+      "loss": 1.27,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2226627261587256,
+      "grad_norm": 0.1313561693046036,
+      "learning_rate": 0.00018116436946311057,
+      "loss": 1.3083,
+      "step": 3089
+    },
+    {
+      "epoch": 0.22273480862106249,
+      "grad_norm": 0.10936064071665162,
+      "learning_rate": 0.0001811507289470644,
+      "loss": 1.3274,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2228068910833994,
+      "grad_norm": 0.10327530016868566,
+      "learning_rate": 0.0001811370840075775,
+      "loss": 1.2554,
+      "step": 3091
+    },
+    {
+      "epoch": 0.22287897354573633,
+      "grad_norm": 0.12227158396747234,
+      "learning_rate": 0.00018112343464539362,
+      "loss": 1.4199,
+      "step": 3092
+    },
+    {
+      "epoch": 0.22295105600807325,
+      "grad_norm": 0.39067998368938184,
+      "learning_rate": 0.00018110978086125683,
+      "loss": 1.3711,
+      "step": 3093
+    },
+    {
+      "epoch": 0.22302313847041014,
+      "grad_norm": 0.1064738722848325,
+      "learning_rate": 0.00018109612265591134,
+      "loss": 1.3383,
+      "step": 3094
+    },
+    {
+      "epoch": 0.22309522093274706,
+      "grad_norm": 0.18114330733437034,
+      "learning_rate": 0.00018108246003010166,
+      "loss": 1.268,
+      "step": 3095
+    },
+    {
+      "epoch": 0.22316730339508398,
+      "grad_norm": 0.10690873839996048,
+      "learning_rate": 0.0001810687929845725,
+      "loss": 1.239,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2232393858574209,
+      "grad_norm": 0.11481403196241453,
+      "learning_rate": 0.00018105512152006892,
+      "loss": 1.2522,
+      "step": 3097
+    },
+    {
+      "epoch": 0.22331146831975782,
+      "grad_norm": 0.100517202960385,
+      "learning_rate": 0.00018104144563733604,
+      "loss": 1.3231,
+      "step": 3098
+    },
+    {
+      "epoch": 0.2233835507820947,
+      "grad_norm": 0.11419350023721805,
+      "learning_rate": 0.00018102776533711936,
+      "loss": 1.3692,
+      "step": 3099
+    },
+    {
+      "epoch": 0.22345563324443163,
+      "grad_norm": 0.12017840007118762,
+      "learning_rate": 0.00018101408062016455,
+      "loss": 1.2816,
+      "step": 3100
+    },
+    {
+      "epoch": 0.22352771570676855,
+      "grad_norm": 0.11354023070517157,
+      "learning_rate": 0.00018100039148721763,
+      "loss": 1.3572,
+      "step": 3101
+    },
+    {
+      "epoch": 0.22359979816910547,
+      "grad_norm": 0.12209273856606975,
+      "learning_rate": 0.00018098669793902467,
+      "loss": 1.411,
+      "step": 3102
+    },
+    {
+      "epoch": 0.22367188063144236,
+      "grad_norm": 0.105165099466102,
+      "learning_rate": 0.00018097299997633218,
+      "loss": 1.3225,
+      "step": 3103
+    },
+    {
+      "epoch": 0.22374396309377928,
+      "grad_norm": 0.1442146572398029,
+      "learning_rate": 0.00018095929759988675,
+      "loss": 1.3904,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2238160455561162,
+      "grad_norm": 0.16997685671423518,
+      "learning_rate": 0.00018094559081043533,
+      "loss": 1.2795,
+      "step": 3105
+    },
+    {
+      "epoch": 0.22388812801845312,
+      "grad_norm": 0.1155193250105558,
+      "learning_rate": 0.00018093187960872506,
+      "loss": 1.3584,
+      "step": 3106
+    },
+    {
+      "epoch": 0.22396021048079,
+      "grad_norm": 0.11064079531858405,
+      "learning_rate": 0.0001809181639955033,
+      "loss": 1.2882,
+      "step": 3107
+    },
+    {
+      "epoch": 0.22403229294312693,
+      "grad_norm": 0.2321202582918886,
+      "learning_rate": 0.00018090444397151774,
+      "loss": 1.3109,
+      "step": 3108
+    },
+    {
+      "epoch": 0.22410437540546385,
+      "grad_norm": 0.10590273361066375,
+      "learning_rate": 0.00018089071953751614,
+      "loss": 1.2802,
+      "step": 3109
+    },
+    {
+      "epoch": 0.22417645786780077,
+      "grad_norm": 0.10653991843445879,
+      "learning_rate": 0.00018087699069424667,
+      "loss": 1.3367,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2242485403301377,
+      "grad_norm": 0.10135655284101468,
+      "learning_rate": 0.00018086325744245767,
+      "loss": 1.2778,
+      "step": 3111
+    },
+    {
+      "epoch": 0.22432062279247458,
+      "grad_norm": 0.09880875667258994,
+      "learning_rate": 0.0001808495197828977,
+      "loss": 1.3755,
+      "step": 3112
+    },
+    {
+      "epoch": 0.2243927052548115,
+      "grad_norm": 0.10048418965183713,
+      "learning_rate": 0.00018083577771631562,
+      "loss": 1.2598,
+      "step": 3113
+    },
+    {
+      "epoch": 0.22446478771714842,
+      "grad_norm": 0.1573048872835811,
+      "learning_rate": 0.00018082203124346045,
+      "loss": 1.3377,
+      "step": 3114
+    },
+    {
+      "epoch": 0.22453687017948534,
+      "grad_norm": 0.10582097864945209,
+      "learning_rate": 0.00018080828036508154,
+      "loss": 1.3257,
+      "step": 3115
+    },
+    {
+      "epoch": 0.22460895264182223,
+      "grad_norm": 0.09966291735183379,
+      "learning_rate": 0.00018079452508192842,
+      "loss": 1.3214,
+      "step": 3116
+    },
+    {
+      "epoch": 0.22468103510415915,
+      "grad_norm": 0.10979963181972553,
+      "learning_rate": 0.00018078076539475085,
+      "loss": 1.3874,
+      "step": 3117
+    },
+    {
+      "epoch": 0.22475311756649607,
+      "grad_norm": 0.09900127231618402,
+      "learning_rate": 0.0001807670013042989,
+      "loss": 1.1977,
+      "step": 3118
+    },
+    {
+      "epoch": 0.224825200028833,
+      "grad_norm": 0.09830598145159662,
+      "learning_rate": 0.00018075323281132284,
+      "loss": 1.3102,
+      "step": 3119
+    },
+    {
+      "epoch": 0.2248972824911699,
+      "grad_norm": 0.10322174568490823,
+      "learning_rate": 0.00018073945991657313,
+      "loss": 1.2819,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2249693649535068,
+      "grad_norm": 0.11298347565703946,
+      "learning_rate": 0.00018072568262080053,
+      "loss": 1.2924,
+      "step": 3121
+    },
+    {
+      "epoch": 0.22504144741584373,
+      "grad_norm": 0.09742166812678953,
+      "learning_rate": 0.00018071190092475603,
+      "loss": 1.1944,
+      "step": 3122
+    },
+    {
+      "epoch": 0.22511352987818065,
+      "grad_norm": 0.10983322776780752,
+      "learning_rate": 0.0001806981148291909,
+      "loss": 1.3756,
+      "step": 3123
+    },
+    {
+      "epoch": 0.22518561234051757,
+      "grad_norm": 0.14972371715258198,
+      "learning_rate": 0.00018068432433485652,
+      "loss": 1.3125,
+      "step": 3124
+    },
+    {
+      "epoch": 0.22525769480285446,
+      "grad_norm": 0.11700302823661461,
+      "learning_rate": 0.00018067052944250467,
+      "loss": 1.3879,
+      "step": 3125
+    },
+    {
+      "epoch": 0.22532977726519138,
+      "grad_norm": 0.11075930910060969,
+      "learning_rate": 0.00018065673015288725,
+      "loss": 1.2525,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2254018597275283,
+      "grad_norm": 0.11703401274047108,
+      "learning_rate": 0.00018064292646675646,
+      "loss": 1.3645,
+      "step": 3127
+    },
+    {
+      "epoch": 0.22547394218986522,
+      "grad_norm": 0.10949066452779312,
+      "learning_rate": 0.00018062911838486473,
+      "loss": 1.3175,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2255460246522021,
+      "grad_norm": 0.10080906930063784,
+      "learning_rate": 0.00018061530590796475,
+      "loss": 1.4213,
+      "step": 3129
+    },
+    {
+      "epoch": 0.22561810711453903,
+      "grad_norm": 0.17074537888888697,
+      "learning_rate": 0.00018060148903680933,
+      "loss": 1.39,
+      "step": 3130
+    },
+    {
+      "epoch": 0.22569018957687595,
+      "grad_norm": 0.12213279989347739,
+      "learning_rate": 0.0001805876677721517,
+      "loss": 1.2194,
+      "step": 3131
+    },
+    {
+      "epoch": 0.22576227203921287,
+      "grad_norm": 0.12887519400521705,
+      "learning_rate": 0.00018057384211474527,
+      "loss": 1.3706,
+      "step": 3132
+    },
+    {
+      "epoch": 0.22583435450154976,
+      "grad_norm": 0.12549766897516096,
+      "learning_rate": 0.00018056001206534356,
+      "loss": 1.3124,
+      "step": 3133
+    },
+    {
+      "epoch": 0.22590643696388668,
+      "grad_norm": 0.10890224233192342,
+      "learning_rate": 0.0001805461776247005,
+      "loss": 1.2894,
+      "step": 3134
+    },
+    {
+      "epoch": 0.2259785194262236,
+      "grad_norm": 0.14886595198036423,
+      "learning_rate": 0.00018053233879357014,
+      "loss": 1.2731,
+      "step": 3135
+    },
+    {
+      "epoch": 0.22605060188856052,
+      "grad_norm": 0.1130520208361066,
+      "learning_rate": 0.00018051849557270688,
+      "loss": 1.2256,
+      "step": 3136
+    },
+    {
+      "epoch": 0.22612268435089744,
+      "grad_norm": 0.12815825264927083,
+      "learning_rate": 0.00018050464796286528,
+      "loss": 1.3521,
+      "step": 3137
+    },
+    {
+      "epoch": 0.22619476681323433,
+      "grad_norm": 0.11917125613410696,
+      "learning_rate": 0.00018049079596480015,
+      "loss": 1.2927,
+      "step": 3138
+    },
+    {
+      "epoch": 0.22626684927557125,
+      "grad_norm": 0.10697658745063662,
+      "learning_rate": 0.00018047693957926652,
+      "loss": 1.3593,
+      "step": 3139
+    },
+    {
+      "epoch": 0.22633893173790817,
+      "grad_norm": 0.10012862329869142,
+      "learning_rate": 0.00018046307880701976,
+      "loss": 1.3312,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2264110142002451,
+      "grad_norm": 0.16101975763702978,
+      "learning_rate": 0.00018044921364881533,
+      "loss": 1.3282,
+      "step": 3141
+    },
+    {
+      "epoch": 0.22648309666258198,
+      "grad_norm": 0.11481581805762958,
+      "learning_rate": 0.00018043534410540905,
+      "loss": 1.3251,
+      "step": 3142
+    },
+    {
+      "epoch": 0.2265551791249189,
+      "grad_norm": 0.10067850565730241,
+      "learning_rate": 0.0001804214701775569,
+      "loss": 1.3141,
+      "step": 3143
+    },
+    {
+      "epoch": 0.22662726158725582,
+      "grad_norm": 0.10865189055079259,
+      "learning_rate": 0.00018040759186601518,
+      "loss": 1.313,
+      "step": 3144
+    },
+    {
+      "epoch": 0.22669934404959274,
+      "grad_norm": 0.10091051556588672,
+      "learning_rate": 0.00018039370917154036,
+      "loss": 1.3663,
+      "step": 3145
+    },
+    {
+      "epoch": 0.22677142651192964,
+      "grad_norm": 0.11512842027825866,
+      "learning_rate": 0.00018037982209488914,
+      "loss": 1.2843,
+      "step": 3146
+    },
+    {
+      "epoch": 0.22684350897426656,
+      "grad_norm": 0.11333891293868012,
+      "learning_rate": 0.00018036593063681852,
+      "loss": 1.303,
+      "step": 3147
+    },
+    {
+      "epoch": 0.22691559143660348,
+      "grad_norm": 0.11965529565075501,
+      "learning_rate": 0.00018035203479808574,
+      "loss": 1.353,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2269876738989404,
+      "grad_norm": 0.11946586977537511,
+      "learning_rate": 0.00018033813457944816,
+      "loss": 1.3989,
+      "step": 3149
+    },
+    {
+      "epoch": 0.22705975636127732,
+      "grad_norm": 0.11126444407486942,
+      "learning_rate": 0.00018032422998166356,
+      "loss": 1.2952,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2271318388236142,
+      "grad_norm": 0.11547355007710684,
+      "learning_rate": 0.0001803103210054898,
+      "loss": 1.3349,
+      "step": 3151
+    },
+    {
+      "epoch": 0.22720392128595113,
+      "grad_norm": 0.12058471187087111,
+      "learning_rate": 0.0001802964076516851,
+      "loss": 1.4772,
+      "step": 3152
+    },
+    {
+      "epoch": 0.22727600374828805,
+      "grad_norm": 0.11021786308977628,
+      "learning_rate": 0.00018028248992100782,
+      "loss": 1.3295,
+      "step": 3153
+    },
+    {
+      "epoch": 0.22734808621062497,
+      "grad_norm": 0.11277434146002512,
+      "learning_rate": 0.0001802685678142166,
+      "loss": 1.3537,
+      "step": 3154
+    },
+    {
+      "epoch": 0.22742016867296186,
+      "grad_norm": 0.12539078421876876,
+      "learning_rate": 0.00018025464133207035,
+      "loss": 1.338,
+      "step": 3155
+    },
+    {
+      "epoch": 0.22749225113529878,
+      "grad_norm": 0.10984461272997749,
+      "learning_rate": 0.00018024071047532817,
+      "loss": 1.1953,
+      "step": 3156
+    },
+    {
+      "epoch": 0.2275643335976357,
+      "grad_norm": 0.10970911278159459,
+      "learning_rate": 0.00018022677524474938,
+      "loss": 1.2815,
+      "step": 3157
+    },
+    {
+      "epoch": 0.22763641605997262,
+      "grad_norm": 0.11757296389868538,
+      "learning_rate": 0.00018021283564109364,
+      "loss": 1.3196,
+      "step": 3158
+    },
+    {
+      "epoch": 0.2277084985223095,
+      "grad_norm": 0.1278798214456731,
+      "learning_rate": 0.00018019889166512074,
+      "loss": 1.4211,
+      "step": 3159
+    },
+    {
+      "epoch": 0.22778058098464643,
+      "grad_norm": 0.13356930376260032,
+      "learning_rate": 0.00018018494331759077,
+      "loss": 1.2344,
+      "step": 3160
+    },
+    {
+      "epoch": 0.22785266344698335,
+      "grad_norm": 0.10781687673134771,
+      "learning_rate": 0.00018017099059926403,
+      "loss": 1.3076,
+      "step": 3161
+    },
+    {
+      "epoch": 0.22792474590932027,
+      "grad_norm": 0.12082200319458303,
+      "learning_rate": 0.00018015703351090107,
+      "loss": 1.3139,
+      "step": 3162
+    },
+    {
+      "epoch": 0.22799682837165716,
+      "grad_norm": 0.11099660240063673,
+      "learning_rate": 0.00018014307205326268,
+      "loss": 1.221,
+      "step": 3163
+    },
+    {
+      "epoch": 0.22806891083399408,
+      "grad_norm": 0.10691898083580852,
+      "learning_rate": 0.0001801291062271099,
+      "loss": 1.3583,
+      "step": 3164
+    },
+    {
+      "epoch": 0.228140993296331,
+      "grad_norm": 0.09773579664208691,
+      "learning_rate": 0.00018011513603320397,
+      "loss": 1.3185,
+      "step": 3165
+    },
+    {
+      "epoch": 0.22821307575866792,
+      "grad_norm": 0.13989909582790813,
+      "learning_rate": 0.00018010116147230637,
+      "loss": 1.2937,
+      "step": 3166
+    },
+    {
+      "epoch": 0.22828515822100484,
+      "grad_norm": 0.10923913905384831,
+      "learning_rate": 0.0001800871825451789,
+      "loss": 1.3407,
+      "step": 3167
+    },
+    {
+      "epoch": 0.22835724068334173,
+      "grad_norm": 0.10241572669136083,
+      "learning_rate": 0.00018007319925258348,
+      "loss": 1.2833,
+      "step": 3168
+    },
+    {
+      "epoch": 0.22842932314567865,
+      "grad_norm": 0.11526229296347383,
+      "learning_rate": 0.00018005921159528233,
+      "loss": 1.327,
+      "step": 3169
+    },
+    {
+      "epoch": 0.22850140560801557,
+      "grad_norm": 0.11085384162106252,
+      "learning_rate": 0.00018004521957403795,
+      "loss": 1.3635,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2285734880703525,
+      "grad_norm": 0.14205170111008178,
+      "learning_rate": 0.000180031223189613,
+      "loss": 1.2407,
+      "step": 3171
+    },
+    {
+      "epoch": 0.22864557053268939,
+      "grad_norm": 0.10008910178594013,
+      "learning_rate": 0.00018001722244277035,
+      "loss": 1.3439,
+      "step": 3172
+    },
+    {
+      "epoch": 0.2287176529950263,
+      "grad_norm": 0.10233272200515171,
+      "learning_rate": 0.00018000321733427328,
+      "loss": 1.2851,
+      "step": 3173
+    },
+    {
+      "epoch": 0.22878973545736322,
+      "grad_norm": 0.12340238145251728,
+      "learning_rate": 0.0001799892078648851,
+      "loss": 1.3282,
+      "step": 3174
+    },
+    {
+      "epoch": 0.22886181791970014,
+      "grad_norm": 0.1003601873016233,
+      "learning_rate": 0.00017997519403536953,
+      "loss": 1.225,
+      "step": 3175
+    },
+    {
+      "epoch": 0.22893390038203704,
+      "grad_norm": 0.10942668072671889,
+      "learning_rate": 0.00017996117584649038,
+      "loss": 1.3746,
+      "step": 3176
+    },
+    {
+      "epoch": 0.22900598284437396,
+      "grad_norm": 0.10501751705188382,
+      "learning_rate": 0.0001799471532990118,
+      "loss": 1.3136,
+      "step": 3177
+    },
+    {
+      "epoch": 0.22907806530671088,
+      "grad_norm": 0.11545081555479852,
+      "learning_rate": 0.00017993312639369812,
+      "loss": 1.3132,
+      "step": 3178
+    },
+    {
+      "epoch": 0.2291501477690478,
+      "grad_norm": 0.11582306651688815,
+      "learning_rate": 0.00017991909513131397,
+      "loss": 1.249,
+      "step": 3179
+    },
+    {
+      "epoch": 0.22922223023138472,
+      "grad_norm": 0.11105783170385515,
+      "learning_rate": 0.00017990505951262416,
+      "loss": 1.2812,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2292943126937216,
+      "grad_norm": 0.11118599608701184,
+      "learning_rate": 0.00017989101953839375,
+      "loss": 1.3223,
+      "step": 3181
+    },
+    {
+      "epoch": 0.22936639515605853,
+      "grad_norm": 0.3657099421454032,
+      "learning_rate": 0.00017987697520938803,
+      "loss": 1.3463,
+      "step": 3182
+    },
+    {
+      "epoch": 0.22943847761839545,
+      "grad_norm": 0.11630422398330373,
+      "learning_rate": 0.00017986292652637259,
+      "loss": 1.3827,
+      "step": 3183
+    },
+    {
+      "epoch": 0.22951056008073237,
+      "grad_norm": 0.11747367959136838,
+      "learning_rate": 0.00017984887349011314,
+      "loss": 1.299,
+      "step": 3184
+    },
+    {
+      "epoch": 0.22958264254306926,
+      "grad_norm": 0.09729169792054104,
+      "learning_rate": 0.00017983481610137577,
+      "loss": 1.4203,
+      "step": 3185
+    },
+    {
+      "epoch": 0.22965472500540618,
+      "grad_norm": 0.10922515575676493,
+      "learning_rate": 0.00017982075436092672,
+      "loss": 1.2508,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2297268074677431,
+      "grad_norm": 0.1078892124990222,
+      "learning_rate": 0.00017980668826953245,
+      "loss": 1.2859,
+      "step": 3187
+    },
+    {
+      "epoch": 0.22979888993008002,
+      "grad_norm": 0.10469262344131308,
+      "learning_rate": 0.00017979261782795965,
+      "loss": 1.2552,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2298709723924169,
+      "grad_norm": 0.11496060289097872,
+      "learning_rate": 0.0001797785430369754,
+      "loss": 1.3069,
+      "step": 3189
+    },
+    {
+      "epoch": 0.22994305485475383,
+      "grad_norm": 0.1242872869252329,
+      "learning_rate": 0.00017976446389734676,
+      "loss": 1.2566,
+      "step": 3190
+    },
+    {
+      "epoch": 0.23001513731709075,
+      "grad_norm": 0.1129359976519368,
+      "learning_rate": 0.0001797503804098413,
+      "loss": 1.3064,
+      "step": 3191
+    },
+    {
+      "epoch": 0.23008721977942767,
+      "grad_norm": 0.1151453531523794,
+      "learning_rate": 0.0001797362925752266,
+      "loss": 1.2903,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2301593022417646,
+      "grad_norm": 0.09579248871999443,
+      "learning_rate": 0.00017972220039427067,
+      "loss": 1.2269,
+      "step": 3193
+    },
+    {
+      "epoch": 0.23023138470410148,
+      "grad_norm": 0.11311794809218512,
+      "learning_rate": 0.00017970810386774155,
+      "loss": 1.3134,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2303034671664384,
+      "grad_norm": 0.10180387362826114,
+      "learning_rate": 0.00017969400299640772,
+      "loss": 1.3393,
+      "step": 3195
+    },
+    {
+      "epoch": 0.23037554962877532,
+      "grad_norm": 0.10949761729846197,
+      "learning_rate": 0.00017967989778103774,
+      "loss": 1.3298,
+      "step": 3196
+    },
+    {
+      "epoch": 0.23044763209111224,
+      "grad_norm": 0.12903566339342287,
+      "learning_rate": 0.0001796657882224005,
+      "loss": 1.3352,
+      "step": 3197
+    },
+    {
+      "epoch": 0.23051971455344913,
+      "grad_norm": 0.3009703116677409,
+      "learning_rate": 0.0001796516743212651,
+      "loss": 1.3005,
+      "step": 3198
+    },
+    {
+      "epoch": 0.23059179701578605,
+      "grad_norm": 0.11562408527671258,
+      "learning_rate": 0.00017963755607840084,
+      "loss": 1.3973,
+      "step": 3199
+    },
+    {
+      "epoch": 0.23066387947812297,
+      "grad_norm": 0.11509725175590618,
+      "learning_rate": 0.00017962343349457735,
+      "loss": 1.4083,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2307359619404599,
+      "grad_norm": 0.1615183619789243,
+      "learning_rate": 0.00017960930657056438,
+      "loss": 1.3098,
+      "step": 3201
+    },
+    {
+      "epoch": 0.23080804440279679,
+      "grad_norm": 0.13222207373154876,
+      "learning_rate": 0.00017959517530713202,
+      "loss": 1.3419,
+      "step": 3202
+    },
+    {
+      "epoch": 0.2308801268651337,
+      "grad_norm": 0.10600288692671647,
+      "learning_rate": 0.0001795810397050505,
+      "loss": 1.3043,
+      "step": 3203
+    },
+    {
+      "epoch": 0.23095220932747063,
+      "grad_norm": 0.14043472000610366,
+      "learning_rate": 0.00017956689976509038,
+      "loss": 1.4329,
+      "step": 3204
+    },
+    {
+      "epoch": 0.23102429178980755,
+      "grad_norm": 0.09631378797662374,
+      "learning_rate": 0.0001795527554880224,
+      "loss": 1.2887,
+      "step": 3205
+    },
+    {
+      "epoch": 0.23109637425214447,
+      "grad_norm": 0.11247295664115438,
+      "learning_rate": 0.0001795386068746175,
+      "loss": 1.2755,
+      "step": 3206
+    },
+    {
+      "epoch": 0.23116845671448136,
+      "grad_norm": 0.18327436743412143,
+      "learning_rate": 0.00017952445392564701,
+      "loss": 1.3106,
+      "step": 3207
+    },
+    {
+      "epoch": 0.23124053917681828,
+      "grad_norm": 0.1100875713525348,
+      "learning_rate": 0.00017951029664188233,
+      "loss": 1.3116,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2313126216391552,
+      "grad_norm": 0.11691044123588616,
+      "learning_rate": 0.0001794961350240951,
+      "loss": 1.2859,
+      "step": 3209
+    },
+    {
+      "epoch": 0.23138470410149212,
+      "grad_norm": 0.1115269502054107,
+      "learning_rate": 0.00017948196907305742,
+      "loss": 1.3021,
+      "step": 3210
+    },
+    {
+      "epoch": 0.231456786563829,
+      "grad_norm": 0.10379908145994156,
+      "learning_rate": 0.0001794677987895413,
+      "loss": 1.3901,
+      "step": 3211
+    },
+    {
+      "epoch": 0.23152886902616593,
+      "grad_norm": 0.12274237775338837,
+      "learning_rate": 0.00017945362417431918,
+      "loss": 1.2258,
+      "step": 3212
+    },
+    {
+      "epoch": 0.23160095148850285,
+      "grad_norm": 0.10057709665664573,
+      "learning_rate": 0.00017943944522816379,
+      "loss": 1.2591,
+      "step": 3213
+    },
+    {
+      "epoch": 0.23167303395083977,
+      "grad_norm": 0.1243259059784087,
+      "learning_rate": 0.0001794252619518479,
+      "loss": 1.383,
+      "step": 3214
+    },
+    {
+      "epoch": 0.23174511641317666,
+      "grad_norm": 0.210854587284376,
+      "learning_rate": 0.0001794110743461447,
+      "loss": 1.2577,
+      "step": 3215
+    },
+    {
+      "epoch": 0.23181719887551358,
+      "grad_norm": 0.11077984743455507,
+      "learning_rate": 0.00017939688241182751,
+      "loss": 1.2185,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2318892813378505,
+      "grad_norm": 0.11512059123263566,
+      "learning_rate": 0.00017938268614966994,
+      "loss": 1.3291,
+      "step": 3217
+    },
+    {
+      "epoch": 0.23196136380018742,
+      "grad_norm": 0.11822454188367935,
+      "learning_rate": 0.00017936848556044578,
+      "loss": 1.3188,
+      "step": 3218
+    },
+    {
+      "epoch": 0.23203344626252434,
+      "grad_norm": 0.12126965950600671,
+      "learning_rate": 0.00017935428064492912,
+      "loss": 1.3397,
+      "step": 3219
+    },
+    {
+      "epoch": 0.23210552872486123,
+      "grad_norm": 0.120572801770632,
+      "learning_rate": 0.00017934007140389425,
+      "loss": 1.3712,
+      "step": 3220
+    },
+    {
+      "epoch": 0.23217761118719815,
+      "grad_norm": 0.11086053422352864,
+      "learning_rate": 0.00017932585783811565,
+      "loss": 1.2461,
+      "step": 3221
+    },
+    {
+      "epoch": 0.23224969364953507,
+      "grad_norm": 0.10019392767377301,
+      "learning_rate": 0.00017931163994836817,
+      "loss": 1.3549,
+      "step": 3222
+    },
+    {
+      "epoch": 0.232321776111872,
+      "grad_norm": 0.13242618086415953,
+      "learning_rate": 0.00017929741773542674,
+      "loss": 1.3738,
+      "step": 3223
+    },
+    {
+      "epoch": 0.23239385857420888,
+      "grad_norm": 0.11070904739310058,
+      "learning_rate": 0.00017928319120006663,
+      "loss": 1.2488,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2324659410365458,
+      "grad_norm": 0.10901986445007296,
+      "learning_rate": 0.00017926896034306332,
+      "loss": 1.3624,
+      "step": 3225
+    },
+    {
+      "epoch": 0.23253802349888272,
+      "grad_norm": 0.1015649209373588,
+      "learning_rate": 0.00017925472516519252,
+      "loss": 1.3401,
+      "step": 3226
+    },
+    {
+      "epoch": 0.23261010596121964,
+      "grad_norm": 0.1065133071992268,
+      "learning_rate": 0.00017924048566723014,
+      "loss": 1.2811,
+      "step": 3227
+    },
+    {
+      "epoch": 0.23268218842355654,
+      "grad_norm": 0.10672429664228426,
+      "learning_rate": 0.0001792262418499524,
+      "loss": 1.3106,
+      "step": 3228
+    },
+    {
+      "epoch": 0.23275427088589346,
+      "grad_norm": 0.10407424677598298,
+      "learning_rate": 0.00017921199371413575,
+      "loss": 1.2725,
+      "step": 3229
+    },
+    {
+      "epoch": 0.23282635334823037,
+      "grad_norm": 0.11058855217814703,
+      "learning_rate": 0.00017919774126055673,
+      "loss": 1.2747,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2328984358105673,
+      "grad_norm": 0.1060376500982521,
+      "learning_rate": 0.0001791834844899923,
+      "loss": 1.2864,
+      "step": 3231
+    },
+    {
+      "epoch": 0.23297051827290421,
+      "grad_norm": 0.11128318278977929,
+      "learning_rate": 0.00017916922340321958,
+      "loss": 1.3262,
+      "step": 3232
+    },
+    {
+      "epoch": 0.2330426007352411,
+      "grad_norm": 0.11588003148789264,
+      "learning_rate": 0.00017915495800101594,
+      "loss": 1.2866,
+      "step": 3233
+    },
+    {
+      "epoch": 0.23311468319757803,
+      "grad_norm": 0.11316551547562707,
+      "learning_rate": 0.0001791406882841589,
+      "loss": 1.4346,
+      "step": 3234
+    },
+    {
+      "epoch": 0.23318676565991495,
+      "grad_norm": 0.2395287072972118,
+      "learning_rate": 0.00017912641425342637,
+      "loss": 1.3028,
+      "step": 3235
+    },
+    {
+      "epoch": 0.23325884812225187,
+      "grad_norm": 0.12287938475365742,
+      "learning_rate": 0.00017911213590959636,
+      "loss": 1.4512,
+      "step": 3236
+    },
+    {
+      "epoch": 0.23333093058458876,
+      "grad_norm": 0.10718764727953449,
+      "learning_rate": 0.00017909785325344717,
+      "loss": 1.3662,
+      "step": 3237
+    },
+    {
+      "epoch": 0.23340301304692568,
+      "grad_norm": 0.11626659541518945,
+      "learning_rate": 0.00017908356628575737,
+      "loss": 1.31,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2334750955092626,
+      "grad_norm": 0.11580499767735539,
+      "learning_rate": 0.0001790692750073057,
+      "loss": 1.3571,
+      "step": 3239
+    },
+    {
+      "epoch": 0.23354717797159952,
+      "grad_norm": 0.10708141753410035,
+      "learning_rate": 0.00017905497941887117,
+      "loss": 1.3185,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2336192604339364,
+      "grad_norm": 0.10541381031901068,
+      "learning_rate": 0.00017904067952123303,
+      "loss": 1.3868,
+      "step": 3241
+    },
+    {
+      "epoch": 0.23369134289627333,
+      "grad_norm": 0.10319172235193196,
+      "learning_rate": 0.00017902637531517072,
+      "loss": 1.2967,
+      "step": 3242
+    },
+    {
+      "epoch": 0.23376342535861025,
+      "grad_norm": 0.10974612126095203,
+      "learning_rate": 0.00017901206680146398,
+      "loss": 1.3309,
+      "step": 3243
+    },
+    {
+      "epoch": 0.23383550782094717,
+      "grad_norm": 0.1097248742840685,
+      "learning_rate": 0.00017899775398089274,
+      "loss": 1.3706,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2339075902832841,
+      "grad_norm": 0.09877086767421121,
+      "learning_rate": 0.00017898343685423716,
+      "loss": 1.3089,
+      "step": 3245
+    },
+    {
+      "epoch": 0.23397967274562098,
+      "grad_norm": 0.0971819435217171,
+      "learning_rate": 0.00017896911542227765,
+      "loss": 1.3738,
+      "step": 3246
+    },
+    {
+      "epoch": 0.2340517552079579,
+      "grad_norm": 0.10812298712800282,
+      "learning_rate": 0.0001789547896857949,
+      "loss": 1.3735,
+      "step": 3247
+    },
+    {
+      "epoch": 0.23412383767029482,
+      "grad_norm": 0.12171917150275552,
+      "learning_rate": 0.00017894045964556975,
+      "loss": 1.3284,
+      "step": 3248
+    },
+    {
+      "epoch": 0.23419592013263174,
+      "grad_norm": 0.11442823655886425,
+      "learning_rate": 0.00017892612530238334,
+      "loss": 1.3359,
+      "step": 3249
+    },
+    {
+      "epoch": 0.23426800259496863,
+      "grad_norm": 0.11498720427891879,
+      "learning_rate": 0.00017891178665701697,
+      "loss": 1.3223,
+      "step": 3250
+    },
+    {
+      "epoch": 0.23434008505730555,
+      "grad_norm": 0.09889678731318004,
+      "learning_rate": 0.0001788974437102523,
+      "loss": 1.2272,
+      "step": 3251
+    },
+    {
+      "epoch": 0.23441216751964247,
+      "grad_norm": 0.11299805701752783,
+      "learning_rate": 0.00017888309646287113,
+      "loss": 1.2805,
+      "step": 3252
+    },
+    {
+      "epoch": 0.2344842499819794,
+      "grad_norm": 0.10932812891543568,
+      "learning_rate": 0.00017886874491565544,
+      "loss": 1.3099,
+      "step": 3253
+    },
+    {
+      "epoch": 0.23455633244431628,
+      "grad_norm": 0.15781697798449734,
+      "learning_rate": 0.0001788543890693876,
+      "loss": 1.3309,
+      "step": 3254
+    },
+    {
+      "epoch": 0.2346284149066532,
+      "grad_norm": 0.1045719440781209,
+      "learning_rate": 0.0001788400289248501,
+      "loss": 1.325,
+      "step": 3255
+    },
+    {
+      "epoch": 0.23470049736899012,
+      "grad_norm": 0.09975241174108138,
+      "learning_rate": 0.0001788256644828257,
+      "loss": 1.3242,
+      "step": 3256
+    },
+    {
+      "epoch": 0.23477257983132704,
+      "grad_norm": 0.29041112510763106,
+      "learning_rate": 0.0001788112957440974,
+      "loss": 1.3099,
+      "step": 3257
+    },
+    {
+      "epoch": 0.23484466229366396,
+      "grad_norm": 0.11765412912452922,
+      "learning_rate": 0.0001787969227094484,
+      "loss": 1.3058,
+      "step": 3258
+    },
+    {
+      "epoch": 0.23491674475600086,
+      "grad_norm": 0.2923207928064044,
+      "learning_rate": 0.00017878254537966216,
+      "loss": 1.3355,
+      "step": 3259
+    },
+    {
+      "epoch": 0.23498882721833778,
+      "grad_norm": 0.11323964804564496,
+      "learning_rate": 0.00017876816375552243,
+      "loss": 1.334,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2350609096806747,
+      "grad_norm": 0.10557564783436986,
+      "learning_rate": 0.0001787537778378131,
+      "loss": 1.4162,
+      "step": 3261
+    },
+    {
+      "epoch": 0.23513299214301162,
+      "grad_norm": 0.10536779433266069,
+      "learning_rate": 0.0001787393876273183,
+      "loss": 1.3437,
+      "step": 3262
+    },
+    {
+      "epoch": 0.2352050746053485,
+      "grad_norm": 0.11096307470412381,
+      "learning_rate": 0.00017872499312482246,
+      "loss": 1.3034,
+      "step": 3263
+    },
+    {
+      "epoch": 0.23527715706768543,
+      "grad_norm": 0.09652034724639735,
+      "learning_rate": 0.00017871059433111018,
+      "loss": 1.4153,
+      "step": 3264
+    },
+    {
+      "epoch": 0.23534923953002235,
+      "grad_norm": 0.11750626331695013,
+      "learning_rate": 0.00017869619124696634,
+      "loss": 1.3074,
+      "step": 3265
+    },
+    {
+      "epoch": 0.23542132199235927,
+      "grad_norm": 0.12703505459144968,
+      "learning_rate": 0.00017868178387317606,
+      "loss": 1.286,
+      "step": 3266
+    },
+    {
+      "epoch": 0.23549340445469616,
+      "grad_norm": 0.10823286593111488,
+      "learning_rate": 0.00017866737221052467,
+      "loss": 1.4236,
+      "step": 3267
+    },
+    {
+      "epoch": 0.23556548691703308,
+      "grad_norm": 0.10909270301820083,
+      "learning_rate": 0.0001786529562597977,
+      "loss": 1.4163,
+      "step": 3268
+    },
+    {
+      "epoch": 0.23563756937937,
+      "grad_norm": 0.150392972864545,
+      "learning_rate": 0.00017863853602178095,
+      "loss": 1.2173,
+      "step": 3269
+    },
+    {
+      "epoch": 0.23570965184170692,
+      "grad_norm": 0.10862120386422479,
+      "learning_rate": 0.00017862411149726048,
+      "loss": 1.2882,
+      "step": 3270
+    },
+    {
+      "epoch": 0.23578173430404384,
+      "grad_norm": 0.11585407119564609,
+      "learning_rate": 0.00017860968268702254,
+      "loss": 1.4041,
+      "step": 3271
+    },
+    {
+      "epoch": 0.23585381676638073,
+      "grad_norm": 0.09907959419380179,
+      "learning_rate": 0.00017859524959185362,
+      "loss": 1.3217,
+      "step": 3272
+    },
+    {
+      "epoch": 0.23592589922871765,
+      "grad_norm": 0.10695590322582976,
+      "learning_rate": 0.00017858081221254048,
+      "loss": 1.3544,
+      "step": 3273
+    },
+    {
+      "epoch": 0.23599798169105457,
+      "grad_norm": 0.10280855801868599,
+      "learning_rate": 0.00017856637054987007,
+      "loss": 1.3359,
+      "step": 3274
+    },
+    {
+      "epoch": 0.2360700641533915,
+      "grad_norm": 0.10632660072712699,
+      "learning_rate": 0.00017855192460462956,
+      "loss": 1.3244,
+      "step": 3275
+    },
+    {
+      "epoch": 0.23614214661572838,
+      "grad_norm": 0.10071264164999133,
+      "learning_rate": 0.00017853747437760647,
+      "loss": 1.3107,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2362142290780653,
+      "grad_norm": 0.1189865956550665,
+      "learning_rate": 0.00017852301986958838,
+      "loss": 1.2988,
+      "step": 3277
+    },
+    {
+      "epoch": 0.23628631154040222,
+      "grad_norm": 0.11413144218347235,
+      "learning_rate": 0.00017850856108136322,
+      "loss": 1.3358,
+      "step": 3278
+    },
+    {
+      "epoch": 0.23635839400273914,
+      "grad_norm": 0.1052074802702204,
+      "learning_rate": 0.00017849409801371912,
+      "loss": 1.326,
+      "step": 3279
+    },
+    {
+      "epoch": 0.23643047646507603,
+      "grad_norm": 0.1876061884423009,
+      "learning_rate": 0.00017847963066744445,
+      "loss": 1.3029,
+      "step": 3280
+    },
+    {
+      "epoch": 0.23650255892741295,
+      "grad_norm": 0.12388276396432296,
+      "learning_rate": 0.00017846515904332782,
+      "loss": 1.245,
+      "step": 3281
+    },
+    {
+      "epoch": 0.23657464138974987,
+      "grad_norm": 0.13018649899968218,
+      "learning_rate": 0.00017845068314215805,
+      "loss": 1.273,
+      "step": 3282
+    },
+    {
+      "epoch": 0.2366467238520868,
+      "grad_norm": 0.13473135076505802,
+      "learning_rate": 0.00017843620296472418,
+      "loss": 1.3888,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2367188063144237,
+      "grad_norm": 0.10262349313614662,
+      "learning_rate": 0.00017842171851181556,
+      "loss": 1.4152,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2367908887767606,
+      "grad_norm": 0.12743344537587478,
+      "learning_rate": 0.00017840722978422174,
+      "loss": 1.2997,
+      "step": 3285
+    },
+    {
+      "epoch": 0.23686297123909753,
+      "grad_norm": 0.10749794529405655,
+      "learning_rate": 0.00017839273678273238,
+      "loss": 1.4254,
+      "step": 3286
+    },
+    {
+      "epoch": 0.23693505370143444,
+      "grad_norm": 0.3087262886947749,
+      "learning_rate": 0.00017837823950813756,
+      "loss": 1.3306,
+      "step": 3287
+    },
+    {
+      "epoch": 0.23700713616377136,
+      "grad_norm": 0.10432995029131366,
+      "learning_rate": 0.0001783637379612275,
+      "loss": 1.3763,
+      "step": 3288
+    },
+    {
+      "epoch": 0.23707921862610826,
+      "grad_norm": 0.11346628663182663,
+      "learning_rate": 0.00017834923214279268,
+      "loss": 1.3326,
+      "step": 3289
+    },
+    {
+      "epoch": 0.23715130108844518,
+      "grad_norm": 0.1175827336701436,
+      "learning_rate": 0.00017833472205362378,
+      "loss": 1.3509,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2372233835507821,
+      "grad_norm": 0.10433049942124396,
+      "learning_rate": 0.0001783202076945117,
+      "loss": 1.4045,
+      "step": 3291
+    },
+    {
+      "epoch": 0.23729546601311902,
+      "grad_norm": 0.12172687012119499,
+      "learning_rate": 0.00017830568906624763,
+      "loss": 1.3277,
+      "step": 3292
+    },
+    {
+      "epoch": 0.2373675484754559,
+      "grad_norm": 0.14907692036701986,
+      "learning_rate": 0.00017829116616962297,
+      "loss": 1.3219,
+      "step": 3293
+    },
+    {
+      "epoch": 0.23743963093779283,
+      "grad_norm": 0.114953974899182,
+      "learning_rate": 0.00017827663900542935,
+      "loss": 1.3865,
+      "step": 3294
+    },
+    {
+      "epoch": 0.23751171340012975,
+      "grad_norm": 0.1067974413245723,
+      "learning_rate": 0.0001782621075744586,
+      "loss": 1.3233,
+      "step": 3295
+    },
+    {
+      "epoch": 0.23758379586246667,
+      "grad_norm": 0.12216103924030647,
+      "learning_rate": 0.00017824757187750285,
+      "loss": 1.3324,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2376558783248036,
+      "grad_norm": 0.09780457878538354,
+      "learning_rate": 0.00017823303191535442,
+      "loss": 1.413,
+      "step": 3297
+    },
+    {
+      "epoch": 0.23772796078714048,
+      "grad_norm": 0.1162254953223505,
+      "learning_rate": 0.00017821848768880582,
+      "loss": 1.3982,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2378000432494774,
+      "grad_norm": 0.15495943305672336,
+      "learning_rate": 0.0001782039391986499,
+      "loss": 1.3104,
+      "step": 3299
+    },
+    {
+      "epoch": 0.23787212571181432,
+      "grad_norm": 0.11922204688383133,
+      "learning_rate": 0.00017818938644567967,
+      "loss": 1.3391,
+      "step": 3300
+    },
+    {
+      "epoch": 0.23794420817415124,
+      "grad_norm": 0.1120181146646133,
+      "learning_rate": 0.00017817482943068838,
+      "loss": 1.3097,
+      "step": 3301
+    },
+    {
+      "epoch": 0.23801629063648813,
+      "grad_norm": 0.0993546355992158,
+      "learning_rate": 0.00017816026815446948,
+      "loss": 1.3342,
+      "step": 3302
+    },
+    {
+      "epoch": 0.23808837309882505,
+      "grad_norm": 0.11099497806494762,
+      "learning_rate": 0.0001781457026178168,
+      "loss": 1.2778,
+      "step": 3303
+    },
+    {
+      "epoch": 0.23816045556116197,
+      "grad_norm": 0.11302290113306031,
+      "learning_rate": 0.00017813113282152414,
+      "loss": 1.4655,
+      "step": 3304
+    },
+    {
+      "epoch": 0.2382325380234989,
+      "grad_norm": 0.09796452906376306,
+      "learning_rate": 0.00017811655876638578,
+      "loss": 1.2528,
+      "step": 3305
+    },
+    {
+      "epoch": 0.23830462048583578,
+      "grad_norm": 0.11332503656685983,
+      "learning_rate": 0.00017810198045319612,
+      "loss": 1.3868,
+      "step": 3306
+    },
+    {
+      "epoch": 0.2383767029481727,
+      "grad_norm": 0.1274295800413343,
+      "learning_rate": 0.00017808739788274978,
+      "loss": 1.265,
+      "step": 3307
+    },
+    {
+      "epoch": 0.23844878541050962,
+      "grad_norm": 0.10653399666708847,
+      "learning_rate": 0.0001780728110558417,
+      "loss": 1.2982,
+      "step": 3308
+    },
+    {
+      "epoch": 0.23852086787284654,
+      "grad_norm": 0.10621779479839197,
+      "learning_rate": 0.00017805821997326696,
+      "loss": 1.3274,
+      "step": 3309
+    },
+    {
+      "epoch": 0.23859295033518346,
+      "grad_norm": 0.10421383078647753,
+      "learning_rate": 0.0001780436246358209,
+      "loss": 1.2975,
+      "step": 3310
+    },
+    {
+      "epoch": 0.23866503279752035,
+      "grad_norm": 0.12323011889593316,
+      "learning_rate": 0.00017802902504429909,
+      "loss": 1.2872,
+      "step": 3311
+    },
+    {
+      "epoch": 0.23873711525985727,
+      "grad_norm": 0.22558732115875113,
+      "learning_rate": 0.00017801442119949737,
+      "loss": 1.284,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2388091977221942,
+      "grad_norm": 0.2389468129363806,
+      "learning_rate": 0.00017799981310221173,
+      "loss": 1.2898,
+      "step": 3313
+    },
+    {
+      "epoch": 0.23888128018453111,
+      "grad_norm": 0.1169766350739788,
+      "learning_rate": 0.00017798520075323847,
+      "loss": 1.3729,
+      "step": 3314
+    },
+    {
+      "epoch": 0.238953362646868,
+      "grad_norm": 0.18457380288414976,
+      "learning_rate": 0.0001779705841533741,
+      "loss": 1.3391,
+      "step": 3315
+    },
+    {
+      "epoch": 0.23902544510920493,
+      "grad_norm": 0.10799203328567238,
+      "learning_rate": 0.00017795596330341532,
+      "loss": 1.3548,
+      "step": 3316
+    },
+    {
+      "epoch": 0.23909752757154185,
+      "grad_norm": 0.10827889432616745,
+      "learning_rate": 0.00017794133820415916,
+      "loss": 1.3019,
+      "step": 3317
+    },
+    {
+      "epoch": 0.23916961003387877,
+      "grad_norm": 0.16289745110074663,
+      "learning_rate": 0.0001779267088564028,
+      "loss": 1.3776,
+      "step": 3318
+    },
+    {
+      "epoch": 0.23924169249621566,
+      "grad_norm": 0.10472594021052054,
+      "learning_rate": 0.00017791207526094362,
+      "loss": 1.2546,
+      "step": 3319
+    },
+    {
+      "epoch": 0.23931377495855258,
+      "grad_norm": 0.1101112215649774,
+      "learning_rate": 0.00017789743741857933,
+      "loss": 1.3528,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2393858574208895,
+      "grad_norm": 0.12498469944986913,
+      "learning_rate": 0.00017788279533010785,
+      "loss": 1.3856,
+      "step": 3321
+    },
+    {
+      "epoch": 0.23945793988322642,
+      "grad_norm": 0.12521067955166984,
+      "learning_rate": 0.0001778681489963272,
+      "loss": 1.3279,
+      "step": 3322
+    },
+    {
+      "epoch": 0.23953002234556334,
+      "grad_norm": 0.12373209187145175,
+      "learning_rate": 0.00017785349841803583,
+      "loss": 1.3248,
+      "step": 3323
+    },
+    {
+      "epoch": 0.23960210480790023,
+      "grad_norm": 0.10288049725759225,
+      "learning_rate": 0.0001778388435960323,
+      "loss": 1.4402,
+      "step": 3324
+    },
+    {
+      "epoch": 0.23967418727023715,
+      "grad_norm": 0.10703673102618759,
+      "learning_rate": 0.0001778241845311154,
+      "loss": 1.3665,
+      "step": 3325
+    },
+    {
+      "epoch": 0.23974626973257407,
+      "grad_norm": 0.11453932600420441,
+      "learning_rate": 0.00017780952122408423,
+      "loss": 1.327,
+      "step": 3326
+    },
+    {
+      "epoch": 0.239818352194911,
+      "grad_norm": 0.24368272089513202,
+      "learning_rate": 0.00017779485367573806,
+      "loss": 1.3434,
+      "step": 3327
+    },
+    {
+      "epoch": 0.23989043465724788,
+      "grad_norm": 0.10042508970904104,
+      "learning_rate": 0.0001777801818868764,
+      "loss": 1.3007,
+      "step": 3328
+    },
+    {
+      "epoch": 0.2399625171195848,
+      "grad_norm": 0.11345496417113955,
+      "learning_rate": 0.00017776550585829896,
+      "loss": 1.3724,
+      "step": 3329
+    },
+    {
+      "epoch": 0.24003459958192172,
+      "grad_norm": 0.10799926350954937,
+      "learning_rate": 0.00017775082559080575,
+      "loss": 1.3742,
+      "step": 3330
+    },
+    {
+      "epoch": 0.24010668204425864,
+      "grad_norm": 0.11900416426428294,
+      "learning_rate": 0.00017773614108519697,
+      "loss": 1.2181,
+      "step": 3331
+    },
+    {
+      "epoch": 0.24017876450659553,
+      "grad_norm": 0.13760038995072393,
+      "learning_rate": 0.00017772145234227306,
+      "loss": 1.2816,
+      "step": 3332
+    },
+    {
+      "epoch": 0.24025084696893245,
+      "grad_norm": 0.11668535983853552,
+      "learning_rate": 0.00017770675936283465,
+      "loss": 1.3113,
+      "step": 3333
+    },
+    {
+      "epoch": 0.24032292943126937,
+      "grad_norm": 0.10579099972185672,
+      "learning_rate": 0.0001776920621476827,
+      "loss": 1.298,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2403950118936063,
+      "grad_norm": 0.10528612313522212,
+      "learning_rate": 0.0001776773606976183,
+      "loss": 1.3625,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2404670943559432,
+      "grad_norm": 0.11139091097325812,
+      "learning_rate": 0.00017766265501344283,
+      "loss": 1.4189,
+      "step": 3336
+    },
+    {
+      "epoch": 0.2405391768182801,
+      "grad_norm": 0.12211648166380609,
+      "learning_rate": 0.00017764794509595786,
+      "loss": 1.2527,
+      "step": 3337
+    },
+    {
+      "epoch": 0.24061125928061702,
+      "grad_norm": 0.12308037346002643,
+      "learning_rate": 0.00017763323094596522,
+      "loss": 1.2165,
+      "step": 3338
+    },
+    {
+      "epoch": 0.24068334174295394,
+      "grad_norm": 0.15488744761260945,
+      "learning_rate": 0.00017761851256426698,
+      "loss": 1.3545,
+      "step": 3339
+    },
+    {
+      "epoch": 0.24075542420529086,
+      "grad_norm": 0.11781857873112107,
+      "learning_rate": 0.00017760378995166541,
+      "loss": 1.2273,
+      "step": 3340
+    },
+    {
+      "epoch": 0.24082750666762776,
+      "grad_norm": 0.14420695102815986,
+      "learning_rate": 0.000177589063108963,
+      "loss": 1.282,
+      "step": 3341
+    },
+    {
+      "epoch": 0.24089958912996468,
+      "grad_norm": 0.12700990113530136,
+      "learning_rate": 0.00017757433203696255,
+      "loss": 1.304,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2409716715923016,
+      "grad_norm": 0.10568559683958015,
+      "learning_rate": 0.00017755959673646697,
+      "loss": 1.2862,
+      "step": 3343
+    },
+    {
+      "epoch": 0.24104375405463851,
+      "grad_norm": 0.11945787118025714,
+      "learning_rate": 0.00017754485720827953,
+      "loss": 1.4229,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2411158365169754,
+      "grad_norm": 0.10995757107200531,
+      "learning_rate": 0.00017753011345320366,
+      "loss": 1.3118,
+      "step": 3345
+    },
+    {
+      "epoch": 0.24118791897931233,
+      "grad_norm": 0.11271095539131536,
+      "learning_rate": 0.00017751536547204295,
+      "loss": 1.3515,
+      "step": 3346
+    },
+    {
+      "epoch": 0.24126000144164925,
+      "grad_norm": 0.11106416469470888,
+      "learning_rate": 0.00017750061326560136,
+      "loss": 1.3901,
+      "step": 3347
+    },
+    {
+      "epoch": 0.24133208390398617,
+      "grad_norm": 0.12609942117415643,
+      "learning_rate": 0.00017748585683468302,
+      "loss": 1.3923,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2414041663663231,
+      "grad_norm": 0.10475848144877156,
+      "learning_rate": 0.00017747109618009227,
+      "loss": 1.3045,
+      "step": 3349
+    },
+    {
+      "epoch": 0.24147624882865998,
+      "grad_norm": 0.11139071637050661,
+      "learning_rate": 0.0001774563313026337,
+      "loss": 1.2553,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2415483312909969,
+      "grad_norm": 0.10773873848672953,
+      "learning_rate": 0.00017744156220311216,
+      "loss": 1.418,
+      "step": 3351
+    },
+    {
+      "epoch": 0.24162041375333382,
+      "grad_norm": 0.11451742910915207,
+      "learning_rate": 0.00017742678888233262,
+      "loss": 1.258,
+      "step": 3352
+    },
+    {
+      "epoch": 0.24169249621567074,
+      "grad_norm": 0.10534433917822598,
+      "learning_rate": 0.00017741201134110042,
+      "loss": 1.313,
+      "step": 3353
+    },
+    {
+      "epoch": 0.24176457867800763,
+      "grad_norm": 0.10648679287647644,
+      "learning_rate": 0.00017739722958022108,
+      "loss": 1.334,
+      "step": 3354
+    },
+    {
+      "epoch": 0.24183666114034455,
+      "grad_norm": 0.10641440078698688,
+      "learning_rate": 0.0001773824436005003,
+      "loss": 1.316,
+      "step": 3355
+    },
+    {
+      "epoch": 0.24190874360268147,
+      "grad_norm": 0.11107742548785625,
+      "learning_rate": 0.00017736765340274407,
+      "loss": 1.3642,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2419808260650184,
+      "grad_norm": 0.10424524957673484,
+      "learning_rate": 0.00017735285898775853,
+      "loss": 1.2502,
+      "step": 3357
+    },
+    {
+      "epoch": 0.24205290852735528,
+      "grad_norm": 0.1082592503855403,
+      "learning_rate": 0.00017733806035635018,
+      "loss": 1.2741,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2421249909896922,
+      "grad_norm": 0.10681763541360698,
+      "learning_rate": 0.00017732325750932567,
+      "loss": 1.3673,
+      "step": 3359
+    },
+    {
+      "epoch": 0.24219707345202912,
+      "grad_norm": 0.11505507517474468,
+      "learning_rate": 0.00017730845044749188,
+      "loss": 1.275,
+      "step": 3360
+    },
+    {
+      "epoch": 0.24226915591436604,
+      "grad_norm": 0.1243509586377002,
+      "learning_rate": 0.0001772936391716559,
+      "loss": 1.2975,
+      "step": 3361
+    },
+    {
+      "epoch": 0.24234123837670296,
+      "grad_norm": 0.10247649991290667,
+      "learning_rate": 0.0001772788236826251,
+      "loss": 1.2664,
+      "step": 3362
+    },
+    {
+      "epoch": 0.24241332083903985,
+      "grad_norm": 0.09057806151010081,
+      "learning_rate": 0.00017726400398120705,
+      "loss": 1.3403,
+      "step": 3363
+    },
+    {
+      "epoch": 0.24248540330137677,
+      "grad_norm": 0.11526368334417975,
+      "learning_rate": 0.0001772491800682096,
+      "loss": 1.3378,
+      "step": 3364
+    },
+    {
+      "epoch": 0.2425574857637137,
+      "grad_norm": 0.13557623012948006,
+      "learning_rate": 0.0001772343519444407,
+      "loss": 1.3122,
+      "step": 3365
+    },
+    {
+      "epoch": 0.2426295682260506,
+      "grad_norm": 0.10604680483539082,
+      "learning_rate": 0.00017721951961070865,
+      "loss": 1.3253,
+      "step": 3366
+    },
+    {
+      "epoch": 0.2427016506883875,
+      "grad_norm": 0.11649189035426522,
+      "learning_rate": 0.00017720468306782196,
+      "loss": 1.3575,
+      "step": 3367
+    },
+    {
+      "epoch": 0.24277373315072442,
+      "grad_norm": 0.11433148467717599,
+      "learning_rate": 0.00017718984231658938,
+      "loss": 1.2429,
+      "step": 3368
+    },
+    {
+      "epoch": 0.24284581561306134,
+      "grad_norm": 0.10227312957221767,
+      "learning_rate": 0.00017717499735781983,
+      "loss": 1.3077,
+      "step": 3369
+    },
+    {
+      "epoch": 0.24291789807539826,
+      "grad_norm": 0.10740114459181382,
+      "learning_rate": 0.0001771601481923225,
+      "loss": 1.3372,
+      "step": 3370
+    },
+    {
+      "epoch": 0.24298998053773516,
+      "grad_norm": 0.10623197519983966,
+      "learning_rate": 0.00017714529482090678,
+      "loss": 1.2885,
+      "step": 3371
+    },
+    {
+      "epoch": 0.24306206300007208,
+      "grad_norm": 0.10306602983637451,
+      "learning_rate": 0.0001771304372443823,
+      "loss": 1.2575,
+      "step": 3372
+    },
+    {
+      "epoch": 0.243134145462409,
+      "grad_norm": 0.10563528408376698,
+      "learning_rate": 0.00017711557546355904,
+      "loss": 1.263,
+      "step": 3373
+    },
+    {
+      "epoch": 0.24320622792474592,
+      "grad_norm": 0.11054018736699738,
+      "learning_rate": 0.00017710070947924697,
+      "loss": 1.3133,
+      "step": 3374
+    },
+    {
+      "epoch": 0.24327831038708284,
+      "grad_norm": 0.11435686083455698,
+      "learning_rate": 0.0001770858392922565,
+      "loss": 1.3217,
+      "step": 3375
+    },
+    {
+      "epoch": 0.24335039284941973,
+      "grad_norm": 0.19339911940158738,
+      "learning_rate": 0.0001770709649033981,
+      "loss": 1.362,
+      "step": 3376
+    },
+    {
+      "epoch": 0.24342247531175665,
+      "grad_norm": 0.11166877052052579,
+      "learning_rate": 0.00017705608631348268,
+      "loss": 1.3752,
+      "step": 3377
+    },
+    {
+      "epoch": 0.24349455777409357,
+      "grad_norm": 0.12270447692483114,
+      "learning_rate": 0.00017704120352332118,
+      "loss": 1.3036,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2435666402364305,
+      "grad_norm": 0.09934565017753684,
+      "learning_rate": 0.00017702631653372485,
+      "loss": 1.2963,
+      "step": 3379
+    },
+    {
+      "epoch": 0.24363872269876738,
+      "grad_norm": 0.12032803773381047,
+      "learning_rate": 0.00017701142534550513,
+      "loss": 1.3624,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2437108051611043,
+      "grad_norm": 0.13969385347173524,
+      "learning_rate": 0.00017699652995947383,
+      "loss": 1.2988,
+      "step": 3381
+    },
+    {
+      "epoch": 0.24378288762344122,
+      "grad_norm": 0.11097760902326173,
+      "learning_rate": 0.00017698163037644277,
+      "loss": 1.2769,
+      "step": 3382
+    },
+    {
+      "epoch": 0.24385497008577814,
+      "grad_norm": 0.10966204293404,
+      "learning_rate": 0.00017696672659722417,
+      "loss": 1.2911,
+      "step": 3383
+    },
+    {
+      "epoch": 0.24392705254811503,
+      "grad_norm": 0.17884997837841732,
+      "learning_rate": 0.00017695181862263044,
+      "loss": 1.3268,
+      "step": 3384
+    },
+    {
+      "epoch": 0.24399913501045195,
+      "grad_norm": 0.11624853751949824,
+      "learning_rate": 0.00017693690645347414,
+      "loss": 1.2936,
+      "step": 3385
+    },
+    {
+      "epoch": 0.24407121747278887,
+      "grad_norm": 0.09676813976111527,
+      "learning_rate": 0.00017692199009056815,
+      "loss": 1.3379,
+      "step": 3386
+    },
+    {
+      "epoch": 0.2441432999351258,
+      "grad_norm": 0.10846957187679639,
+      "learning_rate": 0.00017690706953472552,
+      "loss": 1.2531,
+      "step": 3387
+    },
+    {
+      "epoch": 0.2442153823974627,
+      "grad_norm": 0.13669986407488469,
+      "learning_rate": 0.00017689214478675954,
+      "loss": 1.3577,
+      "step": 3388
+    },
+    {
+      "epoch": 0.2442874648597996,
+      "grad_norm": 0.1163370963191993,
+      "learning_rate": 0.00017687721584748383,
+      "loss": 1.3968,
+      "step": 3389
+    },
+    {
+      "epoch": 0.24435954732213652,
+      "grad_norm": 0.11782357698974585,
+      "learning_rate": 0.00017686228271771205,
+      "loss": 1.2915,
+      "step": 3390
+    },
+    {
+      "epoch": 0.24443162978447344,
+      "grad_norm": 0.10566069121390158,
+      "learning_rate": 0.00017684734539825825,
+      "loss": 1.3749,
+      "step": 3391
+    },
+    {
+      "epoch": 0.24450371224681036,
+      "grad_norm": 0.0968141680455327,
+      "learning_rate": 0.00017683240388993664,
+      "loss": 1.2805,
+      "step": 3392
+    },
+    {
+      "epoch": 0.24457579470914725,
+      "grad_norm": 0.11464851140433961,
+      "learning_rate": 0.00017681745819356164,
+      "loss": 1.3153,
+      "step": 3393
+    },
+    {
+      "epoch": 0.24464787717148417,
+      "grad_norm": 0.1052076868032125,
+      "learning_rate": 0.00017680250830994794,
+      "loss": 1.2998,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2447199596338211,
+      "grad_norm": 0.13956954197567648,
+      "learning_rate": 0.0001767875542399105,
+      "loss": 1.3723,
+      "step": 3395
+    },
+    {
+      "epoch": 0.244792042096158,
+      "grad_norm": 0.10032258783503598,
+      "learning_rate": 0.00017677259598426433,
+      "loss": 1.2836,
+      "step": 3396
+    },
+    {
+      "epoch": 0.2448641245584949,
+      "grad_norm": 0.13577028087575177,
+      "learning_rate": 0.00017675763354382484,
+      "loss": 1.2571,
+      "step": 3397
+    },
+    {
+      "epoch": 0.24493620702083183,
+      "grad_norm": 0.10465356081548526,
+      "learning_rate": 0.00017674266691940766,
+      "loss": 1.3055,
+      "step": 3398
+    },
+    {
+      "epoch": 0.24500828948316875,
+      "grad_norm": 0.11023048824187805,
+      "learning_rate": 0.00017672769611182858,
+      "loss": 1.3908,
+      "step": 3399
+    },
+    {
+      "epoch": 0.24508037194550567,
+      "grad_norm": 0.13367760820642746,
+      "learning_rate": 0.0001767127211219036,
+      "loss": 1.291,
+      "step": 3400
+    },
+    {
+      "epoch": 0.24515245440784258,
+      "grad_norm": 0.10070580992971336,
+      "learning_rate": 0.00017669774195044907,
+      "loss": 1.2753,
+      "step": 3401
+    },
+    {
+      "epoch": 0.24522453687017948,
+      "grad_norm": 0.11740854111378257,
+      "learning_rate": 0.00017668275859828141,
+      "loss": 1.3434,
+      "step": 3402
+    },
+    {
+      "epoch": 0.2452966193325164,
+      "grad_norm": 0.11396118131901511,
+      "learning_rate": 0.00017666777106621742,
+      "loss": 1.3262,
+      "step": 3403
+    },
+    {
+      "epoch": 0.24536870179485332,
+      "grad_norm": 0.11405168895053906,
+      "learning_rate": 0.00017665277935507398,
+      "loss": 1.286,
+      "step": 3404
+    },
+    {
+      "epoch": 0.24544078425719024,
+      "grad_norm": 0.10728710463408167,
+      "learning_rate": 0.00017663778346566831,
+      "loss": 1.2715,
+      "step": 3405
+    },
+    {
+      "epoch": 0.24551286671952713,
+      "grad_norm": 0.12440491659353398,
+      "learning_rate": 0.00017662278339881785,
+      "loss": 1.3248,
+      "step": 3406
+    },
+    {
+      "epoch": 0.24558494918186405,
+      "grad_norm": 0.2365881214122396,
+      "learning_rate": 0.00017660777915534022,
+      "loss": 1.3764,
+      "step": 3407
+    },
+    {
+      "epoch": 0.24565703164420097,
+      "grad_norm": 0.11764196827130582,
+      "learning_rate": 0.00017659277073605323,
+      "loss": 1.3255,
+      "step": 3408
+    },
+    {
+      "epoch": 0.2457291141065379,
+      "grad_norm": 0.11641726359179487,
+      "learning_rate": 0.00017657775814177505,
+      "loss": 1.3385,
+      "step": 3409
+    },
+    {
+      "epoch": 0.24580119656887478,
+      "grad_norm": 0.105363823630147,
+      "learning_rate": 0.0001765627413733239,
+      "loss": 1.2949,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2458732790312117,
+      "grad_norm": 0.11151928907141775,
+      "learning_rate": 0.00017654772043151847,
+      "loss": 1.3408,
+      "step": 3411
+    },
+    {
+      "epoch": 0.24594536149354862,
+      "grad_norm": 0.11001301951158192,
+      "learning_rate": 0.00017653269531717745,
+      "loss": 1.3936,
+      "step": 3412
+    },
+    {
+      "epoch": 0.24601744395588554,
+      "grad_norm": 0.10036486134373396,
+      "learning_rate": 0.00017651766603111983,
+      "loss": 1.2521,
+      "step": 3413
+    },
+    {
+      "epoch": 0.24608952641822246,
+      "grad_norm": 0.1165261758265969,
+      "learning_rate": 0.00017650263257416485,
+      "loss": 1.3474,
+      "step": 3414
+    },
+    {
+      "epoch": 0.24616160888055935,
+      "grad_norm": 0.10437218587639366,
+      "learning_rate": 0.00017648759494713203,
+      "loss": 1.3235,
+      "step": 3415
+    },
+    {
+      "epoch": 0.24623369134289627,
+      "grad_norm": 0.16760499742825466,
+      "learning_rate": 0.000176472553150841,
+      "loss": 1.308,
+      "step": 3416
+    },
+    {
+      "epoch": 0.2463057738052332,
+      "grad_norm": 0.10476665705778601,
+      "learning_rate": 0.00017645750718611164,
+      "loss": 1.407,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2463778562675701,
+      "grad_norm": 0.18729542987390357,
+      "learning_rate": 0.00017644245705376416,
+      "loss": 1.4011,
+      "step": 3418
+    },
+    {
+      "epoch": 0.246449938729907,
+      "grad_norm": 0.10617746630941353,
+      "learning_rate": 0.0001764274027546189,
+      "loss": 1.3401,
+      "step": 3419
+    },
+    {
+      "epoch": 0.24652202119224392,
+      "grad_norm": 0.09711645693408194,
+      "learning_rate": 0.00017641234428949644,
+      "loss": 1.2621,
+      "step": 3420
+    },
+    {
+      "epoch": 0.24659410365458084,
+      "grad_norm": 0.10166361138040507,
+      "learning_rate": 0.00017639728165921764,
+      "loss": 1.3698,
+      "step": 3421
+    },
+    {
+      "epoch": 0.24666618611691776,
+      "grad_norm": 0.1057178204134567,
+      "learning_rate": 0.00017638221486460348,
+      "loss": 1.2473,
+      "step": 3422
+    },
+    {
+      "epoch": 0.24673826857925465,
+      "grad_norm": 0.1458207855740077,
+      "learning_rate": 0.0001763671439064753,
+      "loss": 1.2442,
+      "step": 3423
+    },
+    {
+      "epoch": 0.24681035104159157,
+      "grad_norm": 0.11375819546411095,
+      "learning_rate": 0.00017635206878565456,
+      "loss": 1.3519,
+      "step": 3424
+    },
+    {
+      "epoch": 0.2468824335039285,
+      "grad_norm": 0.10792029221187655,
+      "learning_rate": 0.000176336989502963,
+      "loss": 1.247,
+      "step": 3425
+    },
+    {
+      "epoch": 0.24695451596626541,
+      "grad_norm": 0.12153834789412989,
+      "learning_rate": 0.0001763219060592226,
+      "loss": 1.4093,
+      "step": 3426
+    },
+    {
+      "epoch": 0.24702659842860233,
+      "grad_norm": 0.10685920072265856,
+      "learning_rate": 0.00017630681845525552,
+      "loss": 1.4039,
+      "step": 3427
+    },
+    {
+      "epoch": 0.24709868089093923,
+      "grad_norm": 0.1278095335462827,
+      "learning_rate": 0.00017629172669188418,
+      "loss": 1.4101,
+      "step": 3428
+    },
+    {
+      "epoch": 0.24717076335327615,
+      "grad_norm": 0.10330048077652891,
+      "learning_rate": 0.0001762766307699312,
+      "loss": 1.265,
+      "step": 3429
+    },
+    {
+      "epoch": 0.24724284581561307,
+      "grad_norm": 0.09677369249904458,
+      "learning_rate": 0.00017626153069021946,
+      "loss": 1.3418,
+      "step": 3430
+    },
+    {
+      "epoch": 0.24731492827794999,
+      "grad_norm": 0.10842488716108178,
+      "learning_rate": 0.00017624642645357205,
+      "loss": 1.2773,
+      "step": 3431
+    },
+    {
+      "epoch": 0.24738701074028688,
+      "grad_norm": 0.14029779571345444,
+      "learning_rate": 0.00017623131806081225,
+      "loss": 1.2659,
+      "step": 3432
+    },
+    {
+      "epoch": 0.2474590932026238,
+      "grad_norm": 0.12035869232496293,
+      "learning_rate": 0.00017621620551276366,
+      "loss": 1.3627,
+      "step": 3433
+    },
+    {
+      "epoch": 0.24753117566496072,
+      "grad_norm": 0.1047712909421789,
+      "learning_rate": 0.00017620108881025002,
+      "loss": 1.3711,
+      "step": 3434
+    },
+    {
+      "epoch": 0.24760325812729764,
+      "grad_norm": 0.13758781019416227,
+      "learning_rate": 0.0001761859679540953,
+      "loss": 1.2703,
+      "step": 3435
+    },
+    {
+      "epoch": 0.24767534058963453,
+      "grad_norm": 0.11073509309870251,
+      "learning_rate": 0.00017617084294512377,
+      "loss": 1.2297,
+      "step": 3436
+    },
+    {
+      "epoch": 0.24774742305197145,
+      "grad_norm": 0.1846693677448195,
+      "learning_rate": 0.00017615571378415984,
+      "loss": 1.3008,
+      "step": 3437
+    },
+    {
+      "epoch": 0.24781950551430837,
+      "grad_norm": 0.10935747521488279,
+      "learning_rate": 0.0001761405804720282,
+      "loss": 1.3134,
+      "step": 3438
+    },
+    {
+      "epoch": 0.2478915879766453,
+      "grad_norm": 0.11294598406793936,
+      "learning_rate": 0.00017612544300955375,
+      "loss": 1.3057,
+      "step": 3439
+    },
+    {
+      "epoch": 0.2479636704389822,
+      "grad_norm": 0.11769323298803011,
+      "learning_rate": 0.00017611030139756163,
+      "loss": 1.3852,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2480357529013191,
+      "grad_norm": 0.1201838456481498,
+      "learning_rate": 0.0001760951556368772,
+      "loss": 1.3385,
+      "step": 3441
+    },
+    {
+      "epoch": 0.24810783536365602,
+      "grad_norm": 0.21875747751379634,
+      "learning_rate": 0.00017608000572832597,
+      "loss": 1.2827,
+      "step": 3442
+    },
+    {
+      "epoch": 0.24817991782599294,
+      "grad_norm": 0.0999244786183725,
+      "learning_rate": 0.00017606485167273383,
+      "loss": 1.2032,
+      "step": 3443
+    },
+    {
+      "epoch": 0.24825200028832986,
+      "grad_norm": 0.10216366939270935,
+      "learning_rate": 0.00017604969347092677,
+      "loss": 1.3037,
+      "step": 3444
+    },
+    {
+      "epoch": 0.24832408275066675,
+      "grad_norm": 0.10425563783099687,
+      "learning_rate": 0.00017603453112373108,
+      "loss": 1.3719,
+      "step": 3445
+    },
+    {
+      "epoch": 0.24839616521300367,
+      "grad_norm": 0.09837552429728803,
+      "learning_rate": 0.00017601936463197318,
+      "loss": 1.3699,
+      "step": 3446
+    },
+    {
+      "epoch": 0.2484682476753406,
+      "grad_norm": 0.10506247401416575,
+      "learning_rate": 0.00017600419399647983,
+      "loss": 1.2621,
+      "step": 3447
+    },
+    {
+      "epoch": 0.2485403301376775,
+      "grad_norm": 0.12645780310866933,
+      "learning_rate": 0.000175989019218078,
+      "loss": 1.3504,
+      "step": 3448
+    },
+    {
+      "epoch": 0.2486124126000144,
+      "grad_norm": 0.09306194163794493,
+      "learning_rate": 0.00017597384029759475,
+      "loss": 1.2802,
+      "step": 3449
+    },
+    {
+      "epoch": 0.24868449506235132,
+      "grad_norm": 0.10714820917215674,
+      "learning_rate": 0.00017595865723585755,
+      "loss": 1.2429,
+      "step": 3450
+    },
+    {
+      "epoch": 0.24875657752468824,
+      "grad_norm": 0.10786399645738537,
+      "learning_rate": 0.00017594347003369395,
+      "loss": 1.3117,
+      "step": 3451
+    },
+    {
+      "epoch": 0.24882865998702516,
+      "grad_norm": 0.11807537980363708,
+      "learning_rate": 0.00017592827869193184,
+      "loss": 1.2328,
+      "step": 3452
+    },
+    {
+      "epoch": 0.24890074244936208,
+      "grad_norm": 0.1318439258195862,
+      "learning_rate": 0.00017591308321139929,
+      "loss": 1.2981,
+      "step": 3453
+    },
+    {
+      "epoch": 0.24897282491169898,
+      "grad_norm": 0.10849347335875403,
+      "learning_rate": 0.00017589788359292455,
+      "loss": 1.3394,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2490449073740359,
+      "grad_norm": 0.10745832844697328,
+      "learning_rate": 0.00017588267983733617,
+      "loss": 1.3432,
+      "step": 3455
+    },
+    {
+      "epoch": 0.24911698983637282,
+      "grad_norm": 0.10468289681387671,
+      "learning_rate": 0.00017586747194546284,
+      "loss": 1.3418,
+      "step": 3456
+    },
+    {
+      "epoch": 0.24918907229870974,
+      "grad_norm": 0.10742291926178456,
+      "learning_rate": 0.0001758522599181336,
+      "loss": 1.2805,
+      "step": 3457
+    },
+    {
+      "epoch": 0.24926115476104663,
+      "grad_norm": 0.1056193734495673,
+      "learning_rate": 0.00017583704375617756,
+      "loss": 1.3541,
+      "step": 3458
+    },
+    {
+      "epoch": 0.24933323722338355,
+      "grad_norm": 0.10668705572003555,
+      "learning_rate": 0.00017582182346042422,
+      "loss": 1.3204,
+      "step": 3459
+    },
+    {
+      "epoch": 0.24940531968572047,
+      "grad_norm": 0.11159098739775682,
+      "learning_rate": 0.00017580659903170312,
+      "loss": 1.3223,
+      "step": 3460
+    },
+    {
+      "epoch": 0.2494774021480574,
+      "grad_norm": 0.10093551960985787,
+      "learning_rate": 0.00017579137047084425,
+      "loss": 1.3532,
+      "step": 3461
+    },
+    {
+      "epoch": 0.24954948461039428,
+      "grad_norm": 0.11097363725155719,
+      "learning_rate": 0.00017577613777867762,
+      "loss": 1.2396,
+      "step": 3462
+    },
+    {
+      "epoch": 0.2496215670727312,
+      "grad_norm": 0.11595680190905784,
+      "learning_rate": 0.00017576090095603356,
+      "loss": 1.2732,
+      "step": 3463
+    },
+    {
+      "epoch": 0.24969364953506812,
+      "grad_norm": 0.10665656771305776,
+      "learning_rate": 0.00017574566000374262,
+      "loss": 1.3507,
+      "step": 3464
+    },
+    {
+      "epoch": 0.24976573199740504,
+      "grad_norm": 0.11356763707544437,
+      "learning_rate": 0.00017573041492263558,
+      "loss": 1.2534,
+      "step": 3465
+    },
+    {
+      "epoch": 0.24983781445974196,
+      "grad_norm": 0.09881483733912748,
+      "learning_rate": 0.0001757151657135434,
+      "loss": 1.3686,
+      "step": 3466
+    },
+    {
+      "epoch": 0.24990989692207885,
+      "grad_norm": 0.11300013739358268,
+      "learning_rate": 0.00017569991237729733,
+      "loss": 1.2936,
+      "step": 3467
+    },
+    {
+      "epoch": 0.24998197938441577,
+      "grad_norm": 0.10160969618542709,
+      "learning_rate": 0.00017568465491472882,
+      "loss": 1.3408,
+      "step": 3468
+    },
+    {
+      "epoch": 0.25005406184675266,
+      "grad_norm": 0.1042418909333812,
+      "learning_rate": 0.0001756693933266695,
+      "loss": 1.3107,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2501261443090896,
+      "grad_norm": 0.10885645697372823,
+      "learning_rate": 0.00017565412761395133,
+      "loss": 1.2542,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2501982267714265,
+      "grad_norm": 0.09898704348596435,
+      "learning_rate": 0.00017563885777740636,
+      "loss": 1.3124,
+      "step": 3471
+    },
+    {
+      "epoch": 0.2502703092337634,
+      "grad_norm": 0.10683531774586809,
+      "learning_rate": 0.00017562358381786693,
+      "loss": 1.2762,
+      "step": 3472
+    },
+    {
+      "epoch": 0.25034239169610034,
+      "grad_norm": 0.10289362016065634,
+      "learning_rate": 0.00017560830573616566,
+      "loss": 1.2483,
+      "step": 3473
+    },
+    {
+      "epoch": 0.25041447415843726,
+      "grad_norm": 0.11378398371919361,
+      "learning_rate": 0.0001755930235331353,
+      "loss": 1.4095,
+      "step": 3474
+    },
+    {
+      "epoch": 0.2504865566207742,
+      "grad_norm": 0.11503312744702637,
+      "learning_rate": 0.00017557773720960893,
+      "loss": 1.4015,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2505586390831111,
+      "grad_norm": 0.10789060893671716,
+      "learning_rate": 0.0001755624467664197,
+      "loss": 1.3493,
+      "step": 3476
+    },
+    {
+      "epoch": 0.25063072154544797,
+      "grad_norm": 0.12262656086408673,
+      "learning_rate": 0.00017554715220440112,
+      "loss": 1.4246,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2507028040077849,
+      "grad_norm": 0.11068993582615902,
+      "learning_rate": 0.0001755318535243869,
+      "loss": 1.3378,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2507748864701218,
+      "grad_norm": 0.10127620020413801,
+      "learning_rate": 0.00017551655072721092,
+      "loss": 1.2638,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2508469689324587,
+      "grad_norm": 0.2886842027784239,
+      "learning_rate": 0.00017550124381370734,
+      "loss": 1.1876,
+      "step": 3480
+    },
+    {
+      "epoch": 0.25091905139479564,
+      "grad_norm": 0.11549328816559434,
+      "learning_rate": 0.00017548593278471052,
+      "loss": 1.2153,
+      "step": 3481
+    },
+    {
+      "epoch": 0.25099113385713256,
+      "grad_norm": 0.10460025534649547,
+      "learning_rate": 0.00017547061764105503,
+      "loss": 1.3621,
+      "step": 3482
+    },
+    {
+      "epoch": 0.2510632163194695,
+      "grad_norm": 0.1117609566013144,
+      "learning_rate": 0.00017545529838357572,
+      "loss": 1.3013,
+      "step": 3483
+    },
+    {
+      "epoch": 0.2511352987818064,
+      "grad_norm": 0.10150024620533385,
+      "learning_rate": 0.00017543997501310763,
+      "loss": 1.2961,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2512073812441433,
+      "grad_norm": 0.10145376149464574,
+      "learning_rate": 0.00017542464753048594,
+      "loss": 1.2723,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2512794637064802,
+      "grad_norm": 0.10449970384186906,
+      "learning_rate": 0.00017540931593654623,
+      "loss": 1.2593,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2513515461688171,
+      "grad_norm": 0.09987757407871216,
+      "learning_rate": 0.00017539398023212415,
+      "loss": 1.3061,
+      "step": 3487
+    },
+    {
+      "epoch": 0.25142362863115403,
+      "grad_norm": 0.12356463722357898,
+      "learning_rate": 0.00017537864041805568,
+      "loss": 1.2005,
+      "step": 3488
+    },
+    {
+      "epoch": 0.25149571109349095,
+      "grad_norm": 0.09894750860974175,
+      "learning_rate": 0.00017536329649517693,
+      "loss": 1.2435,
+      "step": 3489
+    },
+    {
+      "epoch": 0.25156779355582787,
+      "grad_norm": 0.10609431600016486,
+      "learning_rate": 0.0001753479484643243,
+      "loss": 1.2976,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2516398760181648,
+      "grad_norm": 0.11698253054784241,
+      "learning_rate": 0.00017533259632633442,
+      "loss": 1.3492,
+      "step": 3491
+    },
+    {
+      "epoch": 0.2517119584805017,
+      "grad_norm": 0.11581224620998466,
+      "learning_rate": 0.00017531724008204406,
+      "loss": 1.2574,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2517840409428386,
+      "grad_norm": 0.10805206174766778,
+      "learning_rate": 0.00017530187973229032,
+      "loss": 1.2815,
+      "step": 3493
+    },
+    {
+      "epoch": 0.25185612340517555,
+      "grad_norm": 0.11483272931781824,
+      "learning_rate": 0.00017528651527791048,
+      "loss": 1.2857,
+      "step": 3494
+    },
+    {
+      "epoch": 0.2519282058675124,
+      "grad_norm": 0.10332906874895108,
+      "learning_rate": 0.000175271146719742,
+      "loss": 1.3005,
+      "step": 3495
+    },
+    {
+      "epoch": 0.25200028832984933,
+      "grad_norm": 0.12179781289354259,
+      "learning_rate": 0.00017525577405862265,
+      "loss": 1.3211,
+      "step": 3496
+    },
+    {
+      "epoch": 0.25207237079218625,
+      "grad_norm": 0.09858370786808542,
+      "learning_rate": 0.00017524039729539036,
+      "loss": 1.2339,
+      "step": 3497
+    },
+    {
+      "epoch": 0.25214445325452317,
+      "grad_norm": 0.12199738755106368,
+      "learning_rate": 0.00017522501643088328,
+      "loss": 1.3138,
+      "step": 3498
+    },
+    {
+      "epoch": 0.2522165357168601,
+      "grad_norm": 0.1120344806085843,
+      "learning_rate": 0.00017520963146593985,
+      "loss": 1.3156,
+      "step": 3499
+    },
+    {
+      "epoch": 0.252288618179197,
+      "grad_norm": 0.11079275710650346,
+      "learning_rate": 0.00017519424240139864,
+      "loss": 1.2506,
+      "step": 3500
+    },
+    {
+      "epoch": 0.25236070064153393,
+      "grad_norm": 0.11361046017860578,
+      "learning_rate": 0.0001751788492380985,
+      "loss": 1.2387,
+      "step": 3501
+    },
+    {
+      "epoch": 0.25243278310387085,
+      "grad_norm": 0.11408061520622631,
+      "learning_rate": 0.00017516345197687852,
+      "loss": 1.2685,
+      "step": 3502
+    },
+    {
+      "epoch": 0.2525048655662077,
+      "grad_norm": 0.10355539889723914,
+      "learning_rate": 0.00017514805061857798,
+      "loss": 1.3075,
+      "step": 3503
+    },
+    {
+      "epoch": 0.25257694802854463,
+      "grad_norm": 0.12357538770912385,
+      "learning_rate": 0.00017513264516403641,
+      "loss": 1.29,
+      "step": 3504
+    },
+    {
+      "epoch": 0.25264903049088155,
+      "grad_norm": 0.11210834759638501,
+      "learning_rate": 0.00017511723561409347,
+      "loss": 1.3853,
+      "step": 3505
+    },
+    {
+      "epoch": 0.2527211129532185,
+      "grad_norm": 0.1863538127074378,
+      "learning_rate": 0.0001751018219695892,
+      "loss": 1.2549,
+      "step": 3506
+    },
+    {
+      "epoch": 0.2527931954155554,
+      "grad_norm": 0.10913383333167778,
+      "learning_rate": 0.00017508640423136377,
+      "loss": 1.4163,
+      "step": 3507
+    },
+    {
+      "epoch": 0.2528652778778923,
+      "grad_norm": 0.10915205555819525,
+      "learning_rate": 0.0001750709824002575,
+      "loss": 1.3516,
+      "step": 3508
+    },
+    {
+      "epoch": 0.25293736034022923,
+      "grad_norm": 0.10670195090582399,
+      "learning_rate": 0.00017505555647711116,
+      "loss": 1.3044,
+      "step": 3509
+    },
+    {
+      "epoch": 0.25300944280256615,
+      "grad_norm": 0.3923171806764554,
+      "learning_rate": 0.0001750401264627655,
+      "loss": 1.3645,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2530815252649031,
+      "grad_norm": 0.10090297933286074,
+      "learning_rate": 0.0001750246923580616,
+      "loss": 1.3064,
+      "step": 3511
+    },
+    {
+      "epoch": 0.25315360772723994,
+      "grad_norm": 0.2780451873081319,
+      "learning_rate": 0.00017500925416384083,
+      "loss": 1.4009,
+      "step": 3512
+    },
+    {
+      "epoch": 0.25322569018957686,
+      "grad_norm": 0.11256215935070703,
+      "learning_rate": 0.0001749938118809446,
+      "loss": 1.3469,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2532977726519138,
+      "grad_norm": 0.11537007212385333,
+      "learning_rate": 0.00017497836551021474,
+      "loss": 1.2727,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2533698551142507,
+      "grad_norm": 0.11746291728135447,
+      "learning_rate": 0.0001749629150524932,
+      "loss": 1.366,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2534419375765876,
+      "grad_norm": 0.1056594493881642,
+      "learning_rate": 0.0001749474605086221,
+      "loss": 1.244,
+      "step": 3516
+    },
+    {
+      "epoch": 0.25351402003892454,
+      "grad_norm": 0.108284008073619,
+      "learning_rate": 0.00017493200187944393,
+      "loss": 1.3087,
+      "step": 3517
+    },
+    {
+      "epoch": 0.25358610250126146,
+      "grad_norm": 0.09831842690898486,
+      "learning_rate": 0.00017491653916580128,
+      "loss": 1.2127,
+      "step": 3518
+    },
+    {
+      "epoch": 0.2536581849635984,
+      "grad_norm": 0.12774691761477264,
+      "learning_rate": 0.00017490107236853706,
+      "loss": 1.3276,
+      "step": 3519
+    },
+    {
+      "epoch": 0.2537302674259353,
+      "grad_norm": 0.09978925372638106,
+      "learning_rate": 0.00017488560148849427,
+      "loss": 1.3068,
+      "step": 3520
+    },
+    {
+      "epoch": 0.25380234988827216,
+      "grad_norm": 0.11416587662828118,
+      "learning_rate": 0.0001748701265265163,
+      "loss": 1.2835,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2538744323506091,
+      "grad_norm": 0.10177840029319236,
+      "learning_rate": 0.00017485464748344658,
+      "loss": 1.3255,
+      "step": 3522
+    },
+    {
+      "epoch": 0.253946514812946,
+      "grad_norm": 0.11250014932364513,
+      "learning_rate": 0.00017483916436012893,
+      "loss": 1.3604,
+      "step": 3523
+    },
+    {
+      "epoch": 0.2540185972752829,
+      "grad_norm": 0.11740413026068311,
+      "learning_rate": 0.00017482367715740727,
+      "loss": 1.4179,
+      "step": 3524
+    },
+    {
+      "epoch": 0.25409067973761984,
+      "grad_norm": 0.11625595762307965,
+      "learning_rate": 0.00017480818587612584,
+      "loss": 1.3523,
+      "step": 3525
+    },
+    {
+      "epoch": 0.25416276219995676,
+      "grad_norm": 0.10483506188289639,
+      "learning_rate": 0.00017479269051712902,
+      "loss": 1.3317,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2542348446622937,
+      "grad_norm": 0.10409450846300566,
+      "learning_rate": 0.00017477719108126145,
+      "loss": 1.1682,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2543069271246306,
+      "grad_norm": 0.11777757891756739,
+      "learning_rate": 0.00017476168756936798,
+      "loss": 1.2894,
+      "step": 3528
+    },
+    {
+      "epoch": 0.25437900958696746,
+      "grad_norm": 0.1264316403502508,
+      "learning_rate": 0.00017474617998229373,
+      "loss": 1.2451,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2544510920493044,
+      "grad_norm": 0.1290624854829654,
+      "learning_rate": 0.00017473066832088393,
+      "loss": 1.3449,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2545231745116413,
+      "grad_norm": 0.10810351504204636,
+      "learning_rate": 0.00017471515258598417,
+      "loss": 1.2977,
+      "step": 3531
+    },
+    {
+      "epoch": 0.2545952569739782,
+      "grad_norm": 0.10902547324963756,
+      "learning_rate": 0.00017469963277844016,
+      "loss": 1.2884,
+      "step": 3532
+    },
+    {
+      "epoch": 0.25466733943631514,
+      "grad_norm": 0.10786579344950314,
+      "learning_rate": 0.0001746841088990979,
+      "loss": 1.3044,
+      "step": 3533
+    },
+    {
+      "epoch": 0.25473942189865206,
+      "grad_norm": 0.12371715968433254,
+      "learning_rate": 0.00017466858094880355,
+      "loss": 1.2924,
+      "step": 3534
+    },
+    {
+      "epoch": 0.254811504360989,
+      "grad_norm": 0.11701628692178762,
+      "learning_rate": 0.00017465304892840355,
+      "loss": 1.212,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2548835868233259,
+      "grad_norm": 0.12677234083205796,
+      "learning_rate": 0.0001746375128387445,
+      "loss": 1.3184,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2549556692856628,
+      "grad_norm": 0.1157858354517134,
+      "learning_rate": 0.00017462197268067328,
+      "loss": 1.3513,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2550277517479997,
+      "grad_norm": 0.10999411277985932,
+      "learning_rate": 0.00017460642845503697,
+      "loss": 1.203,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2550998342103366,
+      "grad_norm": 0.1172396564922106,
+      "learning_rate": 0.00017459088016268286,
+      "loss": 1.2779,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2551719166726735,
+      "grad_norm": 0.1211585278645188,
+      "learning_rate": 0.00017457532780445844,
+      "loss": 1.1899,
+      "step": 3540
+    },
+    {
+      "epoch": 0.25524399913501045,
+      "grad_norm": 0.10532293728495808,
+      "learning_rate": 0.00017455977138121153,
+      "loss": 1.3157,
+      "step": 3541
+    },
+    {
+      "epoch": 0.25531608159734737,
+      "grad_norm": 0.11267215005323253,
+      "learning_rate": 0.00017454421089379004,
+      "loss": 1.3202,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2553881640596843,
+      "grad_norm": 0.10529332414088119,
+      "learning_rate": 0.00017452864634304212,
+      "loss": 1.3744,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2554602465220212,
+      "grad_norm": 0.10976933998231458,
+      "learning_rate": 0.0001745130777298163,
+      "loss": 1.3794,
+      "step": 3544
+    },
+    {
+      "epoch": 0.2555323289843581,
+      "grad_norm": 0.10590635035516452,
+      "learning_rate": 0.00017449750505496106,
+      "loss": 1.2456,
+      "step": 3545
+    },
+    {
+      "epoch": 0.25560441144669505,
+      "grad_norm": 0.11382126148392503,
+      "learning_rate": 0.00017448192831932534,
+      "loss": 1.3651,
+      "step": 3546
+    },
+    {
+      "epoch": 0.2556764939090319,
+      "grad_norm": 0.10830984719599114,
+      "learning_rate": 0.00017446634752375822,
+      "loss": 1.3608,
+      "step": 3547
+    },
+    {
+      "epoch": 0.25574857637136883,
+      "grad_norm": 0.10978535966185736,
+      "learning_rate": 0.00017445076266910892,
+      "loss": 1.2434,
+      "step": 3548
+    },
+    {
+      "epoch": 0.25582065883370575,
+      "grad_norm": 0.10739477039643454,
+      "learning_rate": 0.00017443517375622704,
+      "loss": 1.31,
+      "step": 3549
+    },
+    {
+      "epoch": 0.25589274129604267,
+      "grad_norm": 0.11372978793939541,
+      "learning_rate": 0.00017441958078596225,
+      "loss": 1.3475,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2559648237583796,
+      "grad_norm": 0.10552846730434434,
+      "learning_rate": 0.00017440398375916455,
+      "loss": 1.2672,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2560369062207165,
+      "grad_norm": 0.10255053561955152,
+      "learning_rate": 0.00017438838267668408,
+      "loss": 1.3588,
+      "step": 3552
+    },
+    {
+      "epoch": 0.25610898868305343,
+      "grad_norm": 0.11958044708093807,
+      "learning_rate": 0.00017437277753937127,
+      "loss": 1.3735,
+      "step": 3553
+    },
+    {
+      "epoch": 0.25618107114539035,
+      "grad_norm": 0.11868744992011156,
+      "learning_rate": 0.00017435716834807673,
+      "loss": 1.2423,
+      "step": 3554
+    },
+    {
+      "epoch": 0.2562531536077272,
+      "grad_norm": 0.1059023721413218,
+      "learning_rate": 0.00017434155510365125,
+      "loss": 1.336,
+      "step": 3555
+    },
+    {
+      "epoch": 0.25632523607006413,
+      "grad_norm": 0.1021227642020617,
+      "learning_rate": 0.000174325937806946,
+      "loss": 1.2555,
+      "step": 3556
+    },
+    {
+      "epoch": 0.25639731853240105,
+      "grad_norm": 0.12317081688468895,
+      "learning_rate": 0.00017431031645881215,
+      "loss": 1.2557,
+      "step": 3557
+    },
+    {
+      "epoch": 0.256469400994738,
+      "grad_norm": 0.1112445008355226,
+      "learning_rate": 0.0001742946910601013,
+      "loss": 1.3092,
+      "step": 3558
+    },
+    {
+      "epoch": 0.2565414834570749,
+      "grad_norm": 0.11384529643165514,
+      "learning_rate": 0.0001742790616116651,
+      "loss": 1.3879,
+      "step": 3559
+    },
+    {
+      "epoch": 0.2566135659194118,
+      "grad_norm": 0.11783937215479626,
+      "learning_rate": 0.00017426342811435554,
+      "loss": 1.3304,
+      "step": 3560
+    },
+    {
+      "epoch": 0.25668564838174873,
+      "grad_norm": 0.11302932418471788,
+      "learning_rate": 0.00017424779056902473,
+      "loss": 1.3819,
+      "step": 3561
+    },
+    {
+      "epoch": 0.25675773084408565,
+      "grad_norm": 0.1078128788443005,
+      "learning_rate": 0.00017423214897652515,
+      "loss": 1.3502,
+      "step": 3562
+    },
+    {
+      "epoch": 0.25682981330642257,
+      "grad_norm": 0.10969755722427083,
+      "learning_rate": 0.0001742165033377093,
+      "loss": 1.422,
+      "step": 3563
+    },
+    {
+      "epoch": 0.25690189576875944,
+      "grad_norm": 0.15915450622216912,
+      "learning_rate": 0.0001742008536534301,
+      "loss": 1.3626,
+      "step": 3564
+    },
+    {
+      "epoch": 0.25697397823109636,
+      "grad_norm": 0.12691248173721884,
+      "learning_rate": 0.00017418519992454053,
+      "loss": 1.282,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2570460606934333,
+      "grad_norm": 0.1120291069685479,
+      "learning_rate": 0.0001741695421518939,
+      "loss": 1.3132,
+      "step": 3566
+    },
+    {
+      "epoch": 0.2571181431557702,
+      "grad_norm": 0.11851003513696663,
+      "learning_rate": 0.00017415388033634364,
+      "loss": 1.3029,
+      "step": 3567
+    },
+    {
+      "epoch": 0.2571902256181071,
+      "grad_norm": 0.10004244222563985,
+      "learning_rate": 0.00017413821447874355,
+      "loss": 1.3178,
+      "step": 3568
+    },
+    {
+      "epoch": 0.25726230808044404,
+      "grad_norm": 0.12011518609069775,
+      "learning_rate": 0.0001741225445799475,
+      "loss": 1.2441,
+      "step": 3569
+    },
+    {
+      "epoch": 0.25733439054278096,
+      "grad_norm": 0.1452821153046727,
+      "learning_rate": 0.00017410687064080968,
+      "loss": 1.3933,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2574064730051179,
+      "grad_norm": 0.12124794466167212,
+      "learning_rate": 0.00017409119266218438,
+      "loss": 1.3617,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2574785554674548,
+      "grad_norm": 0.13134001700665998,
+      "learning_rate": 0.00017407551064492626,
+      "loss": 1.4063,
+      "step": 3572
+    },
+    {
+      "epoch": 0.25755063792979166,
+      "grad_norm": 0.1457099188860805,
+      "learning_rate": 0.0001740598245898901,
+      "loss": 1.3611,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2576227203921286,
+      "grad_norm": 0.11895021764023044,
+      "learning_rate": 0.00017404413449793098,
+      "loss": 1.3288,
+      "step": 3574
+    },
+    {
+      "epoch": 0.2576948028544655,
+      "grad_norm": 0.1258358038138702,
+      "learning_rate": 0.0001740284403699041,
+      "loss": 1.2109,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2577668853168024,
+      "grad_norm": 0.10491472335253982,
+      "learning_rate": 0.00017401274220666492,
+      "loss": 1.3258,
+      "step": 3576
+    },
+    {
+      "epoch": 0.25783896777913934,
+      "grad_norm": 0.11863192209700119,
+      "learning_rate": 0.00017399704000906915,
+      "loss": 1.2039,
+      "step": 3577
+    },
+    {
+      "epoch": 0.25791105024147626,
+      "grad_norm": 0.1102161751572644,
+      "learning_rate": 0.0001739813337779727,
+      "loss": 1.3713,
+      "step": 3578
+    },
+    {
+      "epoch": 0.2579831327038132,
+      "grad_norm": 0.1048734603273285,
+      "learning_rate": 0.00017396562351423175,
+      "loss": 1.2405,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2580552151661501,
+      "grad_norm": 0.11889228193089331,
+      "learning_rate": 0.00017394990921870256,
+      "loss": 1.2913,
+      "step": 3580
+    },
+    {
+      "epoch": 0.25812729762848696,
+      "grad_norm": 0.11100177294056934,
+      "learning_rate": 0.00017393419089224176,
+      "loss": 1.2913,
+      "step": 3581
+    },
+    {
+      "epoch": 0.2581993800908239,
+      "grad_norm": 0.11575242676487386,
+      "learning_rate": 0.00017391846853570612,
+      "loss": 1.3656,
+      "step": 3582
+    },
+    {
+      "epoch": 0.2582714625531608,
+      "grad_norm": 0.10218475735981247,
+      "learning_rate": 0.00017390274214995263,
+      "loss": 1.2494,
+      "step": 3583
+    },
+    {
+      "epoch": 0.2583435450154977,
+      "grad_norm": 0.13763946929012635,
+      "learning_rate": 0.00017388701173583858,
+      "loss": 1.3869,
+      "step": 3584
+    },
+    {
+      "epoch": 0.25841562747783464,
+      "grad_norm": 0.10585890239612326,
+      "learning_rate": 0.00017387127729422133,
+      "loss": 1.2574,
+      "step": 3585
+    },
+    {
+      "epoch": 0.25848770994017156,
+      "grad_norm": 0.12356402289294398,
+      "learning_rate": 0.00017385553882595862,
+      "loss": 1.2603,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2585597924025085,
+      "grad_norm": 0.16006003710940758,
+      "learning_rate": 0.00017383979633190833,
+      "loss": 1.2623,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2586318748648454,
+      "grad_norm": 0.10306555826350944,
+      "learning_rate": 0.00017382404981292852,
+      "loss": 1.3025,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2587039573271823,
+      "grad_norm": 0.1104415737363456,
+      "learning_rate": 0.00017380829926987756,
+      "loss": 1.3196,
+      "step": 3589
+    },
+    {
+      "epoch": 0.2587760397895192,
+      "grad_norm": 0.11529555912435388,
+      "learning_rate": 0.00017379254470361398,
+      "loss": 1.3329,
+      "step": 3590
+    },
+    {
+      "epoch": 0.2588481222518561,
+      "grad_norm": 0.12467403335499005,
+      "learning_rate": 0.00017377678611499658,
+      "loss": 1.2964,
+      "step": 3591
+    },
+    {
+      "epoch": 0.258920204714193,
+      "grad_norm": 0.09631413388589226,
+      "learning_rate": 0.00017376102350488427,
+      "loss": 1.2554,
+      "step": 3592
+    },
+    {
+      "epoch": 0.25899228717652995,
+      "grad_norm": 0.11317848455675658,
+      "learning_rate": 0.00017374525687413633,
+      "loss": 1.3791,
+      "step": 3593
+    },
+    {
+      "epoch": 0.25906436963886686,
+      "grad_norm": 0.1232745903864966,
+      "learning_rate": 0.00017372948622361212,
+      "loss": 1.4236,
+      "step": 3594
+    },
+    {
+      "epoch": 0.2591364521012038,
+      "grad_norm": 0.10710202240291304,
+      "learning_rate": 0.00017371371155417128,
+      "loss": 1.3609,
+      "step": 3595
+    },
+    {
+      "epoch": 0.2592085345635407,
+      "grad_norm": 0.12040082753571536,
+      "learning_rate": 0.00017369793286667372,
+      "loss": 1.2366,
+      "step": 3596
+    },
+    {
+      "epoch": 0.2592806170258776,
+      "grad_norm": 0.11696803740994725,
+      "learning_rate": 0.0001736821501619795,
+      "loss": 1.3587,
+      "step": 3597
+    },
+    {
+      "epoch": 0.25935269948821454,
+      "grad_norm": 0.10644560942145354,
+      "learning_rate": 0.00017366636344094894,
+      "loss": 1.271,
+      "step": 3598
+    },
+    {
+      "epoch": 0.2594247819505514,
+      "grad_norm": 0.19318193227305072,
+      "learning_rate": 0.00017365057270444254,
+      "loss": 1.349,
+      "step": 3599
+    },
+    {
+      "epoch": 0.25949686441288833,
+      "grad_norm": 0.12539194978584597,
+      "learning_rate": 0.000173634777953321,
+      "loss": 1.4234,
+      "step": 3600
+    },
+    {
+      "epoch": 0.25956894687522525,
+      "grad_norm": 0.11830568086156167,
+      "learning_rate": 0.0001736189791884453,
+      "loss": 1.3702,
+      "step": 3601
+    },
+    {
+      "epoch": 0.25964102933756217,
+      "grad_norm": 0.19196715959883973,
+      "learning_rate": 0.00017360317641067666,
+      "loss": 1.3525,
+      "step": 3602
+    },
+    {
+      "epoch": 0.2597131117998991,
+      "grad_norm": 0.11797651527233116,
+      "learning_rate": 0.0001735873696208764,
+      "loss": 1.236,
+      "step": 3603
+    },
+    {
+      "epoch": 0.259785194262236,
+      "grad_norm": 0.11426418347778888,
+      "learning_rate": 0.00017357155881990617,
+      "loss": 1.2821,
+      "step": 3604
+    },
+    {
+      "epoch": 0.2598572767245729,
+      "grad_norm": 0.11254222698813265,
+      "learning_rate": 0.00017355574400862778,
+      "loss": 1.2861,
+      "step": 3605
+    },
+    {
+      "epoch": 0.25992935918690985,
+      "grad_norm": 0.11653499102599552,
+      "learning_rate": 0.00017353992518790332,
+      "loss": 1.2203,
+      "step": 3606
+    },
+    {
+      "epoch": 0.2600014416492467,
+      "grad_norm": 0.10633851021425995,
+      "learning_rate": 0.00017352410235859503,
+      "loss": 1.2912,
+      "step": 3607
+    },
+    {
+      "epoch": 0.26007352411158363,
+      "grad_norm": 0.11390613694074911,
+      "learning_rate": 0.0001735082755215654,
+      "loss": 1.4804,
+      "step": 3608
+    },
+    {
+      "epoch": 0.26014560657392055,
+      "grad_norm": 0.11081042420160779,
+      "learning_rate": 0.00017349244467767708,
+      "loss": 1.2895,
+      "step": 3609
+    },
+    {
+      "epoch": 0.26021768903625747,
+      "grad_norm": 0.10153115808359361,
+      "learning_rate": 0.00017347660982779307,
+      "loss": 1.3048,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2602897714985944,
+      "grad_norm": 0.10169621998529599,
+      "learning_rate": 0.00017346077097277652,
+      "loss": 1.2975,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2603618539609313,
+      "grad_norm": 0.1047003968492854,
+      "learning_rate": 0.0001734449281134907,
+      "loss": 1.2896,
+      "step": 3612
+    },
+    {
+      "epoch": 0.26043393642326823,
+      "grad_norm": 0.11956815155459505,
+      "learning_rate": 0.00017342908125079928,
+      "loss": 1.3327,
+      "step": 3613
+    },
+    {
+      "epoch": 0.26050601888560515,
+      "grad_norm": 0.1167059390412135,
+      "learning_rate": 0.000173413230385566,
+      "loss": 1.3049,
+      "step": 3614
+    },
+    {
+      "epoch": 0.26057810134794207,
+      "grad_norm": 0.1514894626247825,
+      "learning_rate": 0.0001733973755186549,
+      "loss": 1.282,
+      "step": 3615
+    },
+    {
+      "epoch": 0.26065018381027893,
+      "grad_norm": 0.1209973250559852,
+      "learning_rate": 0.0001733815166509302,
+      "loss": 1.3675,
+      "step": 3616
+    },
+    {
+      "epoch": 0.26072226627261585,
+      "grad_norm": 0.10760368198751057,
+      "learning_rate": 0.00017336565378325632,
+      "loss": 1.4261,
+      "step": 3617
+    },
+    {
+      "epoch": 0.2607943487349528,
+      "grad_norm": 0.2020627920755438,
+      "learning_rate": 0.00017334978691649802,
+      "loss": 1.3414,
+      "step": 3618
+    },
+    {
+      "epoch": 0.2608664311972897,
+      "grad_norm": 0.1204657662329319,
+      "learning_rate": 0.00017333391605152008,
+      "loss": 1.2923,
+      "step": 3619
+    },
+    {
+      "epoch": 0.2609385136596266,
+      "grad_norm": 0.11424824142297638,
+      "learning_rate": 0.00017331804118918766,
+      "loss": 1.3047,
+      "step": 3620
+    },
+    {
+      "epoch": 0.26101059612196353,
+      "grad_norm": 0.10566845165257285,
+      "learning_rate": 0.00017330216233036612,
+      "loss": 1.3182,
+      "step": 3621
+    },
+    {
+      "epoch": 0.26108267858430045,
+      "grad_norm": 0.12787942618261264,
+      "learning_rate": 0.00017328627947592087,
+      "loss": 1.3337,
+      "step": 3622
+    },
+    {
+      "epoch": 0.2611547610466374,
+      "grad_norm": 0.11183184772461288,
+      "learning_rate": 0.0001732703926267178,
+      "loss": 1.3341,
+      "step": 3623
+    },
+    {
+      "epoch": 0.2612268435089743,
+      "grad_norm": 0.10692740978452228,
+      "learning_rate": 0.00017325450178362286,
+      "loss": 1.3218,
+      "step": 3624
+    },
+    {
+      "epoch": 0.26129892597131116,
+      "grad_norm": 0.10553887686787403,
+      "learning_rate": 0.00017323860694750217,
+      "loss": 1.2909,
+      "step": 3625
+    },
+    {
+      "epoch": 0.2613710084336481,
+      "grad_norm": 0.11489795684195434,
+      "learning_rate": 0.00017322270811922222,
+      "loss": 1.2846,
+      "step": 3626
+    },
+    {
+      "epoch": 0.261443090895985,
+      "grad_norm": 0.11288166422310157,
+      "learning_rate": 0.0001732068052996496,
+      "loss": 1.3587,
+      "step": 3627
+    },
+    {
+      "epoch": 0.2615151733583219,
+      "grad_norm": 0.1139512458651665,
+      "learning_rate": 0.00017319089848965118,
+      "loss": 1.2819,
+      "step": 3628
+    },
+    {
+      "epoch": 0.26158725582065884,
+      "grad_norm": 0.10823614130743597,
+      "learning_rate": 0.00017317498769009397,
+      "loss": 1.3325,
+      "step": 3629
+    },
+    {
+      "epoch": 0.26165933828299576,
+      "grad_norm": 0.10301837765931178,
+      "learning_rate": 0.0001731590729018453,
+      "loss": 1.4074,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2617314207453327,
+      "grad_norm": 0.10629227574377946,
+      "learning_rate": 0.0001731431541257727,
+      "loss": 1.4045,
+      "step": 3631
+    },
+    {
+      "epoch": 0.2618035032076696,
+      "grad_norm": 0.10901868197968291,
+      "learning_rate": 0.0001731272313627438,
+      "loss": 1.3437,
+      "step": 3632
+    },
+    {
+      "epoch": 0.26187558567000646,
+      "grad_norm": 0.11740218414534269,
+      "learning_rate": 0.00017311130461362658,
+      "loss": 1.1901,
+      "step": 3633
+    },
+    {
+      "epoch": 0.2619476681323434,
+      "grad_norm": 0.11160048999679452,
+      "learning_rate": 0.0001730953738792892,
+      "loss": 1.3447,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2620197505946803,
+      "grad_norm": 0.10965033580595174,
+      "learning_rate": 0.0001730794391606,
+      "loss": 1.2824,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2620918330570172,
+      "grad_norm": 0.10695307202202117,
+      "learning_rate": 0.0001730635004584276,
+      "loss": 1.395,
+      "step": 3636
+    },
+    {
+      "epoch": 0.26216391551935414,
+      "grad_norm": 0.11550668888411458,
+      "learning_rate": 0.00017304755777364074,
+      "loss": 1.3418,
+      "step": 3637
+    },
+    {
+      "epoch": 0.26223599798169106,
+      "grad_norm": 0.12381302569934499,
+      "learning_rate": 0.0001730316111071085,
+      "loss": 1.3235,
+      "step": 3638
+    },
+    {
+      "epoch": 0.262308080444028,
+      "grad_norm": 0.11851950644181682,
+      "learning_rate": 0.0001730156604597001,
+      "loss": 1.3124,
+      "step": 3639
+    },
+    {
+      "epoch": 0.2623801629063649,
+      "grad_norm": 0.11223899475476684,
+      "learning_rate": 0.000172999705832285,
+      "loss": 1.3701,
+      "step": 3640
+    },
+    {
+      "epoch": 0.2624522453687018,
+      "grad_norm": 0.10881722120275919,
+      "learning_rate": 0.00017298374722573282,
+      "loss": 1.3494,
+      "step": 3641
+    },
+    {
+      "epoch": 0.2625243278310387,
+      "grad_norm": 0.10796374888973777,
+      "learning_rate": 0.00017296778464091352,
+      "loss": 1.3081,
+      "step": 3642
+    },
+    {
+      "epoch": 0.2625964102933756,
+      "grad_norm": 0.11394313969006771,
+      "learning_rate": 0.00017295181807869717,
+      "loss": 1.316,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2626684927557125,
+      "grad_norm": 0.1183162595817394,
+      "learning_rate": 0.00017293584753995405,
+      "loss": 1.3612,
+      "step": 3644
+    },
+    {
+      "epoch": 0.26274057521804944,
+      "grad_norm": 0.10982167082076334,
+      "learning_rate": 0.0001729198730255548,
+      "loss": 1.329,
+      "step": 3645
+    },
+    {
+      "epoch": 0.26281265768038636,
+      "grad_norm": 0.10917859797744489,
+      "learning_rate": 0.00017290389453637005,
+      "loss": 1.3133,
+      "step": 3646
+    },
+    {
+      "epoch": 0.2628847401427233,
+      "grad_norm": 0.11052409135790607,
+      "learning_rate": 0.00017288791207327087,
+      "loss": 1.2693,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2629568226050602,
+      "grad_norm": 0.11945313662786648,
+      "learning_rate": 0.0001728719256371284,
+      "loss": 1.4104,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2630289050673971,
+      "grad_norm": 0.11447782012384147,
+      "learning_rate": 0.000172855935228814,
+      "loss": 1.3053,
+      "step": 3649
+    },
+    {
+      "epoch": 0.26310098752973404,
+      "grad_norm": 0.10440036128805237,
+      "learning_rate": 0.0001728399408491994,
+      "loss": 1.384,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2631730699920709,
+      "grad_norm": 0.10275950029898749,
+      "learning_rate": 0.0001728239424991564,
+      "loss": 1.2257,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2632451524544078,
+      "grad_norm": 0.1075123307057025,
+      "learning_rate": 0.00017280794017955698,
+      "loss": 1.3469,
+      "step": 3652
+    },
+    {
+      "epoch": 0.26331723491674475,
+      "grad_norm": 0.11093617563726296,
+      "learning_rate": 0.00017279193389127347,
+      "loss": 1.3037,
+      "step": 3653
+    },
+    {
+      "epoch": 0.26338931737908167,
+      "grad_norm": 0.11581414459007816,
+      "learning_rate": 0.0001727759236351784,
+      "loss": 1.3098,
+      "step": 3654
+    },
+    {
+      "epoch": 0.2634613998414186,
+      "grad_norm": 0.10776370435880304,
+      "learning_rate": 0.00017275990941214436,
+      "loss": 1.4021,
+      "step": 3655
+    },
+    {
+      "epoch": 0.2635334823037555,
+      "grad_norm": 0.10417400316804894,
+      "learning_rate": 0.00017274389122304436,
+      "loss": 1.2411,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2636055647660924,
+      "grad_norm": 0.1050028882540327,
+      "learning_rate": 0.00017272786906875147,
+      "loss": 1.3179,
+      "step": 3657
+    },
+    {
+      "epoch": 0.26367764722842935,
+      "grad_norm": 0.1150930676399442,
+      "learning_rate": 0.0001727118429501391,
+      "loss": 1.3327,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2637497296907662,
+      "grad_norm": 0.1015562366824098,
+      "learning_rate": 0.00017269581286808076,
+      "loss": 1.2895,
+      "step": 3659
+    },
+    {
+      "epoch": 0.26382181215310313,
+      "grad_norm": 0.1028932846556109,
+      "learning_rate": 0.0001726797788234503,
+      "loss": 1.2504,
+      "step": 3660
+    },
+    {
+      "epoch": 0.26389389461544005,
+      "grad_norm": 0.11590348771186887,
+      "learning_rate": 0.00017266374081712166,
+      "loss": 1.3873,
+      "step": 3661
+    },
+    {
+      "epoch": 0.26396597707777697,
+      "grad_norm": 0.10573735374802519,
+      "learning_rate": 0.0001726476988499691,
+      "loss": 1.3029,
+      "step": 3662
+    },
+    {
+      "epoch": 0.2640380595401139,
+      "grad_norm": 0.10917105415685202,
+      "learning_rate": 0.000172631652922867,
+      "loss": 1.3561,
+      "step": 3663
+    },
+    {
+      "epoch": 0.2641101420024508,
+      "grad_norm": 0.11389954215457851,
+      "learning_rate": 0.00017261560303669007,
+      "loss": 1.3,
+      "step": 3664
+    },
+    {
+      "epoch": 0.26418222446478773,
+      "grad_norm": 0.10907592309465056,
+      "learning_rate": 0.0001725995491923131,
+      "loss": 1.3689,
+      "step": 3665
+    },
+    {
+      "epoch": 0.26425430692712465,
+      "grad_norm": 0.11350348894206898,
+      "learning_rate": 0.0001725834913906112,
+      "loss": 1.3361,
+      "step": 3666
+    },
+    {
+      "epoch": 0.26432638938946157,
+      "grad_norm": 0.1066732420704875,
+      "learning_rate": 0.0001725674296324597,
+      "loss": 1.261,
+      "step": 3667
+    },
+    {
+      "epoch": 0.26439847185179843,
+      "grad_norm": 0.10917155048867197,
+      "learning_rate": 0.00017255136391873408,
+      "loss": 1.241,
+      "step": 3668
+    },
+    {
+      "epoch": 0.26447055431413535,
+      "grad_norm": 0.1011759789676654,
+      "learning_rate": 0.00017253529425031005,
+      "loss": 1.3645,
+      "step": 3669
+    },
+    {
+      "epoch": 0.2645426367764723,
+      "grad_norm": 0.11332482272081736,
+      "learning_rate": 0.00017251922062806355,
+      "loss": 1.3166,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2646147192388092,
+      "grad_norm": 0.10571413334004473,
+      "learning_rate": 0.00017250314305287078,
+      "loss": 1.2235,
+      "step": 3671
+    },
+    {
+      "epoch": 0.2646868017011461,
+      "grad_norm": 0.10619200083809952,
+      "learning_rate": 0.00017248706152560804,
+      "loss": 1.2344,
+      "step": 3672
+    },
+    {
+      "epoch": 0.26475888416348303,
+      "grad_norm": 0.1065895777408642,
+      "learning_rate": 0.000172470976047152,
+      "loss": 1.3173,
+      "step": 3673
+    },
+    {
+      "epoch": 0.26483096662581995,
+      "grad_norm": 0.12110999031416379,
+      "learning_rate": 0.0001724548866183794,
+      "loss": 1.3445,
+      "step": 3674
+    },
+    {
+      "epoch": 0.26490304908815687,
+      "grad_norm": 0.10459474119567452,
+      "learning_rate": 0.0001724387932401673,
+      "loss": 1.2751,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2649751315504938,
+      "grad_norm": 0.0992805656570834,
+      "learning_rate": 0.0001724226959133929,
+      "loss": 1.2758,
+      "step": 3676
+    },
+    {
+      "epoch": 0.26504721401283066,
+      "grad_norm": 0.1159207308810915,
+      "learning_rate": 0.00017240659463893364,
+      "loss": 1.257,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2651192964751676,
+      "grad_norm": 0.14076971996191973,
+      "learning_rate": 0.00017239048941766725,
+      "loss": 1.3281,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2651913789375045,
+      "grad_norm": 0.11618966953082263,
+      "learning_rate": 0.00017237438025047153,
+      "loss": 1.2775,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2652634613998414,
+      "grad_norm": 0.10267804318340308,
+      "learning_rate": 0.0001723582671382246,
+      "loss": 1.2757,
+      "step": 3680
+    },
+    {
+      "epoch": 0.26533554386217834,
+      "grad_norm": 0.11028740469867693,
+      "learning_rate": 0.00017234215008180478,
+      "loss": 1.3493,
+      "step": 3681
+    },
+    {
+      "epoch": 0.26540762632451526,
+      "grad_norm": 0.11073435761034608,
+      "learning_rate": 0.00017232602908209062,
+      "loss": 1.221,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2654797087868522,
+      "grad_norm": 0.11828574834614067,
+      "learning_rate": 0.0001723099041399608,
+      "loss": 1.1822,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2655517912491891,
+      "grad_norm": 0.11662569363440943,
+      "learning_rate": 0.0001722937752562943,
+      "loss": 1.2702,
+      "step": 3684
+    },
+    {
+      "epoch": 0.26562387371152596,
+      "grad_norm": 0.10728187925569001,
+      "learning_rate": 0.0001722776424319703,
+      "loss": 1.2979,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2656959561738629,
+      "grad_norm": 0.10557306387563181,
+      "learning_rate": 0.00017226150566786817,
+      "loss": 1.3429,
+      "step": 3686
+    },
+    {
+      "epoch": 0.2657680386361998,
+      "grad_norm": 0.10345487230427895,
+      "learning_rate": 0.0001722453649648675,
+      "loss": 1.1679,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2658401210985367,
+      "grad_norm": 0.1024257803891183,
+      "learning_rate": 0.00017222922032384812,
+      "loss": 1.3375,
+      "step": 3688
+    },
+    {
+      "epoch": 0.26591220356087364,
+      "grad_norm": 0.10369716713908388,
+      "learning_rate": 0.0001722130717456901,
+      "loss": 1.3825,
+      "step": 3689
+    },
+    {
+      "epoch": 0.26598428602321056,
+      "grad_norm": 0.11415582161450125,
+      "learning_rate": 0.00017219691923127357,
+      "loss": 1.325,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2660563684855475,
+      "grad_norm": 0.10577094190040791,
+      "learning_rate": 0.00017218076278147906,
+      "loss": 1.2693,
+      "step": 3691
+    },
+    {
+      "epoch": 0.2661284509478844,
+      "grad_norm": 0.11652203948954439,
+      "learning_rate": 0.00017216460239718724,
+      "loss": 1.304,
+      "step": 3692
+    },
+    {
+      "epoch": 0.2662005334102213,
+      "grad_norm": 0.11460297017581172,
+      "learning_rate": 0.000172148438079279,
+      "loss": 1.2229,
+      "step": 3693
+    },
+    {
+      "epoch": 0.2662726158725582,
+      "grad_norm": 0.12631177065316376,
+      "learning_rate": 0.0001721322698286354,
+      "loss": 1.3688,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2663446983348951,
+      "grad_norm": 0.11293436579595884,
+      "learning_rate": 0.00017211609764613777,
+      "loss": 1.2703,
+      "step": 3695
+    },
+    {
+      "epoch": 0.266416780797232,
+      "grad_norm": 0.10774309854438556,
+      "learning_rate": 0.0001720999215326677,
+      "loss": 1.383,
+      "step": 3696
+    },
+    {
+      "epoch": 0.26648886325956894,
+      "grad_norm": 0.11461446105911677,
+      "learning_rate": 0.00017208374148910686,
+      "loss": 1.4214,
+      "step": 3697
+    },
+    {
+      "epoch": 0.26656094572190586,
+      "grad_norm": 0.12494686719674346,
+      "learning_rate": 0.00017206755751633722,
+      "loss": 1.2843,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2666330281842428,
+      "grad_norm": 0.11953956180992986,
+      "learning_rate": 0.00017205136961524094,
+      "loss": 1.1899,
+      "step": 3699
+    },
+    {
+      "epoch": 0.2667051106465797,
+      "grad_norm": 0.114426189924985,
+      "learning_rate": 0.00017203517778670047,
+      "loss": 1.2198,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2667771931089166,
+      "grad_norm": 0.11412894014538986,
+      "learning_rate": 0.00017201898203159837,
+      "loss": 1.3734,
+      "step": 3701
+    },
+    {
+      "epoch": 0.26684927557125354,
+      "grad_norm": 0.11395150762959888,
+      "learning_rate": 0.00017200278235081741,
+      "loss": 1.2041,
+      "step": 3702
+    },
+    {
+      "epoch": 0.2669213580335904,
+      "grad_norm": 0.10740726769849364,
+      "learning_rate": 0.00017198657874524067,
+      "loss": 1.2424,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2669934404959273,
+      "grad_norm": 0.1067449291612804,
+      "learning_rate": 0.00017197037121575138,
+      "loss": 1.3562,
+      "step": 3704
+    },
+    {
+      "epoch": 0.26706552295826425,
+      "grad_norm": 0.111470350099817,
+      "learning_rate": 0.00017195415976323298,
+      "loss": 1.2919,
+      "step": 3705
+    },
+    {
+      "epoch": 0.26713760542060117,
+      "grad_norm": 0.11098362625922331,
+      "learning_rate": 0.00017193794438856918,
+      "loss": 1.2623,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2672096878829381,
+      "grad_norm": 0.1561656540792497,
+      "learning_rate": 0.00017192172509264384,
+      "loss": 1.3247,
+      "step": 3707
+    },
+    {
+      "epoch": 0.267281770345275,
+      "grad_norm": 0.10523385804679447,
+      "learning_rate": 0.000171905501876341,
+      "loss": 1.3171,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2673538528076119,
+      "grad_norm": 0.11023058878853731,
+      "learning_rate": 0.0001718892747405451,
+      "loss": 1.1892,
+      "step": 3709
+    },
+    {
+      "epoch": 0.26742593526994884,
+      "grad_norm": 0.10760939952298205,
+      "learning_rate": 0.00017187304368614052,
+      "loss": 1.4153,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2674980177322857,
+      "grad_norm": 0.12496827207206691,
+      "learning_rate": 0.0001718568087140121,
+      "loss": 1.3028,
+      "step": 3711
+    },
+    {
+      "epoch": 0.26757010019462263,
+      "grad_norm": 0.10480480859574076,
+      "learning_rate": 0.00017184056982504472,
+      "loss": 1.302,
+      "step": 3712
+    },
+    {
+      "epoch": 0.26764218265695955,
+      "grad_norm": 0.11076605721302464,
+      "learning_rate": 0.00017182432702012363,
+      "loss": 1.2829,
+      "step": 3713
+    },
+    {
+      "epoch": 0.26771426511929647,
+      "grad_norm": 0.09570063961566015,
+      "learning_rate": 0.0001718080803001341,
+      "loss": 1.2924,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2677863475816334,
+      "grad_norm": 0.09327251655095503,
+      "learning_rate": 0.00017179182966596182,
+      "loss": 1.4032,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2678584300439703,
+      "grad_norm": 0.15291955483506067,
+      "learning_rate": 0.00017177557511849256,
+      "loss": 1.2819,
+      "step": 3716
+    },
+    {
+      "epoch": 0.26793051250630723,
+      "grad_norm": 0.2397426433980104,
+      "learning_rate": 0.00017175931665861232,
+      "loss": 1.3443,
+      "step": 3717
+    },
+    {
+      "epoch": 0.26800259496864415,
+      "grad_norm": 0.10385955751660253,
+      "learning_rate": 0.00017174305428720735,
+      "loss": 1.2919,
+      "step": 3718
+    },
+    {
+      "epoch": 0.26807467743098107,
+      "grad_norm": 0.11016816551225714,
+      "learning_rate": 0.00017172678800516408,
+      "loss": 1.2472,
+      "step": 3719
+    },
+    {
+      "epoch": 0.26814675989331793,
+      "grad_norm": 0.12556013191614124,
+      "learning_rate": 0.0001717105178133692,
+      "loss": 1.2631,
+      "step": 3720
+    },
+    {
+      "epoch": 0.26821884235565485,
+      "grad_norm": 0.1626394655926821,
+      "learning_rate": 0.00017169424371270956,
+      "loss": 1.271,
+      "step": 3721
+    },
+    {
+      "epoch": 0.26829092481799177,
+      "grad_norm": 0.10816258233522241,
+      "learning_rate": 0.0001716779657040722,
+      "loss": 1.3127,
+      "step": 3722
+    },
+    {
+      "epoch": 0.2683630072803287,
+      "grad_norm": 0.10615611846885968,
+      "learning_rate": 0.00017166168378834448,
+      "loss": 1.3103,
+      "step": 3723
+    },
+    {
+      "epoch": 0.2684350897426656,
+      "grad_norm": 0.12812622444862634,
+      "learning_rate": 0.0001716453979664139,
+      "loss": 1.3123,
+      "step": 3724
+    },
+    {
+      "epoch": 0.26850717220500253,
+      "grad_norm": 0.10765369500554642,
+      "learning_rate": 0.00017162910823916816,
+      "loss": 1.2442,
+      "step": 3725
+    },
+    {
+      "epoch": 0.26857925466733945,
+      "grad_norm": 0.1044505685324306,
+      "learning_rate": 0.0001716128146074952,
+      "loss": 1.2872,
+      "step": 3726
+    },
+    {
+      "epoch": 0.26865133712967637,
+      "grad_norm": 0.11405797128824156,
+      "learning_rate": 0.00017159651707228323,
+      "loss": 1.3167,
+      "step": 3727
+    },
+    {
+      "epoch": 0.2687234195920133,
+      "grad_norm": 0.10961588858887963,
+      "learning_rate": 0.0001715802156344205,
+      "loss": 1.3561,
+      "step": 3728
+    },
+    {
+      "epoch": 0.26879550205435015,
+      "grad_norm": 0.10671874642705759,
+      "learning_rate": 0.0001715639102947957,
+      "loss": 1.3009,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2688675845166871,
+      "grad_norm": 0.10311261098494619,
+      "learning_rate": 0.00017154760105429748,
+      "loss": 1.3053,
+      "step": 3730
+    },
+    {
+      "epoch": 0.268939666979024,
+      "grad_norm": 0.1564124124191631,
+      "learning_rate": 0.000171531287913815,
+      "loss": 1.2901,
+      "step": 3731
+    },
+    {
+      "epoch": 0.2690117494413609,
+      "grad_norm": 0.17585670960646632,
+      "learning_rate": 0.00017151497087423734,
+      "loss": 1.3102,
+      "step": 3732
+    },
+    {
+      "epoch": 0.26908383190369783,
+      "grad_norm": 0.1213951425122753,
+      "learning_rate": 0.00017149864993645398,
+      "loss": 1.3007,
+      "step": 3733
+    },
+    {
+      "epoch": 0.26915591436603475,
+      "grad_norm": 0.1154469349521673,
+      "learning_rate": 0.0001714823251013546,
+      "loss": 1.3157,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2692279968283717,
+      "grad_norm": 0.10370348693900941,
+      "learning_rate": 0.00017146599636982894,
+      "loss": 1.3336,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2693000792907086,
+      "grad_norm": 0.11622023488575424,
+      "learning_rate": 0.00017144966374276716,
+      "loss": 1.3058,
+      "step": 3736
+    },
+    {
+      "epoch": 0.26937216175304546,
+      "grad_norm": 0.12808118768211352,
+      "learning_rate": 0.0001714333272210595,
+      "loss": 1.2666,
+      "step": 3737
+    },
+    {
+      "epoch": 0.2694442442153824,
+      "grad_norm": 0.09785764852073696,
+      "learning_rate": 0.0001714169868055964,
+      "loss": 1.319,
+      "step": 3738
+    },
+    {
+      "epoch": 0.2695163266777193,
+      "grad_norm": 0.10535987196248137,
+      "learning_rate": 0.00017140064249726865,
+      "loss": 1.3004,
+      "step": 3739
+    },
+    {
+      "epoch": 0.2695884091400562,
+      "grad_norm": 0.11385192899170749,
+      "learning_rate": 0.00017138429429696712,
+      "loss": 1.3126,
+      "step": 3740
+    },
+    {
+      "epoch": 0.26966049160239314,
+      "grad_norm": 0.10964363340339878,
+      "learning_rate": 0.0001713679422055829,
+      "loss": 1.2194,
+      "step": 3741
+    },
+    {
+      "epoch": 0.26973257406473006,
+      "grad_norm": 0.10959477475453404,
+      "learning_rate": 0.00017135158622400739,
+      "loss": 1.3784,
+      "step": 3742
+    },
+    {
+      "epoch": 0.269804656527067,
+      "grad_norm": 0.11921771020358324,
+      "learning_rate": 0.0001713352263531321,
+      "loss": 1.1812,
+      "step": 3743
+    },
+    {
+      "epoch": 0.2698767389894039,
+      "grad_norm": 0.12111221720595611,
+      "learning_rate": 0.00017131886259384875,
+      "loss": 1.3623,
+      "step": 3744
+    },
+    {
+      "epoch": 0.2699488214517408,
+      "grad_norm": 0.11151156770850106,
+      "learning_rate": 0.00017130249494704938,
+      "loss": 1.2234,
+      "step": 3745
+    },
+    {
+      "epoch": 0.2700209039140777,
+      "grad_norm": 0.11820586826061107,
+      "learning_rate": 0.00017128612341362612,
+      "loss": 1.3543,
+      "step": 3746
+    },
+    {
+      "epoch": 0.2700929863764146,
+      "grad_norm": 0.10457172707589867,
+      "learning_rate": 0.00017126974799447142,
+      "loss": 1.259,
+      "step": 3747
+    },
+    {
+      "epoch": 0.2701650688387515,
+      "grad_norm": 0.1922037461472288,
+      "learning_rate": 0.00017125336869047787,
+      "loss": 1.436,
+      "step": 3748
+    },
+    {
+      "epoch": 0.27023715130108844,
+      "grad_norm": 0.10680741105998562,
+      "learning_rate": 0.00017123698550253823,
+      "loss": 1.3373,
+      "step": 3749
+    },
+    {
+      "epoch": 0.27030923376342536,
+      "grad_norm": 0.10297264271697504,
+      "learning_rate": 0.0001712205984315456,
+      "loss": 1.3687,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2703813162257623,
+      "grad_norm": 0.20793177325108142,
+      "learning_rate": 0.00017120420747839319,
+      "loss": 1.3124,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2704533986880992,
+      "grad_norm": 0.1101458021274568,
+      "learning_rate": 0.00017118781264397446,
+      "loss": 1.269,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2705254811504361,
+      "grad_norm": 0.10714837688650018,
+      "learning_rate": 0.00017117141392918308,
+      "loss": 1.2992,
+      "step": 3753
+    },
+    {
+      "epoch": 0.27059756361277304,
+      "grad_norm": 0.11831346564628384,
+      "learning_rate": 0.00017115501133491294,
+      "loss": 1.2782,
+      "step": 3754
+    },
+    {
+      "epoch": 0.2706696460751099,
+      "grad_norm": 0.10445447807694075,
+      "learning_rate": 0.0001711386048620581,
+      "loss": 1.2365,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2707417285374468,
+      "grad_norm": 0.13212821495983054,
+      "learning_rate": 0.00017112219451151286,
+      "loss": 1.3069,
+      "step": 3756
+    },
+    {
+      "epoch": 0.27081381099978374,
+      "grad_norm": 0.11504972215175262,
+      "learning_rate": 0.00017110578028417176,
+      "loss": 1.2091,
+      "step": 3757
+    },
+    {
+      "epoch": 0.27088589346212066,
+      "grad_norm": 0.1209805802139519,
+      "learning_rate": 0.00017108936218092948,
+      "loss": 1.3498,
+      "step": 3758
+    },
+    {
+      "epoch": 0.2709579759244576,
+      "grad_norm": 0.1107554189966811,
+      "learning_rate": 0.000171072940202681,
+      "loss": 1.3359,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2710300583867945,
+      "grad_norm": 0.1640550759876475,
+      "learning_rate": 0.00017105651435032142,
+      "loss": 1.2613,
+      "step": 3760
+    },
+    {
+      "epoch": 0.2711021408491314,
+      "grad_norm": 0.11248164745195308,
+      "learning_rate": 0.00017104008462474614,
+      "loss": 1.2936,
+      "step": 3761
+    },
+    {
+      "epoch": 0.27117422331146834,
+      "grad_norm": 0.09740089029081946,
+      "learning_rate": 0.0001710236510268507,
+      "loss": 1.2868,
+      "step": 3762
+    },
+    {
+      "epoch": 0.2712463057738052,
+      "grad_norm": 0.10823624003366518,
+      "learning_rate": 0.00017100721355753086,
+      "loss": 1.3528,
+      "step": 3763
+    },
+    {
+      "epoch": 0.2713183882361421,
+      "grad_norm": 0.1538221680558147,
+      "learning_rate": 0.00017099077221768266,
+      "loss": 1.28,
+      "step": 3764
+    },
+    {
+      "epoch": 0.27139047069847905,
+      "grad_norm": 0.10995230464716382,
+      "learning_rate": 0.0001709743270082023,
+      "loss": 1.3312,
+      "step": 3765
+    },
+    {
+      "epoch": 0.27146255316081597,
+      "grad_norm": 0.12832017406264867,
+      "learning_rate": 0.00017095787792998613,
+      "loss": 1.2584,
+      "step": 3766
+    },
+    {
+      "epoch": 0.2715346356231529,
+      "grad_norm": 0.14546195192700861,
+      "learning_rate": 0.00017094142498393078,
+      "loss": 1.3279,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2716067180854898,
+      "grad_norm": 0.11086888378397956,
+      "learning_rate": 0.00017092496817093316,
+      "loss": 1.2972,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2716788005478267,
+      "grad_norm": 0.1077451293662402,
+      "learning_rate": 0.00017090850749189022,
+      "loss": 1.2769,
+      "step": 3769
+    },
+    {
+      "epoch": 0.27175088301016365,
+      "grad_norm": 0.1166938669378458,
+      "learning_rate": 0.00017089204294769926,
+      "loss": 1.2858,
+      "step": 3770
+    },
+    {
+      "epoch": 0.27182296547250057,
+      "grad_norm": 0.11660863041504653,
+      "learning_rate": 0.00017087557453925774,
+      "loss": 1.422,
+      "step": 3771
+    },
+    {
+      "epoch": 0.27189504793483743,
+      "grad_norm": 0.11013216770383272,
+      "learning_rate": 0.00017085910226746336,
+      "loss": 1.2492,
+      "step": 3772
+    },
+    {
+      "epoch": 0.27196713039717435,
+      "grad_norm": 0.11789322183606875,
+      "learning_rate": 0.00017084262613321398,
+      "loss": 1.2837,
+      "step": 3773
+    },
+    {
+      "epoch": 0.27203921285951127,
+      "grad_norm": 0.10085279086916941,
+      "learning_rate": 0.00017082614613740768,
+      "loss": 1.3541,
+      "step": 3774
+    },
+    {
+      "epoch": 0.2721112953218482,
+      "grad_norm": 0.1098091685329107,
+      "learning_rate": 0.0001708096622809428,
+      "loss": 1.3601,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2721833777841851,
+      "grad_norm": 0.13317283271820796,
+      "learning_rate": 0.00017079317456471785,
+      "loss": 1.3386,
+      "step": 3776
+    },
+    {
+      "epoch": 0.27225546024652203,
+      "grad_norm": 0.1299041801422694,
+      "learning_rate": 0.00017077668298963155,
+      "loss": 1.3166,
+      "step": 3777
+    },
+    {
+      "epoch": 0.27232754270885895,
+      "grad_norm": 0.11681509955378994,
+      "learning_rate": 0.00017076018755658284,
+      "loss": 1.1946,
+      "step": 3778
+    },
+    {
+      "epoch": 0.27239962517119587,
+      "grad_norm": 0.11299843806079307,
+      "learning_rate": 0.00017074368826647084,
+      "loss": 1.2846,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2724717076335328,
+      "grad_norm": 0.24633885539414152,
+      "learning_rate": 0.000170727185120195,
+      "loss": 1.2339,
+      "step": 3780
+    },
+    {
+      "epoch": 0.27254379009586965,
+      "grad_norm": 0.17089968359427882,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 1.2443,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2726158725582066,
+      "grad_norm": 0.24591070282830632,
+      "learning_rate": 0.00017069416726275,
+      "loss": 1.2886,
+      "step": 3782
+    },
+    {
+      "epoch": 0.2726879550205435,
+      "grad_norm": 0.10367790032487846,
+      "learning_rate": 0.00017067765255338068,
+      "loss": 1.2711,
+      "step": 3783
+    },
+    {
+      "epoch": 0.2727600374828804,
+      "grad_norm": 0.12001599711507938,
+      "learning_rate": 0.000170661133991447,
+      "loss": 1.3698,
+      "step": 3784
+    },
+    {
+      "epoch": 0.27283211994521733,
+      "grad_norm": 0.10350037816536127,
+      "learning_rate": 0.00017064461157784933,
+      "loss": 1.3427,
+      "step": 3785
+    },
+    {
+      "epoch": 0.27290420240755425,
+      "grad_norm": 0.1164119672996431,
+      "learning_rate": 0.00017062808531348838,
+      "loss": 1.2794,
+      "step": 3786
+    },
+    {
+      "epoch": 0.2729762848698912,
+      "grad_norm": 0.12169690633636414,
+      "learning_rate": 0.00017061155519926485,
+      "loss": 1.3665,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2730483673322281,
+      "grad_norm": 0.10849050351512543,
+      "learning_rate": 0.0001705950212360799,
+      "loss": 1.3298,
+      "step": 3788
+    },
+    {
+      "epoch": 0.27312044979456496,
+      "grad_norm": 0.10467492356958734,
+      "learning_rate": 0.00017057848342483473,
+      "loss": 1.3,
+      "step": 3789
+    },
+    {
+      "epoch": 0.2731925322569019,
+      "grad_norm": 0.11188600093862439,
+      "learning_rate": 0.0001705619417664308,
+      "loss": 1.3262,
+      "step": 3790
+    },
+    {
+      "epoch": 0.2732646147192388,
+      "grad_norm": 0.10632709112990411,
+      "learning_rate": 0.00017054539626176976,
+      "loss": 1.3152,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2733366971815757,
+      "grad_norm": 0.11068657966184081,
+      "learning_rate": 0.00017052884691175354,
+      "loss": 1.4107,
+      "step": 3792
+    },
+    {
+      "epoch": 0.27340877964391264,
+      "grad_norm": 0.12347419332272101,
+      "learning_rate": 0.0001705122937172842,
+      "loss": 1.3687,
+      "step": 3793
+    },
+    {
+      "epoch": 0.27348086210624956,
+      "grad_norm": 0.1154786963675164,
+      "learning_rate": 0.00017049573667926398,
+      "loss": 1.233,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2735529445685865,
+      "grad_norm": 0.1055171177817803,
+      "learning_rate": 0.00017047917579859548,
+      "loss": 1.4032,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2736250270309234,
+      "grad_norm": 0.10825169925059876,
+      "learning_rate": 0.00017046261107618132,
+      "loss": 1.4264,
+      "step": 3796
+    },
+    {
+      "epoch": 0.2736971094932603,
+      "grad_norm": 0.1172498805645455,
+      "learning_rate": 0.00017044604251292454,
+      "loss": 1.3596,
+      "step": 3797
+    },
+    {
+      "epoch": 0.2737691919555972,
+      "grad_norm": 0.27944158920345213,
+      "learning_rate": 0.00017042947010972818,
+      "loss": 1.2959,
+      "step": 3798
+    },
+    {
+      "epoch": 0.2738412744179341,
+      "grad_norm": 0.13117201492139047,
+      "learning_rate": 0.00017041289386749564,
+      "loss": 1.2707,
+      "step": 3799
+    },
+    {
+      "epoch": 0.273913356880271,
+      "grad_norm": 0.12985721795097085,
+      "learning_rate": 0.00017039631378713046,
+      "loss": 1.1487,
+      "step": 3800
+    },
+    {
+      "epoch": 0.27398543934260794,
+      "grad_norm": 0.10640236814456334,
+      "learning_rate": 0.00017037972986953635,
+      "loss": 1.3207,
+      "step": 3801
+    },
+    {
+      "epoch": 0.27405752180494486,
+      "grad_norm": 0.10045177449128967,
+      "learning_rate": 0.00017036314211561738,
+      "loss": 1.204,
+      "step": 3802
+    },
+    {
+      "epoch": 0.2741296042672818,
+      "grad_norm": 0.11234854755435285,
+      "learning_rate": 0.00017034655052627765,
+      "loss": 1.3268,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2742016867296187,
+      "grad_norm": 0.10896113803444366,
+      "learning_rate": 0.0001703299551024216,
+      "loss": 1.3739,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2742737691919556,
+      "grad_norm": 0.1504535098793347,
+      "learning_rate": 0.00017031335584495377,
+      "loss": 1.3964,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2743458516542925,
+      "grad_norm": 0.1316319437899555,
+      "learning_rate": 0.00017029675275477904,
+      "loss": 1.3792,
+      "step": 3806
+    },
+    {
+      "epoch": 0.2744179341166294,
+      "grad_norm": 0.11811962937604985,
+      "learning_rate": 0.00017028014583280235,
+      "loss": 1.2764,
+      "step": 3807
+    },
+    {
+      "epoch": 0.2744900165789663,
+      "grad_norm": 0.11829700430423934,
+      "learning_rate": 0.000170263535079929,
+      "loss": 1.3438,
+      "step": 3808
+    },
+    {
+      "epoch": 0.27456209904130324,
+      "grad_norm": 0.11667597052056378,
+      "learning_rate": 0.00017024692049706438,
+      "loss": 1.3318,
+      "step": 3809
+    },
+    {
+      "epoch": 0.27463418150364016,
+      "grad_norm": 0.1123903937625842,
+      "learning_rate": 0.0001702303020851142,
+      "loss": 1.2185,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2747062639659771,
+      "grad_norm": 0.09616210486285312,
+      "learning_rate": 0.0001702136798449842,
+      "loss": 1.3211,
+      "step": 3811
+    },
+    {
+      "epoch": 0.274778346428314,
+      "grad_norm": 0.11192552506402272,
+      "learning_rate": 0.0001701970537775805,
+      "loss": 1.1697,
+      "step": 3812
+    },
+    {
+      "epoch": 0.2748504288906509,
+      "grad_norm": 0.11461103040102498,
+      "learning_rate": 0.00017018042388380937,
+      "loss": 1.315,
+      "step": 3813
+    },
+    {
+      "epoch": 0.27492251135298784,
+      "grad_norm": 0.2348302593342839,
+      "learning_rate": 0.0001701637901645773,
+      "loss": 1.2301,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2749945938153247,
+      "grad_norm": 0.11576725158293624,
+      "learning_rate": 0.00017014715262079095,
+      "loss": 1.3139,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2750666762776616,
+      "grad_norm": 0.12040811994578006,
+      "learning_rate": 0.00017013051125335725,
+      "loss": 1.2442,
+      "step": 3816
+    },
+    {
+      "epoch": 0.27513875873999855,
+      "grad_norm": 0.10893825660979743,
+      "learning_rate": 0.00017011386606318328,
+      "loss": 1.2593,
+      "step": 3817
+    },
+    {
+      "epoch": 0.27521084120233547,
+      "grad_norm": 0.11730825181232456,
+      "learning_rate": 0.00017009721705117632,
+      "loss": 1.3547,
+      "step": 3818
+    },
+    {
+      "epoch": 0.2752829236646724,
+      "grad_norm": 0.12927722875630213,
+      "learning_rate": 0.00017008056421824396,
+      "loss": 1.3122,
+      "step": 3819
+    },
+    {
+      "epoch": 0.2753550061270093,
+      "grad_norm": 0.11126578871529161,
+      "learning_rate": 0.00017006390756529383,
+      "loss": 1.4578,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2754270885893462,
+      "grad_norm": 0.10471444213095862,
+      "learning_rate": 0.000170047247093234,
+      "loss": 1.2691,
+      "step": 3821
+    },
+    {
+      "epoch": 0.27549917105168314,
+      "grad_norm": 0.10490515702240213,
+      "learning_rate": 0.0001700305828029725,
+      "loss": 1.238,
+      "step": 3822
+    },
+    {
+      "epoch": 0.27557125351402006,
+      "grad_norm": 0.13252839861534457,
+      "learning_rate": 0.00017001391469541777,
+      "loss": 1.2626,
+      "step": 3823
+    },
+    {
+      "epoch": 0.27564333597635693,
+      "grad_norm": 0.10481435809984038,
+      "learning_rate": 0.0001699972427714783,
+      "loss": 1.3448,
+      "step": 3824
+    },
+    {
+      "epoch": 0.27571541843869385,
+      "grad_norm": 0.13460645940044694,
+      "learning_rate": 0.00016998056703206287,
+      "loss": 1.2125,
+      "step": 3825
+    },
+    {
+      "epoch": 0.27578750090103077,
+      "grad_norm": 0.13098229875897704,
+      "learning_rate": 0.0001699638874780805,
+      "loss": 1.2966,
+      "step": 3826
+    },
+    {
+      "epoch": 0.2758595833633677,
+      "grad_norm": 0.10314988628998738,
+      "learning_rate": 0.00016994720411044032,
+      "loss": 1.2853,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2759316658257046,
+      "grad_norm": 0.10803878972455869,
+      "learning_rate": 0.00016993051693005176,
+      "loss": 1.3345,
+      "step": 3828
+    },
+    {
+      "epoch": 0.27600374828804153,
+      "grad_norm": 0.11314534097406477,
+      "learning_rate": 0.00016991382593782442,
+      "loss": 1.3028,
+      "step": 3829
+    },
+    {
+      "epoch": 0.27607583075037845,
+      "grad_norm": 0.10844377460885686,
+      "learning_rate": 0.0001698971311346681,
+      "loss": 1.2736,
+      "step": 3830
+    },
+    {
+      "epoch": 0.27614791321271537,
+      "grad_norm": 0.113969962427997,
+      "learning_rate": 0.0001698804325214928,
+      "loss": 1.2873,
+      "step": 3831
+    },
+    {
+      "epoch": 0.27621999567505223,
+      "grad_norm": 0.10501448761690609,
+      "learning_rate": 0.0001698637300992088,
+      "loss": 1.3198,
+      "step": 3832
+    },
+    {
+      "epoch": 0.27629207813738915,
+      "grad_norm": 0.11190285512188876,
+      "learning_rate": 0.00016984702386872643,
+      "loss": 1.2894,
+      "step": 3833
+    },
+    {
+      "epoch": 0.27636416059972607,
+      "grad_norm": 0.12808507791060358,
+      "learning_rate": 0.00016983031383095644,
+      "loss": 1.2447,
+      "step": 3834
+    },
+    {
+      "epoch": 0.276436243062063,
+      "grad_norm": 0.1255761220918572,
+      "learning_rate": 0.00016981359998680958,
+      "loss": 1.3042,
+      "step": 3835
+    },
+    {
+      "epoch": 0.2765083255243999,
+      "grad_norm": 0.11095556490497115,
+      "learning_rate": 0.00016979688233719698,
+      "loss": 1.2141,
+      "step": 3836
+    },
+    {
+      "epoch": 0.27658040798673683,
+      "grad_norm": 0.11808474973709578,
+      "learning_rate": 0.00016978016088302987,
+      "loss": 1.422,
+      "step": 3837
+    },
+    {
+      "epoch": 0.27665249044907375,
+      "grad_norm": 0.11049241802823402,
+      "learning_rate": 0.00016976343562521976,
+      "loss": 1.2567,
+      "step": 3838
+    },
+    {
+      "epoch": 0.27672457291141067,
+      "grad_norm": 0.10408017035137442,
+      "learning_rate": 0.00016974670656467824,
+      "loss": 1.2765,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2767966553737476,
+      "grad_norm": 0.1369701065478582,
+      "learning_rate": 0.00016972997370231724,
+      "loss": 1.3333,
+      "step": 3840
+    },
+    {
+      "epoch": 0.27686873783608446,
+      "grad_norm": 0.11007133532377152,
+      "learning_rate": 0.00016971323703904884,
+      "loss": 1.3017,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2769408202984214,
+      "grad_norm": 0.11400152814135686,
+      "learning_rate": 0.00016969649657578537,
+      "loss": 1.2645,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2770129027607583,
+      "grad_norm": 0.11511563341817517,
+      "learning_rate": 0.00016967975231343933,
+      "loss": 1.2665,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2770849852230952,
+      "grad_norm": 0.10542028513597233,
+      "learning_rate": 0.0001696630042529234,
+      "loss": 1.3776,
+      "step": 3844
+    },
+    {
+      "epoch": 0.27715706768543213,
+      "grad_norm": 0.11701179823191174,
+      "learning_rate": 0.0001696462523951505,
+      "loss": 1.38,
+      "step": 3845
+    },
+    {
+      "epoch": 0.27722915014776905,
+      "grad_norm": 0.11295439710724699,
+      "learning_rate": 0.00016962949674103378,
+      "loss": 1.4624,
+      "step": 3846
+    },
+    {
+      "epoch": 0.277301232610106,
+      "grad_norm": 0.0952837293413391,
+      "learning_rate": 0.0001696127372914866,
+      "loss": 1.3368,
+      "step": 3847
+    },
+    {
+      "epoch": 0.2773733150724429,
+      "grad_norm": 0.11379024937193541,
+      "learning_rate": 0.0001695959740474224,
+      "loss": 1.3998,
+      "step": 3848
+    },
+    {
+      "epoch": 0.2774453975347798,
+      "grad_norm": 0.11638079104344907,
+      "learning_rate": 0.00016957920700975504,
+      "loss": 1.282,
+      "step": 3849
+    },
+    {
+      "epoch": 0.2775174799971167,
+      "grad_norm": 0.15176648893453093,
+      "learning_rate": 0.00016956243617939843,
+      "loss": 1.2771,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2775895624594536,
+      "grad_norm": 0.30354179210336757,
+      "learning_rate": 0.00016954566155726668,
+      "loss": 1.2595,
+      "step": 3851
+    },
+    {
+      "epoch": 0.2776616449217905,
+      "grad_norm": 0.09808238519229787,
+      "learning_rate": 0.00016952888314427422,
+      "loss": 1.3116,
+      "step": 3852
+    },
+    {
+      "epoch": 0.27773372738412744,
+      "grad_norm": 0.11134953753108931,
+      "learning_rate": 0.0001695121009413356,
+      "loss": 1.2686,
+      "step": 3853
+    },
+    {
+      "epoch": 0.27780580984646436,
+      "grad_norm": 0.10076307686982279,
+      "learning_rate": 0.0001694953149493656,
+      "loss": 1.2684,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2778778923088013,
+      "grad_norm": 0.11082670743739663,
+      "learning_rate": 0.00016947852516927925,
+      "loss": 1.3783,
+      "step": 3855
+    },
+    {
+      "epoch": 0.2779499747711382,
+      "grad_norm": 0.10866033428792224,
+      "learning_rate": 0.0001694617316019917,
+      "loss": 1.306,
+      "step": 3856
+    },
+    {
+      "epoch": 0.2780220572334751,
+      "grad_norm": 0.18175745497971596,
+      "learning_rate": 0.00016944493424841833,
+      "loss": 1.1739,
+      "step": 3857
+    },
+    {
+      "epoch": 0.278094139695812,
+      "grad_norm": 0.11182762994648894,
+      "learning_rate": 0.00016942813310947478,
+      "loss": 1.324,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2781662221581489,
+      "grad_norm": 0.11511929061568689,
+      "learning_rate": 0.00016941132818607685,
+      "loss": 1.2109,
+      "step": 3859
+    },
+    {
+      "epoch": 0.2782383046204858,
+      "grad_norm": 0.1163817218000834,
+      "learning_rate": 0.00016939451947914057,
+      "loss": 1.4046,
+      "step": 3860
+    },
+    {
+      "epoch": 0.27831038708282274,
+      "grad_norm": 0.11447837955719778,
+      "learning_rate": 0.00016937770698958215,
+      "loss": 1.2991,
+      "step": 3861
+    },
+    {
+      "epoch": 0.27838246954515966,
+      "grad_norm": 0.11410956487474831,
+      "learning_rate": 0.00016936089071831805,
+      "loss": 1.3378,
+      "step": 3862
+    },
+    {
+      "epoch": 0.2784545520074966,
+      "grad_norm": 0.1057920152558361,
+      "learning_rate": 0.00016934407066626484,
+      "loss": 1.3107,
+      "step": 3863
+    },
+    {
+      "epoch": 0.2785266344698335,
+      "grad_norm": 0.17054859602958009,
+      "learning_rate": 0.00016932724683433944,
+      "loss": 1.2459,
+      "step": 3864
+    },
+    {
+      "epoch": 0.2785987169321704,
+      "grad_norm": 0.10339527143841787,
+      "learning_rate": 0.00016931041922345886,
+      "loss": 1.2929,
+      "step": 3865
+    },
+    {
+      "epoch": 0.27867079939450734,
+      "grad_norm": 0.106628006897469,
+      "learning_rate": 0.0001692935878345404,
+      "loss": 1.3395,
+      "step": 3866
+    },
+    {
+      "epoch": 0.2787428818568442,
+      "grad_norm": 0.11165610217547738,
+      "learning_rate": 0.00016927675266850141,
+      "loss": 1.245,
+      "step": 3867
+    },
+    {
+      "epoch": 0.2788149643191811,
+      "grad_norm": 0.11405745552241221,
+      "learning_rate": 0.0001692599137262597,
+      "loss": 1.37,
+      "step": 3868
+    },
+    {
+      "epoch": 0.27888704678151804,
+      "grad_norm": 0.1065902155944277,
+      "learning_rate": 0.00016924307100873307,
+      "loss": 1.2845,
+      "step": 3869
+    },
+    {
+      "epoch": 0.27895912924385496,
+      "grad_norm": 0.11068631096475179,
+      "learning_rate": 0.00016922622451683958,
+      "loss": 1.3323,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2790312117061919,
+      "grad_norm": 0.10817911767084809,
+      "learning_rate": 0.00016920937425149752,
+      "loss": 1.3618,
+      "step": 3871
+    },
+    {
+      "epoch": 0.2791032941685288,
+      "grad_norm": 0.11834201588648856,
+      "learning_rate": 0.00016919252021362543,
+      "loss": 1.3006,
+      "step": 3872
+    },
+    {
+      "epoch": 0.2791753766308657,
+      "grad_norm": 1.2482267353213747,
+      "learning_rate": 0.000169175662404142,
+      "loss": 1.2952,
+      "step": 3873
+    },
+    {
+      "epoch": 0.27924745909320264,
+      "grad_norm": 0.12194339977190952,
+      "learning_rate": 0.00016915880082396604,
+      "loss": 1.302,
+      "step": 3874
+    },
+    {
+      "epoch": 0.27931954155553956,
+      "grad_norm": 0.1172658312819781,
+      "learning_rate": 0.00016914193547401676,
+      "loss": 1.4024,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2793916240178764,
+      "grad_norm": 0.11212361004002874,
+      "learning_rate": 0.00016912506635521343,
+      "loss": 1.3983,
+      "step": 3876
+    },
+    {
+      "epoch": 0.27946370648021335,
+      "grad_norm": 0.1079858508520899,
+      "learning_rate": 0.00016910819346847554,
+      "loss": 1.2503,
+      "step": 3877
+    },
+    {
+      "epoch": 0.27953578894255027,
+      "grad_norm": 0.11165362842696805,
+      "learning_rate": 0.00016909131681472293,
+      "loss": 1.2272,
+      "step": 3878
+    },
+    {
+      "epoch": 0.2796078714048872,
+      "grad_norm": 0.1102266497912813,
+      "learning_rate": 0.00016907443639487536,
+      "loss": 1.2987,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2796799538672241,
+      "grad_norm": 0.09232399406358258,
+      "learning_rate": 0.0001690575522098531,
+      "loss": 1.3235,
+      "step": 3880
+    },
+    {
+      "epoch": 0.279752036329561,
+      "grad_norm": 0.10661315170826144,
+      "learning_rate": 0.00016904066426057644,
+      "loss": 1.2697,
+      "step": 3881
+    },
+    {
+      "epoch": 0.27982411879189795,
+      "grad_norm": 0.11497744028746935,
+      "learning_rate": 0.0001690237725479659,
+      "loss": 1.2248,
+      "step": 3882
+    },
+    {
+      "epoch": 0.27989620125423487,
+      "grad_norm": 0.117823058836885,
+      "learning_rate": 0.00016900687707294224,
+      "loss": 1.3403,
+      "step": 3883
+    },
+    {
+      "epoch": 0.27996828371657173,
+      "grad_norm": 0.10338035004919939,
+      "learning_rate": 0.00016898997783642642,
+      "loss": 1.3288,
+      "step": 3884
+    },
+    {
+      "epoch": 0.28004036617890865,
+      "grad_norm": 0.10871263343312251,
+      "learning_rate": 0.00016897307483933966,
+      "loss": 1.2953,
+      "step": 3885
+    },
+    {
+      "epoch": 0.28011244864124557,
+      "grad_norm": 0.16574501756673604,
+      "learning_rate": 0.00016895616808260323,
+      "loss": 1.3354,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2801845311035825,
+      "grad_norm": 0.4414399254587264,
+      "learning_rate": 0.00016893925756713874,
+      "loss": 1.3623,
+      "step": 3887
+    },
+    {
+      "epoch": 0.2802566135659194,
+      "grad_norm": 0.11270912066378613,
+      "learning_rate": 0.00016892234329386798,
+      "loss": 1.3487,
+      "step": 3888
+    },
+    {
+      "epoch": 0.28032869602825633,
+      "grad_norm": 0.11297762048133887,
+      "learning_rate": 0.00016890542526371293,
+      "loss": 1.3323,
+      "step": 3889
+    },
+    {
+      "epoch": 0.28040077849059325,
+      "grad_norm": 0.10959432894355688,
+      "learning_rate": 0.0001688885034775957,
+      "loss": 1.3597,
+      "step": 3890
+    },
+    {
+      "epoch": 0.28047286095293017,
+      "grad_norm": 0.13329618853485772,
+      "learning_rate": 0.00016887157793643877,
+      "loss": 1.3498,
+      "step": 3891
+    },
+    {
+      "epoch": 0.2805449434152671,
+      "grad_norm": 0.13790168706609449,
+      "learning_rate": 0.00016885464864116473,
+      "loss": 1.2562,
+      "step": 3892
+    },
+    {
+      "epoch": 0.28061702587760395,
+      "grad_norm": 0.10195641490161246,
+      "learning_rate": 0.0001688377155926963,
+      "loss": 1.3391,
+      "step": 3893
+    },
+    {
+      "epoch": 0.2806891083399409,
+      "grad_norm": 0.10420857003580786,
+      "learning_rate": 0.00016882077879195654,
+      "loss": 1.3443,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2807611908022778,
+      "grad_norm": 0.11380506045313353,
+      "learning_rate": 0.00016880383823986866,
+      "loss": 1.1995,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2808332732646147,
+      "grad_norm": 0.11634019755460111,
+      "learning_rate": 0.00016878689393735606,
+      "loss": 1.3028,
+      "step": 3896
+    },
+    {
+      "epoch": 0.28090535572695163,
+      "grad_norm": 0.10696739171586139,
+      "learning_rate": 0.00016876994588534234,
+      "loss": 1.3505,
+      "step": 3897
+    },
+    {
+      "epoch": 0.28097743818928855,
+      "grad_norm": 0.14971090254541802,
+      "learning_rate": 0.00016875299408475132,
+      "loss": 1.2724,
+      "step": 3898
+    },
+    {
+      "epoch": 0.2810495206516255,
+      "grad_norm": 0.27302556495458985,
+      "learning_rate": 0.0001687360385365071,
+      "loss": 1.3271,
+      "step": 3899
+    },
+    {
+      "epoch": 0.2811216031139624,
+      "grad_norm": 0.1268652177103704,
+      "learning_rate": 0.0001687190792415338,
+      "loss": 1.3696,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2811936855762993,
+      "grad_norm": 0.16132456305015158,
+      "learning_rate": 0.0001687021162007559,
+      "loss": 1.3389,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2812657680386362,
+      "grad_norm": 0.2546325042310612,
+      "learning_rate": 0.00016868514941509807,
+      "loss": 1.3109,
+      "step": 3902
+    },
+    {
+      "epoch": 0.2813378505009731,
+      "grad_norm": 0.12085249320073989,
+      "learning_rate": 0.00016866817888548514,
+      "loss": 1.2912,
+      "step": 3903
+    },
+    {
+      "epoch": 0.28140993296331,
+      "grad_norm": 0.12059004850003119,
+      "learning_rate": 0.00016865120461284207,
+      "loss": 1.3464,
+      "step": 3904
+    },
+    {
+      "epoch": 0.28148201542564694,
+      "grad_norm": 0.1236003736498376,
+      "learning_rate": 0.00016863422659809424,
+      "loss": 1.3554,
+      "step": 3905
+    },
+    {
+      "epoch": 0.28155409788798386,
+      "grad_norm": 0.106570361626139,
+      "learning_rate": 0.00016861724484216702,
+      "loss": 1.2972,
+      "step": 3906
+    },
+    {
+      "epoch": 0.2816261803503208,
+      "grad_norm": 0.10497937945482176,
+      "learning_rate": 0.0001686002593459861,
+      "loss": 1.3421,
+      "step": 3907
+    },
+    {
+      "epoch": 0.2816982628126577,
+      "grad_norm": 0.10490144765834508,
+      "learning_rate": 0.0001685832701104773,
+      "loss": 1.2204,
+      "step": 3908
+    },
+    {
+      "epoch": 0.2817703452749946,
+      "grad_norm": 0.1255923265625201,
+      "learning_rate": 0.00016856627713656672,
+      "loss": 1.4056,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2818424277373315,
+      "grad_norm": 0.13145625989022267,
+      "learning_rate": 0.00016854928042518062,
+      "loss": 1.3814,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2819145101996684,
+      "grad_norm": 0.11085555031967695,
+      "learning_rate": 0.00016853227997724554,
+      "loss": 1.2944,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2819865926620053,
+      "grad_norm": 0.12040238056249919,
+      "learning_rate": 0.000168515275793688,
+      "loss": 1.3358,
+      "step": 3912
+    },
+    {
+      "epoch": 0.28205867512434224,
+      "grad_norm": 0.10851427000960906,
+      "learning_rate": 0.00016849826787543502,
+      "loss": 1.2493,
+      "step": 3913
+    },
+    {
+      "epoch": 0.28213075758667916,
+      "grad_norm": 0.11247936021426684,
+      "learning_rate": 0.00016848125622341366,
+      "loss": 1.2916,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2822028400490161,
+      "grad_norm": 0.11035642952571653,
+      "learning_rate": 0.00016846424083855117,
+      "loss": 1.3532,
+      "step": 3915
+    },
+    {
+      "epoch": 0.282274922511353,
+      "grad_norm": 0.1009413429967976,
+      "learning_rate": 0.00016844722172177506,
+      "loss": 1.2461,
+      "step": 3916
+    },
+    {
+      "epoch": 0.2823470049736899,
+      "grad_norm": 0.11104922889723445,
+      "learning_rate": 0.000168430198874013,
+      "loss": 1.3672,
+      "step": 3917
+    },
+    {
+      "epoch": 0.28241908743602684,
+      "grad_norm": 0.1188835924040387,
+      "learning_rate": 0.00016841317229619293,
+      "loss": 1.3758,
+      "step": 3918
+    },
+    {
+      "epoch": 0.2824911698983637,
+      "grad_norm": 0.11169020250883427,
+      "learning_rate": 0.0001683961419892429,
+      "loss": 1.294,
+      "step": 3919
+    },
+    {
+      "epoch": 0.2825632523607006,
+      "grad_norm": 0.10808158181503573,
+      "learning_rate": 0.00016837910795409127,
+      "loss": 1.4369,
+      "step": 3920
+    },
+    {
+      "epoch": 0.28263533482303754,
+      "grad_norm": 0.1259121651661248,
+      "learning_rate": 0.00016836207019166647,
+      "loss": 1.369,
+      "step": 3921
+    },
+    {
+      "epoch": 0.28270741728537446,
+      "grad_norm": 0.12772102906993454,
+      "learning_rate": 0.00016834502870289732,
+      "loss": 1.3197,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2827794997477114,
+      "grad_norm": 0.1107957398418312,
+      "learning_rate": 0.00016832798348871269,
+      "loss": 1.2605,
+      "step": 3923
+    },
+    {
+      "epoch": 0.2828515822100483,
+      "grad_norm": 0.10921694090110338,
+      "learning_rate": 0.00016831093455004162,
+      "loss": 1.2935,
+      "step": 3924
+    },
+    {
+      "epoch": 0.2829236646723852,
+      "grad_norm": 0.10799906954986332,
+      "learning_rate": 0.00016829388188781353,
+      "loss": 1.3655,
+      "step": 3925
+    },
+    {
+      "epoch": 0.28299574713472214,
+      "grad_norm": 0.18886894321306857,
+      "learning_rate": 0.00016827682550295785,
+      "loss": 1.2749,
+      "step": 3926
+    },
+    {
+      "epoch": 0.28306782959705906,
+      "grad_norm": 0.10415470305630223,
+      "learning_rate": 0.0001682597653964044,
+      "loss": 1.339,
+      "step": 3927
+    },
+    {
+      "epoch": 0.2831399120593959,
+      "grad_norm": 0.11355547286229196,
+      "learning_rate": 0.00016824270156908305,
+      "loss": 1.2467,
+      "step": 3928
+    },
+    {
+      "epoch": 0.28321199452173285,
+      "grad_norm": 0.12057440273567223,
+      "learning_rate": 0.00016822563402192397,
+      "loss": 1.2402,
+      "step": 3929
+    },
+    {
+      "epoch": 0.28328407698406977,
+      "grad_norm": 0.10748811107741504,
+      "learning_rate": 0.00016820856275585744,
+      "loss": 1.3374,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2833561594464067,
+      "grad_norm": 0.10371545262952318,
+      "learning_rate": 0.00016819148777181406,
+      "loss": 1.2402,
+      "step": 3931
+    },
+    {
+      "epoch": 0.2834282419087436,
+      "grad_norm": 0.3596316782850399,
+      "learning_rate": 0.0001681744090707245,
+      "loss": 1.223,
+      "step": 3932
+    },
+    {
+      "epoch": 0.2835003243710805,
+      "grad_norm": 0.11148719342865891,
+      "learning_rate": 0.00016815732665351977,
+      "loss": 1.3818,
+      "step": 3933
+    },
+    {
+      "epoch": 0.28357240683341745,
+      "grad_norm": 0.10986619677249539,
+      "learning_rate": 0.000168140240521131,
+      "loss": 1.4448,
+      "step": 3934
+    },
+    {
+      "epoch": 0.28364448929575437,
+      "grad_norm": 0.11539879298326243,
+      "learning_rate": 0.0001681231506744895,
+      "loss": 1.4036,
+      "step": 3935
+    },
+    {
+      "epoch": 0.28371657175809123,
+      "grad_norm": 0.10456727380036684,
+      "learning_rate": 0.00016810605711452685,
+      "loss": 1.3471,
+      "step": 3936
+    },
+    {
+      "epoch": 0.28378865422042815,
+      "grad_norm": 0.10808929167218004,
+      "learning_rate": 0.00016808895984217476,
+      "loss": 1.3577,
+      "step": 3937
+    },
+    {
+      "epoch": 0.28386073668276507,
+      "grad_norm": 0.11026754747913962,
+      "learning_rate": 0.00016807185885836526,
+      "loss": 1.3374,
+      "step": 3938
+    },
+    {
+      "epoch": 0.283932819145102,
+      "grad_norm": 0.09908214417728915,
+      "learning_rate": 0.00016805475416403046,
+      "loss": 1.2576,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2840049016074389,
+      "grad_norm": 0.12027039806319773,
+      "learning_rate": 0.00016803764576010268,
+      "loss": 1.3944,
+      "step": 3940
+    },
+    {
+      "epoch": 0.28407698406977583,
+      "grad_norm": 0.12137059618981706,
+      "learning_rate": 0.00016802053364751456,
+      "loss": 1.3191,
+      "step": 3941
+    },
+    {
+      "epoch": 0.28414906653211275,
+      "grad_norm": 0.10620103502316589,
+      "learning_rate": 0.00016800341782719882,
+      "loss": 1.3604,
+      "step": 3942
+    },
+    {
+      "epoch": 0.28422114899444967,
+      "grad_norm": 0.1362978895214766,
+      "learning_rate": 0.00016798629830008842,
+      "loss": 1.2236,
+      "step": 3943
+    },
+    {
+      "epoch": 0.2842932314567866,
+      "grad_norm": 0.13918548549357948,
+      "learning_rate": 0.00016796917506711656,
+      "loss": 1.2932,
+      "step": 3944
+    },
+    {
+      "epoch": 0.28436531391912345,
+      "grad_norm": 0.10352111740802135,
+      "learning_rate": 0.0001679520481292166,
+      "loss": 1.3348,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2844373963814604,
+      "grad_norm": 0.12312511876498573,
+      "learning_rate": 0.00016793491748732207,
+      "loss": 1.2414,
+      "step": 3946
+    },
+    {
+      "epoch": 0.2845094788437973,
+      "grad_norm": 0.10492848696971059,
+      "learning_rate": 0.0001679177831423668,
+      "loss": 1.329,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2845815613061342,
+      "grad_norm": 0.09764375230580725,
+      "learning_rate": 0.0001679006450952847,
+      "loss": 1.3363,
+      "step": 3948
+    },
+    {
+      "epoch": 0.28465364376847113,
+      "grad_norm": 0.23883955449074037,
+      "learning_rate": 0.00016788350334701,
+      "loss": 1.2991,
+      "step": 3949
+    },
+    {
+      "epoch": 0.28472572623080805,
+      "grad_norm": 0.10016103998845384,
+      "learning_rate": 0.0001678663578984771,
+      "loss": 1.2934,
+      "step": 3950
+    },
+    {
+      "epoch": 0.28479780869314497,
+      "grad_norm": 0.10650600459361854,
+      "learning_rate": 0.00016784920875062052,
+      "loss": 1.2492,
+      "step": 3951
+    },
+    {
+      "epoch": 0.2848698911554819,
+      "grad_norm": 0.11262859138037272,
+      "learning_rate": 0.00016783205590437504,
+      "loss": 1.3337,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2849419736178188,
+      "grad_norm": 0.09970743670303822,
+      "learning_rate": 0.0001678148993606757,
+      "loss": 1.2061,
+      "step": 3953
+    },
+    {
+      "epoch": 0.2850140560801557,
+      "grad_norm": 0.11678246338876096,
+      "learning_rate": 0.00016779773912045767,
+      "loss": 1.2676,
+      "step": 3954
+    },
+    {
+      "epoch": 0.2850861385424926,
+      "grad_norm": 0.10894710317961788,
+      "learning_rate": 0.0001677805751846563,
+      "loss": 1.2386,
+      "step": 3955
+    },
+    {
+      "epoch": 0.2851582210048295,
+      "grad_norm": 0.16358174382462223,
+      "learning_rate": 0.0001677634075542072,
+      "loss": 1.337,
+      "step": 3956
+    },
+    {
+      "epoch": 0.28523030346716644,
+      "grad_norm": 0.11938688236790596,
+      "learning_rate": 0.00016774623623004616,
+      "loss": 1.2426,
+      "step": 3957
+    },
+    {
+      "epoch": 0.28530238592950335,
+      "grad_norm": 0.12458396154984598,
+      "learning_rate": 0.00016772906121310919,
+      "loss": 1.2484,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2853744683918403,
+      "grad_norm": 0.09775307646327937,
+      "learning_rate": 0.00016771188250433246,
+      "loss": 1.3301,
+      "step": 3959
+    },
+    {
+      "epoch": 0.2854465508541772,
+      "grad_norm": 0.10908875967709097,
+      "learning_rate": 0.00016769470010465234,
+      "loss": 1.4362,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2855186333165141,
+      "grad_norm": 0.11622802440579161,
+      "learning_rate": 0.00016767751401500545,
+      "loss": 1.3105,
+      "step": 3961
+    },
+    {
+      "epoch": 0.285590715778851,
+      "grad_norm": 0.11771268834471309,
+      "learning_rate": 0.0001676603242363286,
+      "loss": 1.2987,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2856627982411879,
+      "grad_norm": 0.14784180511198122,
+      "learning_rate": 0.00016764313076955876,
+      "loss": 1.2569,
+      "step": 3963
+    },
+    {
+      "epoch": 0.2857348807035248,
+      "grad_norm": 0.1367369374929619,
+      "learning_rate": 0.00016762593361563318,
+      "loss": 1.2978,
+      "step": 3964
+    },
+    {
+      "epoch": 0.28580696316586174,
+      "grad_norm": 0.11083420948174681,
+      "learning_rate": 0.00016760873277548916,
+      "loss": 1.3296,
+      "step": 3965
+    },
+    {
+      "epoch": 0.28587904562819866,
+      "grad_norm": 0.1145069595572024,
+      "learning_rate": 0.0001675915282500644,
+      "loss": 1.3262,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2859511280905356,
+      "grad_norm": 0.10332253146915102,
+      "learning_rate": 0.00016757432004029665,
+      "loss": 1.3665,
+      "step": 3967
+    },
+    {
+      "epoch": 0.2860232105528725,
+      "grad_norm": 0.11897601861304277,
+      "learning_rate": 0.00016755710814712394,
+      "loss": 1.3427,
+      "step": 3968
+    },
+    {
+      "epoch": 0.2860952930152094,
+      "grad_norm": 0.10810268413343334,
+      "learning_rate": 0.00016753989257148438,
+      "loss": 1.268,
+      "step": 3969
+    },
+    {
+      "epoch": 0.28616737547754634,
+      "grad_norm": 0.09979071878532397,
+      "learning_rate": 0.00016752267331431652,
+      "loss": 1.3503,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2862394579398832,
+      "grad_norm": 0.10470227037304615,
+      "learning_rate": 0.00016750545037655886,
+      "loss": 1.3952,
+      "step": 3971
+    },
+    {
+      "epoch": 0.2863115404022201,
+      "grad_norm": 0.10358800562107595,
+      "learning_rate": 0.00016748822375915027,
+      "loss": 1.2485,
+      "step": 3972
+    },
+    {
+      "epoch": 0.28638362286455704,
+      "grad_norm": 0.10178117195522864,
+      "learning_rate": 0.0001674709934630297,
+      "loss": 1.2853,
+      "step": 3973
+    },
+    {
+      "epoch": 0.28645570532689396,
+      "grad_norm": 0.10558339886684635,
+      "learning_rate": 0.00016745375948913632,
+      "loss": 1.2856,
+      "step": 3974
+    },
+    {
+      "epoch": 0.2865277877892309,
+      "grad_norm": 0.11665364765503279,
+      "learning_rate": 0.00016743652183840964,
+      "loss": 1.3466,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2865998702515678,
+      "grad_norm": 0.10256494743611384,
+      "learning_rate": 0.0001674192805117892,
+      "loss": 1.3598,
+      "step": 3976
+    },
+    {
+      "epoch": 0.2866719527139047,
+      "grad_norm": 0.11254808935391726,
+      "learning_rate": 0.00016740203551021484,
+      "loss": 1.2626,
+      "step": 3977
+    },
+    {
+      "epoch": 0.28674403517624164,
+      "grad_norm": 0.11292262370891991,
+      "learning_rate": 0.00016738478683462656,
+      "loss": 1.297,
+      "step": 3978
+    },
+    {
+      "epoch": 0.28681611763857856,
+      "grad_norm": 0.10158276836355672,
+      "learning_rate": 0.00016736753448596452,
+      "loss": 1.3362,
+      "step": 3979
+    },
+    {
+      "epoch": 0.2868882001009154,
+      "grad_norm": 0.11117409495254728,
+      "learning_rate": 0.00016735027846516917,
+      "loss": 1.358,
+      "step": 3980
+    },
+    {
+      "epoch": 0.28696028256325234,
+      "grad_norm": 0.11736618281359196,
+      "learning_rate": 0.0001673330187731811,
+      "loss": 1.3476,
+      "step": 3981
+    },
+    {
+      "epoch": 0.28703236502558926,
+      "grad_norm": 0.10230011537381437,
+      "learning_rate": 0.00016731575541094115,
+      "loss": 1.311,
+      "step": 3982
+    },
+    {
+      "epoch": 0.2871044474879262,
+      "grad_norm": 0.10816562524918992,
+      "learning_rate": 0.0001672984883793903,
+      "loss": 1.2693,
+      "step": 3983
+    },
+    {
+      "epoch": 0.2871765299502631,
+      "grad_norm": 0.14749389220531012,
+      "learning_rate": 0.00016728121767946977,
+      "loss": 1.3453,
+      "step": 3984
+    },
+    {
+      "epoch": 0.2872486124126,
+      "grad_norm": 0.10689185664649459,
+      "learning_rate": 0.00016726394331212095,
+      "loss": 1.3339,
+      "step": 3985
+    },
+    {
+      "epoch": 0.28732069487493694,
+      "grad_norm": 0.11935334324419343,
+      "learning_rate": 0.00016724666527828547,
+      "loss": 1.2385,
+      "step": 3986
+    },
+    {
+      "epoch": 0.28739277733727386,
+      "grad_norm": 0.1193302838420886,
+      "learning_rate": 0.00016722938357890512,
+      "loss": 1.3948,
+      "step": 3987
+    },
+    {
+      "epoch": 0.28746485979961073,
+      "grad_norm": 0.10565780083368517,
+      "learning_rate": 0.00016721209821492192,
+      "loss": 1.2568,
+      "step": 3988
+    },
+    {
+      "epoch": 0.28753694226194765,
+      "grad_norm": 0.12530788567360857,
+      "learning_rate": 0.00016719480918727807,
+      "loss": 1.292,
+      "step": 3989
+    },
+    {
+      "epoch": 0.28760902472428457,
+      "grad_norm": 0.10876450933218053,
+      "learning_rate": 0.00016717751649691597,
+      "loss": 1.2907,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2876811071866215,
+      "grad_norm": 0.12546100252325557,
+      "learning_rate": 0.00016716022014477827,
+      "loss": 1.3034,
+      "step": 3991
+    },
+    {
+      "epoch": 0.2877531896489584,
+      "grad_norm": 0.10171350901267101,
+      "learning_rate": 0.00016714292013180767,
+      "loss": 1.2956,
+      "step": 3992
+    },
+    {
+      "epoch": 0.2878252721112953,
+      "grad_norm": 0.10238567635488828,
+      "learning_rate": 0.0001671256164589473,
+      "loss": 1.3667,
+      "step": 3993
+    },
+    {
+      "epoch": 0.28789735457363225,
+      "grad_norm": 0.17029823991878187,
+      "learning_rate": 0.0001671083091271403,
+      "loss": 1.3273,
+      "step": 3994
+    },
+    {
+      "epoch": 0.28796943703596917,
+      "grad_norm": 0.10416972556329833,
+      "learning_rate": 0.00016709099813733006,
+      "loss": 1.2945,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2880415194983061,
+      "grad_norm": 0.11631101525438248,
+      "learning_rate": 0.00016707368349046022,
+      "loss": 1.2409,
+      "step": 3996
+    },
+    {
+      "epoch": 0.28811360196064295,
+      "grad_norm": 0.107685265720983,
+      "learning_rate": 0.0001670563651874746,
+      "loss": 1.3343,
+      "step": 3997
+    },
+    {
+      "epoch": 0.28818568442297987,
+      "grad_norm": 0.11250816742793932,
+      "learning_rate": 0.00016703904322931714,
+      "loss": 1.3728,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2882577668853168,
+      "grad_norm": 0.10930685247470201,
+      "learning_rate": 0.00016702171761693208,
+      "loss": 1.2863,
+      "step": 3999
+    },
+    {
+      "epoch": 0.2883298493476537,
+      "grad_norm": 0.11478096359183969,
+      "learning_rate": 0.00016700438835126383,
+      "loss": 1.2325,
+      "step": 4000
+    },
+    {
+      "epoch": 0.28840193180999063,
+      "grad_norm": 0.10645421224154687,
+      "learning_rate": 0.00016698705543325698,
+      "loss": 1.2971,
+      "step": 4001
+    },
+    {
+      "epoch": 0.28847401427232755,
+      "grad_norm": 0.10167973594915307,
+      "learning_rate": 0.00016696971886385632,
+      "loss": 1.257,
+      "step": 4002
+    },
+    {
+      "epoch": 0.28854609673466447,
+      "grad_norm": 0.10119409947928626,
+      "learning_rate": 0.00016695237864400682,
+      "loss": 1.3337,
+      "step": 4003
+    },
+    {
+      "epoch": 0.2886181791970014,
+      "grad_norm": 0.11360446300742484,
+      "learning_rate": 0.00016693503477465375,
+      "loss": 1.3092,
+      "step": 4004
+    },
+    {
+      "epoch": 0.2886902616593383,
+      "grad_norm": 0.11087724000502813,
+      "learning_rate": 0.00016691768725674247,
+      "loss": 1.4078,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2887623441216752,
+      "grad_norm": 0.11141513582247438,
+      "learning_rate": 0.00016690033609121857,
+      "loss": 1.3314,
+      "step": 4006
+    },
+    {
+      "epoch": 0.2888344265840121,
+      "grad_norm": 0.09669714054794099,
+      "learning_rate": 0.0001668829812790278,
+      "loss": 1.3941,
+      "step": 4007
+    },
+    {
+      "epoch": 0.288906509046349,
+      "grad_norm": 0.09777617018557935,
+      "learning_rate": 0.00016686562282111625,
+      "loss": 1.3321,
+      "step": 4008
+    },
+    {
+      "epoch": 0.28897859150868593,
+      "grad_norm": 0.11356470837012213,
+      "learning_rate": 0.00016684826071843004,
+      "loss": 1.3987,
+      "step": 4009
+    },
+    {
+      "epoch": 0.28905067397102285,
+      "grad_norm": 0.12208103360709113,
+      "learning_rate": 0.0001668308949719156,
+      "loss": 1.3321,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2891227564333598,
+      "grad_norm": 0.118354561194082,
+      "learning_rate": 0.00016681352558251946,
+      "loss": 1.3047,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2891948388956967,
+      "grad_norm": 0.12762868209194508,
+      "learning_rate": 0.00016679615255118846,
+      "loss": 1.3271,
+      "step": 4012
+    },
+    {
+      "epoch": 0.2892669213580336,
+      "grad_norm": 0.10513407309859224,
+      "learning_rate": 0.00016677877587886956,
+      "loss": 1.3337,
+      "step": 4013
+    },
+    {
+      "epoch": 0.2893390038203705,
+      "grad_norm": 0.1309676951831124,
+      "learning_rate": 0.00016676139556651,
+      "loss": 1.2789,
+      "step": 4014
+    },
+    {
+      "epoch": 0.2894110862827074,
+      "grad_norm": 0.1005895252290724,
+      "learning_rate": 0.00016674401161505706,
+      "loss": 1.3732,
+      "step": 4015
+    },
+    {
+      "epoch": 0.2894831687450443,
+      "grad_norm": 0.11525160859499453,
+      "learning_rate": 0.0001667266240254584,
+      "loss": 1.3101,
+      "step": 4016
+    },
+    {
+      "epoch": 0.28955525120738124,
+      "grad_norm": 0.13179257848393558,
+      "learning_rate": 0.0001667092327986618,
+      "loss": 1.2659,
+      "step": 4017
+    },
+    {
+      "epoch": 0.28962733366971816,
+      "grad_norm": 0.11044895289781863,
+      "learning_rate": 0.0001666918379356152,
+      "loss": 1.2923,
+      "step": 4018
+    },
+    {
+      "epoch": 0.2896994161320551,
+      "grad_norm": 0.10399979628105055,
+      "learning_rate": 0.00016667443943726678,
+      "loss": 1.3266,
+      "step": 4019
+    },
+    {
+      "epoch": 0.289771498594392,
+      "grad_norm": 0.1200551025560587,
+      "learning_rate": 0.00016665703730456491,
+      "loss": 1.3187,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2898435810567289,
+      "grad_norm": 0.10741401246031242,
+      "learning_rate": 0.0001666396315384582,
+      "loss": 1.341,
+      "step": 4021
+    },
+    {
+      "epoch": 0.28991566351906584,
+      "grad_norm": 0.10328474886815829,
+      "learning_rate": 0.0001666222221398954,
+      "loss": 1.3088,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2899877459814027,
+      "grad_norm": 0.11882485316575395,
+      "learning_rate": 0.00016660480910982546,
+      "loss": 1.2056,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2900598284437396,
+      "grad_norm": 0.1060414928666547,
+      "learning_rate": 0.0001665873924491976,
+      "loss": 1.3232,
+      "step": 4024
+    },
+    {
+      "epoch": 0.29013191090607654,
+      "grad_norm": 0.12061997548084169,
+      "learning_rate": 0.0001665699721589611,
+      "loss": 1.2914,
+      "step": 4025
+    },
+    {
+      "epoch": 0.29020399336841346,
+      "grad_norm": 0.11518316109944268,
+      "learning_rate": 0.00016655254824006556,
+      "loss": 1.3721,
+      "step": 4026
+    },
+    {
+      "epoch": 0.2902760758307504,
+      "grad_norm": 0.09905845536343406,
+      "learning_rate": 0.00016653512069346075,
+      "loss": 1.3294,
+      "step": 4027
+    },
+    {
+      "epoch": 0.2903481582930873,
+      "grad_norm": 0.10562659191904532,
+      "learning_rate": 0.00016651768952009662,
+      "loss": 1.2238,
+      "step": 4028
+    },
+    {
+      "epoch": 0.2904202407554242,
+      "grad_norm": 0.11216474044317433,
+      "learning_rate": 0.00016650025472092334,
+      "loss": 1.3673,
+      "step": 4029
+    },
+    {
+      "epoch": 0.29049232321776114,
+      "grad_norm": 0.10968850565581946,
+      "learning_rate": 0.00016648281629689127,
+      "loss": 1.2991,
+      "step": 4030
+    },
+    {
+      "epoch": 0.29056440568009806,
+      "grad_norm": 0.11628131245091491,
+      "learning_rate": 0.0001664653742489509,
+      "loss": 1.3398,
+      "step": 4031
+    },
+    {
+      "epoch": 0.2906364881424349,
+      "grad_norm": 0.10685990682590348,
+      "learning_rate": 0.00016644792857805306,
+      "loss": 1.3111,
+      "step": 4032
+    },
+    {
+      "epoch": 0.29070857060477184,
+      "grad_norm": 0.12376847657026963,
+      "learning_rate": 0.00016643047928514861,
+      "loss": 1.2585,
+      "step": 4033
+    },
+    {
+      "epoch": 0.29078065306710876,
+      "grad_norm": 0.11796885543517598,
+      "learning_rate": 0.00016641302637118873,
+      "loss": 1.3296,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2908527355294457,
+      "grad_norm": 0.10611730558917472,
+      "learning_rate": 0.00016639556983712481,
+      "loss": 1.3185,
+      "step": 4035
+    },
+    {
+      "epoch": 0.2909248179917826,
+      "grad_norm": 0.21464409449553362,
+      "learning_rate": 0.00016637810968390832,
+      "loss": 1.2925,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2909969004541195,
+      "grad_norm": 0.11065929419753318,
+      "learning_rate": 0.000166360645912491,
+      "loss": 1.3087,
+      "step": 4037
+    },
+    {
+      "epoch": 0.29106898291645644,
+      "grad_norm": 0.2060971848046189,
+      "learning_rate": 0.00016634317852382482,
+      "loss": 1.2854,
+      "step": 4038
+    },
+    {
+      "epoch": 0.29114106537879336,
+      "grad_norm": 0.11109943020848205,
+      "learning_rate": 0.00016632570751886187,
+      "loss": 1.3784,
+      "step": 4039
+    },
+    {
+      "epoch": 0.2912131478411302,
+      "grad_norm": 0.10121604773434939,
+      "learning_rate": 0.0001663082328985545,
+      "loss": 1.2761,
+      "step": 4040
+    },
+    {
+      "epoch": 0.29128523030346715,
+      "grad_norm": 0.1624586080081058,
+      "learning_rate": 0.00016629075466385523,
+      "loss": 1.3365,
+      "step": 4041
+    },
+    {
+      "epoch": 0.29135731276580407,
+      "grad_norm": 0.10257167641920152,
+      "learning_rate": 0.00016627327281571678,
+      "loss": 1.3032,
+      "step": 4042
+    },
+    {
+      "epoch": 0.291429395228141,
+      "grad_norm": 0.10456178438094324,
+      "learning_rate": 0.00016625578735509207,
+      "loss": 1.2686,
+      "step": 4043
+    },
+    {
+      "epoch": 0.2915014776904779,
+      "grad_norm": 0.13133027209320533,
+      "learning_rate": 0.0001662382982829342,
+      "loss": 1.329,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2915735601528148,
+      "grad_norm": 0.30292813844700583,
+      "learning_rate": 0.00016622080560019651,
+      "loss": 1.3719,
+      "step": 4045
+    },
+    {
+      "epoch": 0.29164564261515175,
+      "grad_norm": 0.13116998999755888,
+      "learning_rate": 0.00016620330930783243,
+      "loss": 1.3186,
+      "step": 4046
+    },
+    {
+      "epoch": 0.29171772507748867,
+      "grad_norm": 0.1085171424522554,
+      "learning_rate": 0.00016618580940679577,
+      "loss": 1.3139,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2917898075398256,
+      "grad_norm": 0.10460357661859856,
+      "learning_rate": 0.0001661683058980404,
+      "loss": 1.2721,
+      "step": 4048
+    },
+    {
+      "epoch": 0.29186189000216245,
+      "grad_norm": 0.1175268820080676,
+      "learning_rate": 0.00016615079878252033,
+      "loss": 1.4285,
+      "step": 4049
+    },
+    {
+      "epoch": 0.29193397246449937,
+      "grad_norm": 0.11153802944341235,
+      "learning_rate": 0.00016613328806118998,
+      "loss": 1.3322,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2920060549268363,
+      "grad_norm": 0.11099008643900277,
+      "learning_rate": 0.00016611577373500377,
+      "loss": 1.3032,
+      "step": 4051
+    },
+    {
+      "epoch": 0.2920781373891732,
+      "grad_norm": 0.10369321554301526,
+      "learning_rate": 0.00016609825580491641,
+      "loss": 1.2715,
+      "step": 4052
+    },
+    {
+      "epoch": 0.29215021985151013,
+      "grad_norm": 0.10821704557090274,
+      "learning_rate": 0.00016608073427188273,
+      "loss": 1.3212,
+      "step": 4053
+    },
+    {
+      "epoch": 0.29222230231384705,
+      "grad_norm": 0.17183232460853196,
+      "learning_rate": 0.0001660632091368579,
+      "loss": 1.3543,
+      "step": 4054
+    },
+    {
+      "epoch": 0.29229438477618397,
+      "grad_norm": 0.1255960948060806,
+      "learning_rate": 0.00016604568040079717,
+      "loss": 1.3057,
+      "step": 4055
+    },
+    {
+      "epoch": 0.2923664672385209,
+      "grad_norm": 0.10151701731395094,
+      "learning_rate": 0.00016602814806465597,
+      "loss": 1.2261,
+      "step": 4056
+    },
+    {
+      "epoch": 0.2924385497008578,
+      "grad_norm": 0.2701508585701256,
+      "learning_rate": 0.00016601061212939,
+      "loss": 1.4236,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2925106321631947,
+      "grad_norm": 0.11180146229061713,
+      "learning_rate": 0.00016599307259595512,
+      "loss": 1.2717,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2925827146255316,
+      "grad_norm": 0.11968709194136157,
+      "learning_rate": 0.00016597552946530738,
+      "loss": 1.4327,
+      "step": 4059
+    },
+    {
+      "epoch": 0.2926547970878685,
+      "grad_norm": 0.12472730193724722,
+      "learning_rate": 0.00016595798273840305,
+      "loss": 1.3659,
+      "step": 4060
+    },
+    {
+      "epoch": 0.29272687955020543,
+      "grad_norm": 0.10512812642269702,
+      "learning_rate": 0.0001659404324161986,
+      "loss": 1.2693,
+      "step": 4061
+    },
+    {
+      "epoch": 0.29279896201254235,
+      "grad_norm": 0.11405501916088524,
+      "learning_rate": 0.00016592287849965067,
+      "loss": 1.26,
+      "step": 4062
+    },
+    {
+      "epoch": 0.29287104447487927,
+      "grad_norm": 0.09993133910594083,
+      "learning_rate": 0.0001659053209897161,
+      "loss": 1.2989,
+      "step": 4063
+    },
+    {
+      "epoch": 0.2929431269372162,
+      "grad_norm": 0.1253673008154665,
+      "learning_rate": 0.00016588775988735191,
+      "loss": 1.3764,
+      "step": 4064
+    },
+    {
+      "epoch": 0.2930152093995531,
+      "grad_norm": 0.1255458779054605,
+      "learning_rate": 0.00016587019519351539,
+      "loss": 1.2448,
+      "step": 4065
+    },
+    {
+      "epoch": 0.29308729186189,
+      "grad_norm": 0.1160051780190791,
+      "learning_rate": 0.00016585262690916393,
+      "loss": 1.2471,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2931593743242269,
+      "grad_norm": 0.12411577599535425,
+      "learning_rate": 0.00016583505503525514,
+      "loss": 1.2236,
+      "step": 4067
+    },
+    {
+      "epoch": 0.2932314567865638,
+      "grad_norm": 0.11317164874068125,
+      "learning_rate": 0.00016581747957274693,
+      "loss": 1.3665,
+      "step": 4068
+    },
+    {
+      "epoch": 0.29330353924890074,
+      "grad_norm": 0.10532945352297234,
+      "learning_rate": 0.0001657999005225972,
+      "loss": 1.2882,
+      "step": 4069
+    },
+    {
+      "epoch": 0.29337562171123766,
+      "grad_norm": 0.11365532404037393,
+      "learning_rate": 0.00016578231788576427,
+      "loss": 1.2587,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2934477041735746,
+      "grad_norm": 0.10969672443603258,
+      "learning_rate": 0.00016576473166320644,
+      "loss": 1.2841,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2935197866359115,
+      "grad_norm": 0.1362869719295866,
+      "learning_rate": 0.00016574714185588248,
+      "loss": 1.3596,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2935918690982484,
+      "grad_norm": 0.11415117267823148,
+      "learning_rate": 0.000165729548464751,
+      "loss": 1.3432,
+      "step": 4073
+    },
+    {
+      "epoch": 0.29366395156058533,
+      "grad_norm": 0.1095844791562553,
+      "learning_rate": 0.00016571195149077118,
+      "loss": 1.4038,
+      "step": 4074
+    },
+    {
+      "epoch": 0.2937360340229222,
+      "grad_norm": 0.10270988015865831,
+      "learning_rate": 0.0001656943509349021,
+      "loss": 1.46,
+      "step": 4075
+    },
+    {
+      "epoch": 0.2938081164852591,
+      "grad_norm": 0.10784728710499249,
+      "learning_rate": 0.0001656767467981032,
+      "loss": 1.1624,
+      "step": 4076
+    },
+    {
+      "epoch": 0.29388019894759604,
+      "grad_norm": 0.12039497691488538,
+      "learning_rate": 0.000165659139081334,
+      "loss": 1.2778,
+      "step": 4077
+    },
+    {
+      "epoch": 0.29395228140993296,
+      "grad_norm": 0.09972308930430122,
+      "learning_rate": 0.00016564152778555434,
+      "loss": 1.2652,
+      "step": 4078
+    },
+    {
+      "epoch": 0.2940243638722699,
+      "grad_norm": 0.10318160498660801,
+      "learning_rate": 0.00016562391291172416,
+      "loss": 1.2927,
+      "step": 4079
+    },
+    {
+      "epoch": 0.2940964463346068,
+      "grad_norm": 0.11779635982047157,
+      "learning_rate": 0.0001656062944608036,
+      "loss": 1.3151,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2941685287969437,
+      "grad_norm": 0.11551527509426954,
+      "learning_rate": 0.00016558867243375314,
+      "loss": 1.217,
+      "step": 4081
+    },
+    {
+      "epoch": 0.29424061125928064,
+      "grad_norm": 0.1152830558086536,
+      "learning_rate": 0.0001655710468315332,
+      "loss": 1.4211,
+      "step": 4082
+    },
+    {
+      "epoch": 0.29431269372161756,
+      "grad_norm": 0.10764904988765607,
+      "learning_rate": 0.00016555341765510463,
+      "loss": 1.3407,
+      "step": 4083
+    },
+    {
+      "epoch": 0.2943847761839544,
+      "grad_norm": 0.10394815703133652,
+      "learning_rate": 0.00016553578490542832,
+      "loss": 1.2776,
+      "step": 4084
+    },
+    {
+      "epoch": 0.29445685864629134,
+      "grad_norm": 0.10349180893210602,
+      "learning_rate": 0.00016551814858346543,
+      "loss": 1.2886,
+      "step": 4085
+    },
+    {
+      "epoch": 0.29452894110862826,
+      "grad_norm": 0.10885862520543792,
+      "learning_rate": 0.0001655005086901773,
+      "loss": 1.369,
+      "step": 4086
+    },
+    {
+      "epoch": 0.2946010235709652,
+      "grad_norm": 0.11593283867779036,
+      "learning_rate": 0.00016548286522652546,
+      "loss": 1.3218,
+      "step": 4087
+    },
+    {
+      "epoch": 0.2946731060333021,
+      "grad_norm": 0.11605706693039185,
+      "learning_rate": 0.00016546521819347167,
+      "loss": 1.364,
+      "step": 4088
+    },
+    {
+      "epoch": 0.294745188495639,
+      "grad_norm": 0.11749043897215974,
+      "learning_rate": 0.00016544756759197778,
+      "loss": 1.3381,
+      "step": 4089
+    },
+    {
+      "epoch": 0.29481727095797594,
+      "grad_norm": 0.1077950124496459,
+      "learning_rate": 0.00016542991342300597,
+      "loss": 1.2455,
+      "step": 4090
+    },
+    {
+      "epoch": 0.29488935342031286,
+      "grad_norm": 0.12211419985389758,
+      "learning_rate": 0.00016541225568751853,
+      "loss": 1.2814,
+      "step": 4091
+    },
+    {
+      "epoch": 0.2949614358826497,
+      "grad_norm": 0.1068421431236686,
+      "learning_rate": 0.00016539459438647794,
+      "loss": 1.3165,
+      "step": 4092
+    },
+    {
+      "epoch": 0.29503351834498664,
+      "grad_norm": 0.1139000218719387,
+      "learning_rate": 0.0001653769295208469,
+      "loss": 1.3014,
+      "step": 4093
+    },
+    {
+      "epoch": 0.29510560080732356,
+      "grad_norm": 0.11122037765013348,
+      "learning_rate": 0.00016535926109158836,
+      "loss": 1.3614,
+      "step": 4094
+    },
+    {
+      "epoch": 0.2951776832696605,
+      "grad_norm": 0.09930471300226015,
+      "learning_rate": 0.00016534158909966537,
+      "loss": 1.3006,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2952497657319974,
+      "grad_norm": 0.11203823844049299,
+      "learning_rate": 0.00016532391354604119,
+      "loss": 1.265,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2953218481943343,
+      "grad_norm": 0.09973597304051211,
+      "learning_rate": 0.00016530623443167933,
+      "loss": 1.3355,
+      "step": 4097
+    },
+    {
+      "epoch": 0.29539393065667124,
+      "grad_norm": 0.11051496031789282,
+      "learning_rate": 0.00016528855175754344,
+      "loss": 1.2956,
+      "step": 4098
+    },
+    {
+      "epoch": 0.29546601311900816,
+      "grad_norm": 0.11327504016239341,
+      "learning_rate": 0.0001652708655245974,
+      "loss": 1.2889,
+      "step": 4099
+    },
+    {
+      "epoch": 0.2955380955813451,
+      "grad_norm": 0.11152637510289376,
+      "learning_rate": 0.00016525317573380525,
+      "loss": 1.352,
+      "step": 4100
+    },
+    {
+      "epoch": 0.29561017804368195,
+      "grad_norm": 0.12132143231420336,
+      "learning_rate": 0.00016523548238613127,
+      "loss": 1.2301,
+      "step": 4101
+    },
+    {
+      "epoch": 0.29568226050601887,
+      "grad_norm": 0.10656314096342147,
+      "learning_rate": 0.00016521778548253985,
+      "loss": 1.2325,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2957543429683558,
+      "grad_norm": 0.10578073113156632,
+      "learning_rate": 0.00016520008502399571,
+      "loss": 1.3552,
+      "step": 4103
+    },
+    {
+      "epoch": 0.2958264254306927,
+      "grad_norm": 0.11343726537673977,
+      "learning_rate": 0.0001651823810114636,
+      "loss": 1.329,
+      "step": 4104
+    },
+    {
+      "epoch": 0.2958985078930296,
+      "grad_norm": 0.11182533464106023,
+      "learning_rate": 0.0001651646734459086,
+      "loss": 1.3051,
+      "step": 4105
+    },
+    {
+      "epoch": 0.29597059035536655,
+      "grad_norm": 0.09497017474769473,
+      "learning_rate": 0.00016514696232829593,
+      "loss": 1.2747,
+      "step": 4106
+    },
+    {
+      "epoch": 0.29604267281770347,
+      "grad_norm": 0.10651688171152743,
+      "learning_rate": 0.000165129247659591,
+      "loss": 1.3104,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2961147552800404,
+      "grad_norm": 0.10587610285473661,
+      "learning_rate": 0.0001651115294407594,
+      "loss": 1.2679,
+      "step": 4108
+    },
+    {
+      "epoch": 0.2961868377423773,
+      "grad_norm": 0.11200415047761718,
+      "learning_rate": 0.0001650938076727669,
+      "loss": 1.3223,
+      "step": 4109
+    },
+    {
+      "epoch": 0.29625892020471417,
+      "grad_norm": 0.10965024790787993,
+      "learning_rate": 0.00016507608235657962,
+      "loss": 1.329,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2963310026670511,
+      "grad_norm": 0.09868150669901261,
+      "learning_rate": 0.00016505835349316358,
+      "loss": 1.279,
+      "step": 4111
+    },
+    {
+      "epoch": 0.296403085129388,
+      "grad_norm": 0.11270892988158135,
+      "learning_rate": 0.0001650406210834853,
+      "loss": 1.2606,
+      "step": 4112
+    },
+    {
+      "epoch": 0.29647516759172493,
+      "grad_norm": 0.10426589160142627,
+      "learning_rate": 0.00016502288512851125,
+      "loss": 1.3233,
+      "step": 4113
+    },
+    {
+      "epoch": 0.29654725005406185,
+      "grad_norm": 0.11099905017601921,
+      "learning_rate": 0.00016500514562920831,
+      "loss": 1.2453,
+      "step": 4114
+    },
+    {
+      "epoch": 0.29661933251639877,
+      "grad_norm": 0.11300125046835745,
+      "learning_rate": 0.00016498740258654335,
+      "loss": 1.3156,
+      "step": 4115
+    },
+    {
+      "epoch": 0.2966914149787357,
+      "grad_norm": 0.11521580392035154,
+      "learning_rate": 0.00016496965600148355,
+      "loss": 1.3034,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2967634974410726,
+      "grad_norm": 0.11615274076575059,
+      "learning_rate": 0.00016495190587499626,
+      "loss": 1.3337,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2968355799034095,
+      "grad_norm": 0.11745615163231467,
+      "learning_rate": 0.00016493415220804905,
+      "loss": 1.2197,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2969076623657464,
+      "grad_norm": 0.11412764328130189,
+      "learning_rate": 0.00016491639500160958,
+      "loss": 1.2715,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2969797448280833,
+      "grad_norm": 0.11300284780951644,
+      "learning_rate": 0.00016489863425664587,
+      "loss": 1.2361,
+      "step": 4120
+    },
+    {
+      "epoch": 0.29705182729042023,
+      "grad_norm": 0.11368076957939165,
+      "learning_rate": 0.00016488086997412598,
+      "loss": 1.3189,
+      "step": 4121
+    },
+    {
+      "epoch": 0.29712390975275715,
+      "grad_norm": 0.10866952723672152,
+      "learning_rate": 0.00016486310215501823,
+      "loss": 1.2611,
+      "step": 4122
+    },
+    {
+      "epoch": 0.2971959922150941,
+      "grad_norm": 0.10163610120399766,
+      "learning_rate": 0.00016484533080029113,
+      "loss": 1.3035,
+      "step": 4123
+    },
+    {
+      "epoch": 0.297268074677431,
+      "grad_norm": 0.10600110210450685,
+      "learning_rate": 0.00016482755591091337,
+      "loss": 1.2735,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2973401571397679,
+      "grad_norm": 0.18939591292553024,
+      "learning_rate": 0.00016480977748785386,
+      "loss": 1.2759,
+      "step": 4125
+    },
+    {
+      "epoch": 0.29741223960210483,
+      "grad_norm": 0.2514775717840848,
+      "learning_rate": 0.0001647919955320817,
+      "loss": 1.2764,
+      "step": 4126
+    },
+    {
+      "epoch": 0.2974843220644417,
+      "grad_norm": 0.12484478692646889,
+      "learning_rate": 0.00016477421004456607,
+      "loss": 1.2574,
+      "step": 4127
+    },
+    {
+      "epoch": 0.2975564045267786,
+      "grad_norm": 0.12395934119386315,
+      "learning_rate": 0.00016475642102627657,
+      "loss": 1.3045,
+      "step": 4128
+    },
+    {
+      "epoch": 0.29762848698911554,
+      "grad_norm": 0.11797822651019486,
+      "learning_rate": 0.00016473862847818277,
+      "loss": 1.3556,
+      "step": 4129
+    },
+    {
+      "epoch": 0.29770056945145246,
+      "grad_norm": 0.12268658988009208,
+      "learning_rate": 0.00016472083240125454,
+      "loss": 1.3532,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2977726519137894,
+      "grad_norm": 0.1144288218840905,
+      "learning_rate": 0.00016470303279646196,
+      "loss": 1.2774,
+      "step": 4131
+    },
+    {
+      "epoch": 0.2978447343761263,
+      "grad_norm": 0.2826610827667437,
+      "learning_rate": 0.00016468522966477523,
+      "loss": 1.3801,
+      "step": 4132
+    },
+    {
+      "epoch": 0.2979168168384632,
+      "grad_norm": 0.1934858359324507,
+      "learning_rate": 0.0001646674230071648,
+      "loss": 1.2455,
+      "step": 4133
+    },
+    {
+      "epoch": 0.29798889930080014,
+      "grad_norm": 0.10169606991289257,
+      "learning_rate": 0.00016464961282460126,
+      "loss": 1.3387,
+      "step": 4134
+    },
+    {
+      "epoch": 0.29806098176313706,
+      "grad_norm": 0.12837860585120112,
+      "learning_rate": 0.0001646317991180555,
+      "loss": 1.3048,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2981330642254739,
+      "grad_norm": 0.1123796761468635,
+      "learning_rate": 0.0001646139818884984,
+      "loss": 1.2108,
+      "step": 4136
+    },
+    {
+      "epoch": 0.29820514668781084,
+      "grad_norm": 0.101775954131216,
+      "learning_rate": 0.0001645961611369013,
+      "loss": 1.2693,
+      "step": 4137
+    },
+    {
+      "epoch": 0.29827722915014776,
+      "grad_norm": 0.1019745737314242,
+      "learning_rate": 0.0001645783368642355,
+      "loss": 1.2574,
+      "step": 4138
+    },
+    {
+      "epoch": 0.2983493116124847,
+      "grad_norm": 0.10629638116002153,
+      "learning_rate": 0.0001645605090714726,
+      "loss": 1.2888,
+      "step": 4139
+    },
+    {
+      "epoch": 0.2984213940748216,
+      "grad_norm": 0.19787049137234872,
+      "learning_rate": 0.00016454267775958437,
+      "loss": 1.4144,
+      "step": 4140
+    },
+    {
+      "epoch": 0.2984934765371585,
+      "grad_norm": 0.1033462603638141,
+      "learning_rate": 0.00016452484292954283,
+      "loss": 1.3839,
+      "step": 4141
+    },
+    {
+      "epoch": 0.29856555899949544,
+      "grad_norm": 0.12722617395495764,
+      "learning_rate": 0.00016450700458232008,
+      "loss": 1.3481,
+      "step": 4142
+    },
+    {
+      "epoch": 0.29863764146183236,
+      "grad_norm": 0.10829627598259879,
+      "learning_rate": 0.00016448916271888847,
+      "loss": 1.3538,
+      "step": 4143
+    },
+    {
+      "epoch": 0.2987097239241692,
+      "grad_norm": 0.11567667642230572,
+      "learning_rate": 0.00016447131734022055,
+      "loss": 1.2618,
+      "step": 4144
+    },
+    {
+      "epoch": 0.29878180638650614,
+      "grad_norm": 0.11226020789492658,
+      "learning_rate": 0.00016445346844728905,
+      "loss": 1.364,
+      "step": 4145
+    },
+    {
+      "epoch": 0.29885388884884306,
+      "grad_norm": 0.18705467084836747,
+      "learning_rate": 0.00016443561604106695,
+      "loss": 1.2176,
+      "step": 4146
+    },
+    {
+      "epoch": 0.29892597131118,
+      "grad_norm": 0.12223366302572017,
+      "learning_rate": 0.00016441776012252727,
+      "loss": 1.339,
+      "step": 4147
+    },
+    {
+      "epoch": 0.2989980537735169,
+      "grad_norm": 0.11364224396324593,
+      "learning_rate": 0.00016439990069264338,
+      "loss": 1.2702,
+      "step": 4148
+    },
+    {
+      "epoch": 0.2990701362358538,
+      "grad_norm": 0.11534506136348845,
+      "learning_rate": 0.00016438203775238875,
+      "loss": 1.2552,
+      "step": 4149
+    },
+    {
+      "epoch": 0.29914221869819074,
+      "grad_norm": 0.10396079004288068,
+      "learning_rate": 0.00016436417130273712,
+      "loss": 1.2921,
+      "step": 4150
+    },
+    {
+      "epoch": 0.29921430116052766,
+      "grad_norm": 0.10918632670341585,
+      "learning_rate": 0.0001643463013446623,
+      "loss": 1.3331,
+      "step": 4151
+    },
+    {
+      "epoch": 0.2992863836228646,
+      "grad_norm": 0.11305939982389297,
+      "learning_rate": 0.00016432842787913841,
+      "loss": 1.3058,
+      "step": 4152
+    },
+    {
+      "epoch": 0.29935846608520145,
+      "grad_norm": 0.11732400861405738,
+      "learning_rate": 0.00016431055090713974,
+      "loss": 1.3272,
+      "step": 4153
+    },
+    {
+      "epoch": 0.29943054854753837,
+      "grad_norm": 0.10755763595288215,
+      "learning_rate": 0.00016429267042964067,
+      "loss": 1.374,
+      "step": 4154
+    },
+    {
+      "epoch": 0.2995026310098753,
+      "grad_norm": 0.11963127244136136,
+      "learning_rate": 0.00016427478644761586,
+      "loss": 1.2858,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2995747134722122,
+      "grad_norm": 0.14990015301947418,
+      "learning_rate": 0.00016425689896204025,
+      "loss": 1.3072,
+      "step": 4156
+    },
+    {
+      "epoch": 0.2996467959345491,
+      "grad_norm": 0.11365085443018728,
+      "learning_rate": 0.0001642390079738887,
+      "loss": 1.2688,
+      "step": 4157
+    },
+    {
+      "epoch": 0.29971887839688605,
+      "grad_norm": 0.1052786048548931,
+      "learning_rate": 0.00016422111348413657,
+      "loss": 1.3659,
+      "step": 4158
+    },
+    {
+      "epoch": 0.29979096085922297,
+      "grad_norm": 0.10262557846680448,
+      "learning_rate": 0.00016420321549375918,
+      "loss": 1.2972,
+      "step": 4159
+    },
+    {
+      "epoch": 0.2998630433215599,
+      "grad_norm": 0.09808152510508317,
+      "learning_rate": 0.0001641853140037322,
+      "loss": 1.291,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2999351257838968,
+      "grad_norm": 0.11614724680703088,
+      "learning_rate": 0.00016416740901503135,
+      "loss": 1.2239,
+      "step": 4161
+    },
+    {
+      "epoch": 0.30000720824623367,
+      "grad_norm": 0.1799550104800392,
+      "learning_rate": 0.0001641495005286327,
+      "loss": 1.3789,
+      "step": 4162
+    },
+    {
+      "epoch": 0.3000792907085706,
+      "grad_norm": 0.2105638650448506,
+      "learning_rate": 0.00016413158854551232,
+      "loss": 1.2854,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3001513731709075,
+      "grad_norm": 0.12992828480513016,
+      "learning_rate": 0.00016411367306664668,
+      "loss": 1.2964,
+      "step": 4164
+    },
+    {
+      "epoch": 0.30022345563324443,
+      "grad_norm": 0.12802859033337147,
+      "learning_rate": 0.00016409575409301227,
+      "loss": 1.2363,
+      "step": 4165
+    },
+    {
+      "epoch": 0.30029553809558135,
+      "grad_norm": 0.12984295914681784,
+      "learning_rate": 0.0001640778316255858,
+      "loss": 1.2544,
+      "step": 4166
+    },
+    {
+      "epoch": 0.30036762055791827,
+      "grad_norm": 0.11172275155188287,
+      "learning_rate": 0.0001640599056653443,
+      "loss": 1.3323,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3004397030202552,
+      "grad_norm": 0.11303046304251034,
+      "learning_rate": 0.00016404197621326482,
+      "loss": 1.3236,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3005117854825921,
+      "grad_norm": 0.11147345100255067,
+      "learning_rate": 0.0001640240432703247,
+      "loss": 1.4118,
+      "step": 4169
+    },
+    {
+      "epoch": 0.300583867944929,
+      "grad_norm": 0.1457670504693144,
+      "learning_rate": 0.00016400610683750143,
+      "loss": 1.2315,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3006559504072659,
+      "grad_norm": 0.21791777087522982,
+      "learning_rate": 0.0001639881669157728,
+      "loss": 1.3489,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3007280328696028,
+      "grad_norm": 0.10889480867973787,
+      "learning_rate": 0.00016397022350611652,
+      "loss": 1.2936,
+      "step": 4172
+    },
+    {
+      "epoch": 0.30080011533193973,
+      "grad_norm": 0.10812248271939683,
+      "learning_rate": 0.00016395227660951083,
+      "loss": 1.3219,
+      "step": 4173
+    },
+    {
+      "epoch": 0.30087219779427665,
+      "grad_norm": 0.1188536753344531,
+      "learning_rate": 0.0001639343262269339,
+      "loss": 1.2256,
+      "step": 4174
+    },
+    {
+      "epoch": 0.30094428025661357,
+      "grad_norm": 0.11243578568118229,
+      "learning_rate": 0.00016391637235936424,
+      "loss": 1.4062,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3010163627189505,
+      "grad_norm": 0.13067540891977597,
+      "learning_rate": 0.00016389841500778047,
+      "loss": 1.2098,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3010884451812874,
+      "grad_norm": 0.10829063256656363,
+      "learning_rate": 0.0001638804541731614,
+      "loss": 1.3097,
+      "step": 4177
+    },
+    {
+      "epoch": 0.30116052764362433,
+      "grad_norm": 0.11160121102932911,
+      "learning_rate": 0.00016386248985648613,
+      "loss": 1.2791,
+      "step": 4178
+    },
+    {
+      "epoch": 0.3012326101059612,
+      "grad_norm": 0.10479474788868448,
+      "learning_rate": 0.00016384452205873377,
+      "loss": 1.3222,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3013046925682981,
+      "grad_norm": 0.1195258426258948,
+      "learning_rate": 0.00016382655078088387,
+      "loss": 1.3368,
+      "step": 4180
+    },
+    {
+      "epoch": 0.30137677503063504,
+      "grad_norm": 0.13884663146040963,
+      "learning_rate": 0.0001638085760239159,
+      "loss": 1.3363,
+      "step": 4181
+    },
+    {
+      "epoch": 0.30144885749297196,
+      "grad_norm": 0.11547853630859556,
+      "learning_rate": 0.00016379059778880967,
+      "loss": 1.3084,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3015209399553089,
+      "grad_norm": 0.1278317792039085,
+      "learning_rate": 0.00016377261607654518,
+      "loss": 1.3183,
+      "step": 4183
+    },
+    {
+      "epoch": 0.3015930224176458,
+      "grad_norm": 0.10502009249970189,
+      "learning_rate": 0.00016375463088810264,
+      "loss": 1.2653,
+      "step": 4184
+    },
+    {
+      "epoch": 0.3016651048799827,
+      "grad_norm": 0.1465495281582494,
+      "learning_rate": 0.0001637366422244623,
+      "loss": 1.3569,
+      "step": 4185
+    },
+    {
+      "epoch": 0.30173718734231963,
+      "grad_norm": 0.11166370052391554,
+      "learning_rate": 0.00016371865008660475,
+      "loss": 1.4,
+      "step": 4186
+    },
+    {
+      "epoch": 0.30180926980465655,
+      "grad_norm": 0.10885807589122601,
+      "learning_rate": 0.00016370065447551078,
+      "loss": 1.2883,
+      "step": 4187
+    },
+    {
+      "epoch": 0.3018813522669934,
+      "grad_norm": 0.11039341132441013,
+      "learning_rate": 0.0001636826553921612,
+      "loss": 1.2935,
+      "step": 4188
+    },
+    {
+      "epoch": 0.30195343472933034,
+      "grad_norm": 0.10292976981289062,
+      "learning_rate": 0.00016366465283753723,
+      "loss": 1.234,
+      "step": 4189
+    },
+    {
+      "epoch": 0.30202551719166726,
+      "grad_norm": 0.12507857451608068,
+      "learning_rate": 0.0001636466468126201,
+      "loss": 1.2244,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3020975996540042,
+      "grad_norm": 0.1191668209105516,
+      "learning_rate": 0.00016362863731839128,
+      "loss": 1.4149,
+      "step": 4191
+    },
+    {
+      "epoch": 0.3021696821163411,
+      "grad_norm": 0.11538597402085041,
+      "learning_rate": 0.00016361062435583254,
+      "loss": 1.2806,
+      "step": 4192
+    },
+    {
+      "epoch": 0.302241764578678,
+      "grad_norm": 0.12192246142124787,
+      "learning_rate": 0.00016359260792592567,
+      "loss": 1.2853,
+      "step": 4193
+    },
+    {
+      "epoch": 0.30231384704101494,
+      "grad_norm": 0.12367040371515334,
+      "learning_rate": 0.00016357458802965276,
+      "loss": 1.3402,
+      "step": 4194
+    },
+    {
+      "epoch": 0.30238592950335186,
+      "grad_norm": 0.10642456629666597,
+      "learning_rate": 0.00016355656466799604,
+      "loss": 1.2903,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3024580119656887,
+      "grad_norm": 0.11348857997090323,
+      "learning_rate": 0.00016353853784193797,
+      "loss": 1.2375,
+      "step": 4196
+    },
+    {
+      "epoch": 0.30253009442802564,
+      "grad_norm": 0.12077752046056704,
+      "learning_rate": 0.00016352050755246117,
+      "loss": 1.2403,
+      "step": 4197
+    },
+    {
+      "epoch": 0.30260217689036256,
+      "grad_norm": 0.09618804610646417,
+      "learning_rate": 0.0001635024738005484,
+      "loss": 1.1929,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3026742593526995,
+      "grad_norm": 0.14122224242491266,
+      "learning_rate": 0.0001634844365871827,
+      "loss": 1.2395,
+      "step": 4199
+    },
+    {
+      "epoch": 0.3027463418150364,
+      "grad_norm": 0.12185115214788193,
+      "learning_rate": 0.0001634663959133473,
+      "loss": 1.267,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3028184242773733,
+      "grad_norm": 0.09977640156458024,
+      "learning_rate": 0.00016344835178002548,
+      "loss": 1.303,
+      "step": 4201
+    },
+    {
+      "epoch": 0.30289050673971024,
+      "grad_norm": 0.10808024969795443,
+      "learning_rate": 0.0001634303041882009,
+      "loss": 1.345,
+      "step": 4202
+    },
+    {
+      "epoch": 0.30296258920204716,
+      "grad_norm": 0.11209387247573886,
+      "learning_rate": 0.00016341225313885727,
+      "loss": 1.3345,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3030346716643841,
+      "grad_norm": 0.10748415183187619,
+      "learning_rate": 0.00016339419863297853,
+      "loss": 1.3977,
+      "step": 4204
+    },
+    {
+      "epoch": 0.30310675412672095,
+      "grad_norm": 0.11201539395351938,
+      "learning_rate": 0.00016337614067154883,
+      "loss": 1.2593,
+      "step": 4205
+    },
+    {
+      "epoch": 0.30317883658905787,
+      "grad_norm": 0.11673947504636568,
+      "learning_rate": 0.00016335807925555248,
+      "loss": 1.3298,
+      "step": 4206
+    },
+    {
+      "epoch": 0.3032509190513948,
+      "grad_norm": 0.10978296375740113,
+      "learning_rate": 0.000163340014385974,
+      "loss": 1.4407,
+      "step": 4207
+    },
+    {
+      "epoch": 0.3033230015137317,
+      "grad_norm": 0.09652953759763559,
+      "learning_rate": 0.0001633219460637981,
+      "loss": 1.3001,
+      "step": 4208
+    },
+    {
+      "epoch": 0.3033950839760686,
+      "grad_norm": 0.1004110636146533,
+      "learning_rate": 0.00016330387429000962,
+      "loss": 1.1917,
+      "step": 4209
+    },
+    {
+      "epoch": 0.30346716643840554,
+      "grad_norm": 0.10271693615897079,
+      "learning_rate": 0.00016328579906559366,
+      "loss": 1.3253,
+      "step": 4210
+    },
+    {
+      "epoch": 0.30353924890074246,
+      "grad_norm": 0.11477612480778068,
+      "learning_rate": 0.0001632677203915355,
+      "loss": 1.2169,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3036113313630794,
+      "grad_norm": 0.10566250007523094,
+      "learning_rate": 0.00016324963826882055,
+      "loss": 1.283,
+      "step": 4212
+    },
+    {
+      "epoch": 0.30368341382541625,
+      "grad_norm": 0.11985779372840837,
+      "learning_rate": 0.00016323155269843447,
+      "loss": 1.2156,
+      "step": 4213
+    },
+    {
+      "epoch": 0.30375549628775317,
+      "grad_norm": 0.11367407269774653,
+      "learning_rate": 0.00016321346368136305,
+      "loss": 1.2837,
+      "step": 4214
+    },
+    {
+      "epoch": 0.3038275787500901,
+      "grad_norm": 0.11874739232823231,
+      "learning_rate": 0.0001631953712185924,
+      "loss": 1.2888,
+      "step": 4215
+    },
+    {
+      "epoch": 0.303899661212427,
+      "grad_norm": 0.12699938312393594,
+      "learning_rate": 0.0001631772753111086,
+      "loss": 1.149,
+      "step": 4216
+    },
+    {
+      "epoch": 0.30397174367476393,
+      "grad_norm": 0.11396938418018508,
+      "learning_rate": 0.00016315917595989814,
+      "loss": 1.2531,
+      "step": 4217
+    },
+    {
+      "epoch": 0.30404382613710085,
+      "grad_norm": 0.10507370489924267,
+      "learning_rate": 0.00016314107316594753,
+      "loss": 1.3475,
+      "step": 4218
+    },
+    {
+      "epoch": 0.30411590859943777,
+      "grad_norm": 0.10939108011990273,
+      "learning_rate": 0.00016312296693024356,
+      "loss": 1.2774,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3041879910617747,
+      "grad_norm": 0.11696997353118319,
+      "learning_rate": 0.0001631048572537732,
+      "loss": 1.2872,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3042600735241116,
+      "grad_norm": 0.11057118980541873,
+      "learning_rate": 0.00016308674413752352,
+      "loss": 1.2828,
+      "step": 4221
+    },
+    {
+      "epoch": 0.30433215598644847,
+      "grad_norm": 0.1051384561544175,
+      "learning_rate": 0.00016306862758248193,
+      "loss": 1.3577,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3044042384487854,
+      "grad_norm": 0.10883449967581173,
+      "learning_rate": 0.00016305050758963588,
+      "loss": 1.2893,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3044763209111223,
+      "grad_norm": 0.11700025741899717,
+      "learning_rate": 0.00016303238415997314,
+      "loss": 1.3143,
+      "step": 4224
+    },
+    {
+      "epoch": 0.30454840337345923,
+      "grad_norm": 0.11035298456860337,
+      "learning_rate": 0.0001630142572944815,
+      "loss": 1.2804,
+      "step": 4225
+    },
+    {
+      "epoch": 0.30462048583579615,
+      "grad_norm": 0.17577022392078262,
+      "learning_rate": 0.00016299612699414916,
+      "loss": 1.3146,
+      "step": 4226
+    },
+    {
+      "epoch": 0.30469256829813307,
+      "grad_norm": 0.11232752197256043,
+      "learning_rate": 0.0001629779932599643,
+      "loss": 1.3179,
+      "step": 4227
+    },
+    {
+      "epoch": 0.30476465076047,
+      "grad_norm": 0.10685715714269635,
+      "learning_rate": 0.00016295985609291534,
+      "loss": 1.2561,
+      "step": 4228
+    },
+    {
+      "epoch": 0.3048367332228069,
+      "grad_norm": 0.11451648748777991,
+      "learning_rate": 0.00016294171549399104,
+      "loss": 1.2712,
+      "step": 4229
+    },
+    {
+      "epoch": 0.30490881568514383,
+      "grad_norm": 0.10894289523661371,
+      "learning_rate": 0.0001629235714641801,
+      "loss": 1.375,
+      "step": 4230
+    },
+    {
+      "epoch": 0.3049808981474807,
+      "grad_norm": 0.10023908249878605,
+      "learning_rate": 0.00016290542400447164,
+      "loss": 1.3143,
+      "step": 4231
+    },
+    {
+      "epoch": 0.3050529806098176,
+      "grad_norm": 0.11540284610251125,
+      "learning_rate": 0.00016288727311585473,
+      "loss": 1.343,
+      "step": 4232
+    },
+    {
+      "epoch": 0.30512506307215453,
+      "grad_norm": 0.11864709788246183,
+      "learning_rate": 0.0001628691187993189,
+      "loss": 1.2588,
+      "step": 4233
+    },
+    {
+      "epoch": 0.30519714553449145,
+      "grad_norm": 0.11731607868178698,
+      "learning_rate": 0.00016285096105585363,
+      "loss": 1.3298,
+      "step": 4234
+    },
+    {
+      "epoch": 0.3052692279968284,
+      "grad_norm": 0.10311786260503453,
+      "learning_rate": 0.00016283279988644868,
+      "loss": 1.3133,
+      "step": 4235
+    },
+    {
+      "epoch": 0.3053413104591653,
+      "grad_norm": 0.10565178624998718,
+      "learning_rate": 0.00016281463529209405,
+      "loss": 1.3755,
+      "step": 4236
+    },
+    {
+      "epoch": 0.3054133929215022,
+      "grad_norm": 0.11982696826133972,
+      "learning_rate": 0.00016279646727377983,
+      "loss": 1.2994,
+      "step": 4237
+    },
+    {
+      "epoch": 0.30548547538383913,
+      "grad_norm": 0.10463251199351636,
+      "learning_rate": 0.00016277829583249634,
+      "loss": 1.2792,
+      "step": 4238
+    },
+    {
+      "epoch": 0.305557557846176,
+      "grad_norm": 0.10316905972461873,
+      "learning_rate": 0.0001627601209692341,
+      "loss": 1.2366,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3056296403085129,
+      "grad_norm": 0.1197314186015759,
+      "learning_rate": 0.0001627419426849838,
+      "loss": 1.3433,
+      "step": 4240
+    },
+    {
+      "epoch": 0.30570172277084984,
+      "grad_norm": 0.1022423340239507,
+      "learning_rate": 0.0001627237609807363,
+      "loss": 1.2326,
+      "step": 4241
+    },
+    {
+      "epoch": 0.30577380523318676,
+      "grad_norm": 0.10610198166482605,
+      "learning_rate": 0.0001627055758574827,
+      "loss": 1.3248,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3058458876955237,
+      "grad_norm": 0.09873889592314727,
+      "learning_rate": 0.00016268738731621424,
+      "loss": 1.3054,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3059179701578606,
+      "grad_norm": 0.10246488361294045,
+      "learning_rate": 0.00016266919535792238,
+      "loss": 1.2801,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3059900526201975,
+      "grad_norm": 0.11252104073168834,
+      "learning_rate": 0.00016265099998359866,
+      "loss": 1.4017,
+      "step": 4245
+    },
+    {
+      "epoch": 0.30606213508253444,
+      "grad_norm": 0.10637475890925246,
+      "learning_rate": 0.000162632801194235,
+      "loss": 1.3093,
+      "step": 4246
+    },
+    {
+      "epoch": 0.30613421754487136,
+      "grad_norm": 0.10295936872607425,
+      "learning_rate": 0.0001626145989908233,
+      "loss": 1.2676,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3062063000072082,
+      "grad_norm": 0.10790572918604699,
+      "learning_rate": 0.00016259639337435583,
+      "loss": 1.3592,
+      "step": 4248
+    },
+    {
+      "epoch": 0.30627838246954514,
+      "grad_norm": 0.10664711161187564,
+      "learning_rate": 0.0001625781843458249,
+      "loss": 1.4152,
+      "step": 4249
+    },
+    {
+      "epoch": 0.30635046493188206,
+      "grad_norm": 0.10502862843832177,
+      "learning_rate": 0.0001625599719062231,
+      "loss": 1.3386,
+      "step": 4250
+    },
+    {
+      "epoch": 0.306422547394219,
+      "grad_norm": 0.1017369167855477,
+      "learning_rate": 0.00016254175605654312,
+      "loss": 1.3379,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3064946298565559,
+      "grad_norm": 0.14050719849083004,
+      "learning_rate": 0.00016252353679777794,
+      "loss": 1.3156,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3065667123188928,
+      "grad_norm": 0.1114587751520081,
+      "learning_rate": 0.00016250531413092065,
+      "loss": 1.2674,
+      "step": 4253
+    },
+    {
+      "epoch": 0.30663879478122974,
+      "grad_norm": 0.2374064201065268,
+      "learning_rate": 0.00016248708805696456,
+      "loss": 1.3506,
+      "step": 4254
+    },
+    {
+      "epoch": 0.30671087724356666,
+      "grad_norm": 0.10803832296815345,
+      "learning_rate": 0.00016246885857690315,
+      "loss": 1.3203,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3067829597059036,
+      "grad_norm": 0.11249387117597096,
+      "learning_rate": 0.00016245062569173005,
+      "loss": 1.3423,
+      "step": 4256
+    },
+    {
+      "epoch": 0.30685504216824044,
+      "grad_norm": 0.11363397664023414,
+      "learning_rate": 0.0001624323894024392,
+      "loss": 1.2806,
+      "step": 4257
+    },
+    {
+      "epoch": 0.30692712463057736,
+      "grad_norm": 0.12113990175380435,
+      "learning_rate": 0.00016241414971002454,
+      "loss": 1.2792,
+      "step": 4258
+    },
+    {
+      "epoch": 0.3069992070929143,
+      "grad_norm": 0.10481716539050119,
+      "learning_rate": 0.00016239590661548035,
+      "loss": 1.3213,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3070712895552512,
+      "grad_norm": 0.10661151799227384,
+      "learning_rate": 0.0001623776601198011,
+      "loss": 1.2731,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3071433720175881,
+      "grad_norm": 0.11091648846523589,
+      "learning_rate": 0.0001623594102239813,
+      "loss": 1.3041,
+      "step": 4261
+    },
+    {
+      "epoch": 0.30721545447992504,
+      "grad_norm": 0.1058004471730411,
+      "learning_rate": 0.00016234115692901572,
+      "loss": 1.2921,
+      "step": 4262
+    },
+    {
+      "epoch": 0.30728753694226196,
+      "grad_norm": 0.11303349264799481,
+      "learning_rate": 0.0001623229002358994,
+      "loss": 1.372,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3073596194045989,
+      "grad_norm": 0.10555547966915221,
+      "learning_rate": 0.00016230464014562747,
+      "loss": 1.29,
+      "step": 4264
+    },
+    {
+      "epoch": 0.30743170186693575,
+      "grad_norm": 0.10844806676382923,
+      "learning_rate": 0.00016228637665919525,
+      "loss": 1.3107,
+      "step": 4265
+    },
+    {
+      "epoch": 0.30750378432927267,
+      "grad_norm": 0.18119774799184712,
+      "learning_rate": 0.00016226810977759825,
+      "loss": 1.2961,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3075758667916096,
+      "grad_norm": 0.11066587757942838,
+      "learning_rate": 0.00016224983950183222,
+      "loss": 1.285,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3076479492539465,
+      "grad_norm": 0.11576309060521109,
+      "learning_rate": 0.00016223156583289306,
+      "loss": 1.3253,
+      "step": 4268
+    },
+    {
+      "epoch": 0.3077200317162834,
+      "grad_norm": 0.13772920687002463,
+      "learning_rate": 0.00016221328877177678,
+      "loss": 1.3534,
+      "step": 4269
+    },
+    {
+      "epoch": 0.30779211417862035,
+      "grad_norm": 0.09424303151097757,
+      "learning_rate": 0.0001621950083194797,
+      "loss": 1.3735,
+      "step": 4270
+    },
+    {
+      "epoch": 0.30786419664095727,
+      "grad_norm": 0.10895630416619874,
+      "learning_rate": 0.00016217672447699824,
+      "loss": 1.2973,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3079362791032942,
+      "grad_norm": 0.12726105648063127,
+      "learning_rate": 0.00016215843724532908,
+      "loss": 1.2236,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3080083615656311,
+      "grad_norm": 0.10525516170675969,
+      "learning_rate": 0.00016214014662546897,
+      "loss": 1.2489,
+      "step": 4273
+    },
+    {
+      "epoch": 0.30808044402796797,
+      "grad_norm": 0.11084582593253714,
+      "learning_rate": 0.00016212185261841499,
+      "loss": 1.2381,
+      "step": 4274
+    },
+    {
+      "epoch": 0.3081525264903049,
+      "grad_norm": 0.11001948898682522,
+      "learning_rate": 0.00016210355522516424,
+      "loss": 1.28,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3082246089526418,
+      "grad_norm": 0.11078278049018962,
+      "learning_rate": 0.00016208525444671415,
+      "loss": 1.3052,
+      "step": 4276
+    },
+    {
+      "epoch": 0.30829669141497873,
+      "grad_norm": 0.1078405920589287,
+      "learning_rate": 0.00016206695028406229,
+      "loss": 1.2352,
+      "step": 4277
+    },
+    {
+      "epoch": 0.30836877387731565,
+      "grad_norm": 0.09975445554284838,
+      "learning_rate": 0.00016204864273820633,
+      "loss": 1.3256,
+      "step": 4278
+    },
+    {
+      "epoch": 0.30844085633965257,
+      "grad_norm": 0.10353096506077675,
+      "learning_rate": 0.00016203033181014427,
+      "loss": 1.3194,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3085129388019895,
+      "grad_norm": 0.10779816213119722,
+      "learning_rate": 0.00016201201750087418,
+      "loss": 1.2707,
+      "step": 4280
+    },
+    {
+      "epoch": 0.3085850212643264,
+      "grad_norm": 0.11346359527305734,
+      "learning_rate": 0.00016199369981139433,
+      "loss": 1.2979,
+      "step": 4281
+    },
+    {
+      "epoch": 0.30865710372666333,
+      "grad_norm": 0.10337568465659183,
+      "learning_rate": 0.00016197537874270326,
+      "loss": 1.2418,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3087291861890002,
+      "grad_norm": 0.10653654778341944,
+      "learning_rate": 0.00016195705429579958,
+      "loss": 1.3615,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3088012686513371,
+      "grad_norm": 0.11040678514786123,
+      "learning_rate": 0.0001619387264716822,
+      "loss": 1.4368,
+      "step": 4284
+    },
+    {
+      "epoch": 0.30887335111367403,
+      "grad_norm": 0.1143901402684081,
+      "learning_rate": 0.00016192039527135007,
+      "loss": 1.2984,
+      "step": 4285
+    },
+    {
+      "epoch": 0.30894543357601095,
+      "grad_norm": 0.09971933994869014,
+      "learning_rate": 0.00016190206069580248,
+      "loss": 1.2993,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3090175160383479,
+      "grad_norm": 0.10780939782695273,
+      "learning_rate": 0.00016188372274603873,
+      "loss": 1.3193,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3090895985006848,
+      "grad_norm": 0.11684622088052049,
+      "learning_rate": 0.00016186538142305853,
+      "loss": 1.3566,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3091616809630217,
+      "grad_norm": 0.11708808386231016,
+      "learning_rate": 0.0001618470367278615,
+      "loss": 1.3785,
+      "step": 4289
+    },
+    {
+      "epoch": 0.30923376342535863,
+      "grad_norm": 0.1125713990828764,
+      "learning_rate": 0.00016182868866144773,
+      "loss": 1.3155,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3093058458876955,
+      "grad_norm": 0.11028611418443002,
+      "learning_rate": 0.0001618103372248173,
+      "loss": 1.2827,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3093779283500324,
+      "grad_norm": 0.22104197114060128,
+      "learning_rate": 0.00016179198241897047,
+      "loss": 1.2211,
+      "step": 4292
+    },
+    {
+      "epoch": 0.30945001081236934,
+      "grad_norm": 0.1073529369130176,
+      "learning_rate": 0.0001617736242449078,
+      "loss": 1.3198,
+      "step": 4293
+    },
+    {
+      "epoch": 0.30952209327470626,
+      "grad_norm": 0.11334735408485413,
+      "learning_rate": 0.00016175526270363,
+      "loss": 1.3289,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3095941757370432,
+      "grad_norm": 0.10354054595526356,
+      "learning_rate": 0.0001617368977961379,
+      "loss": 1.3892,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3096662581993801,
+      "grad_norm": 0.10963028094683186,
+      "learning_rate": 0.00016171852952343255,
+      "loss": 1.3719,
+      "step": 4296
+    },
+    {
+      "epoch": 0.309738340661717,
+      "grad_norm": 0.1254553762427688,
+      "learning_rate": 0.00016170015788651517,
+      "loss": 1.3555,
+      "step": 4297
+    },
+    {
+      "epoch": 0.30981042312405394,
+      "grad_norm": 0.11269902203012166,
+      "learning_rate": 0.0001616817828863872,
+      "loss": 1.349,
+      "step": 4298
+    },
+    {
+      "epoch": 0.30988250558639086,
+      "grad_norm": 0.11854215035502898,
+      "learning_rate": 0.00016166340452405028,
+      "loss": 1.3065,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3099545880487277,
+      "grad_norm": 0.11033704789696476,
+      "learning_rate": 0.00016164502280050614,
+      "loss": 1.3804,
+      "step": 4300
+    },
+    {
+      "epoch": 0.31002667051106464,
+      "grad_norm": 0.10290128868753952,
+      "learning_rate": 0.00016162663771675674,
+      "loss": 1.2027,
+      "step": 4301
+    },
+    {
+      "epoch": 0.31009875297340156,
+      "grad_norm": 0.11340536093438809,
+      "learning_rate": 0.0001616082492738043,
+      "loss": 1.358,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3101708354357385,
+      "grad_norm": 0.11697828340775378,
+      "learning_rate": 0.00016158985747265108,
+      "loss": 1.2153,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3102429178980754,
+      "grad_norm": 0.1095828986772246,
+      "learning_rate": 0.0001615714623142996,
+      "loss": 1.3708,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3103150003604123,
+      "grad_norm": 0.11544403122919468,
+      "learning_rate": 0.00016155306379975266,
+      "loss": 1.3607,
+      "step": 4305
+    },
+    {
+      "epoch": 0.31038708282274924,
+      "grad_norm": 0.1313784544713779,
+      "learning_rate": 0.00016153466193001304,
+      "loss": 1.2196,
+      "step": 4306
+    },
+    {
+      "epoch": 0.31045916528508616,
+      "grad_norm": 0.1105375156020373,
+      "learning_rate": 0.00016151625670608383,
+      "loss": 1.3501,
+      "step": 4307
+    },
+    {
+      "epoch": 0.3105312477474231,
+      "grad_norm": 0.4000324356554,
+      "learning_rate": 0.00016149784812896832,
+      "loss": 1.3767,
+      "step": 4308
+    },
+    {
+      "epoch": 0.31060333020975994,
+      "grad_norm": 0.10994584894991678,
+      "learning_rate": 0.00016147943619966988,
+      "loss": 1.4048,
+      "step": 4309
+    },
+    {
+      "epoch": 0.31067541267209686,
+      "grad_norm": 0.10897708680867801,
+      "learning_rate": 0.00016146102091919216,
+      "loss": 1.2761,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3107474951344338,
+      "grad_norm": 0.1209058890762227,
+      "learning_rate": 0.00016144260228853895,
+      "loss": 1.3451,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3108195775967707,
+      "grad_norm": 0.10617213923465223,
+      "learning_rate": 0.00016142418030871424,
+      "loss": 1.3424,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3108916600591076,
+      "grad_norm": 0.11613741450479365,
+      "learning_rate": 0.0001614057549807222,
+      "loss": 1.356,
+      "step": 4313
+    },
+    {
+      "epoch": 0.31096374252144454,
+      "grad_norm": 0.10012301027304696,
+      "learning_rate": 0.00016138732630556715,
+      "loss": 1.2981,
+      "step": 4314
+    },
+    {
+      "epoch": 0.31103582498378146,
+      "grad_norm": 0.11942558662107394,
+      "learning_rate": 0.00016136889428425364,
+      "loss": 1.4061,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3111079074461184,
+      "grad_norm": 0.10610604328917282,
+      "learning_rate": 0.0001613504589177864,
+      "loss": 1.3371,
+      "step": 4316
+    },
+    {
+      "epoch": 0.31117998990845525,
+      "grad_norm": 0.10594805446523868,
+      "learning_rate": 0.00016133202020717023,
+      "loss": 1.3204,
+      "step": 4317
+    },
+    {
+      "epoch": 0.31125207237079217,
+      "grad_norm": 0.10570308922926372,
+      "learning_rate": 0.00016131357815341027,
+      "loss": 1.3265,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3113241548331291,
+      "grad_norm": 0.10629971726962553,
+      "learning_rate": 0.0001612951327575118,
+      "loss": 1.349,
+      "step": 4319
+    },
+    {
+      "epoch": 0.311396237295466,
+      "grad_norm": 0.11338857996928113,
+      "learning_rate": 0.00016127668402048025,
+      "loss": 1.2042,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3114683197578029,
+      "grad_norm": 0.11906292914382051,
+      "learning_rate": 0.0001612582319433212,
+      "loss": 1.2456,
+      "step": 4321
+    },
+    {
+      "epoch": 0.31154040222013984,
+      "grad_norm": 0.10374945170971817,
+      "learning_rate": 0.00016123977652704046,
+      "loss": 1.2502,
+      "step": 4322
+    },
+    {
+      "epoch": 0.31161248468247676,
+      "grad_norm": 0.1202759873143955,
+      "learning_rate": 0.00016122131777264405,
+      "loss": 1.2443,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3116845671448137,
+      "grad_norm": 0.1142279075490333,
+      "learning_rate": 0.0001612028556811381,
+      "loss": 1.2961,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3117566496071506,
+      "grad_norm": 0.10932390803474998,
+      "learning_rate": 0.00016118439025352902,
+      "loss": 1.2952,
+      "step": 4325
+    },
+    {
+      "epoch": 0.31182873206948747,
+      "grad_norm": 0.10784701565410773,
+      "learning_rate": 0.00016116592149082327,
+      "loss": 1.2292,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3119008145318244,
+      "grad_norm": 0.10348505744249427,
+      "learning_rate": 0.00016114744939402763,
+      "loss": 1.3113,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3119728969941613,
+      "grad_norm": 0.11480285338458186,
+      "learning_rate": 0.00016112897396414893,
+      "loss": 1.3085,
+      "step": 4328
+    },
+    {
+      "epoch": 0.31204497945649823,
+      "grad_norm": 0.1488880106742264,
+      "learning_rate": 0.00016111049520219428,
+      "loss": 1.3222,
+      "step": 4329
+    },
+    {
+      "epoch": 0.31211706191883515,
+      "grad_norm": 0.11665503890517709,
+      "learning_rate": 0.00016109201310917092,
+      "loss": 1.2692,
+      "step": 4330
+    },
+    {
+      "epoch": 0.31218914438117207,
+      "grad_norm": 0.1164897278842531,
+      "learning_rate": 0.0001610735276860863,
+      "loss": 1.2842,
+      "step": 4331
+    },
+    {
+      "epoch": 0.312261226843509,
+      "grad_norm": 0.10482672009804318,
+      "learning_rate": 0.00016105503893394806,
+      "loss": 1.3483,
+      "step": 4332
+    },
+    {
+      "epoch": 0.3123333093058459,
+      "grad_norm": 0.11347554361325943,
+      "learning_rate": 0.00016103654685376397,
+      "loss": 1.3229,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3124053917681828,
+      "grad_norm": 0.11382341420274333,
+      "learning_rate": 0.00016101805144654204,
+      "loss": 1.2952,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3124774742305197,
+      "grad_norm": 0.13346934566376745,
+      "learning_rate": 0.0001609995527132904,
+      "loss": 1.3581,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3125495566928566,
+      "grad_norm": 0.12312570358605557,
+      "learning_rate": 0.00016098105065501743,
+      "loss": 1.2041,
+      "step": 4336
+    },
+    {
+      "epoch": 0.31262163915519353,
+      "grad_norm": 0.18219623706985402,
+      "learning_rate": 0.00016096254527273165,
+      "loss": 1.2586,
+      "step": 4337
+    },
+    {
+      "epoch": 0.31269372161753045,
+      "grad_norm": 0.09555551532411002,
+      "learning_rate": 0.00016094403656744178,
+      "loss": 1.2964,
+      "step": 4338
+    },
+    {
+      "epoch": 0.31276580407986737,
+      "grad_norm": 0.09320193825813425,
+      "learning_rate": 0.0001609255245401567,
+      "loss": 1.224,
+      "step": 4339
+    },
+    {
+      "epoch": 0.3128378865422043,
+      "grad_norm": 0.10845257503078067,
+      "learning_rate": 0.00016090700919188544,
+      "loss": 1.3994,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3129099690045412,
+      "grad_norm": 0.12052640042993305,
+      "learning_rate": 0.00016088849052363732,
+      "loss": 1.2598,
+      "step": 4341
+    },
+    {
+      "epoch": 0.31298205146687813,
+      "grad_norm": 0.12692042338269138,
+      "learning_rate": 0.0001608699685364217,
+      "loss": 1.2501,
+      "step": 4342
+    },
+    {
+      "epoch": 0.313054133929215,
+      "grad_norm": 0.11847694046781596,
+      "learning_rate": 0.00016085144323124828,
+      "loss": 1.3855,
+      "step": 4343
+    },
+    {
+      "epoch": 0.3131262163915519,
+      "grad_norm": 0.1242451030455015,
+      "learning_rate": 0.0001608329146091268,
+      "loss": 1.2382,
+      "step": 4344
+    },
+    {
+      "epoch": 0.31319829885388883,
+      "grad_norm": 0.17476832106784482,
+      "learning_rate": 0.00016081438267106726,
+      "loss": 1.2224,
+      "step": 4345
+    },
+    {
+      "epoch": 0.31327038131622575,
+      "grad_norm": 0.10436741497708395,
+      "learning_rate": 0.00016079584741807975,
+      "loss": 1.3741,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3133424637785627,
+      "grad_norm": 0.11495450692733683,
+      "learning_rate": 0.00016077730885117472,
+      "loss": 1.2753,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3134145462408996,
+      "grad_norm": 0.11683018558572648,
+      "learning_rate": 0.0001607587669713626,
+      "loss": 1.3644,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3134866287032365,
+      "grad_norm": 0.10904445080901308,
+      "learning_rate": 0.00016074022177965414,
+      "loss": 1.3771,
+      "step": 4349
+    },
+    {
+      "epoch": 0.31355871116557343,
+      "grad_norm": 0.11698455261853985,
+      "learning_rate": 0.00016072167327706017,
+      "loss": 1.3,
+      "step": 4350
+    },
+    {
+      "epoch": 0.31363079362791035,
+      "grad_norm": 0.11636103130597188,
+      "learning_rate": 0.0001607031214645918,
+      "loss": 1.4178,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3137028760902472,
+      "grad_norm": 0.11893789610700407,
+      "learning_rate": 0.0001606845663432602,
+      "loss": 1.2411,
+      "step": 4352
+    },
+    {
+      "epoch": 0.31377495855258414,
+      "grad_norm": 0.12057945797427587,
+      "learning_rate": 0.0001606660079140769,
+      "loss": 1.3016,
+      "step": 4353
+    },
+    {
+      "epoch": 0.31384704101492106,
+      "grad_norm": 0.10656205098662479,
+      "learning_rate": 0.0001606474461780534,
+      "loss": 1.247,
+      "step": 4354
+    },
+    {
+      "epoch": 0.313919123477258,
+      "grad_norm": 0.10285179008885073,
+      "learning_rate": 0.0001606288811362015,
+      "loss": 1.3611,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3139912059395949,
+      "grad_norm": 0.11449334964190355,
+      "learning_rate": 0.00016061031278953318,
+      "loss": 1.2244,
+      "step": 4356
+    },
+    {
+      "epoch": 0.3140632884019318,
+      "grad_norm": 0.12045188391705598,
+      "learning_rate": 0.0001605917411390606,
+      "loss": 1.3225,
+      "step": 4357
+    },
+    {
+      "epoch": 0.31413537086426874,
+      "grad_norm": 0.11138539087337333,
+      "learning_rate": 0.00016057316618579606,
+      "loss": 1.3741,
+      "step": 4358
+    },
+    {
+      "epoch": 0.31420745332660566,
+      "grad_norm": 0.12041373543548185,
+      "learning_rate": 0.00016055458793075202,
+      "loss": 1.3591,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3142795357889426,
+      "grad_norm": 0.12255724474796663,
+      "learning_rate": 0.0001605360063749412,
+      "loss": 1.256,
+      "step": 4360
+    },
+    {
+      "epoch": 0.31435161825127944,
+      "grad_norm": 0.10325945370427203,
+      "learning_rate": 0.00016051742151937655,
+      "loss": 1.29,
+      "step": 4361
+    },
+    {
+      "epoch": 0.31442370071361636,
+      "grad_norm": 0.15572749546754686,
+      "learning_rate": 0.00016049883336507096,
+      "loss": 1.3149,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3144957831759533,
+      "grad_norm": 0.1353351816173783,
+      "learning_rate": 0.0001604802419130377,
+      "loss": 1.1602,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3145678656382902,
+      "grad_norm": 0.11108397536658918,
+      "learning_rate": 0.0001604616471642902,
+      "loss": 1.4114,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3146399481006271,
+      "grad_norm": 0.1092196399729342,
+      "learning_rate": 0.00016044304911984203,
+      "loss": 1.3319,
+      "step": 4365
+    },
+    {
+      "epoch": 0.31471203056296404,
+      "grad_norm": 0.11547264977856234,
+      "learning_rate": 0.0001604244477807069,
+      "loss": 1.2759,
+      "step": 4366
+    },
+    {
+      "epoch": 0.31478411302530096,
+      "grad_norm": 0.12431651897581462,
+      "learning_rate": 0.00016040584314789888,
+      "loss": 1.3053,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3148561954876379,
+      "grad_norm": 0.1004459211663595,
+      "learning_rate": 0.00016038723522243193,
+      "loss": 1.3714,
+      "step": 4368
+    },
+    {
+      "epoch": 0.31492827794997474,
+      "grad_norm": 0.10459503624677184,
+      "learning_rate": 0.00016036862400532046,
+      "loss": 1.3196,
+      "step": 4369
+    },
+    {
+      "epoch": 0.31500036041231166,
+      "grad_norm": 0.10729911997033083,
+      "learning_rate": 0.00016035000949757888,
+      "loss": 1.2419,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3150724428746486,
+      "grad_norm": 0.10593572242456933,
+      "learning_rate": 0.00016033139170022192,
+      "loss": 1.2875,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3151445253369855,
+      "grad_norm": 0.11947181788759925,
+      "learning_rate": 0.00016031277061426434,
+      "loss": 1.265,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3152166077993224,
+      "grad_norm": 0.10576101187676182,
+      "learning_rate": 0.00016029414624072117,
+      "loss": 1.3392,
+      "step": 4373
+    },
+    {
+      "epoch": 0.31528869026165934,
+      "grad_norm": 0.10345290525009991,
+      "learning_rate": 0.00016027551858060766,
+      "loss": 1.3086,
+      "step": 4374
+    },
+    {
+      "epoch": 0.31536077272399626,
+      "grad_norm": 0.12521051379271708,
+      "learning_rate": 0.0001602568876349391,
+      "loss": 1.3203,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3154328551863332,
+      "grad_norm": 0.12936491235345485,
+      "learning_rate": 0.00016023825340473114,
+      "loss": 1.2886,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3155049376486701,
+      "grad_norm": 0.10986231970585562,
+      "learning_rate": 0.00016021961589099943,
+      "loss": 1.3457,
+      "step": 4377
+    },
+    {
+      "epoch": 0.31557702011100697,
+      "grad_norm": 0.12057666786745638,
+      "learning_rate": 0.00016020097509475995,
+      "loss": 1.2427,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3156491025733439,
+      "grad_norm": 0.10471299358559415,
+      "learning_rate": 0.00016018233101702873,
+      "loss": 1.2967,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3157211850356808,
+      "grad_norm": 0.11465039703524398,
+      "learning_rate": 0.00016016368365882205,
+      "loss": 1.2693,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3157932674980177,
+      "grad_norm": 0.11046563241920508,
+      "learning_rate": 0.00016014503302115639,
+      "loss": 1.2647,
+      "step": 4381
+    },
+    {
+      "epoch": 0.31586534996035465,
+      "grad_norm": 0.11224048800462985,
+      "learning_rate": 0.00016012637910504837,
+      "loss": 1.3286,
+      "step": 4382
+    },
+    {
+      "epoch": 0.31593743242269157,
+      "grad_norm": 0.11785953684457397,
+      "learning_rate": 0.00016010772191151474,
+      "loss": 1.3238,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3160095148850285,
+      "grad_norm": 0.10678930016190354,
+      "learning_rate": 0.00016008906144157256,
+      "loss": 1.2933,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3160815973473654,
+      "grad_norm": 0.14560422074858445,
+      "learning_rate": 0.00016007039769623895,
+      "loss": 1.2609,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3161536798097023,
+      "grad_norm": 0.10564171108396153,
+      "learning_rate": 0.00016005173067653127,
+      "loss": 1.3053,
+      "step": 4386
+    },
+    {
+      "epoch": 0.3162257622720392,
+      "grad_norm": 0.11374548131806957,
+      "learning_rate": 0.00016003306038346704,
+      "loss": 1.3347,
+      "step": 4387
+    },
+    {
+      "epoch": 0.3162978447343761,
+      "grad_norm": 0.10415160514525744,
+      "learning_rate": 0.0001600143868180639,
+      "loss": 1.2819,
+      "step": 4388
+    },
+    {
+      "epoch": 0.31636992719671303,
+      "grad_norm": 0.13156271929198887,
+      "learning_rate": 0.00015999570998133985,
+      "loss": 1.343,
+      "step": 4389
+    },
+    {
+      "epoch": 0.31644200965904995,
+      "grad_norm": 0.1086439917344547,
+      "learning_rate": 0.0001599770298743128,
+      "loss": 1.3892,
+      "step": 4390
+    },
+    {
+      "epoch": 0.31651409212138687,
+      "grad_norm": 0.10797026661475465,
+      "learning_rate": 0.00015995834649800106,
+      "loss": 1.2706,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3165861745837238,
+      "grad_norm": 0.19673541088734497,
+      "learning_rate": 0.00015993965985342304,
+      "loss": 1.3561,
+      "step": 4392
+    },
+    {
+      "epoch": 0.3166582570460607,
+      "grad_norm": 0.11615080493291972,
+      "learning_rate": 0.00015992096994159736,
+      "loss": 1.2859,
+      "step": 4393
+    },
+    {
+      "epoch": 0.31673033950839763,
+      "grad_norm": 0.11939288375409723,
+      "learning_rate": 0.00015990227676354274,
+      "loss": 1.2934,
+      "step": 4394
+    },
+    {
+      "epoch": 0.3168024219707345,
+      "grad_norm": 0.11281309160703347,
+      "learning_rate": 0.0001598835803202781,
+      "loss": 1.2188,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3168745044330714,
+      "grad_norm": 0.11301503948743062,
+      "learning_rate": 0.0001598648806128226,
+      "loss": 1.2579,
+      "step": 4396
+    },
+    {
+      "epoch": 0.31694658689540833,
+      "grad_norm": 0.11841079244983739,
+      "learning_rate": 0.00015984617764219555,
+      "loss": 1.3373,
+      "step": 4397
+    },
+    {
+      "epoch": 0.31701866935774525,
+      "grad_norm": 0.10771742847562767,
+      "learning_rate": 0.00015982747140941644,
+      "loss": 1.368,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3170907518200822,
+      "grad_norm": 0.10775893089231113,
+      "learning_rate": 0.00015980876191550486,
+      "loss": 1.2442,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3171628342824191,
+      "grad_norm": 0.12435511147348229,
+      "learning_rate": 0.00015979004916148075,
+      "loss": 1.2072,
+      "step": 4400
+    },
+    {
+      "epoch": 0.317234916744756,
+      "grad_norm": 0.12119997715420809,
+      "learning_rate": 0.00015977133314836402,
+      "loss": 1.4181,
+      "step": 4401
+    },
+    {
+      "epoch": 0.31730699920709293,
+      "grad_norm": 0.11011630605468158,
+      "learning_rate": 0.00015975261387717496,
+      "loss": 1.2218,
+      "step": 4402
+    },
+    {
+      "epoch": 0.31737908166942985,
+      "grad_norm": 0.0970067790187098,
+      "learning_rate": 0.00015973389134893383,
+      "loss": 1.326,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3174511641317667,
+      "grad_norm": 0.09931178708208326,
+      "learning_rate": 0.00015971516556466124,
+      "loss": 1.2641,
+      "step": 4404
+    },
+    {
+      "epoch": 0.31752324659410364,
+      "grad_norm": 0.10040366830890414,
+      "learning_rate": 0.00015969643652537788,
+      "loss": 1.3527,
+      "step": 4405
+    },
+    {
+      "epoch": 0.31759532905644056,
+      "grad_norm": 0.09986305002340566,
+      "learning_rate": 0.00015967770423210467,
+      "loss": 1.2479,
+      "step": 4406
+    },
+    {
+      "epoch": 0.3176674115187775,
+      "grad_norm": 0.11319756947808164,
+      "learning_rate": 0.00015965896868586272,
+      "loss": 1.3386,
+      "step": 4407
+    },
+    {
+      "epoch": 0.3177394939811144,
+      "grad_norm": 0.10688442622957031,
+      "learning_rate": 0.00015964022988767325,
+      "loss": 1.2474,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3178115764434513,
+      "grad_norm": 0.1158392393942522,
+      "learning_rate": 0.00015962148783855767,
+      "loss": 1.3033,
+      "step": 4409
+    },
+    {
+      "epoch": 0.31788365890578824,
+      "grad_norm": 0.09813881439247589,
+      "learning_rate": 0.00015960274253953762,
+      "loss": 1.258,
+      "step": 4410
+    },
+    {
+      "epoch": 0.31795574136812516,
+      "grad_norm": 0.12045629939017735,
+      "learning_rate": 0.0001595839939916349,
+      "loss": 1.3387,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3180278238304621,
+      "grad_norm": 0.10959067754292626,
+      "learning_rate": 0.0001595652421958714,
+      "loss": 1.2437,
+      "step": 4412
+    },
+    {
+      "epoch": 0.31809990629279894,
+      "grad_norm": 0.11839962033687913,
+      "learning_rate": 0.00015954648715326938,
+      "loss": 1.3464,
+      "step": 4413
+    },
+    {
+      "epoch": 0.31817198875513586,
+      "grad_norm": 0.10617953773209268,
+      "learning_rate": 0.0001595277288648511,
+      "loss": 1.3853,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3182440712174728,
+      "grad_norm": 0.21278018511230617,
+      "learning_rate": 0.000159508967331639,
+      "loss": 1.4265,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3183161536798097,
+      "grad_norm": 0.10518533148833323,
+      "learning_rate": 0.0001594902025546558,
+      "loss": 1.2193,
+      "step": 4416
+    },
+    {
+      "epoch": 0.3183882361421466,
+      "grad_norm": 0.12056114790836746,
+      "learning_rate": 0.00015947143453492435,
+      "loss": 1.3546,
+      "step": 4417
+    },
+    {
+      "epoch": 0.31846031860448354,
+      "grad_norm": 0.10828321311078923,
+      "learning_rate": 0.00015945266327346767,
+      "loss": 1.2404,
+      "step": 4418
+    },
+    {
+      "epoch": 0.31853240106682046,
+      "grad_norm": 0.1219525007982299,
+      "learning_rate": 0.000159433888771309,
+      "loss": 1.3471,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3186044835291574,
+      "grad_norm": 0.3037623329469193,
+      "learning_rate": 0.00015941511102947163,
+      "loss": 1.3468,
+      "step": 4420
+    },
+    {
+      "epoch": 0.31867656599149424,
+      "grad_norm": 0.11114328093493796,
+      "learning_rate": 0.00015939633004897918,
+      "loss": 1.2901,
+      "step": 4421
+    },
+    {
+      "epoch": 0.31874864845383116,
+      "grad_norm": 0.1070625772953385,
+      "learning_rate": 0.00015937754583085538,
+      "loss": 1.2118,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3188207309161681,
+      "grad_norm": 0.11458775034563264,
+      "learning_rate": 0.0001593587583761241,
+      "loss": 1.305,
+      "step": 4423
+    },
+    {
+      "epoch": 0.318892813378505,
+      "grad_norm": 0.10749459307400981,
+      "learning_rate": 0.0001593399676858095,
+      "loss": 1.3664,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3189648958408419,
+      "grad_norm": 0.22977333102058128,
+      "learning_rate": 0.00015932117376093574,
+      "loss": 1.3222,
+      "step": 4425
+    },
+    {
+      "epoch": 0.31903697830317884,
+      "grad_norm": 0.10252651024984832,
+      "learning_rate": 0.00015930237660252735,
+      "loss": 1.3738,
+      "step": 4426
+    },
+    {
+      "epoch": 0.31910906076551576,
+      "grad_norm": 0.10933549886097711,
+      "learning_rate": 0.00015928357621160886,
+      "loss": 1.2804,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3191811432278527,
+      "grad_norm": 0.32036793986000317,
+      "learning_rate": 0.00015926477258920515,
+      "loss": 1.2823,
+      "step": 4428
+    },
+    {
+      "epoch": 0.3192532256901896,
+      "grad_norm": 0.11888339837578941,
+      "learning_rate": 0.0001592459657363411,
+      "loss": 1.3419,
+      "step": 4429
+    },
+    {
+      "epoch": 0.31932530815252647,
+      "grad_norm": 0.11391137887797682,
+      "learning_rate": 0.0001592271556540419,
+      "loss": 1.2838,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3193973906148634,
+      "grad_norm": 0.10768925322832375,
+      "learning_rate": 0.00015920834234333284,
+      "loss": 1.3311,
+      "step": 4431
+    },
+    {
+      "epoch": 0.3194694730772003,
+      "grad_norm": 0.1059484261980438,
+      "learning_rate": 0.00015918952580523942,
+      "loss": 1.3579,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3195415555395372,
+      "grad_norm": 0.11352684972963964,
+      "learning_rate": 0.00015917070604078734,
+      "loss": 1.3323,
+      "step": 4433
+    },
+    {
+      "epoch": 0.31961363800187415,
+      "grad_norm": 0.23340161896509581,
+      "learning_rate": 0.00015915188305100245,
+      "loss": 1.3482,
+      "step": 4434
+    },
+    {
+      "epoch": 0.31968572046421107,
+      "grad_norm": 0.11439477129423103,
+      "learning_rate": 0.00015913305683691072,
+      "loss": 1.3628,
+      "step": 4435
+    },
+    {
+      "epoch": 0.319757802926548,
+      "grad_norm": 0.11800542747233671,
+      "learning_rate": 0.00015911422739953835,
+      "loss": 1.2879,
+      "step": 4436
+    },
+    {
+      "epoch": 0.3198298853888849,
+      "grad_norm": 0.10172937587275505,
+      "learning_rate": 0.00015909539473991179,
+      "loss": 1.455,
+      "step": 4437
+    },
+    {
+      "epoch": 0.3199019678512218,
+      "grad_norm": 0.10579176520841026,
+      "learning_rate": 0.00015907655885905747,
+      "loss": 1.3744,
+      "step": 4438
+    },
+    {
+      "epoch": 0.3199740503135587,
+      "grad_norm": 0.14535640719776632,
+      "learning_rate": 0.00015905771975800221,
+      "loss": 1.3306,
+      "step": 4439
+    },
+    {
+      "epoch": 0.3200461327758956,
+      "grad_norm": 0.10932252170024369,
+      "learning_rate": 0.00015903887743777287,
+      "loss": 1.3334,
+      "step": 4440
+    },
+    {
+      "epoch": 0.32011821523823253,
+      "grad_norm": 0.11769552081207213,
+      "learning_rate": 0.00015902003189939654,
+      "loss": 1.2766,
+      "step": 4441
+    },
+    {
+      "epoch": 0.32019029770056945,
+      "grad_norm": 0.1238364868543811,
+      "learning_rate": 0.0001590011831439005,
+      "loss": 1.3557,
+      "step": 4442
+    },
+    {
+      "epoch": 0.32026238016290637,
+      "grad_norm": 0.10713071777267615,
+      "learning_rate": 0.00015898233117231207,
+      "loss": 1.3526,
+      "step": 4443
+    },
+    {
+      "epoch": 0.3203344626252433,
+      "grad_norm": 0.11278967798611908,
+      "learning_rate": 0.00015896347598565896,
+      "loss": 1.3055,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3204065450875802,
+      "grad_norm": 0.16187034776058412,
+      "learning_rate": 0.00015894461758496887,
+      "loss": 1.2595,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3204786275499171,
+      "grad_norm": 0.2348678111882854,
+      "learning_rate": 0.00015892575597126985,
+      "loss": 1.2214,
+      "step": 4446
+    },
+    {
+      "epoch": 0.320550710012254,
+      "grad_norm": 0.11588623440895521,
+      "learning_rate": 0.0001589068911455899,
+      "loss": 1.3223,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3206227924745909,
+      "grad_norm": 0.12311527792622758,
+      "learning_rate": 0.00015888802310895742,
+      "loss": 1.3065,
+      "step": 4448
+    },
+    {
+      "epoch": 0.32069487493692783,
+      "grad_norm": 0.1473207820562749,
+      "learning_rate": 0.0001588691518624009,
+      "loss": 1.192,
+      "step": 4449
+    },
+    {
+      "epoch": 0.32076695739926475,
+      "grad_norm": 0.1190077778366939,
+      "learning_rate": 0.00015885027740694888,
+      "loss": 1.3191,
+      "step": 4450
+    },
+    {
+      "epoch": 0.32083903986160167,
+      "grad_norm": 0.12433337964722668,
+      "learning_rate": 0.00015883139974363027,
+      "loss": 1.216,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3209111223239386,
+      "grad_norm": 0.11567182947598474,
+      "learning_rate": 0.00015881251887347404,
+      "loss": 1.2599,
+      "step": 4452
+    },
+    {
+      "epoch": 0.3209832047862755,
+      "grad_norm": 0.11519285604444784,
+      "learning_rate": 0.0001587936347975094,
+      "loss": 1.3855,
+      "step": 4453
+    },
+    {
+      "epoch": 0.32105528724861243,
+      "grad_norm": 0.11609061371820777,
+      "learning_rate": 0.00015877474751676567,
+      "loss": 1.2333,
+      "step": 4454
+    },
+    {
+      "epoch": 0.32112736971094935,
+      "grad_norm": 0.11244611653511948,
+      "learning_rate": 0.0001587558570322724,
+      "loss": 1.3056,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3211994521732862,
+      "grad_norm": 0.18623169392454678,
+      "learning_rate": 0.00015873696334505922,
+      "loss": 1.3201,
+      "step": 4456
+    },
+    {
+      "epoch": 0.32127153463562313,
+      "grad_norm": 0.11829474661945978,
+      "learning_rate": 0.00015871806645615614,
+      "loss": 1.2855,
+      "step": 4457
+    },
+    {
+      "epoch": 0.32134361709796005,
+      "grad_norm": 0.11294012591315929,
+      "learning_rate": 0.0001586991663665931,
+      "loss": 1.3732,
+      "step": 4458
+    },
+    {
+      "epoch": 0.321415699560297,
+      "grad_norm": 0.12801355403234332,
+      "learning_rate": 0.00015868026307740035,
+      "loss": 1.241,
+      "step": 4459
+    },
+    {
+      "epoch": 0.3214877820226339,
+      "grad_norm": 0.11668462715440386,
+      "learning_rate": 0.0001586613565896083,
+      "loss": 1.2787,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3215598644849708,
+      "grad_norm": 0.17699043838190315,
+      "learning_rate": 0.0001586424469042475,
+      "loss": 1.3468,
+      "step": 4461
+    },
+    {
+      "epoch": 0.32163194694730773,
+      "grad_norm": 0.12407698257750362,
+      "learning_rate": 0.00015862353402234876,
+      "loss": 1.3066,
+      "step": 4462
+    },
+    {
+      "epoch": 0.32170402940964465,
+      "grad_norm": 0.11184015647204956,
+      "learning_rate": 0.00015860461794494287,
+      "loss": 1.3334,
+      "step": 4463
+    },
+    {
+      "epoch": 0.3217761118719816,
+      "grad_norm": 0.11154444541693512,
+      "learning_rate": 0.00015858569867306107,
+      "loss": 1.2918,
+      "step": 4464
+    },
+    {
+      "epoch": 0.32184819433431844,
+      "grad_norm": 0.09950165952047839,
+      "learning_rate": 0.00015856677620773457,
+      "loss": 1.2733,
+      "step": 4465
+    },
+    {
+      "epoch": 0.32192027679665536,
+      "grad_norm": 0.10832042176147355,
+      "learning_rate": 0.0001585478505499948,
+      "loss": 1.3145,
+      "step": 4466
+    },
+    {
+      "epoch": 0.3219923592589923,
+      "grad_norm": 0.11620968933230351,
+      "learning_rate": 0.0001585289217008734,
+      "loss": 1.3178,
+      "step": 4467
+    },
+    {
+      "epoch": 0.3220644417213292,
+      "grad_norm": 0.10256503014285814,
+      "learning_rate": 0.00015850998966140216,
+      "loss": 1.2943,
+      "step": 4468
+    },
+    {
+      "epoch": 0.3221365241836661,
+      "grad_norm": 0.11612350133588158,
+      "learning_rate": 0.000158491054432613,
+      "loss": 1.2461,
+      "step": 4469
+    },
+    {
+      "epoch": 0.32220860664600304,
+      "grad_norm": 0.11420465181570383,
+      "learning_rate": 0.00015847211601553816,
+      "loss": 1.2461,
+      "step": 4470
+    },
+    {
+      "epoch": 0.32228068910833996,
+      "grad_norm": 0.1125268400517778,
+      "learning_rate": 0.00015845317441120984,
+      "loss": 1.3699,
+      "step": 4471
+    },
+    {
+      "epoch": 0.3223527715706769,
+      "grad_norm": 0.1256059798286088,
+      "learning_rate": 0.0001584342296206606,
+      "loss": 1.2583,
+      "step": 4472
+    },
+    {
+      "epoch": 0.32242485403301374,
+      "grad_norm": 0.12425872852204502,
+      "learning_rate": 0.00015841528164492306,
+      "loss": 1.3918,
+      "step": 4473
+    },
+    {
+      "epoch": 0.32249693649535066,
+      "grad_norm": 0.12161916176619351,
+      "learning_rate": 0.0001583963304850301,
+      "loss": 1.2487,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3225690189576876,
+      "grad_norm": 0.11984218678785925,
+      "learning_rate": 0.00015837737614201466,
+      "loss": 1.2855,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3226411014200245,
+      "grad_norm": 0.11463677701981798,
+      "learning_rate": 0.00015835841861690998,
+      "loss": 1.3379,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3227131838823614,
+      "grad_norm": 0.11166315876596734,
+      "learning_rate": 0.00015833945791074943,
+      "loss": 1.273,
+      "step": 4477
+    },
+    {
+      "epoch": 0.32278526634469834,
+      "grad_norm": 0.10004345908456756,
+      "learning_rate": 0.00015832049402456644,
+      "loss": 1.254,
+      "step": 4478
+    },
+    {
+      "epoch": 0.32285734880703526,
+      "grad_norm": 0.10384731350693778,
+      "learning_rate": 0.0001583015269593948,
+      "loss": 1.2442,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3229294312693722,
+      "grad_norm": 0.1505689095161175,
+      "learning_rate": 0.00015828255671626837,
+      "loss": 1.2768,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3230015137317091,
+      "grad_norm": 0.11449436543475458,
+      "learning_rate": 0.0001582635832962212,
+      "loss": 1.2584,
+      "step": 4481
+    },
+    {
+      "epoch": 0.32307359619404596,
+      "grad_norm": 0.10253171593742949,
+      "learning_rate": 0.0001582446067002875,
+      "loss": 1.3074,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3231456786563829,
+      "grad_norm": 0.09909891386466761,
+      "learning_rate": 0.00015822562692950165,
+      "loss": 1.2832,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3232177611187198,
+      "grad_norm": 0.09944433192849811,
+      "learning_rate": 0.00015820664398489824,
+      "loss": 1.2789,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3232898435810567,
+      "grad_norm": 0.10545953751411018,
+      "learning_rate": 0.00015818765786751197,
+      "loss": 1.276,
+      "step": 4485
+    },
+    {
+      "epoch": 0.32336192604339364,
+      "grad_norm": 0.10857258438010671,
+      "learning_rate": 0.0001581686685783778,
+      "loss": 1.354,
+      "step": 4486
+    },
+    {
+      "epoch": 0.32343400850573056,
+      "grad_norm": 0.11986932413547109,
+      "learning_rate": 0.00015814967611853083,
+      "loss": 1.3507,
+      "step": 4487
+    },
+    {
+      "epoch": 0.3235060909680675,
+      "grad_norm": 0.10887914918307583,
+      "learning_rate": 0.00015813068048900627,
+      "loss": 1.3168,
+      "step": 4488
+    },
+    {
+      "epoch": 0.3235781734304044,
+      "grad_norm": 0.10801989448288696,
+      "learning_rate": 0.00015811168169083955,
+      "loss": 1.3544,
+      "step": 4489
+    },
+    {
+      "epoch": 0.3236502558927413,
+      "grad_norm": 0.1059143297592413,
+      "learning_rate": 0.00015809267972506633,
+      "loss": 1.2009,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3237223383550782,
+      "grad_norm": 0.11274868069792325,
+      "learning_rate": 0.00015807367459272234,
+      "loss": 1.2736,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3237944208174151,
+      "grad_norm": 0.10622702769563006,
+      "learning_rate": 0.00015805466629484355,
+      "loss": 1.3758,
+      "step": 4492
+    },
+    {
+      "epoch": 0.323866503279752,
+      "grad_norm": 0.11005229909270714,
+      "learning_rate": 0.00015803565483246606,
+      "loss": 1.2185,
+      "step": 4493
+    },
+    {
+      "epoch": 0.32393858574208895,
+      "grad_norm": 0.10829720414333548,
+      "learning_rate": 0.00015801664020662618,
+      "loss": 1.225,
+      "step": 4494
+    },
+    {
+      "epoch": 0.32401066820442587,
+      "grad_norm": 0.11165769873406169,
+      "learning_rate": 0.0001579976224183604,
+      "loss": 1.213,
+      "step": 4495
+    },
+    {
+      "epoch": 0.3240827506667628,
+      "grad_norm": 0.11922672669405739,
+      "learning_rate": 0.0001579786014687053,
+      "loss": 1.3511,
+      "step": 4496
+    },
+    {
+      "epoch": 0.3241548331290997,
+      "grad_norm": 0.10763803600483597,
+      "learning_rate": 0.00015795957735869774,
+      "loss": 1.2913,
+      "step": 4497
+    },
+    {
+      "epoch": 0.3242269155914366,
+      "grad_norm": 0.10857746688937384,
+      "learning_rate": 0.0001579405500893747,
+      "loss": 1.3308,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3242989980537735,
+      "grad_norm": 0.1045728865900748,
+      "learning_rate": 0.00015792151966177335,
+      "loss": 1.22,
+      "step": 4499
+    },
+    {
+      "epoch": 0.3243710805161104,
+      "grad_norm": 0.11310382511442352,
+      "learning_rate": 0.00015790248607693098,
+      "loss": 1.2889,
+      "step": 4500
+    },
+    {
+      "epoch": 0.32444316297844733,
+      "grad_norm": 0.11068922249562547,
+      "learning_rate": 0.00015788344933588512,
+      "loss": 1.2823,
+      "step": 4501
+    },
+    {
+      "epoch": 0.32451524544078425,
+      "grad_norm": 0.19204722684084144,
+      "learning_rate": 0.0001578644094396734,
+      "loss": 1.3144,
+      "step": 4502
+    },
+    {
+      "epoch": 0.32458732790312117,
+      "grad_norm": 0.10680150375196527,
+      "learning_rate": 0.0001578453663893337,
+      "loss": 1.3096,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3246594103654581,
+      "grad_norm": 0.1128244111557443,
+      "learning_rate": 0.00015782632018590407,
+      "loss": 1.1496,
+      "step": 4504
+    },
+    {
+      "epoch": 0.324731492827795,
+      "grad_norm": 0.09727225063126442,
+      "learning_rate": 0.0001578072708304226,
+      "loss": 1.3469,
+      "step": 4505
+    },
+    {
+      "epoch": 0.32480357529013193,
+      "grad_norm": 0.10531487642722244,
+      "learning_rate": 0.00015778821832392777,
+      "loss": 1.4024,
+      "step": 4506
+    },
+    {
+      "epoch": 0.32487565775246885,
+      "grad_norm": 0.10406510452703287,
+      "learning_rate": 0.00015776916266745805,
+      "loss": 1.2653,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3249477402148057,
+      "grad_norm": 0.2082270776153536,
+      "learning_rate": 0.0001577501038620521,
+      "loss": 1.3623,
+      "step": 4508
+    },
+    {
+      "epoch": 0.32501982267714263,
+      "grad_norm": 0.11879378571160691,
+      "learning_rate": 0.0001577310419087489,
+      "loss": 1.2901,
+      "step": 4509
+    },
+    {
+      "epoch": 0.32509190513947955,
+      "grad_norm": 0.2190485241184293,
+      "learning_rate": 0.00015771197680858738,
+      "loss": 1.2252,
+      "step": 4510
+    },
+    {
+      "epoch": 0.3251639876018165,
+      "grad_norm": 0.1079119296068348,
+      "learning_rate": 0.00015769290856260684,
+      "loss": 1.3202,
+      "step": 4511
+    },
+    {
+      "epoch": 0.3252360700641534,
+      "grad_norm": 0.1259797932917266,
+      "learning_rate": 0.00015767383717184668,
+      "loss": 1.3127,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3253081525264903,
+      "grad_norm": 0.11060959404873796,
+      "learning_rate": 0.0001576547626373464,
+      "loss": 1.4147,
+      "step": 4513
+    },
+    {
+      "epoch": 0.32538023498882723,
+      "grad_norm": 0.11032347910099422,
+      "learning_rate": 0.00015763568496014575,
+      "loss": 1.2373,
+      "step": 4514
+    },
+    {
+      "epoch": 0.32545231745116415,
+      "grad_norm": 0.18297647705398762,
+      "learning_rate": 0.00015761660414128468,
+      "loss": 1.3071,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3255243999135011,
+      "grad_norm": 0.11333498122206453,
+      "learning_rate": 0.0001575975201818032,
+      "loss": 1.2666,
+      "step": 4516
+    },
+    {
+      "epoch": 0.32559648237583794,
+      "grad_norm": 0.11788576795037554,
+      "learning_rate": 0.00015757843308274164,
+      "loss": 1.2629,
+      "step": 4517
+    },
+    {
+      "epoch": 0.32566856483817486,
+      "grad_norm": 0.12437597339102992,
+      "learning_rate": 0.00015755934284514029,
+      "loss": 1.2483,
+      "step": 4518
+    },
+    {
+      "epoch": 0.3257406473005118,
+      "grad_norm": 0.11008351079875778,
+      "learning_rate": 0.00015754024947003985,
+      "loss": 1.3862,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3258127297628487,
+      "grad_norm": 0.12130290866810182,
+      "learning_rate": 0.00015752115295848104,
+      "loss": 1.168,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3258848122251856,
+      "grad_norm": 0.1008203804511937,
+      "learning_rate": 0.0001575020533115048,
+      "loss": 1.2608,
+      "step": 4521
+    },
+    {
+      "epoch": 0.32595689468752254,
+      "grad_norm": 0.10504315587846093,
+      "learning_rate": 0.00015748295053015222,
+      "loss": 1.2948,
+      "step": 4522
+    },
+    {
+      "epoch": 0.32602897714985946,
+      "grad_norm": 0.12319893714668922,
+      "learning_rate": 0.00015746384461546458,
+      "loss": 1.3308,
+      "step": 4523
+    },
+    {
+      "epoch": 0.3261010596121964,
+      "grad_norm": 0.10626233203880195,
+      "learning_rate": 0.0001574447355684834,
+      "loss": 1.2818,
+      "step": 4524
+    },
+    {
+      "epoch": 0.32617314207453324,
+      "grad_norm": 0.10731632644350152,
+      "learning_rate": 0.0001574256233902501,
+      "loss": 1.3232,
+      "step": 4525
+    },
+    {
+      "epoch": 0.32624522453687016,
+      "grad_norm": 0.11831079714284479,
+      "learning_rate": 0.00015740650808180662,
+      "loss": 1.1913,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3263173069992071,
+      "grad_norm": 0.1043074751236657,
+      "learning_rate": 0.00015738738964419488,
+      "loss": 1.2037,
+      "step": 4527
+    },
+    {
+      "epoch": 0.326389389461544,
+      "grad_norm": 0.12122798310191535,
+      "learning_rate": 0.00015736826807845702,
+      "loss": 1.2462,
+      "step": 4528
+    },
+    {
+      "epoch": 0.3264614719238809,
+      "grad_norm": 0.13310395003512682,
+      "learning_rate": 0.0001573491433856353,
+      "loss": 1.2578,
+      "step": 4529
+    },
+    {
+      "epoch": 0.32653355438621784,
+      "grad_norm": 0.120095471643051,
+      "learning_rate": 0.0001573300155667722,
+      "loss": 1.278,
+      "step": 4530
+    },
+    {
+      "epoch": 0.32660563684855476,
+      "grad_norm": 0.11317515456268434,
+      "learning_rate": 0.00015731088462291038,
+      "loss": 1.2613,
+      "step": 4531
+    },
+    {
+      "epoch": 0.3266777193108917,
+      "grad_norm": 0.10793875876894067,
+      "learning_rate": 0.00015729175055509263,
+      "loss": 1.3888,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3267498017732286,
+      "grad_norm": 0.11343746524106064,
+      "learning_rate": 0.0001572726133643619,
+      "loss": 1.2509,
+      "step": 4533
+    },
+    {
+      "epoch": 0.32682188423556546,
+      "grad_norm": 0.11527216872739307,
+      "learning_rate": 0.00015725347305176138,
+      "loss": 1.2819,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3268939666979024,
+      "grad_norm": 0.1091516994846865,
+      "learning_rate": 0.0001572343296183344,
+      "loss": 1.2976,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3269660491602393,
+      "grad_norm": 0.09868236841717122,
+      "learning_rate": 0.00015721518306512438,
+      "loss": 1.2316,
+      "step": 4536
+    },
+    {
+      "epoch": 0.3270381316225762,
+      "grad_norm": 0.11061054244423897,
+      "learning_rate": 0.00015719603339317507,
+      "loss": 1.1934,
+      "step": 4537
+    },
+    {
+      "epoch": 0.32711021408491314,
+      "grad_norm": 0.10473200460048931,
+      "learning_rate": 0.00015717688060353023,
+      "loss": 1.2574,
+      "step": 4538
+    },
+    {
+      "epoch": 0.32718229654725006,
+      "grad_norm": 0.11123585739394054,
+      "learning_rate": 0.00015715772469723386,
+      "loss": 1.2937,
+      "step": 4539
+    },
+    {
+      "epoch": 0.327254379009587,
+      "grad_norm": 0.11531750972335142,
+      "learning_rate": 0.00015713856567533014,
+      "loss": 1.335,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3273264614719239,
+      "grad_norm": 0.11808501573423631,
+      "learning_rate": 0.00015711940353886346,
+      "loss": 1.322,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3273985439342608,
+      "grad_norm": 0.17812794575960736,
+      "learning_rate": 0.00015710023828887824,
+      "loss": 1.3019,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3274706263965977,
+      "grad_norm": 0.13854451931469652,
+      "learning_rate": 0.00015708106992641922,
+      "loss": 1.3083,
+      "step": 4543
+    },
+    {
+      "epoch": 0.3275427088589346,
+      "grad_norm": 0.10334568936708573,
+      "learning_rate": 0.0001570618984525312,
+      "loss": 1.3403,
+      "step": 4544
+    },
+    {
+      "epoch": 0.3276147913212715,
+      "grad_norm": 0.10838510460192037,
+      "learning_rate": 0.00015704272386825926,
+      "loss": 1.2628,
+      "step": 4545
+    },
+    {
+      "epoch": 0.32768687378360845,
+      "grad_norm": 0.15218301429041656,
+      "learning_rate": 0.00015702354617464854,
+      "loss": 1.3098,
+      "step": 4546
+    },
+    {
+      "epoch": 0.32775895624594537,
+      "grad_norm": 0.10584298536070486,
+      "learning_rate": 0.00015700436537274438,
+      "loss": 1.3421,
+      "step": 4547
+    },
+    {
+      "epoch": 0.3278310387082823,
+      "grad_norm": 0.12113299489082702,
+      "learning_rate": 0.0001569851814635924,
+      "loss": 1.3175,
+      "step": 4548
+    },
+    {
+      "epoch": 0.3279031211706192,
+      "grad_norm": 0.09957478045300971,
+      "learning_rate": 0.00015696599444823815,
+      "loss": 1.3476,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3279752036329561,
+      "grad_norm": 0.11713429721840318,
+      "learning_rate": 0.00015694680432772762,
+      "loss": 1.3715,
+      "step": 4550
+    },
+    {
+      "epoch": 0.328047286095293,
+      "grad_norm": 0.10916701720785471,
+      "learning_rate": 0.00015692761110310674,
+      "loss": 1.3933,
+      "step": 4551
+    },
+    {
+      "epoch": 0.3281193685576299,
+      "grad_norm": 0.10228755212394006,
+      "learning_rate": 0.0001569084147754218,
+      "loss": 1.3964,
+      "step": 4552
+    },
+    {
+      "epoch": 0.32819145101996683,
+      "grad_norm": 0.10688567007529223,
+      "learning_rate": 0.00015688921534571913,
+      "loss": 1.3213,
+      "step": 4553
+    },
+    {
+      "epoch": 0.32826353348230375,
+      "grad_norm": 0.10217115287637131,
+      "learning_rate": 0.0001568700128150453,
+      "loss": 1.2473,
+      "step": 4554
+    },
+    {
+      "epoch": 0.32833561594464067,
+      "grad_norm": 0.11184787960517824,
+      "learning_rate": 0.00015685080718444699,
+      "loss": 1.2329,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3284076984069776,
+      "grad_norm": 0.10783187767002858,
+      "learning_rate": 0.0001568315984549711,
+      "loss": 1.3602,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3284797808693145,
+      "grad_norm": 0.1253486566247752,
+      "learning_rate": 0.00015681238662766463,
+      "loss": 1.3004,
+      "step": 4557
+    },
+    {
+      "epoch": 0.32855186333165143,
+      "grad_norm": 0.10660585737615551,
+      "learning_rate": 0.00015679317170357485,
+      "loss": 1.3516,
+      "step": 4558
+    },
+    {
+      "epoch": 0.32862394579398835,
+      "grad_norm": 0.12284622088928313,
+      "learning_rate": 0.00015677395368374913,
+      "loss": 1.2905,
+      "step": 4559
+    },
+    {
+      "epoch": 0.3286960282563252,
+      "grad_norm": 0.1138930621288531,
+      "learning_rate": 0.00015675473256923498,
+      "loss": 1.2904,
+      "step": 4560
+    },
+    {
+      "epoch": 0.32876811071866213,
+      "grad_norm": 0.11887901664736798,
+      "learning_rate": 0.0001567355083610802,
+      "loss": 1.3535,
+      "step": 4561
+    },
+    {
+      "epoch": 0.32884019318099905,
+      "grad_norm": 0.11073045004962433,
+      "learning_rate": 0.00015671628106033264,
+      "loss": 1.2786,
+      "step": 4562
+    },
+    {
+      "epoch": 0.32891227564333597,
+      "grad_norm": 0.12611572758513814,
+      "learning_rate": 0.00015669705066804037,
+      "loss": 1.2731,
+      "step": 4563
+    },
+    {
+      "epoch": 0.3289843581056729,
+      "grad_norm": 0.09976001996192148,
+      "learning_rate": 0.00015667781718525157,
+      "loss": 1.2377,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3290564405680098,
+      "grad_norm": 0.10628250885137369,
+      "learning_rate": 0.00015665858061301476,
+      "loss": 1.2583,
+      "step": 4565
+    },
+    {
+      "epoch": 0.32912852303034673,
+      "grad_norm": 0.12928724658912413,
+      "learning_rate": 0.00015663934095237837,
+      "loss": 1.1783,
+      "step": 4566
+    },
+    {
+      "epoch": 0.32920060549268365,
+      "grad_norm": 0.10229679883785048,
+      "learning_rate": 0.00015662009820439118,
+      "loss": 1.2661,
+      "step": 4567
+    },
+    {
+      "epoch": 0.32927268795502057,
+      "grad_norm": 0.09985832765380165,
+      "learning_rate": 0.00015660085237010214,
+      "loss": 1.1943,
+      "step": 4568
+    },
+    {
+      "epoch": 0.32934477041735744,
+      "grad_norm": 0.10002410556396166,
+      "learning_rate": 0.00015658160345056023,
+      "loss": 1.2835,
+      "step": 4569
+    },
+    {
+      "epoch": 0.32941685287969436,
+      "grad_norm": 0.10005446838202177,
+      "learning_rate": 0.00015656235144681478,
+      "loss": 1.3251,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3294889353420313,
+      "grad_norm": 0.19461484824253258,
+      "learning_rate": 0.00015654309635991512,
+      "loss": 1.2727,
+      "step": 4571
+    },
+    {
+      "epoch": 0.3295610178043682,
+      "grad_norm": 0.13751156976181939,
+      "learning_rate": 0.00015652383819091088,
+      "loss": 1.2876,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3296331002667051,
+      "grad_norm": 0.11035776542396022,
+      "learning_rate": 0.0001565045769408518,
+      "loss": 1.2409,
+      "step": 4573
+    },
+    {
+      "epoch": 0.32970518272904203,
+      "grad_norm": 0.14916726209491915,
+      "learning_rate": 0.00015648531261078775,
+      "loss": 1.1973,
+      "step": 4574
+    },
+    {
+      "epoch": 0.32977726519137895,
+      "grad_norm": 0.12493927482295895,
+      "learning_rate": 0.00015646604520176883,
+      "loss": 1.2293,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3298493476537159,
+      "grad_norm": 0.1139919952282943,
+      "learning_rate": 0.00015644677471484528,
+      "loss": 1.2708,
+      "step": 4576
+    },
+    {
+      "epoch": 0.32992143011605274,
+      "grad_norm": 0.11109175502789423,
+      "learning_rate": 0.00015642750115106753,
+      "loss": 1.2735,
+      "step": 4577
+    },
+    {
+      "epoch": 0.32999351257838966,
+      "grad_norm": 0.11797706159039006,
+      "learning_rate": 0.00015640822451148614,
+      "loss": 1.339,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3300655950407266,
+      "grad_norm": 0.1138259680197833,
+      "learning_rate": 0.0001563889447971519,
+      "loss": 1.2926,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3301376775030635,
+      "grad_norm": 0.10916199687579348,
+      "learning_rate": 0.0001563696620091157,
+      "loss": 1.217,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3302097599654004,
+      "grad_norm": 0.11246002213663926,
+      "learning_rate": 0.00015635037614842857,
+      "loss": 1.3613,
+      "step": 4581
+    },
+    {
+      "epoch": 0.33028184242773734,
+      "grad_norm": 0.12170735601948635,
+      "learning_rate": 0.00015633108721614187,
+      "loss": 1.2721,
+      "step": 4582
+    },
+    {
+      "epoch": 0.33035392489007426,
+      "grad_norm": 0.13061344280879156,
+      "learning_rate": 0.00015631179521330695,
+      "loss": 1.3562,
+      "step": 4583
+    },
+    {
+      "epoch": 0.3304260073524112,
+      "grad_norm": 0.12372824291604081,
+      "learning_rate": 0.00015629250014097542,
+      "loss": 1.2526,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3304980898147481,
+      "grad_norm": 0.11386664156413621,
+      "learning_rate": 0.000156273202000199,
+      "loss": 1.2127,
+      "step": 4585
+    },
+    {
+      "epoch": 0.33057017227708496,
+      "grad_norm": 0.11624929461041333,
+      "learning_rate": 0.00015625390079202961,
+      "loss": 1.2717,
+      "step": 4586
+    },
+    {
+      "epoch": 0.3306422547394219,
+      "grad_norm": 0.10892346235527592,
+      "learning_rate": 0.0001562345965175194,
+      "loss": 1.2801,
+      "step": 4587
+    },
+    {
+      "epoch": 0.3307143372017588,
+      "grad_norm": 0.13314880126588594,
+      "learning_rate": 0.0001562152891777206,
+      "loss": 1.4086,
+      "step": 4588
+    },
+    {
+      "epoch": 0.3307864196640957,
+      "grad_norm": 0.11495667424040529,
+      "learning_rate": 0.00015619597877368564,
+      "loss": 1.3772,
+      "step": 4589
+    },
+    {
+      "epoch": 0.33085850212643264,
+      "grad_norm": 0.10605949420916118,
+      "learning_rate": 0.0001561766653064671,
+      "loss": 1.3627,
+      "step": 4590
+    },
+    {
+      "epoch": 0.33093058458876956,
+      "grad_norm": 0.11429086591575902,
+      "learning_rate": 0.00015615734877711768,
+      "loss": 1.2606,
+      "step": 4591
+    },
+    {
+      "epoch": 0.3310026670511065,
+      "grad_norm": 0.10204391933912305,
+      "learning_rate": 0.0001561380291866904,
+      "loss": 1.365,
+      "step": 4592
+    },
+    {
+      "epoch": 0.3310747495134434,
+      "grad_norm": 0.10640001205419075,
+      "learning_rate": 0.00015611870653623825,
+      "loss": 1.3512,
+      "step": 4593
+    },
+    {
+      "epoch": 0.3311468319757803,
+      "grad_norm": 0.11136616128114898,
+      "learning_rate": 0.00015609938082681457,
+      "loss": 1.3356,
+      "step": 4594
+    },
+    {
+      "epoch": 0.3312189144381172,
+      "grad_norm": 0.11828366639136537,
+      "learning_rate": 0.00015608005205947274,
+      "loss": 1.2507,
+      "step": 4595
+    },
+    {
+      "epoch": 0.3312909969004541,
+      "grad_norm": 0.11805814239837376,
+      "learning_rate": 0.0001560607202352664,
+      "loss": 1.3237,
+      "step": 4596
+    },
+    {
+      "epoch": 0.331363079362791,
+      "grad_norm": 0.10796411174586586,
+      "learning_rate": 0.00015604138535524925,
+      "loss": 1.3398,
+      "step": 4597
+    },
+    {
+      "epoch": 0.33143516182512794,
+      "grad_norm": 0.11716585583113512,
+      "learning_rate": 0.00015602204742047525,
+      "loss": 1.2982,
+      "step": 4598
+    },
+    {
+      "epoch": 0.33150724428746486,
+      "grad_norm": 0.11361104018616207,
+      "learning_rate": 0.00015600270643199846,
+      "loss": 1.3487,
+      "step": 4599
+    },
+    {
+      "epoch": 0.3315793267498018,
+      "grad_norm": 0.10968072579650233,
+      "learning_rate": 0.00015598336239087317,
+      "loss": 1.3059,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3316514092121387,
+      "grad_norm": 0.12090120554765847,
+      "learning_rate": 0.0001559640152981538,
+      "loss": 1.3201,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3317234916744756,
+      "grad_norm": 0.1060923792992949,
+      "learning_rate": 0.0001559446651548949,
+      "loss": 1.2977,
+      "step": 4602
+    },
+    {
+      "epoch": 0.3317955741368125,
+      "grad_norm": 0.11575146833288955,
+      "learning_rate": 0.00015592531196215127,
+      "loss": 1.3218,
+      "step": 4603
+    },
+    {
+      "epoch": 0.3318676565991494,
+      "grad_norm": 0.12041663075270592,
+      "learning_rate": 0.0001559059557209778,
+      "loss": 1.3241,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3319397390614863,
+      "grad_norm": 0.12946853848096831,
+      "learning_rate": 0.0001558865964324296,
+      "loss": 1.3122,
+      "step": 4605
+    },
+    {
+      "epoch": 0.33201182152382325,
+      "grad_norm": 0.11448161184578493,
+      "learning_rate": 0.00015586723409756194,
+      "loss": 1.3232,
+      "step": 4606
+    },
+    {
+      "epoch": 0.33208390398616017,
+      "grad_norm": 0.1250739413186021,
+      "learning_rate": 0.0001558478687174302,
+      "loss": 1.2771,
+      "step": 4607
+    },
+    {
+      "epoch": 0.3321559864484971,
+      "grad_norm": 0.10465418846109498,
+      "learning_rate": 0.00015582850029308997,
+      "loss": 1.3307,
+      "step": 4608
+    },
+    {
+      "epoch": 0.332228068910834,
+      "grad_norm": 0.10831280528273966,
+      "learning_rate": 0.00015580912882559705,
+      "loss": 1.3767,
+      "step": 4609
+    },
+    {
+      "epoch": 0.3323001513731709,
+      "grad_norm": 0.10414886770437713,
+      "learning_rate": 0.00015578975431600732,
+      "loss": 1.3765,
+      "step": 4610
+    },
+    {
+      "epoch": 0.33237223383550785,
+      "grad_norm": 0.11287085634359861,
+      "learning_rate": 0.00015577037676537688,
+      "loss": 1.2648,
+      "step": 4611
+    },
+    {
+      "epoch": 0.3324443162978447,
+      "grad_norm": 0.10342811205834802,
+      "learning_rate": 0.00015575099617476195,
+      "loss": 1.2925,
+      "step": 4612
+    },
+    {
+      "epoch": 0.33251639876018163,
+      "grad_norm": 0.10593278458752686,
+      "learning_rate": 0.00015573161254521896,
+      "loss": 1.4629,
+      "step": 4613
+    },
+    {
+      "epoch": 0.33258848122251855,
+      "grad_norm": 0.11367172480203827,
+      "learning_rate": 0.00015571222587780451,
+      "loss": 1.2605,
+      "step": 4614
+    },
+    {
+      "epoch": 0.33266056368485547,
+      "grad_norm": 0.11809575203847977,
+      "learning_rate": 0.00015569283617357534,
+      "loss": 1.2198,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3327326461471924,
+      "grad_norm": 0.11068741369083007,
+      "learning_rate": 0.00015567344343358835,
+      "loss": 1.2609,
+      "step": 4616
+    },
+    {
+      "epoch": 0.3328047286095293,
+      "grad_norm": 0.1337617213611393,
+      "learning_rate": 0.00015565404765890061,
+      "loss": 1.2891,
+      "step": 4617
+    },
+    {
+      "epoch": 0.33287681107186623,
+      "grad_norm": 0.11499153632150112,
+      "learning_rate": 0.0001556346488505694,
+      "loss": 1.3478,
+      "step": 4618
+    },
+    {
+      "epoch": 0.33294889353420315,
+      "grad_norm": 0.10785277854889283,
+      "learning_rate": 0.00015561524700965208,
+      "loss": 1.3371,
+      "step": 4619
+    },
+    {
+      "epoch": 0.33302097599654007,
+      "grad_norm": 0.10562026678488673,
+      "learning_rate": 0.00015559584213720626,
+      "loss": 1.3493,
+      "step": 4620
+    },
+    {
+      "epoch": 0.33309305845887693,
+      "grad_norm": 0.11669770139933522,
+      "learning_rate": 0.00015557643423428968,
+      "loss": 1.4258,
+      "step": 4621
+    },
+    {
+      "epoch": 0.33316514092121385,
+      "grad_norm": 0.11612434289254274,
+      "learning_rate": 0.00015555702330196023,
+      "loss": 1.329,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3332372233835508,
+      "grad_norm": 0.10478113388758492,
+      "learning_rate": 0.000155537609341276,
+      "loss": 1.3144,
+      "step": 4623
+    },
+    {
+      "epoch": 0.3333093058458877,
+      "grad_norm": 0.11072080771734154,
+      "learning_rate": 0.00015551819235329515,
+      "loss": 1.251,
+      "step": 4624
+    },
+    {
+      "epoch": 0.3333813883082246,
+      "grad_norm": 0.19999710923460276,
+      "learning_rate": 0.00015549877233907622,
+      "loss": 1.3022,
+      "step": 4625
+    },
+    {
+      "epoch": 0.33345347077056153,
+      "grad_norm": 0.1105741320774731,
+      "learning_rate": 0.0001554793492996776,
+      "loss": 1.2347,
+      "step": 4626
+    },
+    {
+      "epoch": 0.33352555323289845,
+      "grad_norm": 0.10653204641480604,
+      "learning_rate": 0.00015545992323615817,
+      "loss": 1.3991,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3335976356952354,
+      "grad_norm": 0.10995259630138102,
+      "learning_rate": 0.00015544049414957674,
+      "loss": 1.266,
+      "step": 4628
+    },
+    {
+      "epoch": 0.33366971815757224,
+      "grad_norm": 0.11079027060334463,
+      "learning_rate": 0.00015542106204099242,
+      "loss": 1.37,
+      "step": 4629
+    },
+    {
+      "epoch": 0.33374180061990916,
+      "grad_norm": 0.09730468265718155,
+      "learning_rate": 0.0001554016269114644,
+      "loss": 1.2541,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3338138830822461,
+      "grad_norm": 0.12966037635002578,
+      "learning_rate": 0.00015538218876205205,
+      "loss": 1.3065,
+      "step": 4631
+    },
+    {
+      "epoch": 0.333885965544583,
+      "grad_norm": 0.1187432362310655,
+      "learning_rate": 0.00015536274759381496,
+      "loss": 1.3269,
+      "step": 4632
+    },
+    {
+      "epoch": 0.3339580480069199,
+      "grad_norm": 0.10439319727679368,
+      "learning_rate": 0.00015534330340781285,
+      "loss": 1.371,
+      "step": 4633
+    },
+    {
+      "epoch": 0.33403013046925684,
+      "grad_norm": 0.1269156431402139,
+      "learning_rate": 0.00015532385620510558,
+      "loss": 1.3244,
+      "step": 4634
+    },
+    {
+      "epoch": 0.33410221293159376,
+      "grad_norm": 0.11338072625850118,
+      "learning_rate": 0.0001553044059867532,
+      "loss": 1.3055,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3341742953939307,
+      "grad_norm": 0.11488676475893986,
+      "learning_rate": 0.00015528495275381597,
+      "loss": 1.3475,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3342463778562676,
+      "grad_norm": 0.11271114182886566,
+      "learning_rate": 0.00015526549650735416,
+      "loss": 1.2792,
+      "step": 4637
+    },
+    {
+      "epoch": 0.33431846031860446,
+      "grad_norm": 0.11226155935477235,
+      "learning_rate": 0.00015524603724842843,
+      "loss": 1.2915,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3343905427809414,
+      "grad_norm": 0.10006448523037065,
+      "learning_rate": 0.00015522657497809935,
+      "loss": 1.3679,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3344626252432783,
+      "grad_norm": 0.11116451358914246,
+      "learning_rate": 0.00015520710969742793,
+      "loss": 1.3351,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3345347077056152,
+      "grad_norm": 0.1080544224766959,
+      "learning_rate": 0.0001551876414074751,
+      "loss": 1.2553,
+      "step": 4641
+    },
+    {
+      "epoch": 0.33460679016795214,
+      "grad_norm": 0.11525417126629642,
+      "learning_rate": 0.00015516817010930212,
+      "loss": 1.2475,
+      "step": 4642
+    },
+    {
+      "epoch": 0.33467887263028906,
+      "grad_norm": 0.10862349992379697,
+      "learning_rate": 0.0001551486958039703,
+      "loss": 1.2213,
+      "step": 4643
+    },
+    {
+      "epoch": 0.334750955092626,
+      "grad_norm": 0.10384678488827419,
+      "learning_rate": 0.0001551292184925412,
+      "loss": 1.2574,
+      "step": 4644
+    },
+    {
+      "epoch": 0.3348230375549629,
+      "grad_norm": 0.10932483558618485,
+      "learning_rate": 0.00015510973817607648,
+      "loss": 1.3337,
+      "step": 4645
+    },
+    {
+      "epoch": 0.33489512001729976,
+      "grad_norm": 0.10019776567152214,
+      "learning_rate": 0.00015509025485563798,
+      "loss": 1.3678,
+      "step": 4646
+    },
+    {
+      "epoch": 0.3349672024796367,
+      "grad_norm": 0.09846739679238993,
+      "learning_rate": 0.00015507076853228777,
+      "loss": 1.285,
+      "step": 4647
+    },
+    {
+      "epoch": 0.3350392849419736,
+      "grad_norm": 0.09633486282454115,
+      "learning_rate": 0.00015505127920708797,
+      "loss": 1.3665,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3351113674043105,
+      "grad_norm": 0.1048868737393167,
+      "learning_rate": 0.000155031786881101,
+      "loss": 1.2675,
+      "step": 4649
+    },
+    {
+      "epoch": 0.33518344986664744,
+      "grad_norm": 0.11322016002447047,
+      "learning_rate": 0.00015501229155538926,
+      "loss": 1.3669,
+      "step": 4650
+    },
+    {
+      "epoch": 0.33525553232898436,
+      "grad_norm": 0.11797787992179919,
+      "learning_rate": 0.0001549927932310155,
+      "loss": 1.3261,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3353276147913213,
+      "grad_norm": 0.11020196087944185,
+      "learning_rate": 0.00015497329190904252,
+      "loss": 1.3325,
+      "step": 4652
+    },
+    {
+      "epoch": 0.3353996972536582,
+      "grad_norm": 0.1123414297246681,
+      "learning_rate": 0.00015495378759053336,
+      "loss": 1.3195,
+      "step": 4653
+    },
+    {
+      "epoch": 0.3354717797159951,
+      "grad_norm": 0.10069736343799353,
+      "learning_rate": 0.0001549342802765511,
+      "loss": 1.272,
+      "step": 4654
+    },
+    {
+      "epoch": 0.335543862178332,
+      "grad_norm": 0.11532935734204221,
+      "learning_rate": 0.00015491476996815914,
+      "loss": 1.1622,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3356159446406689,
+      "grad_norm": 0.09414184109111037,
+      "learning_rate": 0.00015489525666642092,
+      "loss": 1.2272,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3356880271030058,
+      "grad_norm": 0.11055298551145724,
+      "learning_rate": 0.00015487574037240012,
+      "loss": 1.3923,
+      "step": 4657
+    },
+    {
+      "epoch": 0.33576010956534275,
+      "grad_norm": 0.10908905654531821,
+      "learning_rate": 0.00015485622108716052,
+      "loss": 1.3167,
+      "step": 4658
+    },
+    {
+      "epoch": 0.33583219202767967,
+      "grad_norm": 0.10801898792645998,
+      "learning_rate": 0.00015483669881176612,
+      "loss": 1.2595,
+      "step": 4659
+    },
+    {
+      "epoch": 0.3359042744900166,
+      "grad_norm": 0.11410337706368545,
+      "learning_rate": 0.00015481717354728106,
+      "loss": 1.2941,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3359763569523535,
+      "grad_norm": 0.11405483254520825,
+      "learning_rate": 0.0001547976452947696,
+      "loss": 1.375,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3360484394146904,
+      "grad_norm": 0.11457340998083805,
+      "learning_rate": 0.00015477811405529627,
+      "loss": 1.2225,
+      "step": 4662
+    },
+    {
+      "epoch": 0.33612052187702735,
+      "grad_norm": 0.1000107772469063,
+      "learning_rate": 0.00015475857982992568,
+      "loss": 1.3427,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3361926043393642,
+      "grad_norm": 0.11082519555941328,
+      "learning_rate": 0.00015473904261972257,
+      "loss": 1.3057,
+      "step": 4664
+    },
+    {
+      "epoch": 0.33626468680170113,
+      "grad_norm": 0.10414841346307357,
+      "learning_rate": 0.00015471950242575196,
+      "loss": 1.2896,
+      "step": 4665
+    },
+    {
+      "epoch": 0.33633676926403805,
+      "grad_norm": 0.10928049098726973,
+      "learning_rate": 0.0001546999592490789,
+      "loss": 1.2645,
+      "step": 4666
+    },
+    {
+      "epoch": 0.33640885172637497,
+      "grad_norm": 0.10108609145408011,
+      "learning_rate": 0.00015468041309076873,
+      "loss": 1.368,
+      "step": 4667
+    },
+    {
+      "epoch": 0.3364809341887119,
+      "grad_norm": 0.11562150161768267,
+      "learning_rate": 0.00015466086395188682,
+      "loss": 1.281,
+      "step": 4668
+    },
+    {
+      "epoch": 0.3365530166510488,
+      "grad_norm": 0.10964307552415384,
+      "learning_rate": 0.00015464131183349887,
+      "loss": 1.2816,
+      "step": 4669
+    },
+    {
+      "epoch": 0.33662509911338573,
+      "grad_norm": 0.10271890012218503,
+      "learning_rate": 0.00015462175673667053,
+      "loss": 1.3276,
+      "step": 4670
+    },
+    {
+      "epoch": 0.33669718157572265,
+      "grad_norm": 0.13581141363624713,
+      "learning_rate": 0.00015460219866246785,
+      "loss": 1.1709,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3367692640380595,
+      "grad_norm": 0.11447954931411988,
+      "learning_rate": 0.00015458263761195678,
+      "loss": 1.3473,
+      "step": 4672
+    },
+    {
+      "epoch": 0.33684134650039643,
+      "grad_norm": 0.11185638184488758,
+      "learning_rate": 0.00015456307358620373,
+      "loss": 1.3193,
+      "step": 4673
+    },
+    {
+      "epoch": 0.33691342896273335,
+      "grad_norm": 0.12190372491900793,
+      "learning_rate": 0.00015454350658627498,
+      "loss": 1.3516,
+      "step": 4674
+    },
+    {
+      "epoch": 0.33698551142507027,
+      "grad_norm": 0.10725943013133193,
+      "learning_rate": 0.00015452393661323722,
+      "loss": 1.2919,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3370575938874072,
+      "grad_norm": 0.12068209855901309,
+      "learning_rate": 0.00015450436366815705,
+      "loss": 1.2193,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3371296763497441,
+      "grad_norm": 0.11117895707142948,
+      "learning_rate": 0.00015448478775210149,
+      "loss": 1.3194,
+      "step": 4677
+    },
+    {
+      "epoch": 0.33720175881208103,
+      "grad_norm": 0.10339962212750307,
+      "learning_rate": 0.00015446520886613758,
+      "loss": 1.2898,
+      "step": 4678
+    },
+    {
+      "epoch": 0.33727384127441795,
+      "grad_norm": 0.1013894983994576,
+      "learning_rate": 0.0001544456270113325,
+      "loss": 1.2919,
+      "step": 4679
+    },
+    {
+      "epoch": 0.33734592373675487,
+      "grad_norm": 0.11302444387930255,
+      "learning_rate": 0.0001544260421887537,
+      "loss": 1.2681,
+      "step": 4680
+    },
+    {
+      "epoch": 0.33741800619909174,
+      "grad_norm": 0.11707531101949406,
+      "learning_rate": 0.00015440645439946863,
+      "loss": 1.395,
+      "step": 4681
+    },
+    {
+      "epoch": 0.33749008866142866,
+      "grad_norm": 0.10883450434612496,
+      "learning_rate": 0.00015438686364454513,
+      "loss": 1.2928,
+      "step": 4682
+    },
+    {
+      "epoch": 0.3375621711237656,
+      "grad_norm": 0.11384975352564576,
+      "learning_rate": 0.00015436726992505096,
+      "loss": 1.3929,
+      "step": 4683
+    },
+    {
+      "epoch": 0.3376342535861025,
+      "grad_norm": 0.11598596340654309,
+      "learning_rate": 0.00015434767324205422,
+      "loss": 1.2559,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3377063360484394,
+      "grad_norm": 0.12261150185279766,
+      "learning_rate": 0.00015432807359662304,
+      "loss": 1.2478,
+      "step": 4685
+    },
+    {
+      "epoch": 0.33777841851077633,
+      "grad_norm": 0.10658661194825024,
+      "learning_rate": 0.00015430847098982584,
+      "loss": 1.2235,
+      "step": 4686
+    },
+    {
+      "epoch": 0.33785050097311325,
+      "grad_norm": 0.10374571745170705,
+      "learning_rate": 0.0001542888654227311,
+      "loss": 1.3812,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3379225834354502,
+      "grad_norm": 0.10682635030707262,
+      "learning_rate": 0.00015426925689640756,
+      "loss": 1.3504,
+      "step": 4688
+    },
+    {
+      "epoch": 0.3379946658977871,
+      "grad_norm": 0.1141156583122279,
+      "learning_rate": 0.00015424964541192399,
+      "loss": 1.3463,
+      "step": 4689
+    },
+    {
+      "epoch": 0.33806674836012396,
+      "grad_norm": 0.1159143427259434,
+      "learning_rate": 0.0001542300309703494,
+      "loss": 1.3563,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3381388308224609,
+      "grad_norm": 0.11473231761302931,
+      "learning_rate": 0.00015421041357275298,
+      "loss": 1.2453,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3382109132847978,
+      "grad_norm": 0.14556236095727842,
+      "learning_rate": 0.00015419079322020402,
+      "loss": 1.3157,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3382829957471347,
+      "grad_norm": 0.14253919410188695,
+      "learning_rate": 0.00015417116991377203,
+      "loss": 1.2957,
+      "step": 4693
+    },
+    {
+      "epoch": 0.33835507820947164,
+      "grad_norm": 0.1091670412312519,
+      "learning_rate": 0.00015415154365452664,
+      "loss": 1.2747,
+      "step": 4694
+    },
+    {
+      "epoch": 0.33842716067180856,
+      "grad_norm": 0.11760758307629353,
+      "learning_rate": 0.0001541319144435377,
+      "loss": 1.3334,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3384992431341455,
+      "grad_norm": 0.11107585973802642,
+      "learning_rate": 0.0001541122822818751,
+      "loss": 1.3342,
+      "step": 4696
+    },
+    {
+      "epoch": 0.3385713255964824,
+      "grad_norm": 0.1081439198213073,
+      "learning_rate": 0.00015409264717060906,
+      "loss": 1.3132,
+      "step": 4697
+    },
+    {
+      "epoch": 0.33864340805881926,
+      "grad_norm": 0.1144268361501076,
+      "learning_rate": 0.0001540730091108098,
+      "loss": 1.268,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3387154905211562,
+      "grad_norm": 0.1289179028689081,
+      "learning_rate": 0.0001540533681035478,
+      "loss": 1.3138,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3387875729834931,
+      "grad_norm": 0.10496340457131981,
+      "learning_rate": 0.00015403372414989368,
+      "loss": 1.3054,
+      "step": 4700
+    },
+    {
+      "epoch": 0.33885965544583,
+      "grad_norm": 0.11034192702058541,
+      "learning_rate": 0.00015401407725091815,
+      "loss": 1.2992,
+      "step": 4701
+    },
+    {
+      "epoch": 0.33893173790816694,
+      "grad_norm": 0.11046248821558687,
+      "learning_rate": 0.0001539944274076922,
+      "loss": 1.2385,
+      "step": 4702
+    },
+    {
+      "epoch": 0.33900382037050386,
+      "grad_norm": 0.11558173708698448,
+      "learning_rate": 0.00015397477462128693,
+      "loss": 1.3502,
+      "step": 4703
+    },
+    {
+      "epoch": 0.3390759028328408,
+      "grad_norm": 0.13069384718752017,
+      "learning_rate": 0.00015395511889277355,
+      "loss": 1.372,
+      "step": 4704
+    },
+    {
+      "epoch": 0.3391479852951777,
+      "grad_norm": 0.10998674736940203,
+      "learning_rate": 0.00015393546022322353,
+      "loss": 1.363,
+      "step": 4705
+    },
+    {
+      "epoch": 0.3392200677575146,
+      "grad_norm": 0.1351523069155148,
+      "learning_rate": 0.00015391579861370836,
+      "loss": 1.3281,
+      "step": 4706
+    },
+    {
+      "epoch": 0.3392921502198515,
+      "grad_norm": 0.11544920914123431,
+      "learning_rate": 0.00015389613406529984,
+      "loss": 1.2342,
+      "step": 4707
+    },
+    {
+      "epoch": 0.3393642326821884,
+      "grad_norm": 0.10538629628527386,
+      "learning_rate": 0.00015387646657906987,
+      "loss": 1.288,
+      "step": 4708
+    },
+    {
+      "epoch": 0.3394363151445253,
+      "grad_norm": 0.11516008398792175,
+      "learning_rate": 0.00015385679615609042,
+      "loss": 1.3524,
+      "step": 4709
+    },
+    {
+      "epoch": 0.33950839760686224,
+      "grad_norm": 0.14504461583983966,
+      "learning_rate": 0.0001538371227974338,
+      "loss": 1.4209,
+      "step": 4710
+    },
+    {
+      "epoch": 0.33958048006919916,
+      "grad_norm": 0.12740813190298114,
+      "learning_rate": 0.0001538174465041724,
+      "loss": 1.2731,
+      "step": 4711
+    },
+    {
+      "epoch": 0.3396525625315361,
+      "grad_norm": 0.10969104928705896,
+      "learning_rate": 0.0001537977672773786,
+      "loss": 1.2622,
+      "step": 4712
+    },
+    {
+      "epoch": 0.339724644993873,
+      "grad_norm": 0.13941777220317134,
+      "learning_rate": 0.00015377808511812525,
+      "loss": 1.3382,
+      "step": 4713
+    },
+    {
+      "epoch": 0.3397967274562099,
+      "grad_norm": 0.11536643958499139,
+      "learning_rate": 0.0001537584000274851,
+      "loss": 1.2522,
+      "step": 4714
+    },
+    {
+      "epoch": 0.33986880991854684,
+      "grad_norm": 0.10209659792898332,
+      "learning_rate": 0.00015373871200653131,
+      "loss": 1.2601,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3399408923808837,
+      "grad_norm": 0.11633090092253419,
+      "learning_rate": 0.00015371902105633687,
+      "loss": 1.3009,
+      "step": 4716
+    },
+    {
+      "epoch": 0.34001297484322063,
+      "grad_norm": 0.11126561896124974,
+      "learning_rate": 0.00015369932717797523,
+      "loss": 1.2818,
+      "step": 4717
+    },
+    {
+      "epoch": 0.34008505730555755,
+      "grad_norm": 0.12461884438453627,
+      "learning_rate": 0.00015367963037251983,
+      "loss": 1.3171,
+      "step": 4718
+    },
+    {
+      "epoch": 0.34015713976789447,
+      "grad_norm": 0.10617907268654697,
+      "learning_rate": 0.00015365993064104437,
+      "loss": 1.3609,
+      "step": 4719
+    },
+    {
+      "epoch": 0.3402292222302314,
+      "grad_norm": 0.11711206816913108,
+      "learning_rate": 0.0001536402279846226,
+      "loss": 1.369,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3403013046925683,
+      "grad_norm": 0.11043397253103411,
+      "learning_rate": 0.00015362052240432857,
+      "loss": 1.2743,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3403733871549052,
+      "grad_norm": 0.10582791994610986,
+      "learning_rate": 0.00015360081390123639,
+      "loss": 1.2832,
+      "step": 4722
+    },
+    {
+      "epoch": 0.34044546961724215,
+      "grad_norm": 0.14120181060233383,
+      "learning_rate": 0.00015358110247642025,
+      "loss": 1.2589,
+      "step": 4723
+    },
+    {
+      "epoch": 0.340517552079579,
+      "grad_norm": 0.11624392947482759,
+      "learning_rate": 0.0001535613881309547,
+      "loss": 1.273,
+      "step": 4724
+    },
+    {
+      "epoch": 0.34058963454191593,
+      "grad_norm": 0.12556665573469208,
+      "learning_rate": 0.00015354167086591432,
+      "loss": 1.2989,
+      "step": 4725
+    },
+    {
+      "epoch": 0.34066171700425285,
+      "grad_norm": 0.1038412126330174,
+      "learning_rate": 0.0001535219506823739,
+      "loss": 1.3187,
+      "step": 4726
+    },
+    {
+      "epoch": 0.34073379946658977,
+      "grad_norm": 0.10981342402676436,
+      "learning_rate": 0.00015350222758140833,
+      "loss": 1.3903,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3408058819289267,
+      "grad_norm": 0.10874525410975465,
+      "learning_rate": 0.00015348250156409273,
+      "loss": 1.3376,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3408779643912636,
+      "grad_norm": 0.1086852461730026,
+      "learning_rate": 0.0001534627726315023,
+      "loss": 1.3483,
+      "step": 4729
+    },
+    {
+      "epoch": 0.34095004685360053,
+      "grad_norm": 0.11993767335710884,
+      "learning_rate": 0.0001534430407847125,
+      "loss": 1.2812,
+      "step": 4730
+    },
+    {
+      "epoch": 0.34102212931593745,
+      "grad_norm": 0.11154067834481499,
+      "learning_rate": 0.00015342330602479884,
+      "loss": 1.2472,
+      "step": 4731
+    },
+    {
+      "epoch": 0.34109421177827437,
+      "grad_norm": 0.1320765056346942,
+      "learning_rate": 0.00015340356835283708,
+      "loss": 1.3348,
+      "step": 4732
+    },
+    {
+      "epoch": 0.34116629424061123,
+      "grad_norm": 0.1139495590953944,
+      "learning_rate": 0.00015338382776990307,
+      "loss": 1.3554,
+      "step": 4733
+    },
+    {
+      "epoch": 0.34123837670294815,
+      "grad_norm": 0.11451356680167195,
+      "learning_rate": 0.00015336408427707286,
+      "loss": 1.203,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3413104591652851,
+      "grad_norm": 0.12336910523897383,
+      "learning_rate": 0.00015334433787542266,
+      "loss": 1.3639,
+      "step": 4735
+    },
+    {
+      "epoch": 0.341382541627622,
+      "grad_norm": 0.1356335301352883,
+      "learning_rate": 0.00015332458856602877,
+      "loss": 1.4009,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3414546240899589,
+      "grad_norm": 0.10523986061352855,
+      "learning_rate": 0.00015330483634996782,
+      "loss": 1.2872,
+      "step": 4737
+    },
+    {
+      "epoch": 0.34152670655229583,
+      "grad_norm": 0.10859285399877992,
+      "learning_rate": 0.00015328508122831636,
+      "loss": 1.2626,
+      "step": 4738
+    },
+    {
+      "epoch": 0.34159878901463275,
+      "grad_norm": 0.1123454843177674,
+      "learning_rate": 0.0001532653232021513,
+      "loss": 1.3378,
+      "step": 4739
+    },
+    {
+      "epoch": 0.3416708714769697,
+      "grad_norm": 0.10908855948310318,
+      "learning_rate": 0.00015324556227254958,
+      "loss": 1.3314,
+      "step": 4740
+    },
+    {
+      "epoch": 0.3417429539393066,
+      "grad_norm": 0.11420268907190506,
+      "learning_rate": 0.00015322579844058838,
+      "loss": 1.305,
+      "step": 4741
+    },
+    {
+      "epoch": 0.34181503640164346,
+      "grad_norm": 0.12861058018878793,
+      "learning_rate": 0.00015320603170734497,
+      "loss": 1.2865,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3418871188639804,
+      "grad_norm": 0.11338126190508424,
+      "learning_rate": 0.00015318626207389685,
+      "loss": 1.2666,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3419592013263173,
+      "grad_norm": 0.11621346222634021,
+      "learning_rate": 0.00015316648954132164,
+      "loss": 1.2717,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3420312837886542,
+      "grad_norm": 0.11313704343990819,
+      "learning_rate": 0.00015314671411069712,
+      "loss": 1.2921,
+      "step": 4745
+    },
+    {
+      "epoch": 0.34210336625099114,
+      "grad_norm": 0.12862656657782753,
+      "learning_rate": 0.0001531269357831012,
+      "loss": 1.3364,
+      "step": 4746
+    },
+    {
+      "epoch": 0.34217544871332806,
+      "grad_norm": 0.12825286163836772,
+      "learning_rate": 0.00015310715455961197,
+      "loss": 1.2342,
+      "step": 4747
+    },
+    {
+      "epoch": 0.342247531175665,
+      "grad_norm": 0.1088921842532809,
+      "learning_rate": 0.00015308737044130775,
+      "loss": 1.2199,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3423196136380019,
+      "grad_norm": 0.131033734021761,
+      "learning_rate": 0.00015306758342926687,
+      "loss": 1.2828,
+      "step": 4749
+    },
+    {
+      "epoch": 0.34239169610033876,
+      "grad_norm": 0.10643600040846365,
+      "learning_rate": 0.00015304779352456795,
+      "loss": 1.346,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3424637785626757,
+      "grad_norm": 0.1187831427637337,
+      "learning_rate": 0.0001530280007282897,
+      "loss": 1.3857,
+      "step": 4751
+    },
+    {
+      "epoch": 0.3425358610250126,
+      "grad_norm": 0.13978196342456672,
+      "learning_rate": 0.000153008205041511,
+      "loss": 1.3623,
+      "step": 4752
+    },
+    {
+      "epoch": 0.3426079434873495,
+      "grad_norm": 0.10441008399056198,
+      "learning_rate": 0.0001529884064653109,
+      "loss": 1.3049,
+      "step": 4753
+    },
+    {
+      "epoch": 0.34268002594968644,
+      "grad_norm": 0.11158241680170777,
+      "learning_rate": 0.00015296860500076863,
+      "loss": 1.2727,
+      "step": 4754
+    },
+    {
+      "epoch": 0.34275210841202336,
+      "grad_norm": 0.13126324463772338,
+      "learning_rate": 0.00015294880064896352,
+      "loss": 1.2388,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3428241908743603,
+      "grad_norm": 0.11490052947662094,
+      "learning_rate": 0.00015292899341097505,
+      "loss": 1.3182,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3428962733366972,
+      "grad_norm": 0.11104398036935295,
+      "learning_rate": 0.00015290918328788298,
+      "loss": 1.2273,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3429683557990341,
+      "grad_norm": 0.11133302545409092,
+      "learning_rate": 0.000152889370280767,
+      "loss": 1.2899,
+      "step": 4758
+    },
+    {
+      "epoch": 0.343040438261371,
+      "grad_norm": 0.10660779650217325,
+      "learning_rate": 0.00015286955439070723,
+      "loss": 1.321,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3431125207237079,
+      "grad_norm": 0.1024018791278549,
+      "learning_rate": 0.00015284973561878373,
+      "loss": 1.3608,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3431846031860448,
+      "grad_norm": 0.1087457871370613,
+      "learning_rate": 0.00015282991396607686,
+      "loss": 1.3251,
+      "step": 4761
+    },
+    {
+      "epoch": 0.34325668564838174,
+      "grad_norm": 0.10926187653117267,
+      "learning_rate": 0.000152810089433667,
+      "loss": 1.4637,
+      "step": 4762
+    },
+    {
+      "epoch": 0.34332876811071866,
+      "grad_norm": 0.10589033094610617,
+      "learning_rate": 0.00015279026202263488,
+      "loss": 1.2513,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3434008505730556,
+      "grad_norm": 0.11400683247059427,
+      "learning_rate": 0.00015277043173406117,
+      "loss": 1.2607,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3434729330353925,
+      "grad_norm": 0.1250377037396089,
+      "learning_rate": 0.00015275059856902684,
+      "loss": 1.3101,
+      "step": 4765
+    },
+    {
+      "epoch": 0.3435450154977294,
+      "grad_norm": 0.1171523679146189,
+      "learning_rate": 0.00015273076252861296,
+      "loss": 1.2794,
+      "step": 4766
+    },
+    {
+      "epoch": 0.34361709796006634,
+      "grad_norm": 0.11366259397495802,
+      "learning_rate": 0.00015271092361390077,
+      "loss": 1.3108,
+      "step": 4767
+    },
+    {
+      "epoch": 0.3436891804224032,
+      "grad_norm": 0.10742984910629841,
+      "learning_rate": 0.00015269108182597172,
+      "loss": 1.2772,
+      "step": 4768
+    },
+    {
+      "epoch": 0.3437612628847401,
+      "grad_norm": 0.10900606757354225,
+      "learning_rate": 0.00015267123716590728,
+      "loss": 1.3466,
+      "step": 4769
+    },
+    {
+      "epoch": 0.34383334534707705,
+      "grad_norm": 0.11746208067418198,
+      "learning_rate": 0.0001526513896347892,
+      "loss": 1.2896,
+      "step": 4770
+    },
+    {
+      "epoch": 0.34390542780941397,
+      "grad_norm": 0.12970762393875473,
+      "learning_rate": 0.00015263153923369938,
+      "loss": 1.2499,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3439775102717509,
+      "grad_norm": 0.11222319610850016,
+      "learning_rate": 0.0001526116859637198,
+      "loss": 1.2788,
+      "step": 4772
+    },
+    {
+      "epoch": 0.3440495927340878,
+      "grad_norm": 0.11048326061801268,
+      "learning_rate": 0.0001525918298259327,
+      "loss": 1.2382,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3441216751964247,
+      "grad_norm": 0.12293516300053015,
+      "learning_rate": 0.00015257197082142032,
+      "loss": 1.3247,
+      "step": 4774
+    },
+    {
+      "epoch": 0.34419375765876165,
+      "grad_norm": 0.11208878363806093,
+      "learning_rate": 0.00015255210895126525,
+      "loss": 1.3171,
+      "step": 4775
+    },
+    {
+      "epoch": 0.3442658401210985,
+      "grad_norm": 0.11068703016199129,
+      "learning_rate": 0.00015253224421655004,
+      "loss": 1.2548,
+      "step": 4776
+    },
+    {
+      "epoch": 0.34433792258343543,
+      "grad_norm": 0.11781515201857302,
+      "learning_rate": 0.0001525123766183576,
+      "loss": 1.245,
+      "step": 4777
+    },
+    {
+      "epoch": 0.34441000504577235,
+      "grad_norm": 0.12119758354255589,
+      "learning_rate": 0.00015249250615777084,
+      "loss": 1.2017,
+      "step": 4778
+    },
+    {
+      "epoch": 0.34448208750810927,
+      "grad_norm": 0.1140131360167995,
+      "learning_rate": 0.00015247263283587286,
+      "loss": 1.3025,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3445541699704462,
+      "grad_norm": 0.11081458183203045,
+      "learning_rate": 0.00015245275665374698,
+      "loss": 1.2629,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3446262524327831,
+      "grad_norm": 0.10175799109934316,
+      "learning_rate": 0.0001524328776124766,
+      "loss": 1.2912,
+      "step": 4781
+    },
+    {
+      "epoch": 0.34469833489512003,
+      "grad_norm": 0.10660873717275102,
+      "learning_rate": 0.0001524129957131453,
+      "loss": 1.307,
+      "step": 4782
+    },
+    {
+      "epoch": 0.34477041735745695,
+      "grad_norm": 0.11122727085421172,
+      "learning_rate": 0.00015239311095683683,
+      "loss": 1.237,
+      "step": 4783
+    },
+    {
+      "epoch": 0.34484249981979387,
+      "grad_norm": 0.11784826794466144,
+      "learning_rate": 0.0001523732233446351,
+      "loss": 1.2833,
+      "step": 4784
+    },
+    {
+      "epoch": 0.34491458228213073,
+      "grad_norm": 0.11356160434054714,
+      "learning_rate": 0.00015235333287762418,
+      "loss": 1.2578,
+      "step": 4785
+    },
+    {
+      "epoch": 0.34498666474446765,
+      "grad_norm": 0.1134156569539209,
+      "learning_rate": 0.00015233343955688822,
+      "loss": 1.2447,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3450587472068046,
+      "grad_norm": 0.10280298204398805,
+      "learning_rate": 0.00015231354338351164,
+      "loss": 1.2911,
+      "step": 4787
+    },
+    {
+      "epoch": 0.3451308296691415,
+      "grad_norm": 0.10274247310762159,
+      "learning_rate": 0.00015229364435857893,
+      "loss": 1.1987,
+      "step": 4788
+    },
+    {
+      "epoch": 0.3452029121314784,
+      "grad_norm": 0.12227549497174293,
+      "learning_rate": 0.00015227374248317474,
+      "loss": 1.3144,
+      "step": 4789
+    },
+    {
+      "epoch": 0.34527499459381533,
+      "grad_norm": 0.11289035173060223,
+      "learning_rate": 0.00015225383775838397,
+      "loss": 1.2864,
+      "step": 4790
+    },
+    {
+      "epoch": 0.34534707705615225,
+      "grad_norm": 0.11240777254813371,
+      "learning_rate": 0.00015223393018529152,
+      "loss": 1.3511,
+      "step": 4791
+    },
+    {
+      "epoch": 0.34541915951848917,
+      "grad_norm": 0.11938379646565718,
+      "learning_rate": 0.00015221401976498264,
+      "loss": 1.3626,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3454912419808261,
+      "grad_norm": 0.108914993698882,
+      "learning_rate": 0.00015219410649854252,
+      "loss": 1.3064,
+      "step": 4793
+    },
+    {
+      "epoch": 0.34556332444316296,
+      "grad_norm": 0.1214845946384155,
+      "learning_rate": 0.00015217419038705666,
+      "loss": 1.3635,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3456354069054999,
+      "grad_norm": 0.1366017589560892,
+      "learning_rate": 0.00015215427143161066,
+      "loss": 1.3302,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3457074893678368,
+      "grad_norm": 0.12642338407060633,
+      "learning_rate": 0.0001521343496332903,
+      "loss": 1.3354,
+      "step": 4796
+    },
+    {
+      "epoch": 0.3457795718301737,
+      "grad_norm": 0.12075004436301819,
+      "learning_rate": 0.00015211442499318145,
+      "loss": 1.2974,
+      "step": 4797
+    },
+    {
+      "epoch": 0.34585165429251064,
+      "grad_norm": 0.11452107325929436,
+      "learning_rate": 0.00015209449751237025,
+      "loss": 1.2996,
+      "step": 4798
+    },
+    {
+      "epoch": 0.34592373675484756,
+      "grad_norm": 0.12039415721453918,
+      "learning_rate": 0.00015207456719194284,
+      "loss": 1.3736,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3459958192171845,
+      "grad_norm": 0.11575538777541397,
+      "learning_rate": 0.00015205463403298566,
+      "loss": 1.2644,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3460679016795214,
+      "grad_norm": 0.17610549694251548,
+      "learning_rate": 0.00015203469803658521,
+      "loss": 1.3283,
+      "step": 4801
+    },
+    {
+      "epoch": 0.34613998414185826,
+      "grad_norm": 0.2126424548784234,
+      "learning_rate": 0.00015201475920382823,
+      "loss": 1.2405,
+      "step": 4802
+    },
+    {
+      "epoch": 0.3462120666041952,
+      "grad_norm": 0.1113500488627485,
+      "learning_rate": 0.00015199481753580153,
+      "loss": 1.2994,
+      "step": 4803
+    },
+    {
+      "epoch": 0.3462841490665321,
+      "grad_norm": 0.1202629758892065,
+      "learning_rate": 0.00015197487303359212,
+      "loss": 1.3551,
+      "step": 4804
+    },
+    {
+      "epoch": 0.346356231528869,
+      "grad_norm": 0.11730603242827209,
+      "learning_rate": 0.00015195492569828716,
+      "loss": 1.3833,
+      "step": 4805
+    },
+    {
+      "epoch": 0.34642831399120594,
+      "grad_norm": 0.10605897475855941,
+      "learning_rate": 0.0001519349755309739,
+      "loss": 1.2145,
+      "step": 4806
+    },
+    {
+      "epoch": 0.34650039645354286,
+      "grad_norm": 0.11510414258844469,
+      "learning_rate": 0.0001519150225327399,
+      "loss": 1.1816,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3465724789158798,
+      "grad_norm": 0.2529341660717632,
+      "learning_rate": 0.00015189506670467272,
+      "loss": 1.38,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3466445613782167,
+      "grad_norm": 0.11741769251074059,
+      "learning_rate": 0.00015187510804786012,
+      "loss": 1.2944,
+      "step": 4809
+    },
+    {
+      "epoch": 0.3467166438405536,
+      "grad_norm": 0.11126238068651027,
+      "learning_rate": 0.00015185514656339006,
+      "loss": 1.3178,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3467887263028905,
+      "grad_norm": 0.10693493334437522,
+      "learning_rate": 0.0001518351822523506,
+      "loss": 1.2875,
+      "step": 4811
+    },
+    {
+      "epoch": 0.3468608087652274,
+      "grad_norm": 0.10460120957241624,
+      "learning_rate": 0.00015181521511582996,
+      "loss": 1.2535,
+      "step": 4812
+    },
+    {
+      "epoch": 0.3469328912275643,
+      "grad_norm": 0.10984537920172784,
+      "learning_rate": 0.00015179524515491656,
+      "loss": 1.3351,
+      "step": 4813
+    },
+    {
+      "epoch": 0.34700497368990124,
+      "grad_norm": 0.11240202115734861,
+      "learning_rate": 0.0001517752723706989,
+      "loss": 1.406,
+      "step": 4814
+    },
+    {
+      "epoch": 0.34707705615223816,
+      "grad_norm": 0.10590818703846876,
+      "learning_rate": 0.0001517552967642657,
+      "loss": 1.24,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3471491386145751,
+      "grad_norm": 0.10503476766998758,
+      "learning_rate": 0.00015173531833670586,
+      "loss": 1.3234,
+      "step": 4816
+    },
+    {
+      "epoch": 0.347221221076912,
+      "grad_norm": 0.11781634058824988,
+      "learning_rate": 0.00015171533708910827,
+      "loss": 1.3425,
+      "step": 4817
+    },
+    {
+      "epoch": 0.3472933035392489,
+      "grad_norm": 0.1190092549971224,
+      "learning_rate": 0.00015169535302256222,
+      "loss": 1.305,
+      "step": 4818
+    },
+    {
+      "epoch": 0.34736538600158584,
+      "grad_norm": 0.10161969543328349,
+      "learning_rate": 0.00015167536613815692,
+      "loss": 1.2795,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3474374684639227,
+      "grad_norm": 0.11122994893257844,
+      "learning_rate": 0.00015165537643698184,
+      "loss": 1.2992,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3475095509262596,
+      "grad_norm": 0.11129085972890762,
+      "learning_rate": 0.00015163538392012663,
+      "loss": 1.2358,
+      "step": 4821
+    },
+    {
+      "epoch": 0.34758163338859654,
+      "grad_norm": 0.17803824808835625,
+      "learning_rate": 0.00015161538858868108,
+      "loss": 1.2086,
+      "step": 4822
+    },
+    {
+      "epoch": 0.34765371585093346,
+      "grad_norm": 0.18871568534600916,
+      "learning_rate": 0.00015159539044373504,
+      "loss": 1.3924,
+      "step": 4823
+    },
+    {
+      "epoch": 0.3477257983132704,
+      "grad_norm": 0.10864042166028924,
+      "learning_rate": 0.00015157538948637868,
+      "loss": 1.3129,
+      "step": 4824
+    },
+    {
+      "epoch": 0.3477978807756073,
+      "grad_norm": 0.09976568561782109,
+      "learning_rate": 0.00015155538571770218,
+      "loss": 1.1537,
+      "step": 4825
+    },
+    {
+      "epoch": 0.3478699632379442,
+      "grad_norm": 0.1679853082313214,
+      "learning_rate": 0.00015153537913879594,
+      "loss": 1.333,
+      "step": 4826
+    },
+    {
+      "epoch": 0.34794204570028114,
+      "grad_norm": 0.11983006952379462,
+      "learning_rate": 0.00015151536975075046,
+      "loss": 1.3597,
+      "step": 4827
+    },
+    {
+      "epoch": 0.348014128162618,
+      "grad_norm": 0.11474931787193936,
+      "learning_rate": 0.00015149535755465648,
+      "loss": 1.3176,
+      "step": 4828
+    },
+    {
+      "epoch": 0.34808621062495493,
+      "grad_norm": 0.11347817110096302,
+      "learning_rate": 0.0001514753425516048,
+      "loss": 1.3395,
+      "step": 4829
+    },
+    {
+      "epoch": 0.34815829308729185,
+      "grad_norm": 0.10875465352343042,
+      "learning_rate": 0.00015145532474268646,
+      "loss": 1.256,
+      "step": 4830
+    },
+    {
+      "epoch": 0.34823037554962877,
+      "grad_norm": 0.11264546898237116,
+      "learning_rate": 0.00015143530412899255,
+      "loss": 1.3227,
+      "step": 4831
+    },
+    {
+      "epoch": 0.3483024580119657,
+      "grad_norm": 0.12179267662533061,
+      "learning_rate": 0.00015141528071161448,
+      "loss": 1.3675,
+      "step": 4832
+    },
+    {
+      "epoch": 0.3483745404743026,
+      "grad_norm": 0.11296756559088962,
+      "learning_rate": 0.00015139525449164358,
+      "loss": 1.2837,
+      "step": 4833
+    },
+    {
+      "epoch": 0.3484466229366395,
+      "grad_norm": 0.11954693071945002,
+      "learning_rate": 0.00015137522547017157,
+      "loss": 1.3953,
+      "step": 4834
+    },
+    {
+      "epoch": 0.34851870539897645,
+      "grad_norm": 0.12116756292571682,
+      "learning_rate": 0.00015135519364829007,
+      "loss": 1.3216,
+      "step": 4835
+    },
+    {
+      "epoch": 0.34859078786131337,
+      "grad_norm": 0.1298217284702951,
+      "learning_rate": 0.00015133515902709116,
+      "loss": 1.2903,
+      "step": 4836
+    },
+    {
+      "epoch": 0.34866287032365023,
+      "grad_norm": 0.11218236924104659,
+      "learning_rate": 0.0001513151216076668,
+      "loss": 1.4098,
+      "step": 4837
+    },
+    {
+      "epoch": 0.34873495278598715,
+      "grad_norm": 0.11288524632455418,
+      "learning_rate": 0.00015129508139110923,
+      "loss": 1.2633,
+      "step": 4838
+    },
+    {
+      "epoch": 0.34880703524832407,
+      "grad_norm": 0.11843555434732499,
+      "learning_rate": 0.0001512750383785108,
+      "loss": 1.2835,
+      "step": 4839
+    },
+    {
+      "epoch": 0.348879117710661,
+      "grad_norm": 0.20772383689994267,
+      "learning_rate": 0.0001512549925709641,
+      "loss": 1.2376,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3489512001729979,
+      "grad_norm": 0.12324537657120543,
+      "learning_rate": 0.00015123494396956176,
+      "loss": 1.2926,
+      "step": 4841
+    },
+    {
+      "epoch": 0.34902328263533483,
+      "grad_norm": 0.30753184647013887,
+      "learning_rate": 0.00015121489257539662,
+      "loss": 1.2839,
+      "step": 4842
+    },
+    {
+      "epoch": 0.34909536509767175,
+      "grad_norm": 0.11599328159029315,
+      "learning_rate": 0.0001511948383895616,
+      "loss": 1.3241,
+      "step": 4843
+    },
+    {
+      "epoch": 0.34916744756000867,
+      "grad_norm": 0.10952384092557663,
+      "learning_rate": 0.00015117478141314994,
+      "loss": 1.2499,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3492395300223456,
+      "grad_norm": 0.09611627633954711,
+      "learning_rate": 0.00015115472164725486,
+      "loss": 1.2955,
+      "step": 4845
+    },
+    {
+      "epoch": 0.34931161248468245,
+      "grad_norm": 0.11211209974965194,
+      "learning_rate": 0.00015113465909296977,
+      "loss": 1.3437,
+      "step": 4846
+    },
+    {
+      "epoch": 0.3493836949470194,
+      "grad_norm": 0.11768307028714471,
+      "learning_rate": 0.00015111459375138835,
+      "loss": 1.3761,
+      "step": 4847
+    },
+    {
+      "epoch": 0.3494557774093563,
+      "grad_norm": 0.11608264033677632,
+      "learning_rate": 0.00015109452562360424,
+      "loss": 1.3446,
+      "step": 4848
+    },
+    {
+      "epoch": 0.3495278598716932,
+      "grad_norm": 0.10961778858908489,
+      "learning_rate": 0.0001510744547107114,
+      "loss": 1.34,
+      "step": 4849
+    },
+    {
+      "epoch": 0.34959994233403013,
+      "grad_norm": 0.11312346361943665,
+      "learning_rate": 0.00015105438101380383,
+      "loss": 1.3451,
+      "step": 4850
+    },
+    {
+      "epoch": 0.34967202479636705,
+      "grad_norm": 0.1233917030535717,
+      "learning_rate": 0.0001510343045339758,
+      "loss": 1.3015,
+      "step": 4851
+    },
+    {
+      "epoch": 0.349744107258704,
+      "grad_norm": 0.16092124044964265,
+      "learning_rate": 0.00015101422527232156,
+      "loss": 1.2753,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3498161897210409,
+      "grad_norm": 0.23553216157047113,
+      "learning_rate": 0.00015099414322993567,
+      "loss": 1.3929,
+      "step": 4853
+    },
+    {
+      "epoch": 0.34988827218337776,
+      "grad_norm": 0.10989352065961527,
+      "learning_rate": 0.00015097405840791276,
+      "loss": 1.3258,
+      "step": 4854
+    },
+    {
+      "epoch": 0.3499603546457147,
+      "grad_norm": 0.10638608274584488,
+      "learning_rate": 0.00015095397080734768,
+      "loss": 1.2991,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3500324371080516,
+      "grad_norm": 0.10787715263141694,
+      "learning_rate": 0.0001509338804293353,
+      "loss": 1.3005,
+      "step": 4856
+    },
+    {
+      "epoch": 0.3501045195703885,
+      "grad_norm": 0.1390666019028617,
+      "learning_rate": 0.0001509137872749708,
+      "loss": 1.3086,
+      "step": 4857
+    },
+    {
+      "epoch": 0.35017660203272544,
+      "grad_norm": 0.13592025800601282,
+      "learning_rate": 0.00015089369134534938,
+      "loss": 1.225,
+      "step": 4858
+    },
+    {
+      "epoch": 0.35024868449506236,
+      "grad_norm": 0.40870629285367777,
+      "learning_rate": 0.00015087359264156648,
+      "loss": 1.3068,
+      "step": 4859
+    },
+    {
+      "epoch": 0.3503207669573993,
+      "grad_norm": 0.236621928693066,
+      "learning_rate": 0.00015085349116471765,
+      "loss": 1.2801,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3503928494197362,
+      "grad_norm": 0.12136289356704799,
+      "learning_rate": 0.0001508333869158986,
+      "loss": 1.3274,
+      "step": 4861
+    },
+    {
+      "epoch": 0.3504649318820731,
+      "grad_norm": 0.10514750566145832,
+      "learning_rate": 0.00015081327989620522,
+      "loss": 1.2244,
+      "step": 4862
+    },
+    {
+      "epoch": 0.35053701434441,
+      "grad_norm": 0.1765812013791552,
+      "learning_rate": 0.0001507931701067335,
+      "loss": 1.3233,
+      "step": 4863
+    },
+    {
+      "epoch": 0.3506090968067469,
+      "grad_norm": 0.10789437065050933,
+      "learning_rate": 0.00015077305754857955,
+      "loss": 1.2712,
+      "step": 4864
+    },
+    {
+      "epoch": 0.3506811792690838,
+      "grad_norm": 0.11831332144354763,
+      "learning_rate": 0.0001507529422228398,
+      "loss": 1.2791,
+      "step": 4865
+    },
+    {
+      "epoch": 0.35075326173142074,
+      "grad_norm": 0.11971888389025172,
+      "learning_rate": 0.00015073282413061062,
+      "loss": 1.3362,
+      "step": 4866
+    },
+    {
+      "epoch": 0.35082534419375766,
+      "grad_norm": 0.10580012718734759,
+      "learning_rate": 0.00015071270327298865,
+      "loss": 1.2964,
+      "step": 4867
+    },
+    {
+      "epoch": 0.3508974266560946,
+      "grad_norm": 0.11570594789874941,
+      "learning_rate": 0.00015069257965107065,
+      "loss": 1.2132,
+      "step": 4868
+    },
+    {
+      "epoch": 0.3509695091184315,
+      "grad_norm": 0.10744957709280571,
+      "learning_rate": 0.00015067245326595358,
+      "loss": 1.2454,
+      "step": 4869
+    },
+    {
+      "epoch": 0.3510415915807684,
+      "grad_norm": 0.13758949972729115,
+      "learning_rate": 0.00015065232411873444,
+      "loss": 1.3188,
+      "step": 4870
+    },
+    {
+      "epoch": 0.35111367404310534,
+      "grad_norm": 0.12009258890351213,
+      "learning_rate": 0.00015063219221051055,
+      "loss": 1.2405,
+      "step": 4871
+    },
+    {
+      "epoch": 0.3511857565054422,
+      "grad_norm": 0.11504373118057457,
+      "learning_rate": 0.00015061205754237915,
+      "loss": 1.3161,
+      "step": 4872
+    },
+    {
+      "epoch": 0.3512578389677791,
+      "grad_norm": 0.14500442752930778,
+      "learning_rate": 0.00015059192011543786,
+      "loss": 1.245,
+      "step": 4873
+    },
+    {
+      "epoch": 0.35132992143011604,
+      "grad_norm": 0.10978751153397161,
+      "learning_rate": 0.0001505717799307843,
+      "loss": 1.2894,
+      "step": 4874
+    },
+    {
+      "epoch": 0.35140200389245296,
+      "grad_norm": 0.10429542351052459,
+      "learning_rate": 0.0001505516369895163,
+      "loss": 1.2654,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3514740863547899,
+      "grad_norm": 0.16705143450071705,
+      "learning_rate": 0.00015053149129273186,
+      "loss": 1.213,
+      "step": 4876
+    },
+    {
+      "epoch": 0.3515461688171268,
+      "grad_norm": 0.10770313065661419,
+      "learning_rate": 0.00015051134284152907,
+      "loss": 1.3284,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3516182512794637,
+      "grad_norm": 0.10252423268284905,
+      "learning_rate": 0.0001504911916370062,
+      "loss": 1.2567,
+      "step": 4878
+    },
+    {
+      "epoch": 0.35169033374180064,
+      "grad_norm": 0.10792366729873945,
+      "learning_rate": 0.0001504710376802617,
+      "loss": 1.3123,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3517624162041375,
+      "grad_norm": 0.11863873202350897,
+      "learning_rate": 0.00015045088097239407,
+      "loss": 1.3295,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3518344986664744,
+      "grad_norm": 0.12204251753232065,
+      "learning_rate": 0.00015043072151450212,
+      "loss": 1.3983,
+      "step": 4881
+    },
+    {
+      "epoch": 0.35190658112881135,
+      "grad_norm": 0.09132911920116664,
+      "learning_rate": 0.00015041055930768468,
+      "loss": 1.4185,
+      "step": 4882
+    },
+    {
+      "epoch": 0.35197866359114827,
+      "grad_norm": 0.24780181277625513,
+      "learning_rate": 0.00015039039435304078,
+      "loss": 1.2841,
+      "step": 4883
+    },
+    {
+      "epoch": 0.3520507460534852,
+      "grad_norm": 0.11250015877314717,
+      "learning_rate": 0.00015037022665166956,
+      "loss": 1.4483,
+      "step": 4884
+    },
+    {
+      "epoch": 0.3521228285158221,
+      "grad_norm": 0.11672717227675544,
+      "learning_rate": 0.0001503500562046704,
+      "loss": 1.2987,
+      "step": 4885
+    },
+    {
+      "epoch": 0.352194910978159,
+      "grad_norm": 0.10404742741128065,
+      "learning_rate": 0.00015032988301314268,
+      "loss": 1.346,
+      "step": 4886
+    },
+    {
+      "epoch": 0.35226699344049595,
+      "grad_norm": 0.1026809210774597,
+      "learning_rate": 0.00015030970707818612,
+      "loss": 1.3269,
+      "step": 4887
+    },
+    {
+      "epoch": 0.35233907590283287,
+      "grad_norm": 0.12146575367454787,
+      "learning_rate": 0.0001502895284009004,
+      "loss": 1.2593,
+      "step": 4888
+    },
+    {
+      "epoch": 0.35241115836516973,
+      "grad_norm": 0.11554302138683241,
+      "learning_rate": 0.0001502693469823855,
+      "loss": 1.2838,
+      "step": 4889
+    },
+    {
+      "epoch": 0.35248324082750665,
+      "grad_norm": 0.11718957622274023,
+      "learning_rate": 0.00015024916282374147,
+      "loss": 1.2024,
+      "step": 4890
+    },
+    {
+      "epoch": 0.35255532328984357,
+      "grad_norm": 0.10074094879512342,
+      "learning_rate": 0.00015022897592606853,
+      "loss": 1.2992,
+      "step": 4891
+    },
+    {
+      "epoch": 0.3526274057521805,
+      "grad_norm": 0.11417058560674768,
+      "learning_rate": 0.000150208786290467,
+      "loss": 1.283,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3526994882145174,
+      "grad_norm": 0.1477096972772979,
+      "learning_rate": 0.0001501885939180375,
+      "loss": 1.2829,
+      "step": 4893
+    },
+    {
+      "epoch": 0.35277157067685433,
+      "grad_norm": 0.11603361213846275,
+      "learning_rate": 0.00015016839880988055,
+      "loss": 1.354,
+      "step": 4894
+    },
+    {
+      "epoch": 0.35284365313919125,
+      "grad_norm": 0.10511122377423952,
+      "learning_rate": 0.0001501482009670971,
+      "loss": 1.331,
+      "step": 4895
+    },
+    {
+      "epoch": 0.35291573560152817,
+      "grad_norm": 0.12187372462326496,
+      "learning_rate": 0.00015012800039078807,
+      "loss": 1.3081,
+      "step": 4896
+    },
+    {
+      "epoch": 0.3529878180638651,
+      "grad_norm": 0.11032316088408871,
+      "learning_rate": 0.00015010779708205454,
+      "loss": 1.3023,
+      "step": 4897
+    },
+    {
+      "epoch": 0.35305990052620195,
+      "grad_norm": 0.104195289670521,
+      "learning_rate": 0.0001500875910419978,
+      "loss": 1.3317,
+      "step": 4898
+    },
+    {
+      "epoch": 0.3531319829885389,
+      "grad_norm": 0.11626164166278616,
+      "learning_rate": 0.00015006738227171923,
+      "loss": 1.3311,
+      "step": 4899
+    },
+    {
+      "epoch": 0.3532040654508758,
+      "grad_norm": 0.12151667494287866,
+      "learning_rate": 0.00015004717077232043,
+      "loss": 1.2728,
+      "step": 4900
+    },
+    {
+      "epoch": 0.3532761479132127,
+      "grad_norm": 0.10275794895652654,
+      "learning_rate": 0.00015002695654490306,
+      "loss": 1.2984,
+      "step": 4901
+    },
+    {
+      "epoch": 0.35334823037554963,
+      "grad_norm": 0.12045326158445888,
+      "learning_rate": 0.00015000673959056904,
+      "loss": 1.3931,
+      "step": 4902
+    },
+    {
+      "epoch": 0.35342031283788655,
+      "grad_norm": 0.12330101284463012,
+      "learning_rate": 0.0001499865199104203,
+      "loss": 1.2608,
+      "step": 4903
+    },
+    {
+      "epoch": 0.35349239530022347,
+      "grad_norm": 0.10136971128466621,
+      "learning_rate": 0.00014996629750555908,
+      "loss": 1.3106,
+      "step": 4904
+    },
+    {
+      "epoch": 0.3535644777625604,
+      "grad_norm": 0.1152335468658248,
+      "learning_rate": 0.00014994607237708762,
+      "loss": 1.2632,
+      "step": 4905
+    },
+    {
+      "epoch": 0.35363656022489726,
+      "grad_norm": 0.11672276358117131,
+      "learning_rate": 0.0001499258445261084,
+      "loss": 1.2126,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3537086426872342,
+      "grad_norm": 0.120741401020068,
+      "learning_rate": 0.00014990561395372397,
+      "loss": 1.4096,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3537807251495711,
+      "grad_norm": 0.11518811735287161,
+      "learning_rate": 0.00014988538066103713,
+      "loss": 1.2907,
+      "step": 4908
+    },
+    {
+      "epoch": 0.353852807611908,
+      "grad_norm": 0.10646906353550413,
+      "learning_rate": 0.00014986514464915076,
+      "loss": 1.3303,
+      "step": 4909
+    },
+    {
+      "epoch": 0.35392489007424494,
+      "grad_norm": 0.121334400689887,
+      "learning_rate": 0.00014984490591916786,
+      "loss": 1.358,
+      "step": 4910
+    },
+    {
+      "epoch": 0.35399697253658186,
+      "grad_norm": 0.13681185961642262,
+      "learning_rate": 0.00014982466447219171,
+      "loss": 1.2997,
+      "step": 4911
+    },
+    {
+      "epoch": 0.3540690549989188,
+      "grad_norm": 0.1449503974634814,
+      "learning_rate": 0.00014980442030932558,
+      "loss": 1.1737,
+      "step": 4912
+    },
+    {
+      "epoch": 0.3541411374612557,
+      "grad_norm": 0.17725925019859215,
+      "learning_rate": 0.000149784173431673,
+      "loss": 1.3232,
+      "step": 4913
+    },
+    {
+      "epoch": 0.3542132199235926,
+      "grad_norm": 0.1223027711012203,
+      "learning_rate": 0.00014976392384033756,
+      "loss": 1.3764,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3542853023859295,
+      "grad_norm": 0.11933270617576638,
+      "learning_rate": 0.0001497436715364231,
+      "loss": 1.2553,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3543573848482664,
+      "grad_norm": 0.11988613359545662,
+      "learning_rate": 0.00014972341652103347,
+      "loss": 1.3814,
+      "step": 4916
+    },
+    {
+      "epoch": 0.3544294673106033,
+      "grad_norm": 0.27875858076677057,
+      "learning_rate": 0.00014970315879527286,
+      "loss": 1.3312,
+      "step": 4917
+    },
+    {
+      "epoch": 0.35450154977294024,
+      "grad_norm": 0.11399674683538823,
+      "learning_rate": 0.0001496828983602454,
+      "loss": 1.2228,
+      "step": 4918
+    },
+    {
+      "epoch": 0.35457363223527716,
+      "grad_norm": 0.11518614300827956,
+      "learning_rate": 0.00014966263521705553,
+      "loss": 1.3803,
+      "step": 4919
+    },
+    {
+      "epoch": 0.3546457146976141,
+      "grad_norm": 0.11470841912956638,
+      "learning_rate": 0.00014964236936680774,
+      "loss": 1.2354,
+      "step": 4920
+    },
+    {
+      "epoch": 0.354717797159951,
+      "grad_norm": 0.102747112226017,
+      "learning_rate": 0.0001496221008106067,
+      "loss": 1.238,
+      "step": 4921
+    },
+    {
+      "epoch": 0.3547898796222879,
+      "grad_norm": 0.12636294375300094,
+      "learning_rate": 0.0001496018295495572,
+      "loss": 1.2408,
+      "step": 4922
+    },
+    {
+      "epoch": 0.35486196208462484,
+      "grad_norm": 0.1141404209720415,
+      "learning_rate": 0.0001495815555847643,
+      "loss": 1.3661,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3549340445469617,
+      "grad_norm": 0.12829491032797605,
+      "learning_rate": 0.000149561278917333,
+      "loss": 1.2893,
+      "step": 4924
+    },
+    {
+      "epoch": 0.3550061270092986,
+      "grad_norm": 0.10769301009146413,
+      "learning_rate": 0.00014954099954836864,
+      "loss": 1.2532,
+      "step": 4925
+    },
+    {
+      "epoch": 0.35507820947163554,
+      "grad_norm": 0.10604223800318839,
+      "learning_rate": 0.0001495207174789766,
+      "loss": 1.3304,
+      "step": 4926
+    },
+    {
+      "epoch": 0.35515029193397246,
+      "grad_norm": 0.11065466947383952,
+      "learning_rate": 0.00014950043271026243,
+      "loss": 1.3124,
+      "step": 4927
+    },
+    {
+      "epoch": 0.3552223743963094,
+      "grad_norm": 0.11719581689160774,
+      "learning_rate": 0.00014948014524333183,
+      "loss": 1.3056,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3552944568586463,
+      "grad_norm": 0.1104266767921119,
+      "learning_rate": 0.00014945985507929064,
+      "loss": 1.3131,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3553665393209832,
+      "grad_norm": 0.1083688156632212,
+      "learning_rate": 0.00014943956221924486,
+      "loss": 1.2519,
+      "step": 4930
+    },
+    {
+      "epoch": 0.35543862178332014,
+      "grad_norm": 0.11354418444572827,
+      "learning_rate": 0.00014941926666430067,
+      "loss": 1.3326,
+      "step": 4931
+    },
+    {
+      "epoch": 0.355510704245657,
+      "grad_norm": 0.1182303761855961,
+      "learning_rate": 0.00014939896841556431,
+      "loss": 1.3466,
+      "step": 4932
+    },
+    {
+      "epoch": 0.3555827867079939,
+      "grad_norm": 0.10914901619023029,
+      "learning_rate": 0.00014937866747414223,
+      "loss": 1.1381,
+      "step": 4933
+    },
+    {
+      "epoch": 0.35565486917033085,
+      "grad_norm": 0.1200155172175516,
+      "learning_rate": 0.00014935836384114107,
+      "loss": 1.3368,
+      "step": 4934
+    },
+    {
+      "epoch": 0.35572695163266776,
+      "grad_norm": 0.12591921322938157,
+      "learning_rate": 0.00014933805751766744,
+      "loss": 1.3526,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3557990340950047,
+      "grad_norm": 0.0969778417815132,
+      "learning_rate": 0.00014931774850482832,
+      "loss": 1.3969,
+      "step": 4936
+    },
+    {
+      "epoch": 0.3558711165573416,
+      "grad_norm": 0.1078749104438157,
+      "learning_rate": 0.0001492974368037307,
+      "loss": 1.2437,
+      "step": 4937
+    },
+    {
+      "epoch": 0.3559431990196785,
+      "grad_norm": 0.10868957482077597,
+      "learning_rate": 0.00014927712241548174,
+      "loss": 1.2776,
+      "step": 4938
+    },
+    {
+      "epoch": 0.35601528148201544,
+      "grad_norm": 0.12430008515033365,
+      "learning_rate": 0.00014925680534118876,
+      "loss": 1.3272,
+      "step": 4939
+    },
+    {
+      "epoch": 0.35608736394435236,
+      "grad_norm": 0.12036268564609674,
+      "learning_rate": 0.00014923648558195924,
+      "loss": 1.3486,
+      "step": 4940
+    },
+    {
+      "epoch": 0.35615944640668923,
+      "grad_norm": 0.098296378587748,
+      "learning_rate": 0.00014921616313890072,
+      "loss": 1.2945,
+      "step": 4941
+    },
+    {
+      "epoch": 0.35623152886902615,
+      "grad_norm": 0.11110311519147356,
+      "learning_rate": 0.00014919583801312108,
+      "loss": 1.203,
+      "step": 4942
+    },
+    {
+      "epoch": 0.35630361133136307,
+      "grad_norm": 0.12044455470421866,
+      "learning_rate": 0.0001491755102057281,
+      "loss": 1.3145,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3563756937937,
+      "grad_norm": 0.107178879690057,
+      "learning_rate": 0.0001491551797178299,
+      "loss": 1.2011,
+      "step": 4944
+    },
+    {
+      "epoch": 0.3564477762560369,
+      "grad_norm": 0.10813278684763984,
+      "learning_rate": 0.00014913484655053463,
+      "loss": 1.2696,
+      "step": 4945
+    },
+    {
+      "epoch": 0.3565198587183738,
+      "grad_norm": 0.15766314602179257,
+      "learning_rate": 0.00014911451070495068,
+      "loss": 1.282,
+      "step": 4946
+    },
+    {
+      "epoch": 0.35659194118071075,
+      "grad_norm": 0.10279053615980582,
+      "learning_rate": 0.00014909417218218648,
+      "loss": 1.2956,
+      "step": 4947
+    },
+    {
+      "epoch": 0.35666402364304767,
+      "grad_norm": 0.10629659121986473,
+      "learning_rate": 0.00014907383098335073,
+      "loss": 1.2973,
+      "step": 4948
+    },
+    {
+      "epoch": 0.3567361061053846,
+      "grad_norm": 0.11075384010046437,
+      "learning_rate": 0.0001490534871095521,
+      "loss": 1.3304,
+      "step": 4949
+    },
+    {
+      "epoch": 0.35680818856772145,
+      "grad_norm": 0.14150320662897256,
+      "learning_rate": 0.00014903314056189962,
+      "loss": 1.2376,
+      "step": 4950
+    },
+    {
+      "epoch": 0.35688027103005837,
+      "grad_norm": 0.10270574376309305,
+      "learning_rate": 0.00014901279134150233,
+      "loss": 1.2448,
+      "step": 4951
+    },
+    {
+      "epoch": 0.3569523534923953,
+      "grad_norm": 0.10479159477069956,
+      "learning_rate": 0.00014899243944946938,
+      "loss": 1.2559,
+      "step": 4952
+    },
+    {
+      "epoch": 0.3570244359547322,
+      "grad_norm": 0.12043966762959205,
+      "learning_rate": 0.0001489720848869102,
+      "loss": 1.2759,
+      "step": 4953
+    },
+    {
+      "epoch": 0.35709651841706913,
+      "grad_norm": 0.11163804730234551,
+      "learning_rate": 0.0001489517276549343,
+      "loss": 1.3297,
+      "step": 4954
+    },
+    {
+      "epoch": 0.35716860087940605,
+      "grad_norm": 0.1083675303459364,
+      "learning_rate": 0.0001489313677546513,
+      "loss": 1.2962,
+      "step": 4955
+    },
+    {
+      "epoch": 0.35724068334174297,
+      "grad_norm": 0.10600278901224401,
+      "learning_rate": 0.00014891100518717098,
+      "loss": 1.299,
+      "step": 4956
+    },
+    {
+      "epoch": 0.3573127658040799,
+      "grad_norm": 0.09907487084826555,
+      "learning_rate": 0.00014889063995360334,
+      "loss": 1.2989,
+      "step": 4957
+    },
+    {
+      "epoch": 0.35738484826641675,
+      "grad_norm": 0.11459422104851058,
+      "learning_rate": 0.00014887027205505842,
+      "loss": 1.3973,
+      "step": 4958
+    },
+    {
+      "epoch": 0.3574569307287537,
+      "grad_norm": 0.12189791612064096,
+      "learning_rate": 0.00014884990149264645,
+      "loss": 1.4022,
+      "step": 4959
+    },
+    {
+      "epoch": 0.3575290131910906,
+      "grad_norm": 0.09900316956346812,
+      "learning_rate": 0.00014882952826747784,
+      "loss": 1.272,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3576010956534275,
+      "grad_norm": 0.15530200333197672,
+      "learning_rate": 0.00014880915238066312,
+      "loss": 1.2949,
+      "step": 4961
+    },
+    {
+      "epoch": 0.35767317811576443,
+      "grad_norm": 0.2607485700002553,
+      "learning_rate": 0.00014878877383331292,
+      "loss": 1.2961,
+      "step": 4962
+    },
+    {
+      "epoch": 0.35774526057810135,
+      "grad_norm": 0.11847462049975523,
+      "learning_rate": 0.00014876839262653805,
+      "loss": 1.2194,
+      "step": 4963
+    },
+    {
+      "epoch": 0.3578173430404383,
+      "grad_norm": 0.10864978196149648,
+      "learning_rate": 0.00014874800876144953,
+      "loss": 1.4091,
+      "step": 4964
+    },
+    {
+      "epoch": 0.3578894255027752,
+      "grad_norm": 0.11215239649592616,
+      "learning_rate": 0.00014872762223915837,
+      "loss": 1.3245,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3579615079651121,
+      "grad_norm": 0.11172828620428929,
+      "learning_rate": 0.0001487072330607759,
+      "loss": 1.2845,
+      "step": 4966
+    },
+    {
+      "epoch": 0.358033590427449,
+      "grad_norm": 0.1089670765739122,
+      "learning_rate": 0.00014868684122741344,
+      "loss": 1.2993,
+      "step": 4967
+    },
+    {
+      "epoch": 0.3581056728897859,
+      "grad_norm": 0.10657926691834549,
+      "learning_rate": 0.00014866644674018265,
+      "loss": 1.2038,
+      "step": 4968
+    },
+    {
+      "epoch": 0.3581777553521228,
+      "grad_norm": 0.09867156469931106,
+      "learning_rate": 0.00014864604960019504,
+      "loss": 1.3009,
+      "step": 4969
+    },
+    {
+      "epoch": 0.35824983781445974,
+      "grad_norm": 0.10510429546039449,
+      "learning_rate": 0.00014862564980856258,
+      "loss": 1.392,
+      "step": 4970
+    },
+    {
+      "epoch": 0.35832192027679666,
+      "grad_norm": 0.12129645815703333,
+      "learning_rate": 0.00014860524736639717,
+      "loss": 1.3026,
+      "step": 4971
+    },
+    {
+      "epoch": 0.3583940027391336,
+      "grad_norm": 0.10115191043826022,
+      "learning_rate": 0.00014858484227481096,
+      "loss": 1.4033,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3584660852014705,
+      "grad_norm": 0.110239903652423,
+      "learning_rate": 0.00014856443453491617,
+      "loss": 1.1659,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3585381676638074,
+      "grad_norm": 0.11944440538938908,
+      "learning_rate": 0.00014854402414782525,
+      "loss": 1.2868,
+      "step": 4974
+    },
+    {
+      "epoch": 0.35861025012614434,
+      "grad_norm": 0.10745772168537611,
+      "learning_rate": 0.0001485236111146507,
+      "loss": 1.3016,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3586823325884812,
+      "grad_norm": 0.10502633883386306,
+      "learning_rate": 0.00014850319543650525,
+      "loss": 1.3324,
+      "step": 4976
+    },
+    {
+      "epoch": 0.3587544150508181,
+      "grad_norm": 0.1116853299799014,
+      "learning_rate": 0.00014848277711450176,
+      "loss": 1.2333,
+      "step": 4977
+    },
+    {
+      "epoch": 0.35882649751315504,
+      "grad_norm": 0.11955886757804285,
+      "learning_rate": 0.00014846235614975314,
+      "loss": 1.2903,
+      "step": 4978
+    },
+    {
+      "epoch": 0.35889857997549196,
+      "grad_norm": 0.14287064573601205,
+      "learning_rate": 0.00014844193254337257,
+      "loss": 1.3768,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3589706624378289,
+      "grad_norm": 0.10678562151730751,
+      "learning_rate": 0.0001484215062964733,
+      "loss": 1.395,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3590427449001658,
+      "grad_norm": 0.11107747482957848,
+      "learning_rate": 0.00014840107741016873,
+      "loss": 1.2903,
+      "step": 4981
+    },
+    {
+      "epoch": 0.3591148273625027,
+      "grad_norm": 0.11394492815237445,
+      "learning_rate": 0.00014838064588557247,
+      "loss": 1.2025,
+      "step": 4982
+    },
+    {
+      "epoch": 0.35918690982483964,
+      "grad_norm": 0.10472623730909579,
+      "learning_rate": 0.00014836021172379816,
+      "loss": 1.2382,
+      "step": 4983
+    },
+    {
+      "epoch": 0.3592589922871765,
+      "grad_norm": 0.12375592252512327,
+      "learning_rate": 0.00014833977492595968,
+      "loss": 1.3172,
+      "step": 4984
+    },
+    {
+      "epoch": 0.3593310747495134,
+      "grad_norm": 0.11831409234567852,
+      "learning_rate": 0.00014831933549317095,
+      "loss": 1.3384,
+      "step": 4985
+    },
+    {
+      "epoch": 0.35940315721185034,
+      "grad_norm": 0.11722674102492685,
+      "learning_rate": 0.00014829889342654623,
+      "loss": 1.4496,
+      "step": 4986
+    },
+    {
+      "epoch": 0.35947523967418726,
+      "grad_norm": 0.10653911000572182,
+      "learning_rate": 0.00014827844872719966,
+      "loss": 1.2425,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3595473221365242,
+      "grad_norm": 0.11245068404202627,
+      "learning_rate": 0.0001482580013962458,
+      "loss": 1.228,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3596194045988611,
+      "grad_norm": 0.10679477401283785,
+      "learning_rate": 0.00014823755143479907,
+      "loss": 1.3103,
+      "step": 4989
+    },
+    {
+      "epoch": 0.359691487061198,
+      "grad_norm": 0.10956719889000438,
+      "learning_rate": 0.00014821709884397429,
+      "loss": 1.3325,
+      "step": 4990
+    },
+    {
+      "epoch": 0.35976356952353494,
+      "grad_norm": 0.10130265147958088,
+      "learning_rate": 0.0001481966436248862,
+      "loss": 1.3466,
+      "step": 4991
+    },
+    {
+      "epoch": 0.35983565198587186,
+      "grad_norm": 0.19623157630523766,
+      "learning_rate": 0.00014817618577864992,
+      "loss": 1.3125,
+      "step": 4992
+    },
+    {
+      "epoch": 0.3599077344482087,
+      "grad_norm": 0.12059014737268452,
+      "learning_rate": 0.00014815572530638046,
+      "loss": 1.3847,
+      "step": 4993
+    },
+    {
+      "epoch": 0.35997981691054565,
+      "grad_norm": 0.10466704455703302,
+      "learning_rate": 0.0001481352622091932,
+      "loss": 1.2612,
+      "step": 4994
+    },
+    {
+      "epoch": 0.36005189937288257,
+      "grad_norm": 0.11335979148251575,
+      "learning_rate": 0.0001481147964882035,
+      "loss": 1.3193,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3601239818352195,
+      "grad_norm": 0.10623242292415751,
+      "learning_rate": 0.00014809432814452694,
+      "loss": 1.2478,
+      "step": 4996
+    },
+    {
+      "epoch": 0.3601960642975564,
+      "grad_norm": 0.10100234225525909,
+      "learning_rate": 0.00014807385717927925,
+      "loss": 1.3166,
+      "step": 4997
+    },
+    {
+      "epoch": 0.3602681467598933,
+      "grad_norm": 0.1159722804047708,
+      "learning_rate": 0.00014805338359357627,
+      "loss": 1.2473,
+      "step": 4998
+    },
+    {
+      "epoch": 0.36034022922223025,
+      "grad_norm": 0.10246438350819354,
+      "learning_rate": 0.00014803290738853395,
+      "loss": 1.281,
+      "step": 4999
+    },
+    {
+      "epoch": 0.36041231168456717,
+      "grad_norm": 0.12084281093713778,
+      "learning_rate": 0.0001480124285652685,
+      "loss": 1.351,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3604843941469041,
+      "grad_norm": 0.1485883027629828,
+      "learning_rate": 0.00014799194712489613,
+      "loss": 1.1921,
+      "step": 5001
+    },
+    {
+      "epoch": 0.36055647660924095,
+      "grad_norm": 0.10501739283188079,
+      "learning_rate": 0.0001479714630685333,
+      "loss": 1.3074,
+      "step": 5002
+    },
+    {
+      "epoch": 0.36062855907157787,
+      "grad_norm": 0.0972443948555232,
+      "learning_rate": 0.00014795097639729658,
+      "loss": 1.2959,
+      "step": 5003
+    },
+    {
+      "epoch": 0.3607006415339148,
+      "grad_norm": 0.11450146705738916,
+      "learning_rate": 0.00014793048711230267,
+      "loss": 1.3156,
+      "step": 5004
+    },
+    {
+      "epoch": 0.3607727239962517,
+      "grad_norm": 0.1411433921260832,
+      "learning_rate": 0.0001479099952146684,
+      "loss": 1.4164,
+      "step": 5005
+    },
+    {
+      "epoch": 0.36084480645858863,
+      "grad_norm": 0.10115234030285886,
+      "learning_rate": 0.00014788950070551078,
+      "loss": 1.4535,
+      "step": 5006
+    },
+    {
+      "epoch": 0.36091688892092555,
+      "grad_norm": 0.11168936398577242,
+      "learning_rate": 0.00014786900358594693,
+      "loss": 1.3342,
+      "step": 5007
+    },
+    {
+      "epoch": 0.36098897138326247,
+      "grad_norm": 0.10685917203188325,
+      "learning_rate": 0.00014784850385709413,
+      "loss": 1.3191,
+      "step": 5008
+    },
+    {
+      "epoch": 0.3610610538455994,
+      "grad_norm": 0.10638092602022231,
+      "learning_rate": 0.00014782800152006984,
+      "loss": 1.2465,
+      "step": 5009
+    },
+    {
+      "epoch": 0.36113313630793625,
+      "grad_norm": 0.09991363327608402,
+      "learning_rate": 0.00014780749657599155,
+      "loss": 1.3638,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3612052187702732,
+      "grad_norm": 0.116978352743695,
+      "learning_rate": 0.00014778698902597704,
+      "loss": 1.2872,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3612773012326101,
+      "grad_norm": 0.10986731641418324,
+      "learning_rate": 0.00014776647887114407,
+      "loss": 1.3987,
+      "step": 5012
+    },
+    {
+      "epoch": 0.361349383694947,
+      "grad_norm": 0.11113291837901204,
+      "learning_rate": 0.0001477459661126107,
+      "loss": 1.2577,
+      "step": 5013
+    },
+    {
+      "epoch": 0.36142146615728393,
+      "grad_norm": 0.11623120575895025,
+      "learning_rate": 0.000147725450751495,
+      "loss": 1.2527,
+      "step": 5014
+    },
+    {
+      "epoch": 0.36149354861962085,
+      "grad_norm": 0.11074556266327155,
+      "learning_rate": 0.00014770493278891528,
+      "loss": 1.3414,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3615656310819578,
+      "grad_norm": 0.10654390866255678,
+      "learning_rate": 0.00014768441222598997,
+      "loss": 1.3003,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3616377135442947,
+      "grad_norm": 0.12365874154511931,
+      "learning_rate": 0.00014766388906383758,
+      "loss": 1.2777,
+      "step": 5017
+    },
+    {
+      "epoch": 0.3617097960066316,
+      "grad_norm": 0.10415896108845325,
+      "learning_rate": 0.0001476433633035768,
+      "loss": 1.3431,
+      "step": 5018
+    },
+    {
+      "epoch": 0.3617818784689685,
+      "grad_norm": 0.13712154584030128,
+      "learning_rate": 0.00014762283494632655,
+      "loss": 1.3513,
+      "step": 5019
+    },
+    {
+      "epoch": 0.3618539609313054,
+      "grad_norm": 0.09722426882282333,
+      "learning_rate": 0.00014760230399320568,
+      "loss": 1.2131,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3619260433936423,
+      "grad_norm": 0.10546456573292635,
+      "learning_rate": 0.0001475817704453334,
+      "loss": 1.2571,
+      "step": 5021
+    },
+    {
+      "epoch": 0.36199812585597924,
+      "grad_norm": 0.11231646287799985,
+      "learning_rate": 0.00014756123430382896,
+      "loss": 1.361,
+      "step": 5022
+    },
+    {
+      "epoch": 0.36207020831831616,
+      "grad_norm": 0.1137423151841344,
+      "learning_rate": 0.00014754069556981178,
+      "loss": 1.2406,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3621422907806531,
+      "grad_norm": 0.11276461112829438,
+      "learning_rate": 0.00014752015424440132,
+      "loss": 1.3056,
+      "step": 5024
+    },
+    {
+      "epoch": 0.36221437324299,
+      "grad_norm": 0.09950516393303084,
+      "learning_rate": 0.0001474996103287174,
+      "loss": 1.2535,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3622864557053269,
+      "grad_norm": 0.10912238559798848,
+      "learning_rate": 0.0001474790638238797,
+      "loss": 1.3783,
+      "step": 5026
+    },
+    {
+      "epoch": 0.36235853816766384,
+      "grad_norm": 0.10546116566945335,
+      "learning_rate": 0.00014745851473100833,
+      "loss": 1.2948,
+      "step": 5027
+    },
+    {
+      "epoch": 0.3624306206300007,
+      "grad_norm": 0.12356122808737512,
+      "learning_rate": 0.00014743796305122331,
+      "loss": 1.251,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3625027030923376,
+      "grad_norm": 0.14035615055345615,
+      "learning_rate": 0.0001474174087856449,
+      "loss": 1.2316,
+      "step": 5029
+    },
+    {
+      "epoch": 0.36257478555467454,
+      "grad_norm": 0.11262757817929556,
+      "learning_rate": 0.0001473968519353935,
+      "loss": 1.2914,
+      "step": 5030
+    },
+    {
+      "epoch": 0.36264686801701146,
+      "grad_norm": 0.10761660843319931,
+      "learning_rate": 0.00014737629250158963,
+      "loss": 1.2305,
+      "step": 5031
+    },
+    {
+      "epoch": 0.3627189504793484,
+      "grad_norm": 0.1072765149208257,
+      "learning_rate": 0.000147355730485354,
+      "loss": 1.2583,
+      "step": 5032
+    },
+    {
+      "epoch": 0.3627910329416853,
+      "grad_norm": 0.12523497585880147,
+      "learning_rate": 0.0001473351658878074,
+      "loss": 1.3908,
+      "step": 5033
+    },
+    {
+      "epoch": 0.3628631154040222,
+      "grad_norm": 0.10540943997448846,
+      "learning_rate": 0.00014731459871007078,
+      "loss": 1.2719,
+      "step": 5034
+    },
+    {
+      "epoch": 0.36293519786635914,
+      "grad_norm": 0.10160562407796075,
+      "learning_rate": 0.00014729402895326526,
+      "loss": 1.3193,
+      "step": 5035
+    },
+    {
+      "epoch": 0.363007280328696,
+      "grad_norm": 0.10466931676513774,
+      "learning_rate": 0.00014727345661851203,
+      "loss": 1.1771,
+      "step": 5036
+    },
+    {
+      "epoch": 0.3630793627910329,
+      "grad_norm": 0.11535600481194355,
+      "learning_rate": 0.00014725288170693251,
+      "loss": 1.2042,
+      "step": 5037
+    },
+    {
+      "epoch": 0.36315144525336984,
+      "grad_norm": 0.11167688344188934,
+      "learning_rate": 0.00014723230421964821,
+      "loss": 1.3291,
+      "step": 5038
+    },
+    {
+      "epoch": 0.36322352771570676,
+      "grad_norm": 0.10909418525158245,
+      "learning_rate": 0.0001472117241577808,
+      "loss": 1.3351,
+      "step": 5039
+    },
+    {
+      "epoch": 0.3632956101780437,
+      "grad_norm": 0.1941665474081461,
+      "learning_rate": 0.00014719114152245198,
+      "loss": 1.3959,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3633676926403806,
+      "grad_norm": 0.1112057086417525,
+      "learning_rate": 0.00014717055631478383,
+      "loss": 1.2794,
+      "step": 5041
+    },
+    {
+      "epoch": 0.3634397751027175,
+      "grad_norm": 0.12136887020921694,
+      "learning_rate": 0.0001471499685358983,
+      "loss": 1.2729,
+      "step": 5042
+    },
+    {
+      "epoch": 0.36351185756505444,
+      "grad_norm": 0.13756121752307143,
+      "learning_rate": 0.00014712937818691776,
+      "loss": 1.3006,
+      "step": 5043
+    },
+    {
+      "epoch": 0.36358394002739136,
+      "grad_norm": 0.11888652767835427,
+      "learning_rate": 0.0001471087852689644,
+      "loss": 1.3318,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3636560224897282,
+      "grad_norm": 0.11917887454668151,
+      "learning_rate": 0.0001470881897831608,
+      "loss": 1.2098,
+      "step": 5045
+    },
+    {
+      "epoch": 0.36372810495206515,
+      "grad_norm": 0.11161774488815597,
+      "learning_rate": 0.00014706759173062963,
+      "loss": 1.3093,
+      "step": 5046
+    },
+    {
+      "epoch": 0.36380018741440207,
+      "grad_norm": 0.10757533353933166,
+      "learning_rate": 0.00014704699111249363,
+      "loss": 1.2983,
+      "step": 5047
+    },
+    {
+      "epoch": 0.363872269876739,
+      "grad_norm": 0.11101549466334508,
+      "learning_rate": 0.00014702638792987572,
+      "loss": 1.3132,
+      "step": 5048
+    },
+    {
+      "epoch": 0.3639443523390759,
+      "grad_norm": 0.11060833023427204,
+      "learning_rate": 0.00014700578218389892,
+      "loss": 1.2488,
+      "step": 5049
+    },
+    {
+      "epoch": 0.3640164348014128,
+      "grad_norm": 0.10826706795852294,
+      "learning_rate": 0.00014698517387568648,
+      "loss": 1.3907,
+      "step": 5050
+    },
+    {
+      "epoch": 0.36408851726374974,
+      "grad_norm": 0.10840593200843089,
+      "learning_rate": 0.00014696456300636172,
+      "loss": 1.3512,
+      "step": 5051
+    },
+    {
+      "epoch": 0.36416059972608666,
+      "grad_norm": 0.10298420137405122,
+      "learning_rate": 0.00014694394957704814,
+      "loss": 1.2585,
+      "step": 5052
+    },
+    {
+      "epoch": 0.36423268218842353,
+      "grad_norm": 0.1033095159131996,
+      "learning_rate": 0.00014692333358886934,
+      "loss": 1.2694,
+      "step": 5053
+    },
+    {
+      "epoch": 0.36430476465076045,
+      "grad_norm": 0.10528057819852105,
+      "learning_rate": 0.00014690271504294903,
+      "loss": 1.2607,
+      "step": 5054
+    },
+    {
+      "epoch": 0.36437684711309737,
+      "grad_norm": 0.10854095443081503,
+      "learning_rate": 0.0001468820939404112,
+      "loss": 1.2474,
+      "step": 5055
+    },
+    {
+      "epoch": 0.3644489295754343,
+      "grad_norm": 0.09774183677962778,
+      "learning_rate": 0.0001468614702823798,
+      "loss": 1.2566,
+      "step": 5056
+    },
+    {
+      "epoch": 0.3645210120377712,
+      "grad_norm": 0.11037189925907495,
+      "learning_rate": 0.00014684084406997903,
+      "loss": 1.2474,
+      "step": 5057
+    },
+    {
+      "epoch": 0.36459309450010813,
+      "grad_norm": 0.11348887222757101,
+      "learning_rate": 0.0001468202153043332,
+      "loss": 1.257,
+      "step": 5058
+    },
+    {
+      "epoch": 0.36466517696244505,
+      "grad_norm": 0.10575772009278006,
+      "learning_rate": 0.00014679958398656682,
+      "loss": 1.2799,
+      "step": 5059
+    },
+    {
+      "epoch": 0.36473725942478197,
+      "grad_norm": 0.10695407944367401,
+      "learning_rate": 0.0001467789501178044,
+      "loss": 1.2972,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3648093418871189,
+      "grad_norm": 0.11588868058407924,
+      "learning_rate": 0.00014675831369917068,
+      "loss": 1.3161,
+      "step": 5061
+    },
+    {
+      "epoch": 0.36488142434945575,
+      "grad_norm": 0.10447347895362977,
+      "learning_rate": 0.00014673767473179057,
+      "loss": 1.337,
+      "step": 5062
+    },
+    {
+      "epoch": 0.36495350681179267,
+      "grad_norm": 0.23761305178601724,
+      "learning_rate": 0.00014671703321678906,
+      "loss": 1.2009,
+      "step": 5063
+    },
+    {
+      "epoch": 0.3650255892741296,
+      "grad_norm": 0.11794859728527639,
+      "learning_rate": 0.0001466963891552913,
+      "loss": 1.3099,
+      "step": 5064
+    },
+    {
+      "epoch": 0.3650976717364665,
+      "grad_norm": 0.11338294511000375,
+      "learning_rate": 0.00014667574254842257,
+      "loss": 1.246,
+      "step": 5065
+    },
+    {
+      "epoch": 0.36516975419880343,
+      "grad_norm": 0.10543850104183494,
+      "learning_rate": 0.0001466550933973083,
+      "loss": 1.3196,
+      "step": 5066
+    },
+    {
+      "epoch": 0.36524183666114035,
+      "grad_norm": 0.09847865509679252,
+      "learning_rate": 0.00014663444170307405,
+      "loss": 1.237,
+      "step": 5067
+    },
+    {
+      "epoch": 0.36531391912347727,
+      "grad_norm": 0.11220782902680135,
+      "learning_rate": 0.00014661378746684553,
+      "loss": 1.3014,
+      "step": 5068
+    },
+    {
+      "epoch": 0.3653860015858142,
+      "grad_norm": 0.14594624541785456,
+      "learning_rate": 0.00014659313068974858,
+      "loss": 1.3353,
+      "step": 5069
+    },
+    {
+      "epoch": 0.3654580840481511,
+      "grad_norm": 0.10395379734210868,
+      "learning_rate": 0.00014657247137290916,
+      "loss": 1.3236,
+      "step": 5070
+    },
+    {
+      "epoch": 0.365530166510488,
+      "grad_norm": 0.10453615577046776,
+      "learning_rate": 0.0001465518095174534,
+      "loss": 1.1788,
+      "step": 5071
+    },
+    {
+      "epoch": 0.3656022489728249,
+      "grad_norm": 0.10577995949746487,
+      "learning_rate": 0.00014653114512450756,
+      "loss": 1.2973,
+      "step": 5072
+    },
+    {
+      "epoch": 0.3656743314351618,
+      "grad_norm": 0.10805487233169306,
+      "learning_rate": 0.00014651047819519804,
+      "loss": 1.3629,
+      "step": 5073
+    },
+    {
+      "epoch": 0.36574641389749873,
+      "grad_norm": 0.11447794762197604,
+      "learning_rate": 0.00014648980873065135,
+      "loss": 1.2563,
+      "step": 5074
+    },
+    {
+      "epoch": 0.36581849635983565,
+      "grad_norm": 0.14814323259757745,
+      "learning_rate": 0.00014646913673199417,
+      "loss": 1.3395,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3658905788221726,
+      "grad_norm": 0.11800013367819008,
+      "learning_rate": 0.00014644846220035333,
+      "loss": 1.354,
+      "step": 5076
+    },
+    {
+      "epoch": 0.3659626612845095,
+      "grad_norm": 0.1283899919545024,
+      "learning_rate": 0.00014642778513685576,
+      "loss": 1.3212,
+      "step": 5077
+    },
+    {
+      "epoch": 0.3660347437468464,
+      "grad_norm": 0.09927870553336274,
+      "learning_rate": 0.00014640710554262852,
+      "loss": 1.2796,
+      "step": 5078
+    },
+    {
+      "epoch": 0.3661068262091833,
+      "grad_norm": 0.1205253390602585,
+      "learning_rate": 0.0001463864234187989,
+      "loss": 1.3602,
+      "step": 5079
+    },
+    {
+      "epoch": 0.3661789086715202,
+      "grad_norm": 0.10968710101485254,
+      "learning_rate": 0.00014636573876649419,
+      "loss": 1.2589,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3662509911338571,
+      "grad_norm": 0.1519291960512916,
+      "learning_rate": 0.0001463450515868419,
+      "loss": 1.2524,
+      "step": 5081
+    },
+    {
+      "epoch": 0.36632307359619404,
+      "grad_norm": 0.1138209029232539,
+      "learning_rate": 0.00014632436188096972,
+      "loss": 1.3335,
+      "step": 5082
+    },
+    {
+      "epoch": 0.36639515605853096,
+      "grad_norm": 0.11103806861342583,
+      "learning_rate": 0.0001463036696500054,
+      "loss": 1.3615,
+      "step": 5083
+    },
+    {
+      "epoch": 0.3664672385208679,
+      "grad_norm": 0.10064824621695058,
+      "learning_rate": 0.0001462829748950768,
+      "loss": 1.3005,
+      "step": 5084
+    },
+    {
+      "epoch": 0.3665393209832048,
+      "grad_norm": 0.10751257548910989,
+      "learning_rate": 0.000146262277617312,
+      "loss": 1.2632,
+      "step": 5085
+    },
+    {
+      "epoch": 0.3666114034455417,
+      "grad_norm": 0.1056084341537189,
+      "learning_rate": 0.00014624157781783926,
+      "loss": 1.3555,
+      "step": 5086
+    },
+    {
+      "epoch": 0.36668348590787864,
+      "grad_norm": 0.1088995689701028,
+      "learning_rate": 0.00014622087549778678,
+      "loss": 1.2946,
+      "step": 5087
+    },
+    {
+      "epoch": 0.3667555683702155,
+      "grad_norm": 0.11602228575952604,
+      "learning_rate": 0.00014620017065828308,
+      "loss": 1.3507,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3668276508325524,
+      "grad_norm": 0.14428199765491248,
+      "learning_rate": 0.0001461794633004568,
+      "loss": 1.2273,
+      "step": 5089
+    },
+    {
+      "epoch": 0.36689973329488934,
+      "grad_norm": 0.12292877391692122,
+      "learning_rate": 0.00014615875342543661,
+      "loss": 1.2698,
+      "step": 5090
+    },
+    {
+      "epoch": 0.36697181575722626,
+      "grad_norm": 0.11103807731762588,
+      "learning_rate": 0.00014613804103435143,
+      "loss": 1.3014,
+      "step": 5091
+    },
+    {
+      "epoch": 0.3670438982195632,
+      "grad_norm": 0.11531936206739578,
+      "learning_rate": 0.00014611732612833026,
+      "loss": 1.3664,
+      "step": 5092
+    },
+    {
+      "epoch": 0.3671159806819001,
+      "grad_norm": 0.10804865445234835,
+      "learning_rate": 0.0001460966087085022,
+      "loss": 1.2956,
+      "step": 5093
+    },
+    {
+      "epoch": 0.367188063144237,
+      "grad_norm": 0.10839234196888975,
+      "learning_rate": 0.0001460758887759966,
+      "loss": 1.274,
+      "step": 5094
+    },
+    {
+      "epoch": 0.36726014560657394,
+      "grad_norm": 0.10927925444751319,
+      "learning_rate": 0.00014605516633194282,
+      "loss": 1.2929,
+      "step": 5095
+    },
+    {
+      "epoch": 0.36733222806891086,
+      "grad_norm": 0.114479392379474,
+      "learning_rate": 0.0001460344413774705,
+      "loss": 1.2395,
+      "step": 5096
+    },
+    {
+      "epoch": 0.3674043105312477,
+      "grad_norm": 0.13750658480567934,
+      "learning_rate": 0.00014601371391370925,
+      "loss": 1.1189,
+      "step": 5097
+    },
+    {
+      "epoch": 0.36747639299358464,
+      "grad_norm": 0.11624379973846286,
+      "learning_rate": 0.00014599298394178897,
+      "loss": 1.2867,
+      "step": 5098
+    },
+    {
+      "epoch": 0.36754847545592156,
+      "grad_norm": 0.1147338401935533,
+      "learning_rate": 0.00014597225146283958,
+      "loss": 1.3075,
+      "step": 5099
+    },
+    {
+      "epoch": 0.3676205579182585,
+      "grad_norm": 0.1306876884728106,
+      "learning_rate": 0.00014595151647799125,
+      "loss": 1.2966,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3676926403805954,
+      "grad_norm": 0.11151789890010012,
+      "learning_rate": 0.0001459307789883741,
+      "loss": 1.2994,
+      "step": 5101
+    },
+    {
+      "epoch": 0.3677647228429323,
+      "grad_norm": 0.11765218846359204,
+      "learning_rate": 0.00014591003899511865,
+      "loss": 1.3923,
+      "step": 5102
+    },
+    {
+      "epoch": 0.36783680530526924,
+      "grad_norm": 0.12324464032776535,
+      "learning_rate": 0.00014588929649935535,
+      "loss": 1.3095,
+      "step": 5103
+    },
+    {
+      "epoch": 0.36790888776760616,
+      "grad_norm": 0.14671349077192905,
+      "learning_rate": 0.00014586855150221483,
+      "loss": 1.2481,
+      "step": 5104
+    },
+    {
+      "epoch": 0.367980970229943,
+      "grad_norm": 0.10823045236989798,
+      "learning_rate": 0.0001458478040048279,
+      "loss": 1.1849,
+      "step": 5105
+    },
+    {
+      "epoch": 0.36805305269227995,
+      "grad_norm": 0.11311831767543697,
+      "learning_rate": 0.0001458270540083255,
+      "loss": 1.2704,
+      "step": 5106
+    },
+    {
+      "epoch": 0.36812513515461687,
+      "grad_norm": 0.12945497403791373,
+      "learning_rate": 0.00014580630151383867,
+      "loss": 1.2957,
+      "step": 5107
+    },
+    {
+      "epoch": 0.3681972176169538,
+      "grad_norm": 0.11299115612176255,
+      "learning_rate": 0.00014578554652249862,
+      "loss": 1.2755,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3682693000792907,
+      "grad_norm": 0.11481104590859016,
+      "learning_rate": 0.00014576478903543665,
+      "loss": 1.3363,
+      "step": 5109
+    },
+    {
+      "epoch": 0.3683413825416276,
+      "grad_norm": 0.10555432673776778,
+      "learning_rate": 0.0001457440290537843,
+      "loss": 1.2617,
+      "step": 5110
+    },
+    {
+      "epoch": 0.36841346500396455,
+      "grad_norm": 0.10643968967428896,
+      "learning_rate": 0.0001457232665786731,
+      "loss": 1.3078,
+      "step": 5111
+    },
+    {
+      "epoch": 0.36848554746630147,
+      "grad_norm": 0.13931097138060536,
+      "learning_rate": 0.0001457025016112348,
+      "loss": 1.2852,
+      "step": 5112
+    },
+    {
+      "epoch": 0.3685576299286384,
+      "grad_norm": 0.1275277611779112,
+      "learning_rate": 0.0001456817341526013,
+      "loss": 1.2786,
+      "step": 5113
+    },
+    {
+      "epoch": 0.36862971239097525,
+      "grad_norm": 0.12905660268344124,
+      "learning_rate": 0.00014566096420390464,
+      "loss": 1.3099,
+      "step": 5114
+    },
+    {
+      "epoch": 0.36870179485331217,
+      "grad_norm": 0.11610240535443223,
+      "learning_rate": 0.0001456401917662769,
+      "loss": 1.3298,
+      "step": 5115
+    },
+    {
+      "epoch": 0.3687738773156491,
+      "grad_norm": 0.1087445508553651,
+      "learning_rate": 0.00014561941684085045,
+      "loss": 1.2541,
+      "step": 5116
+    },
+    {
+      "epoch": 0.368845959777986,
+      "grad_norm": 0.10060474605234489,
+      "learning_rate": 0.0001455986394287576,
+      "loss": 1.2425,
+      "step": 5117
+    },
+    {
+      "epoch": 0.36891804224032293,
+      "grad_norm": 0.10455820194633425,
+      "learning_rate": 0.000145577859531131,
+      "loss": 1.1552,
+      "step": 5118
+    },
+    {
+      "epoch": 0.36899012470265985,
+      "grad_norm": 0.15872177947174132,
+      "learning_rate": 0.00014555707714910328,
+      "loss": 1.2568,
+      "step": 5119
+    },
+    {
+      "epoch": 0.36906220716499677,
+      "grad_norm": 0.10510299810693388,
+      "learning_rate": 0.00014553629228380733,
+      "loss": 1.307,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3691342896273337,
+      "grad_norm": 0.10588082289279814,
+      "learning_rate": 0.00014551550493637606,
+      "loss": 1.2526,
+      "step": 5121
+    },
+    {
+      "epoch": 0.3692063720896706,
+      "grad_norm": 0.10770369132085887,
+      "learning_rate": 0.00014549471510794256,
+      "loss": 1.337,
+      "step": 5122
+    },
+    {
+      "epoch": 0.3692784545520075,
+      "grad_norm": 0.11086165294666792,
+      "learning_rate": 0.00014547392279964008,
+      "loss": 1.2434,
+      "step": 5123
+    },
+    {
+      "epoch": 0.3693505370143444,
+      "grad_norm": 0.11245629052308238,
+      "learning_rate": 0.00014545312801260202,
+      "loss": 1.2634,
+      "step": 5124
+    },
+    {
+      "epoch": 0.3694226194766813,
+      "grad_norm": 0.16379328207127858,
+      "learning_rate": 0.0001454323307479618,
+      "loss": 1.201,
+      "step": 5125
+    },
+    {
+      "epoch": 0.36949470193901823,
+      "grad_norm": 0.10453950584027323,
+      "learning_rate": 0.00014541153100685312,
+      "loss": 1.2897,
+      "step": 5126
+    },
+    {
+      "epoch": 0.36956678440135515,
+      "grad_norm": 0.12324021409253337,
+      "learning_rate": 0.00014539072879040976,
+      "loss": 1.2439,
+      "step": 5127
+    },
+    {
+      "epoch": 0.3696388668636921,
+      "grad_norm": 0.11619871816057324,
+      "learning_rate": 0.00014536992409976558,
+      "loss": 1.2392,
+      "step": 5128
+    },
+    {
+      "epoch": 0.369710949326029,
+      "grad_norm": 0.12670708965370572,
+      "learning_rate": 0.00014534911693605464,
+      "loss": 1.3417,
+      "step": 5129
+    },
+    {
+      "epoch": 0.3697830317883659,
+      "grad_norm": 0.11259859526380708,
+      "learning_rate": 0.00014532830730041113,
+      "loss": 1.3395,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3698551142507028,
+      "grad_norm": 0.10783027325146018,
+      "learning_rate": 0.00014530749519396936,
+      "loss": 1.2518,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3699271967130397,
+      "grad_norm": 0.20248274272008557,
+      "learning_rate": 0.00014528668061786378,
+      "loss": 1.2875,
+      "step": 5132
+    },
+    {
+      "epoch": 0.3699992791753766,
+      "grad_norm": 0.10604586682438197,
+      "learning_rate": 0.00014526586357322893,
+      "loss": 1.3889,
+      "step": 5133
+    },
+    {
+      "epoch": 0.37007136163771354,
+      "grad_norm": 0.115258108213524,
+      "learning_rate": 0.00014524504406119958,
+      "loss": 1.3058,
+      "step": 5134
+    },
+    {
+      "epoch": 0.37014344410005046,
+      "grad_norm": 0.10679481392070958,
+      "learning_rate": 0.00014522422208291052,
+      "loss": 1.3515,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3702155265623874,
+      "grad_norm": 0.1180564277091201,
+      "learning_rate": 0.00014520339763949682,
+      "loss": 1.225,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3702876090247243,
+      "grad_norm": 0.1359061492714219,
+      "learning_rate": 0.00014518257073209352,
+      "loss": 1.2543,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3703596914870612,
+      "grad_norm": 0.24284987097544292,
+      "learning_rate": 0.00014516174136183592,
+      "loss": 1.3058,
+      "step": 5138
+    },
+    {
+      "epoch": 0.37043177394939814,
+      "grad_norm": 0.11146752646000889,
+      "learning_rate": 0.00014514090952985934,
+      "loss": 1.2971,
+      "step": 5139
+    },
+    {
+      "epoch": 0.370503856411735,
+      "grad_norm": 0.11843637752605432,
+      "learning_rate": 0.0001451200752372994,
+      "loss": 1.315,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3705759388740719,
+      "grad_norm": 0.10469126431445501,
+      "learning_rate": 0.00014509923848529164,
+      "loss": 1.2388,
+      "step": 5141
+    },
+    {
+      "epoch": 0.37064802133640884,
+      "grad_norm": 0.10523092714707522,
+      "learning_rate": 0.000145078399274972,
+      "loss": 1.3041,
+      "step": 5142
+    },
+    {
+      "epoch": 0.37072010379874576,
+      "grad_norm": 0.21633674562840866,
+      "learning_rate": 0.00014505755760747626,
+      "loss": 1.3468,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3707921862610827,
+      "grad_norm": 0.10626824176749951,
+      "learning_rate": 0.00014503671348394057,
+      "loss": 1.3211,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3708642687234196,
+      "grad_norm": 0.11615675766159399,
+      "learning_rate": 0.00014501586690550102,
+      "loss": 1.2451,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3709363511857565,
+      "grad_norm": 0.16018654414409134,
+      "learning_rate": 0.00014499501787329408,
+      "loss": 1.2484,
+      "step": 5146
+    },
+    {
+      "epoch": 0.37100843364809344,
+      "grad_norm": 0.11996060063691771,
+      "learning_rate": 0.0001449741663884561,
+      "loss": 1.2774,
+      "step": 5147
+    },
+    {
+      "epoch": 0.37108051611043036,
+      "grad_norm": 0.11792737253088333,
+      "learning_rate": 0.00014495331245212366,
+      "loss": 1.2852,
+      "step": 5148
+    },
+    {
+      "epoch": 0.3711525985727672,
+      "grad_norm": 0.11632190306620542,
+      "learning_rate": 0.0001449324560654336,
+      "loss": 1.3349,
+      "step": 5149
+    },
+    {
+      "epoch": 0.37122468103510414,
+      "grad_norm": 0.10835710132091993,
+      "learning_rate": 0.00014491159722952265,
+      "loss": 1.2928,
+      "step": 5150
+    },
+    {
+      "epoch": 0.37129676349744106,
+      "grad_norm": 0.10183117208223462,
+      "learning_rate": 0.00014489073594552786,
+      "loss": 1.3304,
+      "step": 5151
+    },
+    {
+      "epoch": 0.371368845959778,
+      "grad_norm": 0.10955532972842001,
+      "learning_rate": 0.0001448698722145864,
+      "loss": 1.2835,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3714409284221149,
+      "grad_norm": 0.1042347232304795,
+      "learning_rate": 0.00014484900603783544,
+      "loss": 1.3284,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3715130108844518,
+      "grad_norm": 0.10957119063163367,
+      "learning_rate": 0.00014482813741641243,
+      "loss": 1.2575,
+      "step": 5154
+    },
+    {
+      "epoch": 0.37158509334678874,
+      "grad_norm": 0.11697468036150678,
+      "learning_rate": 0.00014480726635145491,
+      "loss": 1.2276,
+      "step": 5155
+    },
+    {
+      "epoch": 0.37165717580912566,
+      "grad_norm": 0.11033293756257283,
+      "learning_rate": 0.0001447863928441005,
+      "loss": 1.2418,
+      "step": 5156
+    },
+    {
+      "epoch": 0.3717292582714625,
+      "grad_norm": 0.11323313851577581,
+      "learning_rate": 0.00014476551689548704,
+      "loss": 1.317,
+      "step": 5157
+    },
+    {
+      "epoch": 0.37180134073379945,
+      "grad_norm": 0.12177738959162658,
+      "learning_rate": 0.0001447446385067524,
+      "loss": 1.2951,
+      "step": 5158
+    },
+    {
+      "epoch": 0.37187342319613637,
+      "grad_norm": 0.10622455408463483,
+      "learning_rate": 0.00014472375767903464,
+      "loss": 1.2346,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3719455056584733,
+      "grad_norm": 0.09979499208322903,
+      "learning_rate": 0.00014470287441347202,
+      "loss": 1.3035,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3720175881208102,
+      "grad_norm": 0.15751404241413336,
+      "learning_rate": 0.0001446819887112028,
+      "loss": 1.3516,
+      "step": 5161
+    },
+    {
+      "epoch": 0.3720896705831471,
+      "grad_norm": 0.11006702897839576,
+      "learning_rate": 0.00014466110057336545,
+      "loss": 1.2718,
+      "step": 5162
+    },
+    {
+      "epoch": 0.37216175304548405,
+      "grad_norm": 0.1421707537548997,
+      "learning_rate": 0.00014464021000109857,
+      "loss": 1.2824,
+      "step": 5163
+    },
+    {
+      "epoch": 0.37223383550782096,
+      "grad_norm": 0.11048317754909653,
+      "learning_rate": 0.0001446193169955409,
+      "loss": 1.3525,
+      "step": 5164
+    },
+    {
+      "epoch": 0.3723059179701579,
+      "grad_norm": 0.10718367127679537,
+      "learning_rate": 0.00014459842155783127,
+      "loss": 1.2498,
+      "step": 5165
+    },
+    {
+      "epoch": 0.37237800043249475,
+      "grad_norm": 0.12152818748923215,
+      "learning_rate": 0.00014457752368910866,
+      "loss": 1.3793,
+      "step": 5166
+    },
+    {
+      "epoch": 0.37245008289483167,
+      "grad_norm": 0.09448494951674982,
+      "learning_rate": 0.00014455662339051223,
+      "loss": 1.2924,
+      "step": 5167
+    },
+    {
+      "epoch": 0.3725221653571686,
+      "grad_norm": 0.11170607899869402,
+      "learning_rate": 0.00014453572066318123,
+      "loss": 1.2734,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3725942478195055,
+      "grad_norm": 0.13837268290607346,
+      "learning_rate": 0.00014451481550825502,
+      "loss": 1.3071,
+      "step": 5169
+    },
+    {
+      "epoch": 0.37266633028184243,
+      "grad_norm": 0.10145729429216846,
+      "learning_rate": 0.00014449390792687305,
+      "loss": 1.2882,
+      "step": 5170
+    },
+    {
+      "epoch": 0.37273841274417935,
+      "grad_norm": 0.10801925864526993,
+      "learning_rate": 0.00014447299792017513,
+      "loss": 1.3058,
+      "step": 5171
+    },
+    {
+      "epoch": 0.37281049520651627,
+      "grad_norm": 0.1054922434519699,
+      "learning_rate": 0.0001444520854893009,
+      "loss": 1.3049,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3728825776688532,
+      "grad_norm": 0.15000038154628306,
+      "learning_rate": 0.00014443117063539038,
+      "loss": 1.3109,
+      "step": 5173
+    },
+    {
+      "epoch": 0.3729546601311901,
+      "grad_norm": 0.11397716778785055,
+      "learning_rate": 0.0001444102533595835,
+      "loss": 1.3782,
+      "step": 5174
+    },
+    {
+      "epoch": 0.37302674259352697,
+      "grad_norm": 0.12728398836917407,
+      "learning_rate": 0.00014438933366302058,
+      "loss": 1.2391,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3730988250558639,
+      "grad_norm": 0.1182247788762433,
+      "learning_rate": 0.00014436841154684182,
+      "loss": 1.2984,
+      "step": 5176
+    },
+    {
+      "epoch": 0.3731709075182008,
+      "grad_norm": 0.10937757588577539,
+      "learning_rate": 0.00014434748701218773,
+      "loss": 1.2592,
+      "step": 5177
+    },
+    {
+      "epoch": 0.37324298998053773,
+      "grad_norm": 0.10946648682719462,
+      "learning_rate": 0.0001443265600601988,
+      "loss": 1.332,
+      "step": 5178
+    },
+    {
+      "epoch": 0.37331507244287465,
+      "grad_norm": 0.11279224257600967,
+      "learning_rate": 0.00014430563069201585,
+      "loss": 1.2114,
+      "step": 5179
+    },
+    {
+      "epoch": 0.37338715490521157,
+      "grad_norm": 0.11097160291255624,
+      "learning_rate": 0.0001442846989087796,
+      "loss": 1.2604,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3734592373675485,
+      "grad_norm": 0.10991456174345829,
+      "learning_rate": 0.00014426376471163112,
+      "loss": 1.3304,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3735313198298854,
+      "grad_norm": 0.10506382657235082,
+      "learning_rate": 0.00014424282810171146,
+      "loss": 1.26,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3736034022922223,
+      "grad_norm": 0.11621644513007244,
+      "learning_rate": 0.00014422188908016183,
+      "loss": 1.329,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3736754847545592,
+      "grad_norm": 0.11264194098563764,
+      "learning_rate": 0.00014420094764812366,
+      "loss": 1.247,
+      "step": 5184
+    },
+    {
+      "epoch": 0.3737475672168961,
+      "grad_norm": 0.10887779498917123,
+      "learning_rate": 0.0001441800038067384,
+      "loss": 1.2795,
+      "step": 5185
+    },
+    {
+      "epoch": 0.37381964967923303,
+      "grad_norm": 0.1078406206217315,
+      "learning_rate": 0.0001441590575571477,
+      "loss": 1.2866,
+      "step": 5186
+    },
+    {
+      "epoch": 0.37389173214156995,
+      "grad_norm": 0.12361284145965251,
+      "learning_rate": 0.00014413810890049326,
+      "loss": 1.2549,
+      "step": 5187
+    },
+    {
+      "epoch": 0.3739638146039069,
+      "grad_norm": 0.12808448017456048,
+      "learning_rate": 0.00014411715783791706,
+      "loss": 1.3134,
+      "step": 5188
+    },
+    {
+      "epoch": 0.3740358970662438,
+      "grad_norm": 0.11644236694959548,
+      "learning_rate": 0.00014409620437056105,
+      "loss": 1.2687,
+      "step": 5189
+    },
+    {
+      "epoch": 0.3741079795285807,
+      "grad_norm": 0.11243151672367599,
+      "learning_rate": 0.0001440752484995674,
+      "loss": 1.3351,
+      "step": 5190
+    },
+    {
+      "epoch": 0.37418006199091763,
+      "grad_norm": 0.11482390107498724,
+      "learning_rate": 0.00014405429022607845,
+      "loss": 1.2577,
+      "step": 5191
+    },
+    {
+      "epoch": 0.3742521444532545,
+      "grad_norm": 0.1172635614641322,
+      "learning_rate": 0.00014403332955123651,
+      "loss": 1.3012,
+      "step": 5192
+    },
+    {
+      "epoch": 0.3743242269155914,
+      "grad_norm": 0.11170806115138283,
+      "learning_rate": 0.0001440123664761842,
+      "loss": 1.2463,
+      "step": 5193
+    },
+    {
+      "epoch": 0.37439630937792834,
+      "grad_norm": 0.12720667212228728,
+      "learning_rate": 0.00014399140100206415,
+      "loss": 1.3091,
+      "step": 5194
+    },
+    {
+      "epoch": 0.37446839184026526,
+      "grad_norm": 0.10935993370318553,
+      "learning_rate": 0.0001439704331300192,
+      "loss": 1.3424,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3745404743026022,
+      "grad_norm": 0.13067101949781013,
+      "learning_rate": 0.00014394946286119226,
+      "loss": 1.3809,
+      "step": 5196
+    },
+    {
+      "epoch": 0.3746125567649391,
+      "grad_norm": 0.10370551457835146,
+      "learning_rate": 0.00014392849019672645,
+      "loss": 1.3429,
+      "step": 5197
+    },
+    {
+      "epoch": 0.374684639227276,
+      "grad_norm": 0.10775672428149927,
+      "learning_rate": 0.00014390751513776489,
+      "loss": 1.3453,
+      "step": 5198
+    },
+    {
+      "epoch": 0.37475672168961294,
+      "grad_norm": 0.1243092691021903,
+      "learning_rate": 0.00014388653768545098,
+      "loss": 1.2769,
+      "step": 5199
+    },
+    {
+      "epoch": 0.37482880415194986,
+      "grad_norm": 0.11669015075631584,
+      "learning_rate": 0.0001438655578409281,
+      "loss": 1.2513,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3749008866142867,
+      "grad_norm": 0.19021037935225474,
+      "learning_rate": 0.00014384457560533994,
+      "loss": 1.1757,
+      "step": 5201
+    },
+    {
+      "epoch": 0.37497296907662364,
+      "grad_norm": 0.11411855809527348,
+      "learning_rate": 0.00014382359097983013,
+      "loss": 1.2599,
+      "step": 5202
+    },
+    {
+      "epoch": 0.37504505153896056,
+      "grad_norm": 0.11642357211148933,
+      "learning_rate": 0.00014380260396554254,
+      "loss": 1.376,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3751171340012975,
+      "grad_norm": 0.1459880831768434,
+      "learning_rate": 0.0001437816145636212,
+      "loss": 1.2563,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3751892164636344,
+      "grad_norm": 0.11400898985917084,
+      "learning_rate": 0.00014376062277521016,
+      "loss": 1.2691,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3752612989259713,
+      "grad_norm": 0.12211633041879318,
+      "learning_rate": 0.0001437396286014537,
+      "loss": 1.2624,
+      "step": 5206
+    },
+    {
+      "epoch": 0.37533338138830824,
+      "grad_norm": 0.10907752449989236,
+      "learning_rate": 0.0001437186320434962,
+      "loss": 1.2911,
+      "step": 5207
+    },
+    {
+      "epoch": 0.37540546385064516,
+      "grad_norm": 0.1184136143618081,
+      "learning_rate": 0.00014369763310248206,
+      "loss": 1.383,
+      "step": 5208
+    },
+    {
+      "epoch": 0.375477546312982,
+      "grad_norm": 0.1161241466818645,
+      "learning_rate": 0.00014367663177955605,
+      "loss": 1.3802,
+      "step": 5209
+    },
+    {
+      "epoch": 0.37554962877531894,
+      "grad_norm": 0.1179433339871364,
+      "learning_rate": 0.00014365562807586282,
+      "loss": 1.3249,
+      "step": 5210
+    },
+    {
+      "epoch": 0.37562171123765586,
+      "grad_norm": 0.11092590632784048,
+      "learning_rate": 0.00014363462199254735,
+      "loss": 1.2317,
+      "step": 5211
+    },
+    {
+      "epoch": 0.3756937936999928,
+      "grad_norm": 0.10728501855916174,
+      "learning_rate": 0.0001436136135307546,
+      "loss": 1.2648,
+      "step": 5212
+    },
+    {
+      "epoch": 0.3757658761623297,
+      "grad_norm": 0.10882372948369422,
+      "learning_rate": 0.00014359260269162976,
+      "loss": 1.34,
+      "step": 5213
+    },
+    {
+      "epoch": 0.3758379586246666,
+      "grad_norm": 0.12218468902599118,
+      "learning_rate": 0.00014357158947631802,
+      "loss": 1.2222,
+      "step": 5214
+    },
+    {
+      "epoch": 0.37591004108700354,
+      "grad_norm": 0.11052274908790763,
+      "learning_rate": 0.00014355057388596492,
+      "loss": 1.3739,
+      "step": 5215
+    },
+    {
+      "epoch": 0.37598212354934046,
+      "grad_norm": 0.11084817062906031,
+      "learning_rate": 0.00014352955592171588,
+      "loss": 1.2532,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3760542060116774,
+      "grad_norm": 0.10529114321710607,
+      "learning_rate": 0.00014350853558471664,
+      "loss": 1.2359,
+      "step": 5217
+    },
+    {
+      "epoch": 0.37612628847401425,
+      "grad_norm": 0.10241207222626204,
+      "learning_rate": 0.00014348751287611292,
+      "loss": 1.2504,
+      "step": 5218
+    },
+    {
+      "epoch": 0.37619837093635117,
+      "grad_norm": 0.09701593030804792,
+      "learning_rate": 0.00014346648779705077,
+      "loss": 1.2951,
+      "step": 5219
+    },
+    {
+      "epoch": 0.3762704533986881,
+      "grad_norm": 0.1097147315153328,
+      "learning_rate": 0.0001434454603486761,
+      "loss": 1.3406,
+      "step": 5220
+    },
+    {
+      "epoch": 0.376342535861025,
+      "grad_norm": 0.10361199191164347,
+      "learning_rate": 0.00014342443053213527,
+      "loss": 1.2801,
+      "step": 5221
+    },
+    {
+      "epoch": 0.3764146183233619,
+      "grad_norm": 0.11412882173104355,
+      "learning_rate": 0.0001434033983485744,
+      "loss": 1.3698,
+      "step": 5222
+    },
+    {
+      "epoch": 0.37648670078569885,
+      "grad_norm": 0.10981742173513061,
+      "learning_rate": 0.00014338236379914004,
+      "loss": 1.2928,
+      "step": 5223
+    },
+    {
+      "epoch": 0.37655878324803577,
+      "grad_norm": 0.1130895327949565,
+      "learning_rate": 0.00014336132688497877,
+      "loss": 1.336,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3766308657103727,
+      "grad_norm": 0.1114035671238517,
+      "learning_rate": 0.00014334028760723725,
+      "loss": 1.4073,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3767029481727096,
+      "grad_norm": 0.1276716797289889,
+      "learning_rate": 0.00014331924596706233,
+      "loss": 1.2957,
+      "step": 5226
+    },
+    {
+      "epoch": 0.37677503063504647,
+      "grad_norm": 0.11827644519122685,
+      "learning_rate": 0.00014329820196560094,
+      "loss": 1.2792,
+      "step": 5227
+    },
+    {
+      "epoch": 0.3768471130973834,
+      "grad_norm": 0.10789307953670191,
+      "learning_rate": 0.0001432771556040002,
+      "loss": 1.266,
+      "step": 5228
+    },
+    {
+      "epoch": 0.3769191955597203,
+      "grad_norm": 0.1441392093030435,
+      "learning_rate": 0.0001432561068834073,
+      "loss": 1.2768,
+      "step": 5229
+    },
+    {
+      "epoch": 0.37699127802205723,
+      "grad_norm": 0.11255822587089045,
+      "learning_rate": 0.0001432350558049696,
+      "loss": 1.3084,
+      "step": 5230
+    },
+    {
+      "epoch": 0.37706336048439415,
+      "grad_norm": 0.11001206848795919,
+      "learning_rate": 0.00014321400236983457,
+      "loss": 1.2612,
+      "step": 5231
+    },
+    {
+      "epoch": 0.37713544294673107,
+      "grad_norm": 0.11561160506492077,
+      "learning_rate": 0.0001431929465791498,
+      "loss": 1.2904,
+      "step": 5232
+    },
+    {
+      "epoch": 0.377207525409068,
+      "grad_norm": 0.12990175493510217,
+      "learning_rate": 0.00014317188843406304,
+      "loss": 1.2159,
+      "step": 5233
+    },
+    {
+      "epoch": 0.3772796078714049,
+      "grad_norm": 0.11064094733996266,
+      "learning_rate": 0.00014315082793572213,
+      "loss": 1.201,
+      "step": 5234
+    },
+    {
+      "epoch": 0.3773516903337418,
+      "grad_norm": 0.11305656272659552,
+      "learning_rate": 0.00014312976508527506,
+      "loss": 1.2662,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3774237727960787,
+      "grad_norm": 0.12222774861971832,
+      "learning_rate": 0.00014310869988386993,
+      "loss": 1.3457,
+      "step": 5236
+    },
+    {
+      "epoch": 0.3774958552584156,
+      "grad_norm": 0.11780464983917245,
+      "learning_rate": 0.000143087632332655,
+      "loss": 1.2315,
+      "step": 5237
+    },
+    {
+      "epoch": 0.37756793772075253,
+      "grad_norm": 0.11429033847588485,
+      "learning_rate": 0.00014306656243277862,
+      "loss": 1.2877,
+      "step": 5238
+    },
+    {
+      "epoch": 0.37764002018308945,
+      "grad_norm": 0.23720835829593578,
+      "learning_rate": 0.00014304549018538933,
+      "loss": 1.1614,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3777121026454264,
+      "grad_norm": 0.32445216359209583,
+      "learning_rate": 0.00014302441559163574,
+      "loss": 1.3374,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3777841851077633,
+      "grad_norm": 0.10314521508103242,
+      "learning_rate": 0.00014300333865266656,
+      "loss": 1.3399,
+      "step": 5241
+    },
+    {
+      "epoch": 0.3778562675701002,
+      "grad_norm": 0.12015069065627204,
+      "learning_rate": 0.0001429822593696307,
+      "loss": 1.3254,
+      "step": 5242
+    },
+    {
+      "epoch": 0.37792835003243713,
+      "grad_norm": 0.1026654736222182,
+      "learning_rate": 0.0001429611777436772,
+      "loss": 1.2781,
+      "step": 5243
+    },
+    {
+      "epoch": 0.378000432494774,
+      "grad_norm": 0.1135237624997162,
+      "learning_rate": 0.00014294009377595516,
+      "loss": 1.2875,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3780725149571109,
+      "grad_norm": 0.10340183263862536,
+      "learning_rate": 0.00014291900746761385,
+      "loss": 1.2575,
+      "step": 5245
+    },
+    {
+      "epoch": 0.37814459741944784,
+      "grad_norm": 0.13684568761757682,
+      "learning_rate": 0.0001428979188198027,
+      "loss": 1.2588,
+      "step": 5246
+    },
+    {
+      "epoch": 0.37821667988178476,
+      "grad_norm": 0.11521589139519624,
+      "learning_rate": 0.00014287682783367117,
+      "loss": 1.3726,
+      "step": 5247
+    },
+    {
+      "epoch": 0.3782887623441217,
+      "grad_norm": 0.10906828169328654,
+      "learning_rate": 0.00014285573451036892,
+      "loss": 1.3089,
+      "step": 5248
+    },
+    {
+      "epoch": 0.3783608448064586,
+      "grad_norm": 0.11620096261874069,
+      "learning_rate": 0.00014283463885104576,
+      "loss": 1.294,
+      "step": 5249
+    },
+    {
+      "epoch": 0.3784329272687955,
+      "grad_norm": 0.11549675299934276,
+      "learning_rate": 0.00014281354085685162,
+      "loss": 1.2749,
+      "step": 5250
+    },
+    {
+      "epoch": 0.37850500973113244,
+      "grad_norm": 0.11629732984366695,
+      "learning_rate": 0.0001427924405289364,
+      "loss": 1.3465,
+      "step": 5251
+    },
+    {
+      "epoch": 0.37857709219346936,
+      "grad_norm": 0.12448865732666389,
+      "learning_rate": 0.0001427713378684504,
+      "loss": 1.2305,
+      "step": 5252
+    },
+    {
+      "epoch": 0.3786491746558062,
+      "grad_norm": 0.10678523875136871,
+      "learning_rate": 0.0001427502328765438,
+      "loss": 1.2401,
+      "step": 5253
+    },
+    {
+      "epoch": 0.37872125711814314,
+      "grad_norm": 0.26219970153439887,
+      "learning_rate": 0.00014272912555436708,
+      "loss": 1.3681,
+      "step": 5254
+    },
+    {
+      "epoch": 0.37879333958048006,
+      "grad_norm": 0.11470974818875819,
+      "learning_rate": 0.00014270801590307073,
+      "loss": 1.2055,
+      "step": 5255
+    },
+    {
+      "epoch": 0.378865422042817,
+      "grad_norm": 0.11262873079030372,
+      "learning_rate": 0.00014268690392380546,
+      "loss": 1.2143,
+      "step": 5256
+    },
+    {
+      "epoch": 0.3789375045051539,
+      "grad_norm": 0.12365828616441353,
+      "learning_rate": 0.00014266578961772203,
+      "loss": 1.2688,
+      "step": 5257
+    },
+    {
+      "epoch": 0.3790095869674908,
+      "grad_norm": 0.1386233672201136,
+      "learning_rate": 0.00014264467298597138,
+      "loss": 1.4018,
+      "step": 5258
+    },
+    {
+      "epoch": 0.37908166942982774,
+      "grad_norm": 0.11367017646983689,
+      "learning_rate": 0.0001426235540297045,
+      "loss": 1.2975,
+      "step": 5259
+    },
+    {
+      "epoch": 0.37915375189216466,
+      "grad_norm": 0.11886705602279794,
+      "learning_rate": 0.00014260243275007265,
+      "loss": 1.3014,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3792258343545015,
+      "grad_norm": 0.10644990840546038,
+      "learning_rate": 0.00014258130914822706,
+      "loss": 1.3718,
+      "step": 5261
+    },
+    {
+      "epoch": 0.37929791681683844,
+      "grad_norm": 0.10441383968189785,
+      "learning_rate": 0.00014256018322531918,
+      "loss": 1.3894,
+      "step": 5262
+    },
+    {
+      "epoch": 0.37936999927917536,
+      "grad_norm": 0.11082322398155917,
+      "learning_rate": 0.00014253905498250058,
+      "loss": 1.1815,
+      "step": 5263
+    },
+    {
+      "epoch": 0.3794420817415123,
+      "grad_norm": 0.11341585784658098,
+      "learning_rate": 0.00014251792442092288,
+      "loss": 1.2967,
+      "step": 5264
+    },
+    {
+      "epoch": 0.3795141642038492,
+      "grad_norm": 0.34338602836785187,
+      "learning_rate": 0.00014249679154173796,
+      "loss": 1.2744,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3795862466661861,
+      "grad_norm": 0.10811906067625586,
+      "learning_rate": 0.0001424756563460977,
+      "loss": 1.2832,
+      "step": 5266
+    },
+    {
+      "epoch": 0.37965832912852304,
+      "grad_norm": 0.11620780490348522,
+      "learning_rate": 0.00014245451883515418,
+      "loss": 1.3132,
+      "step": 5267
+    },
+    {
+      "epoch": 0.37973041159085996,
+      "grad_norm": 0.12151188575228754,
+      "learning_rate": 0.00014243337901005958,
+      "loss": 1.319,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3798024940531969,
+      "grad_norm": 0.10441138881271188,
+      "learning_rate": 0.00014241223687196615,
+      "loss": 1.2558,
+      "step": 5269
+    },
+    {
+      "epoch": 0.37987457651553375,
+      "grad_norm": 0.10336722736886392,
+      "learning_rate": 0.00014239109242202645,
+      "loss": 1.2509,
+      "step": 5270
+    },
+    {
+      "epoch": 0.37994665897787067,
+      "grad_norm": 0.20769360726038874,
+      "learning_rate": 0.0001423699456613929,
+      "loss": 1.2924,
+      "step": 5271
+    },
+    {
+      "epoch": 0.3800187414402076,
+      "grad_norm": 0.09871150373473732,
+      "learning_rate": 0.00014234879659121835,
+      "loss": 1.3144,
+      "step": 5272
+    },
+    {
+      "epoch": 0.3800908239025445,
+      "grad_norm": 0.11166933847152846,
+      "learning_rate": 0.00014232764521265542,
+      "loss": 1.2782,
+      "step": 5273
+    },
+    {
+      "epoch": 0.3801629063648814,
+      "grad_norm": 0.09837839703055372,
+      "learning_rate": 0.00014230649152685722,
+      "loss": 1.349,
+      "step": 5274
+    },
+    {
+      "epoch": 0.38023498882721835,
+      "grad_norm": 0.1093997442826519,
+      "learning_rate": 0.00014228533553497673,
+      "loss": 1.3448,
+      "step": 5275
+    },
+    {
+      "epoch": 0.38030707128955527,
+      "grad_norm": 0.11050686951881379,
+      "learning_rate": 0.00014226417723816717,
+      "loss": 1.388,
+      "step": 5276
+    },
+    {
+      "epoch": 0.3803791537518922,
+      "grad_norm": 0.11358008371487478,
+      "learning_rate": 0.00014224301663758182,
+      "loss": 1.3382,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3804512362142291,
+      "grad_norm": 0.10940652035341271,
+      "learning_rate": 0.00014222185373437415,
+      "loss": 1.2924,
+      "step": 5278
+    },
+    {
+      "epoch": 0.38052331867656597,
+      "grad_norm": 0.10562903476474506,
+      "learning_rate": 0.00014220068852969773,
+      "loss": 1.2615,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3805954011389029,
+      "grad_norm": 0.11648922649293124,
+      "learning_rate": 0.00014217952102470627,
+      "loss": 1.4624,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3806674836012398,
+      "grad_norm": 0.11481916318260176,
+      "learning_rate": 0.00014215835122055354,
+      "loss": 1.3505,
+      "step": 5281
+    },
+    {
+      "epoch": 0.38073956606357673,
+      "grad_norm": 0.24272962108780188,
+      "learning_rate": 0.0001421371791183935,
+      "loss": 1.2998,
+      "step": 5282
+    },
+    {
+      "epoch": 0.38081164852591365,
+      "grad_norm": 0.10376557689059715,
+      "learning_rate": 0.00014211600471938026,
+      "loss": 1.3459,
+      "step": 5283
+    },
+    {
+      "epoch": 0.38088373098825057,
+      "grad_norm": 0.11437434648668669,
+      "learning_rate": 0.00014209482802466795,
+      "loss": 1.2582,
+      "step": 5284
+    },
+    {
+      "epoch": 0.3809558134505875,
+      "grad_norm": 0.10929537258423926,
+      "learning_rate": 0.00014207364903541094,
+      "loss": 1.2112,
+      "step": 5285
+    },
+    {
+      "epoch": 0.3810278959129244,
+      "grad_norm": 0.10775090718119043,
+      "learning_rate": 0.00014205246775276367,
+      "loss": 1.2997,
+      "step": 5286
+    },
+    {
+      "epoch": 0.3810999783752613,
+      "grad_norm": 0.12319451600343759,
+      "learning_rate": 0.00014203128417788067,
+      "loss": 1.3234,
+      "step": 5287
+    },
+    {
+      "epoch": 0.3811720608375982,
+      "grad_norm": 0.10594343411393105,
+      "learning_rate": 0.00014201009831191666,
+      "loss": 1.2672,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3812441432999351,
+      "grad_norm": 0.12792850675256562,
+      "learning_rate": 0.00014198891015602646,
+      "loss": 1.2687,
+      "step": 5289
+    },
+    {
+      "epoch": 0.38131622576227203,
+      "grad_norm": 0.10549358681506717,
+      "learning_rate": 0.00014196771971136505,
+      "loss": 1.2199,
+      "step": 5290
+    },
+    {
+      "epoch": 0.38138830822460895,
+      "grad_norm": 0.09858627142563928,
+      "learning_rate": 0.0001419465269790874,
+      "loss": 1.2879,
+      "step": 5291
+    },
+    {
+      "epoch": 0.38146039068694587,
+      "grad_norm": 0.10522205388690324,
+      "learning_rate": 0.0001419253319603488,
+      "loss": 1.3305,
+      "step": 5292
+    },
+    {
+      "epoch": 0.3815324731492828,
+      "grad_norm": 0.11865344238623807,
+      "learning_rate": 0.00014190413465630451,
+      "loss": 1.3274,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3816045556116197,
+      "grad_norm": 0.11768692065139069,
+      "learning_rate": 0.00014188293506811007,
+      "loss": 1.2318,
+      "step": 5294
+    },
+    {
+      "epoch": 0.38167663807395663,
+      "grad_norm": 0.11483903042759837,
+      "learning_rate": 0.0001418617331969209,
+      "loss": 1.2752,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3817487205362935,
+      "grad_norm": 0.10978625164460107,
+      "learning_rate": 0.00014184052904389277,
+      "loss": 1.3022,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3818208029986304,
+      "grad_norm": 0.11143910447264944,
+      "learning_rate": 0.0001418193226101815,
+      "loss": 1.3786,
+      "step": 5297
+    },
+    {
+      "epoch": 0.38189288546096734,
+      "grad_norm": 0.10979001167004085,
+      "learning_rate": 0.00014179811389694304,
+      "loss": 1.3403,
+      "step": 5298
+    },
+    {
+      "epoch": 0.38196496792330425,
+      "grad_norm": 0.11291568793722358,
+      "learning_rate": 0.0001417769029053334,
+      "loss": 1.3341,
+      "step": 5299
+    },
+    {
+      "epoch": 0.3820370503856412,
+      "grad_norm": 0.11590015503576846,
+      "learning_rate": 0.00014175568963650884,
+      "loss": 1.2645,
+      "step": 5300
+    },
+    {
+      "epoch": 0.3821091328479781,
+      "grad_norm": 0.1033033309974075,
+      "learning_rate": 0.00014173447409162563,
+      "loss": 1.3132,
+      "step": 5301
+    },
+    {
+      "epoch": 0.382181215310315,
+      "grad_norm": 0.101688568141992,
+      "learning_rate": 0.00014171325627184022,
+      "loss": 1.2545,
+      "step": 5302
+    },
+    {
+      "epoch": 0.38225329777265193,
+      "grad_norm": 0.11272083173917373,
+      "learning_rate": 0.00014169203617830914,
+      "loss": 1.3442,
+      "step": 5303
+    },
+    {
+      "epoch": 0.38232538023498885,
+      "grad_norm": 0.1065048942561958,
+      "learning_rate": 0.00014167081381218912,
+      "loss": 1.2556,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3823974626973257,
+      "grad_norm": 0.10276918266978395,
+      "learning_rate": 0.00014164958917463696,
+      "loss": 1.3022,
+      "step": 5305
+    },
+    {
+      "epoch": 0.38246954515966264,
+      "grad_norm": 0.11360557440941306,
+      "learning_rate": 0.00014162836226680955,
+      "loss": 1.3165,
+      "step": 5306
+    },
+    {
+      "epoch": 0.38254162762199956,
+      "grad_norm": 0.12464457395024217,
+      "learning_rate": 0.000141607133089864,
+      "loss": 1.2852,
+      "step": 5307
+    },
+    {
+      "epoch": 0.3826137100843365,
+      "grad_norm": 0.11890971412385205,
+      "learning_rate": 0.00014158590164495746,
+      "loss": 1.2287,
+      "step": 5308
+    },
+    {
+      "epoch": 0.3826857925466734,
+      "grad_norm": 0.17292411861731075,
+      "learning_rate": 0.00014156466793324726,
+      "loss": 1.2362,
+      "step": 5309
+    },
+    {
+      "epoch": 0.3827578750090103,
+      "grad_norm": 0.10310293555840262,
+      "learning_rate": 0.00014154343195589078,
+      "loss": 1.2854,
+      "step": 5310
+    },
+    {
+      "epoch": 0.38282995747134724,
+      "grad_norm": 0.15884019790938836,
+      "learning_rate": 0.00014152219371404564,
+      "loss": 1.1783,
+      "step": 5311
+    },
+    {
+      "epoch": 0.38290203993368416,
+      "grad_norm": 0.11831568077164507,
+      "learning_rate": 0.00014150095320886942,
+      "loss": 1.3804,
+      "step": 5312
+    },
+    {
+      "epoch": 0.382974122396021,
+      "grad_norm": 0.11241146270414505,
+      "learning_rate": 0.00014147971044152003,
+      "loss": 1.3408,
+      "step": 5313
+    },
+    {
+      "epoch": 0.38304620485835794,
+      "grad_norm": 0.09977139397575757,
+      "learning_rate": 0.0001414584654131553,
+      "loss": 1.2803,
+      "step": 5314
+    },
+    {
+      "epoch": 0.38311828732069486,
+      "grad_norm": 0.1031766155382484,
+      "learning_rate": 0.00014143721812493333,
+      "loss": 1.3683,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3831903697830318,
+      "grad_norm": 0.10804091451100756,
+      "learning_rate": 0.00014141596857801222,
+      "loss": 1.1774,
+      "step": 5316
+    },
+    {
+      "epoch": 0.3832624522453687,
+      "grad_norm": 0.10977979084976693,
+      "learning_rate": 0.00014139471677355033,
+      "loss": 1.2593,
+      "step": 5317
+    },
+    {
+      "epoch": 0.3833345347077056,
+      "grad_norm": 0.11951232788953797,
+      "learning_rate": 0.00014137346271270604,
+      "loss": 1.3575,
+      "step": 5318
+    },
+    {
+      "epoch": 0.38340661717004254,
+      "grad_norm": 0.107712179275847,
+      "learning_rate": 0.0001413522063966379,
+      "loss": 1.2587,
+      "step": 5319
+    },
+    {
+      "epoch": 0.38347869963237946,
+      "grad_norm": 0.1179795591293683,
+      "learning_rate": 0.00014133094782650457,
+      "loss": 1.2409,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3835507820947164,
+      "grad_norm": 0.212781527170118,
+      "learning_rate": 0.00014130968700346483,
+      "loss": 1.3753,
+      "step": 5321
+    },
+    {
+      "epoch": 0.38362286455705324,
+      "grad_norm": 0.1045204095671103,
+      "learning_rate": 0.00014128842392867758,
+      "loss": 1.3467,
+      "step": 5322
+    },
+    {
+      "epoch": 0.38369494701939016,
+      "grad_norm": 0.10970888522687305,
+      "learning_rate": 0.00014126715860330187,
+      "loss": 1.2228,
+      "step": 5323
+    },
+    {
+      "epoch": 0.3837670294817271,
+      "grad_norm": 0.42322604153159465,
+      "learning_rate": 0.00014124589102849677,
+      "loss": 1.2832,
+      "step": 5324
+    },
+    {
+      "epoch": 0.383839111944064,
+      "grad_norm": 0.12482908443804135,
+      "learning_rate": 0.00014122462120542167,
+      "loss": 1.2924,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3839111944064009,
+      "grad_norm": 0.11239074897242803,
+      "learning_rate": 0.00014120334913523587,
+      "loss": 1.3398,
+      "step": 5326
+    },
+    {
+      "epoch": 0.38398327686873784,
+      "grad_norm": 0.12223656989454293,
+      "learning_rate": 0.00014118207481909895,
+      "loss": 1.3519,
+      "step": 5327
+    },
+    {
+      "epoch": 0.38405535933107476,
+      "grad_norm": 0.10294054904102473,
+      "learning_rate": 0.0001411607982581705,
+      "loss": 1.1716,
+      "step": 5328
+    },
+    {
+      "epoch": 0.3841274417934117,
+      "grad_norm": 0.16331060773778455,
+      "learning_rate": 0.00014113951945361034,
+      "loss": 1.3132,
+      "step": 5329
+    },
+    {
+      "epoch": 0.3841995242557486,
+      "grad_norm": 0.11533623560846462,
+      "learning_rate": 0.0001411182384065783,
+      "loss": 1.3073,
+      "step": 5330
+    },
+    {
+      "epoch": 0.38427160671808547,
+      "grad_norm": 0.12123176227991367,
+      "learning_rate": 0.00014109695511823445,
+      "loss": 1.2023,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3843436891804224,
+      "grad_norm": 0.12275073102287065,
+      "learning_rate": 0.00014107566958973885,
+      "loss": 1.3306,
+      "step": 5332
+    },
+    {
+      "epoch": 0.3844157716427593,
+      "grad_norm": 0.11242531463828649,
+      "learning_rate": 0.0001410543818222518,
+      "loss": 1.3565,
+      "step": 5333
+    },
+    {
+      "epoch": 0.3844878541050962,
+      "grad_norm": 0.14515519340413943,
+      "learning_rate": 0.00014103309181693366,
+      "loss": 1.3649,
+      "step": 5334
+    },
+    {
+      "epoch": 0.38455993656743315,
+      "grad_norm": 0.1008852422852911,
+      "learning_rate": 0.0001410117995749449,
+      "loss": 1.3158,
+      "step": 5335
+    },
+    {
+      "epoch": 0.38463201902977007,
+      "grad_norm": 0.12417588268817,
+      "learning_rate": 0.00014099050509744617,
+      "loss": 1.3177,
+      "step": 5336
+    },
+    {
+      "epoch": 0.384704101492107,
+      "grad_norm": 0.12272442393299723,
+      "learning_rate": 0.0001409692083855982,
+      "loss": 1.2879,
+      "step": 5337
+    },
+    {
+      "epoch": 0.3847761839544439,
+      "grad_norm": 0.12704602307440319,
+      "learning_rate": 0.00014094790944056186,
+      "loss": 1.3072,
+      "step": 5338
+    },
+    {
+      "epoch": 0.38484826641678077,
+      "grad_norm": 0.11288636858213798,
+      "learning_rate": 0.00014092660826349812,
+      "loss": 1.2707,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3849203488791177,
+      "grad_norm": 0.1162038370710555,
+      "learning_rate": 0.0001409053048555681,
+      "loss": 1.3472,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3849924313414546,
+      "grad_norm": 0.13211442686368535,
+      "learning_rate": 0.00014088399921793302,
+      "loss": 1.2508,
+      "step": 5341
+    },
+    {
+      "epoch": 0.38506451380379153,
+      "grad_norm": 0.11006296451940151,
+      "learning_rate": 0.0001408626913517542,
+      "loss": 1.3362,
+      "step": 5342
+    },
+    {
+      "epoch": 0.38513659626612845,
+      "grad_norm": 0.10706174668821965,
+      "learning_rate": 0.00014084138125819318,
+      "loss": 1.3977,
+      "step": 5343
+    },
+    {
+      "epoch": 0.38520867872846537,
+      "grad_norm": 0.10434041014548401,
+      "learning_rate": 0.00014082006893841145,
+      "loss": 1.279,
+      "step": 5344
+    },
+    {
+      "epoch": 0.3852807611908023,
+      "grad_norm": 0.11102056134174741,
+      "learning_rate": 0.0001407987543935708,
+      "loss": 1.3153,
+      "step": 5345
+    },
+    {
+      "epoch": 0.3853528436531392,
+      "grad_norm": 0.24833041528764044,
+      "learning_rate": 0.00014077743762483302,
+      "loss": 1.2727,
+      "step": 5346
+    },
+    {
+      "epoch": 0.38542492611547613,
+      "grad_norm": 0.14920791109083906,
+      "learning_rate": 0.0001407561186333601,
+      "loss": 1.2019,
+      "step": 5347
+    },
+    {
+      "epoch": 0.385497008577813,
+      "grad_norm": 0.10296918931625913,
+      "learning_rate": 0.0001407347974203141,
+      "loss": 1.2155,
+      "step": 5348
+    },
+    {
+      "epoch": 0.3855690910401499,
+      "grad_norm": 0.11539238464670128,
+      "learning_rate": 0.00014071347398685724,
+      "loss": 1.276,
+      "step": 5349
+    },
+    {
+      "epoch": 0.38564117350248683,
+      "grad_norm": 0.11321620921835808,
+      "learning_rate": 0.00014069214833415175,
+      "loss": 1.314,
+      "step": 5350
+    },
+    {
+      "epoch": 0.38571325596482375,
+      "grad_norm": 0.11510102125634519,
+      "learning_rate": 0.0001406708204633602,
+      "loss": 1.3684,
+      "step": 5351
+    },
+    {
+      "epoch": 0.3857853384271607,
+      "grad_norm": 0.11047836916405275,
+      "learning_rate": 0.00014064949037564503,
+      "loss": 1.1764,
+      "step": 5352
+    },
+    {
+      "epoch": 0.3858574208894976,
+      "grad_norm": 0.1970923066514857,
+      "learning_rate": 0.00014062815807216902,
+      "loss": 1.2583,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3859295033518345,
+      "grad_norm": 0.11694728281355486,
+      "learning_rate": 0.00014060682355409492,
+      "loss": 1.2725,
+      "step": 5354
+    },
+    {
+      "epoch": 0.38600158581417143,
+      "grad_norm": 0.0993517768691023,
+      "learning_rate": 0.00014058548682258564,
+      "loss": 1.2504,
+      "step": 5355
+    },
+    {
+      "epoch": 0.38607366827650835,
+      "grad_norm": 0.11120036111317577,
+      "learning_rate": 0.00014056414787880423,
+      "loss": 1.2164,
+      "step": 5356
+    },
+    {
+      "epoch": 0.3861457507388452,
+      "grad_norm": 0.11758824856473818,
+      "learning_rate": 0.0001405428067239139,
+      "loss": 1.3129,
+      "step": 5357
+    },
+    {
+      "epoch": 0.38621783320118214,
+      "grad_norm": 0.11771867787759081,
+      "learning_rate": 0.00014052146335907786,
+      "loss": 1.2683,
+      "step": 5358
+    },
+    {
+      "epoch": 0.38628991566351906,
+      "grad_norm": 0.11204959467672668,
+      "learning_rate": 0.00014050011778545955,
+      "loss": 1.3266,
+      "step": 5359
+    },
+    {
+      "epoch": 0.386361998125856,
+      "grad_norm": 0.11482235360396234,
+      "learning_rate": 0.00014047877000422253,
+      "loss": 1.3225,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3864340805881929,
+      "grad_norm": 0.1097363647558213,
+      "learning_rate": 0.00014045742001653037,
+      "loss": 1.2393,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3865061630505298,
+      "grad_norm": 0.11119144694038204,
+      "learning_rate": 0.00014043606782354693,
+      "loss": 1.3813,
+      "step": 5362
+    },
+    {
+      "epoch": 0.38657824551286674,
+      "grad_norm": 0.20804417650552137,
+      "learning_rate": 0.00014041471342643598,
+      "loss": 1.3417,
+      "step": 5363
+    },
+    {
+      "epoch": 0.38665032797520366,
+      "grad_norm": 0.1256399248093355,
+      "learning_rate": 0.00014039335682636164,
+      "loss": 1.3709,
+      "step": 5364
+    },
+    {
+      "epoch": 0.3867224104375405,
+      "grad_norm": 0.11293186930891039,
+      "learning_rate": 0.00014037199802448798,
+      "loss": 1.3605,
+      "step": 5365
+    },
+    {
+      "epoch": 0.38679449289987744,
+      "grad_norm": 0.10346779863184515,
+      "learning_rate": 0.00014035063702197922,
+      "loss": 1.254,
+      "step": 5366
+    },
+    {
+      "epoch": 0.38686657536221436,
+      "grad_norm": 0.12882423385441166,
+      "learning_rate": 0.00014032927381999976,
+      "loss": 1.3204,
+      "step": 5367
+    },
+    {
+      "epoch": 0.3869386578245513,
+      "grad_norm": 0.6016756762116624,
+      "learning_rate": 0.0001403079084197141,
+      "loss": 1.2394,
+      "step": 5368
+    },
+    {
+      "epoch": 0.3870107402868882,
+      "grad_norm": 0.12360344595630096,
+      "learning_rate": 0.00014028654082228682,
+      "loss": 1.2494,
+      "step": 5369
+    },
+    {
+      "epoch": 0.3870828227492251,
+      "grad_norm": 0.1178016917337927,
+      "learning_rate": 0.0001402651710288826,
+      "loss": 1.3258,
+      "step": 5370
+    },
+    {
+      "epoch": 0.38715490521156204,
+      "grad_norm": 0.10583358894208729,
+      "learning_rate": 0.00014024379904066642,
+      "loss": 1.3673,
+      "step": 5371
+    },
+    {
+      "epoch": 0.38722698767389896,
+      "grad_norm": 0.10999346221064885,
+      "learning_rate": 0.0001402224248588031,
+      "loss": 1.3505,
+      "step": 5372
+    },
+    {
+      "epoch": 0.3872990701362359,
+      "grad_norm": 0.09776142598686682,
+      "learning_rate": 0.00014020104848445783,
+      "loss": 1.4372,
+      "step": 5373
+    },
+    {
+      "epoch": 0.38737115259857274,
+      "grad_norm": 0.12261794905643794,
+      "learning_rate": 0.00014017966991879574,
+      "loss": 1.2891,
+      "step": 5374
+    },
+    {
+      "epoch": 0.38744323506090966,
+      "grad_norm": 0.10945907142416243,
+      "learning_rate": 0.0001401582891629822,
+      "loss": 1.1796,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3875153175232466,
+      "grad_norm": 0.1151058571713858,
+      "learning_rate": 0.00014013690621818262,
+      "loss": 1.2765,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3875873999855835,
+      "grad_norm": 0.10670983112710242,
+      "learning_rate": 0.0001401155210855626,
+      "loss": 1.3125,
+      "step": 5377
+    },
+    {
+      "epoch": 0.3876594824479204,
+      "grad_norm": 0.10437395003831015,
+      "learning_rate": 0.00014009413376628776,
+      "loss": 1.2714,
+      "step": 5378
+    },
+    {
+      "epoch": 0.38773156491025734,
+      "grad_norm": 0.11056332207176225,
+      "learning_rate": 0.00014007274426152395,
+      "loss": 1.2637,
+      "step": 5379
+    },
+    {
+      "epoch": 0.38780364737259426,
+      "grad_norm": 0.10657743524217808,
+      "learning_rate": 0.0001400513525724371,
+      "loss": 1.3177,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3878757298349312,
+      "grad_norm": 0.11797977274136971,
+      "learning_rate": 0.0001400299587001932,
+      "loss": 1.3171,
+      "step": 5381
+    },
+    {
+      "epoch": 0.3879478122972681,
+      "grad_norm": 0.11394352014289415,
+      "learning_rate": 0.00014000856264595842,
+      "loss": 1.3121,
+      "step": 5382
+    },
+    {
+      "epoch": 0.38801989475960497,
+      "grad_norm": 0.11535088013152457,
+      "learning_rate": 0.00013998716441089907,
+      "loss": 1.2308,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3880919772219419,
+      "grad_norm": 0.3256760825325778,
+      "learning_rate": 0.00013996576399618154,
+      "loss": 1.2787,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3881640596842788,
+      "grad_norm": 0.10655712029721348,
+      "learning_rate": 0.0001399443614029723,
+      "loss": 1.3092,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3882361421466157,
+      "grad_norm": 0.1069495196486042,
+      "learning_rate": 0.000139922956632438,
+      "loss": 1.2957,
+      "step": 5386
+    },
+    {
+      "epoch": 0.38830822460895265,
+      "grad_norm": 0.1144832854395184,
+      "learning_rate": 0.00013990154968574544,
+      "loss": 1.349,
+      "step": 5387
+    },
+    {
+      "epoch": 0.38838030707128957,
+      "grad_norm": 0.12250061637477762,
+      "learning_rate": 0.00013988014056406144,
+      "loss": 1.3317,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3884523895336265,
+      "grad_norm": 0.10757030093524188,
+      "learning_rate": 0.000139858729268553,
+      "loss": 1.3747,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3885244719959634,
+      "grad_norm": 0.11622664652400407,
+      "learning_rate": 0.00013983731580038722,
+      "loss": 1.2403,
+      "step": 5390
+    },
+    {
+      "epoch": 0.38859655445830027,
+      "grad_norm": 0.11130551332640551,
+      "learning_rate": 0.00013981590016073133,
+      "loss": 1.3011,
+      "step": 5391
+    },
+    {
+      "epoch": 0.3886686369206372,
+      "grad_norm": 0.10325904241596257,
+      "learning_rate": 0.0001397944823507527,
+      "loss": 1.3543,
+      "step": 5392
+    },
+    {
+      "epoch": 0.3887407193829741,
+      "grad_norm": 0.12233509189220876,
+      "learning_rate": 0.00013977306237161877,
+      "loss": 1.2242,
+      "step": 5393
+    },
+    {
+      "epoch": 0.38881280184531103,
+      "grad_norm": 0.12307804903061254,
+      "learning_rate": 0.00013975164022449709,
+      "loss": 1.3574,
+      "step": 5394
+    },
+    {
+      "epoch": 0.38888488430764795,
+      "grad_norm": 0.11348537585055558,
+      "learning_rate": 0.0001397302159105554,
+      "loss": 1.2564,
+      "step": 5395
+    },
+    {
+      "epoch": 0.38895696676998487,
+      "grad_norm": 0.11014627076716545,
+      "learning_rate": 0.00013970878943096152,
+      "loss": 1.2791,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3890290492323218,
+      "grad_norm": 0.16821342265106454,
+      "learning_rate": 0.00013968736078688338,
+      "loss": 1.3225,
+      "step": 5397
+    },
+    {
+      "epoch": 0.3891011316946587,
+      "grad_norm": 0.11468207735807225,
+      "learning_rate": 0.00013966592997948903,
+      "loss": 1.2376,
+      "step": 5398
+    },
+    {
+      "epoch": 0.38917321415699563,
+      "grad_norm": 0.11762154474363512,
+      "learning_rate": 0.0001396444970099466,
+      "loss": 1.2247,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3892452966193325,
+      "grad_norm": 0.11329133622954883,
+      "learning_rate": 0.00013962306187942445,
+      "loss": 1.3059,
+      "step": 5400
+    },
+    {
+      "epoch": 0.3893173790816694,
+      "grad_norm": 0.09968981130144372,
+      "learning_rate": 0.0001396016245890909,
+      "loss": 1.2884,
+      "step": 5401
+    },
+    {
+      "epoch": 0.38938946154400633,
+      "grad_norm": 0.2735999758702169,
+      "learning_rate": 0.00013958018514011457,
+      "loss": 1.3017,
+      "step": 5402
+    },
+    {
+      "epoch": 0.38946154400634325,
+      "grad_norm": 0.10701584625662773,
+      "learning_rate": 0.000139558743533664,
+      "loss": 1.2545,
+      "step": 5403
+    },
+    {
+      "epoch": 0.38953362646868017,
+      "grad_norm": 0.12002921859911123,
+      "learning_rate": 0.0001395372997709081,
+      "loss": 1.3663,
+      "step": 5404
+    },
+    {
+      "epoch": 0.3896057089310171,
+      "grad_norm": 0.09685174481745486,
+      "learning_rate": 0.00013951585385301555,
+      "loss": 1.2694,
+      "step": 5405
+    },
+    {
+      "epoch": 0.389677791393354,
+      "grad_norm": 0.10763395926700518,
+      "learning_rate": 0.0001394944057811555,
+      "loss": 1.281,
+      "step": 5406
+    },
+    {
+      "epoch": 0.38974987385569093,
+      "grad_norm": 0.10941287111368168,
+      "learning_rate": 0.000139472955556497,
+      "loss": 1.2627,
+      "step": 5407
+    },
+    {
+      "epoch": 0.38982195631802785,
+      "grad_norm": 0.11167520800065892,
+      "learning_rate": 0.00013945150318020927,
+      "loss": 1.3442,
+      "step": 5408
+    },
+    {
+      "epoch": 0.3898940387803647,
+      "grad_norm": 0.1015974932580734,
+      "learning_rate": 0.0001394300486534617,
+      "loss": 1.3914,
+      "step": 5409
+    },
+    {
+      "epoch": 0.38996612124270164,
+      "grad_norm": 0.11041036106554711,
+      "learning_rate": 0.0001394085919774237,
+      "loss": 1.3404,
+      "step": 5410
+    },
+    {
+      "epoch": 0.39003820370503856,
+      "grad_norm": 0.10994512139201584,
+      "learning_rate": 0.0001393871331532649,
+      "loss": 1.3185,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3901102861673755,
+      "grad_norm": 0.1004221814269221,
+      "learning_rate": 0.00013936567218215498,
+      "loss": 1.2955,
+      "step": 5412
+    },
+    {
+      "epoch": 0.3901823686297124,
+      "grad_norm": 0.09752038157388923,
+      "learning_rate": 0.00013934420906526373,
+      "loss": 1.2534,
+      "step": 5413
+    },
+    {
+      "epoch": 0.3902544510920493,
+      "grad_norm": 0.12655658987375895,
+      "learning_rate": 0.00013932274380376111,
+      "loss": 1.3393,
+      "step": 5414
+    },
+    {
+      "epoch": 0.39032653355438623,
+      "grad_norm": 0.11218696406862577,
+      "learning_rate": 0.00013930127639881715,
+      "loss": 1.29,
+      "step": 5415
+    },
+    {
+      "epoch": 0.39039861601672315,
+      "grad_norm": 0.12652292844931357,
+      "learning_rate": 0.00013927980685160204,
+      "loss": 1.3525,
+      "step": 5416
+    },
+    {
+      "epoch": 0.39047069847906,
+      "grad_norm": 0.1037559578470304,
+      "learning_rate": 0.00013925833516328606,
+      "loss": 1.3148,
+      "step": 5417
+    },
+    {
+      "epoch": 0.39054278094139694,
+      "grad_norm": 0.11083381359128323,
+      "learning_rate": 0.00013923686133503958,
+      "loss": 1.358,
+      "step": 5418
+    },
+    {
+      "epoch": 0.39061486340373386,
+      "grad_norm": 0.10431121807054822,
+      "learning_rate": 0.00013921538536803313,
+      "loss": 1.3404,
+      "step": 5419
+    },
+    {
+      "epoch": 0.3906869458660708,
+      "grad_norm": 0.11415005292462915,
+      "learning_rate": 0.00013919390726343739,
+      "loss": 1.2805,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3907590283284077,
+      "grad_norm": 0.10120495296775776,
+      "learning_rate": 0.00013917242702242303,
+      "loss": 1.3172,
+      "step": 5421
+    },
+    {
+      "epoch": 0.3908311107907446,
+      "grad_norm": 0.10967588905560169,
+      "learning_rate": 0.00013915094464616096,
+      "loss": 1.3166,
+      "step": 5422
+    },
+    {
+      "epoch": 0.39090319325308154,
+      "grad_norm": 0.13517622139289773,
+      "learning_rate": 0.00013912946013582212,
+      "loss": 1.3303,
+      "step": 5423
+    },
+    {
+      "epoch": 0.39097527571541846,
+      "grad_norm": 0.10919678470022502,
+      "learning_rate": 0.0001391079734925777,
+      "loss": 1.3146,
+      "step": 5424
+    },
+    {
+      "epoch": 0.3910473581777554,
+      "grad_norm": 0.10463142874802965,
+      "learning_rate": 0.00013908648471759878,
+      "loss": 1.3433,
+      "step": 5425
+    },
+    {
+      "epoch": 0.39111944064009224,
+      "grad_norm": 0.17181076064730713,
+      "learning_rate": 0.00013906499381205684,
+      "loss": 1.3046,
+      "step": 5426
+    },
+    {
+      "epoch": 0.39119152310242916,
+      "grad_norm": 0.1196938070504818,
+      "learning_rate": 0.0001390435007771232,
+      "loss": 1.265,
+      "step": 5427
+    },
+    {
+      "epoch": 0.3912636055647661,
+      "grad_norm": 0.10684612877227784,
+      "learning_rate": 0.0001390220056139695,
+      "loss": 1.2888,
+      "step": 5428
+    },
+    {
+      "epoch": 0.391335688027103,
+      "grad_norm": 0.10328172919237029,
+      "learning_rate": 0.00013900050832376737,
+      "loss": 1.2447,
+      "step": 5429
+    },
+    {
+      "epoch": 0.3914077704894399,
+      "grad_norm": 0.11820853144478115,
+      "learning_rate": 0.00013897900890768867,
+      "loss": 1.2849,
+      "step": 5430
+    },
+    {
+      "epoch": 0.39147985295177684,
+      "grad_norm": 0.10134302162035923,
+      "learning_rate": 0.00013895750736690526,
+      "loss": 1.2175,
+      "step": 5431
+    },
+    {
+      "epoch": 0.39155193541411376,
+      "grad_norm": 0.10194565192183166,
+      "learning_rate": 0.00013893600370258914,
+      "loss": 1.3534,
+      "step": 5432
+    },
+    {
+      "epoch": 0.3916240178764507,
+      "grad_norm": 0.10736832613159839,
+      "learning_rate": 0.0001389144979159125,
+      "loss": 1.3113,
+      "step": 5433
+    },
+    {
+      "epoch": 0.3916961003387876,
+      "grad_norm": 0.11086586242273076,
+      "learning_rate": 0.0001388929900080476,
+      "loss": 1.2656,
+      "step": 5434
+    },
+    {
+      "epoch": 0.39176818280112446,
+      "grad_norm": 0.1072620222243994,
+      "learning_rate": 0.0001388714799801668,
+      "loss": 1.3315,
+      "step": 5435
+    },
+    {
+      "epoch": 0.3918402652634614,
+      "grad_norm": 0.1039436874313552,
+      "learning_rate": 0.00013884996783344256,
+      "loss": 1.2167,
+      "step": 5436
+    },
+    {
+      "epoch": 0.3919123477257983,
+      "grad_norm": 0.12155481790275431,
+      "learning_rate": 0.00013882845356904752,
+      "loss": 1.3052,
+      "step": 5437
+    },
+    {
+      "epoch": 0.3919844301881352,
+      "grad_norm": 0.11203203749380425,
+      "learning_rate": 0.00013880693718815442,
+      "loss": 1.3064,
+      "step": 5438
+    },
+    {
+      "epoch": 0.39205651265047214,
+      "grad_norm": 0.121199917881158,
+      "learning_rate": 0.00013878541869193601,
+      "loss": 1.1674,
+      "step": 5439
+    },
+    {
+      "epoch": 0.39212859511280906,
+      "grad_norm": 0.10962582098826291,
+      "learning_rate": 0.00013876389808156533,
+      "loss": 1.2998,
+      "step": 5440
+    },
+    {
+      "epoch": 0.392200677575146,
+      "grad_norm": 0.11334926073746733,
+      "learning_rate": 0.00013874237535821542,
+      "loss": 1.3333,
+      "step": 5441
+    },
+    {
+      "epoch": 0.3922727600374829,
+      "grad_norm": 0.24027975866775686,
+      "learning_rate": 0.00013872085052305945,
+      "loss": 1.2708,
+      "step": 5442
+    },
+    {
+      "epoch": 0.39234484249981977,
+      "grad_norm": 0.10864025521100534,
+      "learning_rate": 0.0001386993235772707,
+      "loss": 1.2484,
+      "step": 5443
+    },
+    {
+      "epoch": 0.3924169249621567,
+      "grad_norm": 0.10968275659208827,
+      "learning_rate": 0.00013867779452202266,
+      "loss": 1.291,
+      "step": 5444
+    },
+    {
+      "epoch": 0.3924890074244936,
+      "grad_norm": 0.11272285843807892,
+      "learning_rate": 0.00013865626335848874,
+      "loss": 1.3818,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3925610898868305,
+      "grad_norm": 0.11047749035057328,
+      "learning_rate": 0.00013863473008784264,
+      "loss": 1.3363,
+      "step": 5446
+    },
+    {
+      "epoch": 0.39263317234916745,
+      "grad_norm": 0.11694010532560283,
+      "learning_rate": 0.0001386131947112581,
+      "loss": 1.3105,
+      "step": 5447
+    },
+    {
+      "epoch": 0.39270525481150437,
+      "grad_norm": 0.11527273886422985,
+      "learning_rate": 0.0001385916572299091,
+      "loss": 1.3154,
+      "step": 5448
+    },
+    {
+      "epoch": 0.3927773372738413,
+      "grad_norm": 0.10852475735313294,
+      "learning_rate": 0.00013857011764496942,
+      "loss": 1.2759,
+      "step": 5449
+    },
+    {
+      "epoch": 0.3928494197361782,
+      "grad_norm": 0.10791934957352718,
+      "learning_rate": 0.00013854857595761334,
+      "loss": 1.311,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3929215021985151,
+      "grad_norm": 0.10341857309618131,
+      "learning_rate": 0.00013852703216901495,
+      "loss": 1.3002,
+      "step": 5451
+    },
+    {
+      "epoch": 0.392993584660852,
+      "grad_norm": 0.11666636752506525,
+      "learning_rate": 0.00013850548628034867,
+      "loss": 1.3049,
+      "step": 5452
+    },
+    {
+      "epoch": 0.3930656671231889,
+      "grad_norm": 0.11365267571936268,
+      "learning_rate": 0.0001384839382927889,
+      "loss": 1.2379,
+      "step": 5453
+    },
+    {
+      "epoch": 0.39313774958552583,
+      "grad_norm": 0.10709464824172889,
+      "learning_rate": 0.0001384623882075102,
+      "loss": 1.3626,
+      "step": 5454
+    },
+    {
+      "epoch": 0.39320983204786275,
+      "grad_norm": 0.11383491082288424,
+      "learning_rate": 0.00013844083602568725,
+      "loss": 1.3123,
+      "step": 5455
+    },
+    {
+      "epoch": 0.39328191451019967,
+      "grad_norm": 0.14436211733890605,
+      "learning_rate": 0.00013841928174849485,
+      "loss": 1.3411,
+      "step": 5456
+    },
+    {
+      "epoch": 0.3933539969725366,
+      "grad_norm": 0.12043327513023928,
+      "learning_rate": 0.00013839772537710786,
+      "loss": 1.3221,
+      "step": 5457
+    },
+    {
+      "epoch": 0.3934260794348735,
+      "grad_norm": 0.11120297315282023,
+      "learning_rate": 0.00013837616691270136,
+      "loss": 1.3613,
+      "step": 5458
+    },
+    {
+      "epoch": 0.39349816189721043,
+      "grad_norm": 0.12026512321393133,
+      "learning_rate": 0.00013835460635645039,
+      "loss": 1.2682,
+      "step": 5459
+    },
+    {
+      "epoch": 0.39357024435954735,
+      "grad_norm": 0.11353744411140607,
+      "learning_rate": 0.00013833304370953028,
+      "loss": 1.3533,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3936423268218842,
+      "grad_norm": 0.10446567594331657,
+      "learning_rate": 0.00013831147897311633,
+      "loss": 1.2963,
+      "step": 5461
+    },
+    {
+      "epoch": 0.39371440928422113,
+      "grad_norm": 0.12266516598949444,
+      "learning_rate": 0.00013828991214838405,
+      "loss": 1.2993,
+      "step": 5462
+    },
+    {
+      "epoch": 0.39378649174655805,
+      "grad_norm": 0.10029306636616772,
+      "learning_rate": 0.000138268343236509,
+      "loss": 1.2697,
+      "step": 5463
+    },
+    {
+      "epoch": 0.393858574208895,
+      "grad_norm": 0.12384423539111561,
+      "learning_rate": 0.00013824677223866686,
+      "loss": 1.3372,
+      "step": 5464
+    },
+    {
+      "epoch": 0.3939306566712319,
+      "grad_norm": 0.11123055410111268,
+      "learning_rate": 0.0001382251991560335,
+      "loss": 1.3007,
+      "step": 5465
+    },
+    {
+      "epoch": 0.3940027391335688,
+      "grad_norm": 0.10999913811655346,
+      "learning_rate": 0.0001382036239897848,
+      "loss": 1.2138,
+      "step": 5466
+    },
+    {
+      "epoch": 0.39407482159590573,
+      "grad_norm": 0.14686807274611893,
+      "learning_rate": 0.00013818204674109682,
+      "loss": 1.3401,
+      "step": 5467
+    },
+    {
+      "epoch": 0.39414690405824265,
+      "grad_norm": 0.11316318968288129,
+      "learning_rate": 0.0001381604674111457,
+      "loss": 1.3295,
+      "step": 5468
+    },
+    {
+      "epoch": 0.3942189865205795,
+      "grad_norm": 0.11709154554253381,
+      "learning_rate": 0.0001381388860011077,
+      "loss": 1.2767,
+      "step": 5469
+    },
+    {
+      "epoch": 0.39429106898291644,
+      "grad_norm": 0.12736587661196955,
+      "learning_rate": 0.00013811730251215925,
+      "loss": 1.3829,
+      "step": 5470
+    },
+    {
+      "epoch": 0.39436315144525336,
+      "grad_norm": 0.11253447236015843,
+      "learning_rate": 0.00013809571694547678,
+      "loss": 1.2175,
+      "step": 5471
+    },
+    {
+      "epoch": 0.3944352339075903,
+      "grad_norm": 0.12738604850254537,
+      "learning_rate": 0.00013807412930223693,
+      "loss": 1.3412,
+      "step": 5472
+    },
+    {
+      "epoch": 0.3945073163699272,
+      "grad_norm": 0.1517299926498002,
+      "learning_rate": 0.00013805253958361642,
+      "loss": 1.3621,
+      "step": 5473
+    },
+    {
+      "epoch": 0.3945793988322641,
+      "grad_norm": 0.11589179380588308,
+      "learning_rate": 0.00013803094779079204,
+      "loss": 1.3248,
+      "step": 5474
+    },
+    {
+      "epoch": 0.39465148129460104,
+      "grad_norm": 0.10986793335757185,
+      "learning_rate": 0.00013800935392494084,
+      "loss": 1.2414,
+      "step": 5475
+    },
+    {
+      "epoch": 0.39472356375693796,
+      "grad_norm": 0.10943246868270251,
+      "learning_rate": 0.00013798775798723974,
+      "loss": 1.2839,
+      "step": 5476
+    },
+    {
+      "epoch": 0.3947956462192749,
+      "grad_norm": 0.10646991772091594,
+      "learning_rate": 0.000137966159978866,
+      "loss": 1.3474,
+      "step": 5477
+    },
+    {
+      "epoch": 0.39486772868161174,
+      "grad_norm": 0.10715580122571398,
+      "learning_rate": 0.00013794455990099688,
+      "loss": 1.3173,
+      "step": 5478
+    },
+    {
+      "epoch": 0.39493981114394866,
+      "grad_norm": 0.11879761684072535,
+      "learning_rate": 0.00013792295775480983,
+      "loss": 1.2708,
+      "step": 5479
+    },
+    {
+      "epoch": 0.3950118936062856,
+      "grad_norm": 0.11328853757840837,
+      "learning_rate": 0.00013790135354148226,
+      "loss": 1.2651,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3950839760686225,
+      "grad_norm": 0.10664232560395173,
+      "learning_rate": 0.0001378797472621919,
+      "loss": 1.4132,
+      "step": 5481
+    },
+    {
+      "epoch": 0.3951560585309594,
+      "grad_norm": 0.11421951695470653,
+      "learning_rate": 0.00013785813891811636,
+      "loss": 1.3002,
+      "step": 5482
+    },
+    {
+      "epoch": 0.39522814099329634,
+      "grad_norm": 0.11203909539697315,
+      "learning_rate": 0.00013783652851043362,
+      "loss": 1.295,
+      "step": 5483
+    },
+    {
+      "epoch": 0.39530022345563326,
+      "grad_norm": 0.1011954129867873,
+      "learning_rate": 0.00013781491604032152,
+      "loss": 1.2069,
+      "step": 5484
+    },
+    {
+      "epoch": 0.3953723059179702,
+      "grad_norm": 0.136725372217431,
+      "learning_rate": 0.00013779330150895825,
+      "loss": 1.3993,
+      "step": 5485
+    },
+    {
+      "epoch": 0.39544438838030704,
+      "grad_norm": 0.297105645583597,
+      "learning_rate": 0.00013777168491752194,
+      "loss": 1.2294,
+      "step": 5486
+    },
+    {
+      "epoch": 0.39551647084264396,
+      "grad_norm": 0.11926639108078074,
+      "learning_rate": 0.00013775006626719086,
+      "loss": 1.2988,
+      "step": 5487
+    },
+    {
+      "epoch": 0.3955885533049809,
+      "grad_norm": 0.1970299423048667,
+      "learning_rate": 0.00013772844555914347,
+      "loss": 1.3518,
+      "step": 5488
+    },
+    {
+      "epoch": 0.3956606357673178,
+      "grad_norm": 0.110056809818147,
+      "learning_rate": 0.00013770682279455823,
+      "loss": 1.3639,
+      "step": 5489
+    },
+    {
+      "epoch": 0.3957327182296547,
+      "grad_norm": 0.11634239650058427,
+      "learning_rate": 0.00013768519797461386,
+      "loss": 1.4205,
+      "step": 5490
+    },
+    {
+      "epoch": 0.39580480069199164,
+      "grad_norm": 0.10609302878163145,
+      "learning_rate": 0.00013766357110048903,
+      "loss": 1.2613,
+      "step": 5491
+    },
+    {
+      "epoch": 0.39587688315432856,
+      "grad_norm": 0.11380403885822976,
+      "learning_rate": 0.00013764194217336264,
+      "loss": 1.3665,
+      "step": 5492
+    },
+    {
+      "epoch": 0.3959489656166655,
+      "grad_norm": 0.10568793370043704,
+      "learning_rate": 0.00013762031119441366,
+      "loss": 1.212,
+      "step": 5493
+    },
+    {
+      "epoch": 0.3960210480790024,
+      "grad_norm": 0.1175641228069479,
+      "learning_rate": 0.00013759867816482113,
+      "loss": 1.3901,
+      "step": 5494
+    },
+    {
+      "epoch": 0.39609313054133927,
+      "grad_norm": 0.10989694804383882,
+      "learning_rate": 0.00013757704308576426,
+      "loss": 1.2474,
+      "step": 5495
+    },
+    {
+      "epoch": 0.3961652130036762,
+      "grad_norm": 0.17805170494006062,
+      "learning_rate": 0.0001375554059584224,
+      "loss": 1.2862,
+      "step": 5496
+    },
+    {
+      "epoch": 0.3962372954660131,
+      "grad_norm": 0.19933580670858844,
+      "learning_rate": 0.00013753376678397495,
+      "loss": 1.3607,
+      "step": 5497
+    },
+    {
+      "epoch": 0.39630937792835,
+      "grad_norm": 0.11127135712723449,
+      "learning_rate": 0.00013751212556360138,
+      "loss": 1.2321,
+      "step": 5498
+    },
+    {
+      "epoch": 0.39638146039068695,
+      "grad_norm": 0.11564209515148753,
+      "learning_rate": 0.0001374904822984814,
+      "loss": 1.2384,
+      "step": 5499
+    },
+    {
+      "epoch": 0.39645354285302387,
+      "grad_norm": 0.11432058485762908,
+      "learning_rate": 0.00013746883698979474,
+      "loss": 1.3369,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3965256253153608,
+      "grad_norm": 0.15616130663507083,
+      "learning_rate": 0.00013744718963872124,
+      "loss": 1.2809,
+      "step": 5501
+    },
+    {
+      "epoch": 0.3965977077776977,
+      "grad_norm": 0.11409654360267928,
+      "learning_rate": 0.0001374255402464409,
+      "loss": 1.33,
+      "step": 5502
+    },
+    {
+      "epoch": 0.3966697902400346,
+      "grad_norm": 0.11267687391813676,
+      "learning_rate": 0.0001374038888141338,
+      "loss": 1.4279,
+      "step": 5503
+    },
+    {
+      "epoch": 0.3967418727023715,
+      "grad_norm": 0.11657103272841067,
+      "learning_rate": 0.00013738223534298012,
+      "loss": 1.3616,
+      "step": 5504
+    },
+    {
+      "epoch": 0.3968139551647084,
+      "grad_norm": 0.10037980403118583,
+      "learning_rate": 0.00013736057983416023,
+      "loss": 1.1624,
+      "step": 5505
+    },
+    {
+      "epoch": 0.39688603762704533,
+      "grad_norm": 0.1235788263563538,
+      "learning_rate": 0.00013733892228885442,
+      "loss": 1.3745,
+      "step": 5506
+    },
+    {
+      "epoch": 0.39695812008938225,
+      "grad_norm": 0.11435523679529312,
+      "learning_rate": 0.0001373172627082434,
+      "loss": 1.3046,
+      "step": 5507
+    },
+    {
+      "epoch": 0.39703020255171917,
+      "grad_norm": 0.10241897433430368,
+      "learning_rate": 0.00013729560109350763,
+      "loss": 1.3455,
+      "step": 5508
+    },
+    {
+      "epoch": 0.3971022850140561,
+      "grad_norm": 0.1470871790444726,
+      "learning_rate": 0.00013727393744582798,
+      "loss": 1.2858,
+      "step": 5509
+    },
+    {
+      "epoch": 0.397174367476393,
+      "grad_norm": 0.10729698599996537,
+      "learning_rate": 0.00013725227176638528,
+      "loss": 1.1938,
+      "step": 5510
+    },
+    {
+      "epoch": 0.39724644993872993,
+      "grad_norm": 0.1118236851256817,
+      "learning_rate": 0.0001372306040563605,
+      "loss": 1.3873,
+      "step": 5511
+    },
+    {
+      "epoch": 0.3973185324010668,
+      "grad_norm": 0.11778548744819685,
+      "learning_rate": 0.0001372089343169347,
+      "loss": 1.3435,
+      "step": 5512
+    },
+    {
+      "epoch": 0.3973906148634037,
+      "grad_norm": 0.11286436494330446,
+      "learning_rate": 0.00013718726254928913,
+      "loss": 1.2419,
+      "step": 5513
+    },
+    {
+      "epoch": 0.39746269732574063,
+      "grad_norm": 0.11446407393860643,
+      "learning_rate": 0.00013716558875460507,
+      "loss": 1.3147,
+      "step": 5514
+    },
+    {
+      "epoch": 0.39753477978807755,
+      "grad_norm": 0.11480365642525839,
+      "learning_rate": 0.00013714391293406387,
+      "loss": 1.3379,
+      "step": 5515
+    },
+    {
+      "epoch": 0.3976068622504145,
+      "grad_norm": 0.10797824323249373,
+      "learning_rate": 0.00013712223508884717,
+      "loss": 1.32,
+      "step": 5516
+    },
+    {
+      "epoch": 0.3976789447127514,
+      "grad_norm": 0.10873568211777417,
+      "learning_rate": 0.00013710055522013655,
+      "loss": 1.3061,
+      "step": 5517
+    },
+    {
+      "epoch": 0.3977510271750883,
+      "grad_norm": 0.10932042302880042,
+      "learning_rate": 0.00013707887332911373,
+      "loss": 1.2247,
+      "step": 5518
+    },
+    {
+      "epoch": 0.39782310963742523,
+      "grad_norm": 0.11520130465255846,
+      "learning_rate": 0.00013705718941696063,
+      "loss": 1.2812,
+      "step": 5519
+    },
+    {
+      "epoch": 0.39789519209976215,
+      "grad_norm": 0.1167610112865233,
+      "learning_rate": 0.00013703550348485917,
+      "loss": 1.2097,
+      "step": 5520
+    },
+    {
+      "epoch": 0.397967274562099,
+      "grad_norm": 0.1106933075702686,
+      "learning_rate": 0.00013701381553399145,
+      "loss": 1.2798,
+      "step": 5521
+    },
+    {
+      "epoch": 0.39803935702443594,
+      "grad_norm": 0.10755384995162248,
+      "learning_rate": 0.0001369921255655396,
+      "loss": 1.3262,
+      "step": 5522
+    },
+    {
+      "epoch": 0.39811143948677286,
+      "grad_norm": 0.12765841729223426,
+      "learning_rate": 0.00013697043358068607,
+      "loss": 1.2447,
+      "step": 5523
+    },
+    {
+      "epoch": 0.3981835219491098,
+      "grad_norm": 0.23793128263236887,
+      "learning_rate": 0.00013694873958061305,
+      "loss": 1.4085,
+      "step": 5524
+    },
+    {
+      "epoch": 0.3982556044114467,
+      "grad_norm": 0.12924026566954977,
+      "learning_rate": 0.00013692704356650325,
+      "loss": 1.2564,
+      "step": 5525
+    },
+    {
+      "epoch": 0.3983276868737836,
+      "grad_norm": 0.10121626551716166,
+      "learning_rate": 0.00013690534553953917,
+      "loss": 1.2131,
+      "step": 5526
+    },
+    {
+      "epoch": 0.39839976933612054,
+      "grad_norm": 0.11466727097162241,
+      "learning_rate": 0.00013688364550090362,
+      "loss": 1.2312,
+      "step": 5527
+    },
+    {
+      "epoch": 0.39847185179845745,
+      "grad_norm": 0.12806793520132148,
+      "learning_rate": 0.00013686194345177938,
+      "loss": 1.2916,
+      "step": 5528
+    },
+    {
+      "epoch": 0.3985439342607944,
+      "grad_norm": 0.11287731855105514,
+      "learning_rate": 0.0001368402393933495,
+      "loss": 1.3609,
+      "step": 5529
+    },
+    {
+      "epoch": 0.39861601672313124,
+      "grad_norm": 0.13685071409658214,
+      "learning_rate": 0.00013681853332679698,
+      "loss": 1.2229,
+      "step": 5530
+    },
+    {
+      "epoch": 0.39868809918546816,
+      "grad_norm": 0.12130261168754476,
+      "learning_rate": 0.00013679682525330497,
+      "loss": 1.2989,
+      "step": 5531
+    },
+    {
+      "epoch": 0.3987601816478051,
+      "grad_norm": 0.11008967021556945,
+      "learning_rate": 0.00013677511517405682,
+      "loss": 1.2201,
+      "step": 5532
+    },
+    {
+      "epoch": 0.398832264110142,
+      "grad_norm": 0.11203973971462053,
+      "learning_rate": 0.00013675340309023587,
+      "loss": 1.2813,
+      "step": 5533
+    },
+    {
+      "epoch": 0.3989043465724789,
+      "grad_norm": 0.12328966734359534,
+      "learning_rate": 0.00013673168900302566,
+      "loss": 1.3684,
+      "step": 5534
+    },
+    {
+      "epoch": 0.39897642903481584,
+      "grad_norm": 0.10988010054675526,
+      "learning_rate": 0.00013670997291360978,
+      "loss": 1.3459,
+      "step": 5535
+    },
+    {
+      "epoch": 0.39904851149715276,
+      "grad_norm": 0.29303910818985207,
+      "learning_rate": 0.00013668825482317197,
+      "loss": 1.2746,
+      "step": 5536
+    },
+    {
+      "epoch": 0.3991205939594897,
+      "grad_norm": 0.11203854570632357,
+      "learning_rate": 0.00013666653473289606,
+      "loss": 1.2233,
+      "step": 5537
+    },
+    {
+      "epoch": 0.39919267642182654,
+      "grad_norm": 0.10841189499731683,
+      "learning_rate": 0.00013664481264396596,
+      "loss": 1.2835,
+      "step": 5538
+    },
+    {
+      "epoch": 0.39926475888416346,
+      "grad_norm": 0.1323748756497132,
+      "learning_rate": 0.00013662308855756574,
+      "loss": 1.2695,
+      "step": 5539
+    },
+    {
+      "epoch": 0.3993368413465004,
+      "grad_norm": 0.11674674646603467,
+      "learning_rate": 0.00013660136247487957,
+      "loss": 1.3368,
+      "step": 5540
+    },
+    {
+      "epoch": 0.3994089238088373,
+      "grad_norm": 0.13090562400486513,
+      "learning_rate": 0.00013657963439709167,
+      "loss": 1.3654,
+      "step": 5541
+    },
+    {
+      "epoch": 0.3994810062711742,
+      "grad_norm": 0.10767689643483792,
+      "learning_rate": 0.00013655790432538645,
+      "loss": 1.1367,
+      "step": 5542
+    },
+    {
+      "epoch": 0.39955308873351114,
+      "grad_norm": 0.1076704506354015,
+      "learning_rate": 0.0001365361722609484,
+      "loss": 1.2212,
+      "step": 5543
+    },
+    {
+      "epoch": 0.39962517119584806,
+      "grad_norm": 0.11020677281019822,
+      "learning_rate": 0.0001365144382049621,
+      "loss": 1.3208,
+      "step": 5544
+    },
+    {
+      "epoch": 0.399697253658185,
+      "grad_norm": 0.12771249139383345,
+      "learning_rate": 0.00013649270215861225,
+      "loss": 1.2609,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3997693361205219,
+      "grad_norm": 0.16463835811668084,
+      "learning_rate": 0.00013647096412308367,
+      "loss": 1.3264,
+      "step": 5546
+    },
+    {
+      "epoch": 0.39984141858285877,
+      "grad_norm": 0.12004412515060531,
+      "learning_rate": 0.00013644922409956125,
+      "loss": 1.3654,
+      "step": 5547
+    },
+    {
+      "epoch": 0.3999135010451957,
+      "grad_norm": 0.14195764725665933,
+      "learning_rate": 0.00013642748208923005,
+      "loss": 1.2352,
+      "step": 5548
+    },
+    {
+      "epoch": 0.3999855835075326,
+      "grad_norm": 0.12374724074939047,
+      "learning_rate": 0.00013640573809327514,
+      "loss": 1.3216,
+      "step": 5549
+    },
+    {
+      "epoch": 0.4000576659698695,
+      "grad_norm": 0.12634228711050674,
+      "learning_rate": 0.00013638399211288188,
+      "loss": 1.2916,
+      "step": 5550
+    },
+    {
+      "epoch": 0.40012974843220644,
+      "grad_norm": 0.11301143613567162,
+      "learning_rate": 0.0001363622441492355,
+      "loss": 1.2974,
+      "step": 5551
+    },
+    {
+      "epoch": 0.40020183089454336,
+      "grad_norm": 0.10110369243958067,
+      "learning_rate": 0.00013634049420352152,
+      "loss": 1.1735,
+      "step": 5552
+    },
+    {
+      "epoch": 0.4002739133568803,
+      "grad_norm": 0.10703040383903527,
+      "learning_rate": 0.0001363187422769255,
+      "loss": 1.2661,
+      "step": 5553
+    },
+    {
+      "epoch": 0.4003459958192172,
+      "grad_norm": 0.13876165666024348,
+      "learning_rate": 0.0001362969883706331,
+      "loss": 1.1768,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4004180782815541,
+      "grad_norm": 0.13155941525446954,
+      "learning_rate": 0.0001362752324858301,
+      "loss": 1.3161,
+      "step": 5555
+    },
+    {
+      "epoch": 0.400490160743891,
+      "grad_norm": 0.1125473997792638,
+      "learning_rate": 0.00013625347462370243,
+      "loss": 1.3543,
+      "step": 5556
+    },
+    {
+      "epoch": 0.4005622432062279,
+      "grad_norm": 0.16649321641040263,
+      "learning_rate": 0.00013623171478543603,
+      "loss": 1.2685,
+      "step": 5557
+    },
+    {
+      "epoch": 0.40063432566856483,
+      "grad_norm": 0.10071773224164256,
+      "learning_rate": 0.00013620995297221707,
+      "loss": 1.2527,
+      "step": 5558
+    },
+    {
+      "epoch": 0.40070640813090175,
+      "grad_norm": 0.10910545982348226,
+      "learning_rate": 0.0001361881891852317,
+      "loss": 1.3194,
+      "step": 5559
+    },
+    {
+      "epoch": 0.40077849059323867,
+      "grad_norm": 0.12485121902304697,
+      "learning_rate": 0.0001361664234256663,
+      "loss": 1.3329,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4008505730555756,
+      "grad_norm": 0.1719453229521704,
+      "learning_rate": 0.00013614465569470723,
+      "loss": 1.2737,
+      "step": 5561
+    },
+    {
+      "epoch": 0.4009226555179125,
+      "grad_norm": 0.10782389446783985,
+      "learning_rate": 0.00013612288599354112,
+      "loss": 1.2954,
+      "step": 5562
+    },
+    {
+      "epoch": 0.4009947379802494,
+      "grad_norm": 0.21253293931544345,
+      "learning_rate": 0.00013610111432335454,
+      "loss": 1.2749,
+      "step": 5563
+    },
+    {
+      "epoch": 0.4010668204425863,
+      "grad_norm": 0.11198609674170555,
+      "learning_rate": 0.00013607934068533424,
+      "loss": 1.3884,
+      "step": 5564
+    },
+    {
+      "epoch": 0.4011389029049232,
+      "grad_norm": 0.11807810012494463,
+      "learning_rate": 0.0001360575650806671,
+      "loss": 1.3203,
+      "step": 5565
+    },
+    {
+      "epoch": 0.40121098536726013,
+      "grad_norm": 0.29198534042857677,
+      "learning_rate": 0.0001360357875105401,
+      "loss": 1.3263,
+      "step": 5566
+    },
+    {
+      "epoch": 0.40128306782959705,
+      "grad_norm": 0.11545063266203291,
+      "learning_rate": 0.00013601400797614029,
+      "loss": 1.3119,
+      "step": 5567
+    },
+    {
+      "epoch": 0.40135515029193397,
+      "grad_norm": 0.1090767078824182,
+      "learning_rate": 0.00013599222647865485,
+      "loss": 1.3036,
+      "step": 5568
+    },
+    {
+      "epoch": 0.4014272327542709,
+      "grad_norm": 0.11206773655942164,
+      "learning_rate": 0.00013597044301927105,
+      "loss": 1.2448,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4014993152166078,
+      "grad_norm": 0.1710093014863151,
+      "learning_rate": 0.00013594865759917633,
+      "loss": 1.269,
+      "step": 5570
+    },
+    {
+      "epoch": 0.40157139767894473,
+      "grad_norm": 0.12606377934396937,
+      "learning_rate": 0.00013592687021955814,
+      "loss": 1.3015,
+      "step": 5571
+    },
+    {
+      "epoch": 0.40164348014128165,
+      "grad_norm": 0.11888464939005092,
+      "learning_rate": 0.00013590508088160414,
+      "loss": 1.4291,
+      "step": 5572
+    },
+    {
+      "epoch": 0.4017155626036185,
+      "grad_norm": 0.1257694541898121,
+      "learning_rate": 0.000135883289586502,
+      "loss": 1.2868,
+      "step": 5573
+    },
+    {
+      "epoch": 0.40178764506595543,
+      "grad_norm": 0.11259628548018495,
+      "learning_rate": 0.00013586149633543958,
+      "loss": 1.3312,
+      "step": 5574
+    },
+    {
+      "epoch": 0.40185972752829235,
+      "grad_norm": 0.11812551735489289,
+      "learning_rate": 0.00013583970112960474,
+      "loss": 1.2458,
+      "step": 5575
+    },
+    {
+      "epoch": 0.4019318099906293,
+      "grad_norm": 0.09924089596336964,
+      "learning_rate": 0.00013581790397018562,
+      "loss": 1.3416,
+      "step": 5576
+    },
+    {
+      "epoch": 0.4020038924529662,
+      "grad_norm": 0.1148319110389973,
+      "learning_rate": 0.00013579610485837024,
+      "loss": 1.3394,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4020759749153031,
+      "grad_norm": 0.13926201956501189,
+      "learning_rate": 0.00013577430379534694,
+      "loss": 1.3496,
+      "step": 5578
+    },
+    {
+      "epoch": 0.40214805737764003,
+      "grad_norm": 0.11115355211398054,
+      "learning_rate": 0.000135752500782304,
+      "loss": 1.2364,
+      "step": 5579
+    },
+    {
+      "epoch": 0.40222013983997695,
+      "grad_norm": 0.11737265754583494,
+      "learning_rate": 0.00013573069582042998,
+      "loss": 1.3648,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4022922223023139,
+      "grad_norm": 0.09955174504556479,
+      "learning_rate": 0.0001357088889109133,
+      "loss": 1.3271,
+      "step": 5581
+    },
+    {
+      "epoch": 0.40236430476465074,
+      "grad_norm": 0.11784675683211075,
+      "learning_rate": 0.00013568708005494277,
+      "loss": 1.2795,
+      "step": 5582
+    },
+    {
+      "epoch": 0.40243638722698766,
+      "grad_norm": 0.11536533209253418,
+      "learning_rate": 0.0001356652692537071,
+      "loss": 1.2675,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4025084696893246,
+      "grad_norm": 0.17632250809736078,
+      "learning_rate": 0.00013564345650839517,
+      "loss": 1.3678,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4025805521516615,
+      "grad_norm": 0.12237999304798443,
+      "learning_rate": 0.000135621641820196,
+      "loss": 1.2349,
+      "step": 5585
+    },
+    {
+      "epoch": 0.4026526346139984,
+      "grad_norm": 0.12144936547899399,
+      "learning_rate": 0.0001355998251902987,
+      "loss": 1.2585,
+      "step": 5586
+    },
+    {
+      "epoch": 0.40272471707633534,
+      "grad_norm": 0.1121102893413987,
+      "learning_rate": 0.0001355780066198924,
+      "loss": 1.3615,
+      "step": 5587
+    },
+    {
+      "epoch": 0.40279679953867226,
+      "grad_norm": 0.11360907472592412,
+      "learning_rate": 0.00013555618611016645,
+      "loss": 1.2814,
+      "step": 5588
+    },
+    {
+      "epoch": 0.4028688820010092,
+      "grad_norm": 0.10449553887499209,
+      "learning_rate": 0.00013553436366231026,
+      "loss": 1.2476,
+      "step": 5589
+    },
+    {
+      "epoch": 0.40294096446334604,
+      "grad_norm": 0.10396658216927583,
+      "learning_rate": 0.00013551253927751334,
+      "loss": 1.3859,
+      "step": 5590
+    },
+    {
+      "epoch": 0.40301304692568296,
+      "grad_norm": 0.11026790815206912,
+      "learning_rate": 0.00013549071295696535,
+      "loss": 1.3376,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4030851293880199,
+      "grad_norm": 0.10830667058501514,
+      "learning_rate": 0.00013546888470185602,
+      "loss": 1.2353,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4031572118503568,
+      "grad_norm": 0.12051846442125998,
+      "learning_rate": 0.00013544705451337512,
+      "loss": 1.3139,
+      "step": 5593
+    },
+    {
+      "epoch": 0.4032292943126937,
+      "grad_norm": 0.12125998846565958,
+      "learning_rate": 0.00013542522239271264,
+      "loss": 1.2901,
+      "step": 5594
+    },
+    {
+      "epoch": 0.40330137677503064,
+      "grad_norm": 0.11747380964653438,
+      "learning_rate": 0.00013540338834105861,
+      "loss": 1.1993,
+      "step": 5595
+    },
+    {
+      "epoch": 0.40337345923736756,
+      "grad_norm": 0.12141405853023136,
+      "learning_rate": 0.00013538155235960322,
+      "loss": 1.3185,
+      "step": 5596
+    },
+    {
+      "epoch": 0.4034455416997045,
+      "grad_norm": 0.11788698335254753,
+      "learning_rate": 0.00013535971444953666,
+      "loss": 1.3455,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4035176241620414,
+      "grad_norm": 0.18372469299470062,
+      "learning_rate": 0.00013533787461204934,
+      "loss": 1.3094,
+      "step": 5598
+    },
+    {
+      "epoch": 0.40358970662437826,
+      "grad_norm": 0.12367761552919168,
+      "learning_rate": 0.0001353160328483317,
+      "loss": 1.2706,
+      "step": 5599
+    },
+    {
+      "epoch": 0.4036617890867152,
+      "grad_norm": 0.1818730545349984,
+      "learning_rate": 0.00013529418915957434,
+      "loss": 1.2283,
+      "step": 5600
+    },
+    {
+      "epoch": 0.4037338715490521,
+      "grad_norm": 0.10188821890100097,
+      "learning_rate": 0.00013527234354696787,
+      "loss": 1.3112,
+      "step": 5601
+    },
+    {
+      "epoch": 0.403805954011389,
+      "grad_norm": 0.1223126525193079,
+      "learning_rate": 0.00013525049601170318,
+      "loss": 1.3283,
+      "step": 5602
+    },
+    {
+      "epoch": 0.40387803647372594,
+      "grad_norm": 0.11824172600632901,
+      "learning_rate": 0.00013522864655497106,
+      "loss": 1.1784,
+      "step": 5603
+    },
+    {
+      "epoch": 0.40395011893606286,
+      "grad_norm": 0.10646972722079558,
+      "learning_rate": 0.00013520679517796259,
+      "loss": 1.2863,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4040222013983998,
+      "grad_norm": 0.10888811616019703,
+      "learning_rate": 0.00013518494188186878,
+      "loss": 1.2122,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4040942838607367,
+      "grad_norm": 0.12326298611467967,
+      "learning_rate": 0.00013516308666788087,
+      "loss": 1.277,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4041663663230736,
+      "grad_norm": 0.11057954539915418,
+      "learning_rate": 0.00013514122953719014,
+      "loss": 1.2405,
+      "step": 5607
+    },
+    {
+      "epoch": 0.4042384487854105,
+      "grad_norm": 0.10211080337820462,
+      "learning_rate": 0.00013511937049098805,
+      "loss": 1.2656,
+      "step": 5608
+    },
+    {
+      "epoch": 0.4043105312477474,
+      "grad_norm": 0.11472035509402569,
+      "learning_rate": 0.00013509750953046606,
+      "loss": 1.2803,
+      "step": 5609
+    },
+    {
+      "epoch": 0.4043826137100843,
+      "grad_norm": 0.11119979209183564,
+      "learning_rate": 0.0001350756466568158,
+      "loss": 1.3752,
+      "step": 5610
+    },
+    {
+      "epoch": 0.40445469617242125,
+      "grad_norm": 0.11597330810443784,
+      "learning_rate": 0.00013505378187122903,
+      "loss": 1.1532,
+      "step": 5611
+    },
+    {
+      "epoch": 0.40452677863475817,
+      "grad_norm": 0.11482276511561092,
+      "learning_rate": 0.0001350319151748975,
+      "loss": 1.3211,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4045988610970951,
+      "grad_norm": 0.10876694460263854,
+      "learning_rate": 0.00013501004656901324,
+      "loss": 1.4042,
+      "step": 5613
+    },
+    {
+      "epoch": 0.404670943559432,
+      "grad_norm": 0.11651501837807299,
+      "learning_rate": 0.0001349881760547682,
+      "loss": 1.3177,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4047430260217689,
+      "grad_norm": 0.1362969642043395,
+      "learning_rate": 0.00013496630363335455,
+      "loss": 1.2676,
+      "step": 5615
+    },
+    {
+      "epoch": 0.4048151084841058,
+      "grad_norm": 0.11865153512960919,
+      "learning_rate": 0.00013494442930596457,
+      "loss": 1.3378,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4048871909464427,
+      "grad_norm": 0.10405031841906817,
+      "learning_rate": 0.00013492255307379056,
+      "loss": 1.3015,
+      "step": 5617
+    },
+    {
+      "epoch": 0.40495927340877963,
+      "grad_norm": 0.1104833930350516,
+      "learning_rate": 0.00013490067493802496,
+      "loss": 1.3709,
+      "step": 5618
+    },
+    {
+      "epoch": 0.40503135587111655,
+      "grad_norm": 0.10815700953378588,
+      "learning_rate": 0.00013487879489986035,
+      "loss": 1.2521,
+      "step": 5619
+    },
+    {
+      "epoch": 0.40510343833345347,
+      "grad_norm": 0.10826025503606486,
+      "learning_rate": 0.0001348569129604894,
+      "loss": 1.2576,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4051755207957904,
+      "grad_norm": 0.11989428780822302,
+      "learning_rate": 0.00013483502912110484,
+      "loss": 1.4309,
+      "step": 5621
+    },
+    {
+      "epoch": 0.4052476032581273,
+      "grad_norm": 0.10922040241260035,
+      "learning_rate": 0.00013481314338289956,
+      "loss": 1.315,
+      "step": 5622
+    },
+    {
+      "epoch": 0.40531968572046423,
+      "grad_norm": 0.12157881928820742,
+      "learning_rate": 0.00013479125574706653,
+      "loss": 1.2953,
+      "step": 5623
+    },
+    {
+      "epoch": 0.40539176818280115,
+      "grad_norm": 0.12135499221328683,
+      "learning_rate": 0.00013476936621479877,
+      "loss": 1.2538,
+      "step": 5624
+    },
+    {
+      "epoch": 0.405463850645138,
+      "grad_norm": 0.1106550208657966,
+      "learning_rate": 0.00013474747478728954,
+      "loss": 1.2838,
+      "step": 5625
+    },
+    {
+      "epoch": 0.40553593310747493,
+      "grad_norm": 0.11071634140318996,
+      "learning_rate": 0.00013472558146573204,
+      "loss": 1.3583,
+      "step": 5626
+    },
+    {
+      "epoch": 0.40560801556981185,
+      "grad_norm": 0.12036026767937114,
+      "learning_rate": 0.0001347036862513197,
+      "loss": 1.3414,
+      "step": 5627
+    },
+    {
+      "epoch": 0.4056800980321488,
+      "grad_norm": 0.11440209782902566,
+      "learning_rate": 0.000134681789145246,
+      "loss": 1.2497,
+      "step": 5628
+    },
+    {
+      "epoch": 0.4057521804944857,
+      "grad_norm": 0.11159188331428939,
+      "learning_rate": 0.00013465989014870453,
+      "loss": 1.3705,
+      "step": 5629
+    },
+    {
+      "epoch": 0.4058242629568226,
+      "grad_norm": 0.11076666989409313,
+      "learning_rate": 0.00013463798926288892,
+      "loss": 1.2357,
+      "step": 5630
+    },
+    {
+      "epoch": 0.40589634541915953,
+      "grad_norm": 0.1058145681469488,
+      "learning_rate": 0.00013461608648899308,
+      "loss": 1.2637,
+      "step": 5631
+    },
+    {
+      "epoch": 0.40596842788149645,
+      "grad_norm": 0.12240991233457114,
+      "learning_rate": 0.00013459418182821078,
+      "loss": 1.2999,
+      "step": 5632
+    },
+    {
+      "epoch": 0.40604051034383337,
+      "grad_norm": 0.10800784172873974,
+      "learning_rate": 0.00013457227528173614,
+      "loss": 1.3539,
+      "step": 5633
+    },
+    {
+      "epoch": 0.40611259280617024,
+      "grad_norm": 0.10926536588758747,
+      "learning_rate": 0.00013455036685076314,
+      "loss": 1.2892,
+      "step": 5634
+    },
+    {
+      "epoch": 0.40618467526850716,
+      "grad_norm": 0.11034283003201598,
+      "learning_rate": 0.0001345284565364861,
+      "loss": 1.2321,
+      "step": 5635
+    },
+    {
+      "epoch": 0.4062567577308441,
+      "grad_norm": 0.11119439124785334,
+      "learning_rate": 0.00013450654434009923,
+      "loss": 1.27,
+      "step": 5636
+    },
+    {
+      "epoch": 0.406328840193181,
+      "grad_norm": 0.10026141840644451,
+      "learning_rate": 0.00013448463026279704,
+      "loss": 1.28,
+      "step": 5637
+    },
+    {
+      "epoch": 0.4064009226555179,
+      "grad_norm": 0.11694714604156929,
+      "learning_rate": 0.000134462714305774,
+      "loss": 1.4025,
+      "step": 5638
+    },
+    {
+      "epoch": 0.40647300511785484,
+      "grad_norm": 0.11166116125445463,
+      "learning_rate": 0.0001344407964702247,
+      "loss": 1.2963,
+      "step": 5639
+    },
+    {
+      "epoch": 0.40654508758019176,
+      "grad_norm": 0.1173519930116992,
+      "learning_rate": 0.00013441887675734386,
+      "loss": 1.3336,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4066171700425287,
+      "grad_norm": 0.1328791754744317,
+      "learning_rate": 0.0001343969551683263,
+      "loss": 1.3217,
+      "step": 5641
+    },
+    {
+      "epoch": 0.40668925250486554,
+      "grad_norm": 0.11326321398495037,
+      "learning_rate": 0.000134375031704367,
+      "loss": 1.257,
+      "step": 5642
+    },
+    {
+      "epoch": 0.40676133496720246,
+      "grad_norm": 0.11697605556704445,
+      "learning_rate": 0.00013435310636666093,
+      "loss": 1.3258,
+      "step": 5643
+    },
+    {
+      "epoch": 0.4068334174295394,
+      "grad_norm": 0.11295713594685211,
+      "learning_rate": 0.00013433117915640324,
+      "loss": 1.1927,
+      "step": 5644
+    },
+    {
+      "epoch": 0.4069054998918763,
+      "grad_norm": 0.11079031426340885,
+      "learning_rate": 0.00013430925007478913,
+      "loss": 1.325,
+      "step": 5645
+    },
+    {
+      "epoch": 0.4069775823542132,
+      "grad_norm": 0.11199563856862867,
+      "learning_rate": 0.00013428731912301398,
+      "loss": 1.273,
+      "step": 5646
+    },
+    {
+      "epoch": 0.40704966481655014,
+      "grad_norm": 0.11422252462284639,
+      "learning_rate": 0.00013426538630227316,
+      "loss": 1.3281,
+      "step": 5647
+    },
+    {
+      "epoch": 0.40712174727888706,
+      "grad_norm": 0.1191473271781177,
+      "learning_rate": 0.00013424345161376226,
+      "loss": 1.323,
+      "step": 5648
+    },
+    {
+      "epoch": 0.407193829741224,
+      "grad_norm": 0.10545517943749212,
+      "learning_rate": 0.00013422151505867693,
+      "loss": 1.2772,
+      "step": 5649
+    },
+    {
+      "epoch": 0.4072659122035609,
+      "grad_norm": 0.1130918822508231,
+      "learning_rate": 0.0001341995766382128,
+      "loss": 1.2602,
+      "step": 5650
+    },
+    {
+      "epoch": 0.40733799466589776,
+      "grad_norm": 0.14605516670376872,
+      "learning_rate": 0.00013417763635356584,
+      "loss": 1.3036,
+      "step": 5651
+    },
+    {
+      "epoch": 0.4074100771282347,
+      "grad_norm": 0.12068387465537835,
+      "learning_rate": 0.0001341556942059319,
+      "loss": 1.275,
+      "step": 5652
+    },
+    {
+      "epoch": 0.4074821595905716,
+      "grad_norm": 0.14409085284753884,
+      "learning_rate": 0.0001341337501965071,
+      "loss": 1.3639,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4075542420529085,
+      "grad_norm": 0.12182603398832262,
+      "learning_rate": 0.0001341118043264875,
+      "loss": 1.2671,
+      "step": 5654
+    },
+    {
+      "epoch": 0.40762632451524544,
+      "grad_norm": 0.10774408253915153,
+      "learning_rate": 0.00013408985659706945,
+      "loss": 1.2876,
+      "step": 5655
+    },
+    {
+      "epoch": 0.40769840697758236,
+      "grad_norm": 0.1098578559178572,
+      "learning_rate": 0.00013406790700944918,
+      "loss": 1.2234,
+      "step": 5656
+    },
+    {
+      "epoch": 0.4077704894399193,
+      "grad_norm": 0.11641281093631377,
+      "learning_rate": 0.00013404595556482324,
+      "loss": 1.3151,
+      "step": 5657
+    },
+    {
+      "epoch": 0.4078425719022562,
+      "grad_norm": 0.11863084778267023,
+      "learning_rate": 0.00013402400226438812,
+      "loss": 1.3774,
+      "step": 5658
+    },
+    {
+      "epoch": 0.4079146543645931,
+      "grad_norm": 0.10836175414377132,
+      "learning_rate": 0.0001340020471093405,
+      "loss": 1.2642,
+      "step": 5659
+    },
+    {
+      "epoch": 0.40798673682693,
+      "grad_norm": 0.11366944275727957,
+      "learning_rate": 0.0001339800901008771,
+      "loss": 1.2245,
+      "step": 5660
+    },
+    {
+      "epoch": 0.4080588192892669,
+      "grad_norm": 0.1119811668311904,
+      "learning_rate": 0.00013395813124019484,
+      "loss": 1.3199,
+      "step": 5661
+    },
+    {
+      "epoch": 0.4081309017516038,
+      "grad_norm": 0.12161344581465437,
+      "learning_rate": 0.00013393617052849058,
+      "loss": 1.2883,
+      "step": 5662
+    },
+    {
+      "epoch": 0.40820298421394074,
+      "grad_norm": 0.11251481817195386,
+      "learning_rate": 0.00013391420796696145,
+      "loss": 1.3367,
+      "step": 5663
+    },
+    {
+      "epoch": 0.40827506667627766,
+      "grad_norm": 0.17109562378113527,
+      "learning_rate": 0.00013389224355680454,
+      "loss": 1.3185,
+      "step": 5664
+    },
+    {
+      "epoch": 0.4083471491386146,
+      "grad_norm": 0.12105820393726571,
+      "learning_rate": 0.0001338702772992172,
+      "loss": 1.3329,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4084192316009515,
+      "grad_norm": 0.10911892666991314,
+      "learning_rate": 0.0001338483091953967,
+      "loss": 1.3128,
+      "step": 5666
+    },
+    {
+      "epoch": 0.4084913140632884,
+      "grad_norm": 0.11338087858215219,
+      "learning_rate": 0.00013382633924654056,
+      "loss": 1.3104,
+      "step": 5667
+    },
+    {
+      "epoch": 0.4085633965256253,
+      "grad_norm": 0.15721274141185207,
+      "learning_rate": 0.00013380436745384628,
+      "loss": 1.3182,
+      "step": 5668
+    },
+    {
+      "epoch": 0.4086354789879622,
+      "grad_norm": 0.10976433822855448,
+      "learning_rate": 0.0001337823938185116,
+      "loss": 1.2684,
+      "step": 5669
+    },
+    {
+      "epoch": 0.40870756145029913,
+      "grad_norm": 0.11195660235400774,
+      "learning_rate": 0.00013376041834173419,
+      "loss": 1.2841,
+      "step": 5670
+    },
+    {
+      "epoch": 0.40877964391263605,
+      "grad_norm": 0.10476831890548276,
+      "learning_rate": 0.00013373844102471198,
+      "loss": 1.317,
+      "step": 5671
+    },
+    {
+      "epoch": 0.40885172637497297,
+      "grad_norm": 0.108692580290917,
+      "learning_rate": 0.00013371646186864286,
+      "loss": 1.2695,
+      "step": 5672
+    },
+    {
+      "epoch": 0.4089238088373099,
+      "grad_norm": 0.11556957809720257,
+      "learning_rate": 0.00013369448087472498,
+      "loss": 1.191,
+      "step": 5673
+    },
+    {
+      "epoch": 0.4089958912996468,
+      "grad_norm": 0.11980868056000148,
+      "learning_rate": 0.0001336724980441564,
+      "loss": 1.2781,
+      "step": 5674
+    },
+    {
+      "epoch": 0.4090679737619837,
+      "grad_norm": 0.15332802899818243,
+      "learning_rate": 0.0001336505133781355,
+      "loss": 1.184,
+      "step": 5675
+    },
+    {
+      "epoch": 0.40914005622432065,
+      "grad_norm": 0.09799997518068783,
+      "learning_rate": 0.0001336285268778605,
+      "loss": 1.2705,
+      "step": 5676
+    },
+    {
+      "epoch": 0.4092121386866575,
+      "grad_norm": 0.12306729554148686,
+      "learning_rate": 0.00013360653854453,
+      "loss": 1.2619,
+      "step": 5677
+    },
+    {
+      "epoch": 0.40928422114899443,
+      "grad_norm": 0.1070022317008985,
+      "learning_rate": 0.00013358454837934244,
+      "loss": 1.2817,
+      "step": 5678
+    },
+    {
+      "epoch": 0.40935630361133135,
+      "grad_norm": 0.12037255122132975,
+      "learning_rate": 0.0001335625563834966,
+      "loss": 1.1212,
+      "step": 5679
+    },
+    {
+      "epoch": 0.40942838607366827,
+      "grad_norm": 0.10716919726378195,
+      "learning_rate": 0.0001335405625581911,
+      "loss": 1.3491,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4095004685360052,
+      "grad_norm": 0.12050129251644381,
+      "learning_rate": 0.00013351856690462497,
+      "loss": 1.3162,
+      "step": 5681
+    },
+    {
+      "epoch": 0.4095725509983421,
+      "grad_norm": 0.10914068005910679,
+      "learning_rate": 0.00013349656942399703,
+      "loss": 1.2483,
+      "step": 5682
+    },
+    {
+      "epoch": 0.40964463346067903,
+      "grad_norm": 0.10495985271262381,
+      "learning_rate": 0.00013347457011750636,
+      "loss": 1.2365,
+      "step": 5683
+    },
+    {
+      "epoch": 0.40971671592301595,
+      "grad_norm": 0.10802953277104177,
+      "learning_rate": 0.0001334525689863522,
+      "loss": 1.3306,
+      "step": 5684
+    },
+    {
+      "epoch": 0.40978879838535287,
+      "grad_norm": 0.10458498102533452,
+      "learning_rate": 0.00013343056603173375,
+      "loss": 1.2367,
+      "step": 5685
+    },
+    {
+      "epoch": 0.40986088084768973,
+      "grad_norm": 0.11788584739240673,
+      "learning_rate": 0.00013340856125485033,
+      "loss": 1.2261,
+      "step": 5686
+    },
+    {
+      "epoch": 0.40993296331002665,
+      "grad_norm": 0.0978875902232578,
+      "learning_rate": 0.00013338655465690148,
+      "loss": 1.3106,
+      "step": 5687
+    },
+    {
+      "epoch": 0.4100050457723636,
+      "grad_norm": 0.10579743348144734,
+      "learning_rate": 0.00013336454623908667,
+      "loss": 1.358,
+      "step": 5688
+    },
+    {
+      "epoch": 0.4100771282347005,
+      "grad_norm": 0.11072430966678967,
+      "learning_rate": 0.00013334253600260563,
+      "loss": 1.2695,
+      "step": 5689
+    },
+    {
+      "epoch": 0.4101492106970374,
+      "grad_norm": 0.13013734103319183,
+      "learning_rate": 0.0001333205239486581,
+      "loss": 1.1708,
+      "step": 5690
+    },
+    {
+      "epoch": 0.41022129315937433,
+      "grad_norm": 0.13264905843265537,
+      "learning_rate": 0.00013329851007844393,
+      "loss": 1.3157,
+      "step": 5691
+    },
+    {
+      "epoch": 0.41029337562171125,
+      "grad_norm": 0.10637010863019353,
+      "learning_rate": 0.00013327649439316304,
+      "loss": 1.2461,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4103654580840482,
+      "grad_norm": 0.10574455631508932,
+      "learning_rate": 0.00013325447689401548,
+      "loss": 1.2763,
+      "step": 5693
+    },
+    {
+      "epoch": 0.41043754054638504,
+      "grad_norm": 0.15851624209908657,
+      "learning_rate": 0.00013323245758220146,
+      "loss": 1.3186,
+      "step": 5694
+    },
+    {
+      "epoch": 0.41050962300872196,
+      "grad_norm": 0.10533192864288703,
+      "learning_rate": 0.0001332104364589212,
+      "loss": 1.1932,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4105817054710589,
+      "grad_norm": 0.11536148061924036,
+      "learning_rate": 0.00013318841352537503,
+      "loss": 1.2168,
+      "step": 5696
+    },
+    {
+      "epoch": 0.4106537879333958,
+      "grad_norm": 0.1048950646088059,
+      "learning_rate": 0.0001331663887827634,
+      "loss": 1.3208,
+      "step": 5697
+    },
+    {
+      "epoch": 0.4107258703957327,
+      "grad_norm": 0.10781021078766895,
+      "learning_rate": 0.0001331443622322869,
+      "loss": 1.326,
+      "step": 5698
+    },
+    {
+      "epoch": 0.41079795285806964,
+      "grad_norm": 0.12667356382978567,
+      "learning_rate": 0.00013312233387514612,
+      "loss": 1.2568,
+      "step": 5699
+    },
+    {
+      "epoch": 0.41087003532040656,
+      "grad_norm": 0.1137007331082986,
+      "learning_rate": 0.00013310030371254183,
+      "loss": 1.2571,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4109421177827435,
+      "grad_norm": 0.1182913850317735,
+      "learning_rate": 0.00013307827174567486,
+      "loss": 1.3686,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4110142002450804,
+      "grad_norm": 0.11638697849299985,
+      "learning_rate": 0.00013305623797574615,
+      "loss": 1.194,
+      "step": 5702
+    },
+    {
+      "epoch": 0.41108628270741726,
+      "grad_norm": 0.10824052608869236,
+      "learning_rate": 0.00013303420240395674,
+      "loss": 1.3069,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4111583651697542,
+      "grad_norm": 0.12065173132982018,
+      "learning_rate": 0.0001330121650315078,
+      "loss": 1.3299,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4112304476320911,
+      "grad_norm": 0.11942874816969073,
+      "learning_rate": 0.00013299012585960048,
+      "loss": 1.2949,
+      "step": 5705
+    },
+    {
+      "epoch": 0.411302530094428,
+      "grad_norm": 0.11608587681842103,
+      "learning_rate": 0.00013296808488943618,
+      "loss": 1.3253,
+      "step": 5706
+    },
+    {
+      "epoch": 0.41137461255676494,
+      "grad_norm": 0.11757105639400324,
+      "learning_rate": 0.0001329460421222163,
+      "loss": 1.2655,
+      "step": 5707
+    },
+    {
+      "epoch": 0.41144669501910186,
+      "grad_norm": 0.12108696053670363,
+      "learning_rate": 0.00013292399755914246,
+      "loss": 1.3629,
+      "step": 5708
+    },
+    {
+      "epoch": 0.4115187774814388,
+      "grad_norm": 0.10805178993489543,
+      "learning_rate": 0.00013290195120141607,
+      "loss": 1.2954,
+      "step": 5709
+    },
+    {
+      "epoch": 0.4115908599437757,
+      "grad_norm": 0.11109568264159322,
+      "learning_rate": 0.0001328799030502391,
+      "loss": 1.2644,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4116629424061126,
+      "grad_norm": 0.2832343048857961,
+      "learning_rate": 0.0001328578531068132,
+      "loss": 1.2418,
+      "step": 5711
+    },
+    {
+      "epoch": 0.4117350248684495,
+      "grad_norm": 0.268277169143338,
+      "learning_rate": 0.0001328358013723404,
+      "loss": 1.2961,
+      "step": 5712
+    },
+    {
+      "epoch": 0.4118071073307864,
+      "grad_norm": 0.11015818833427049,
+      "learning_rate": 0.00013281374784802262,
+      "loss": 1.296,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4118791897931233,
+      "grad_norm": 0.1097190654158983,
+      "learning_rate": 0.00013279169253506206,
+      "loss": 1.1956,
+      "step": 5714
+    },
+    {
+      "epoch": 0.41195127225546024,
+      "grad_norm": 0.11132267044981346,
+      "learning_rate": 0.0001327696354346609,
+      "loss": 1.2304,
+      "step": 5715
+    },
+    {
+      "epoch": 0.41202335471779716,
+      "grad_norm": 0.12660566150906918,
+      "learning_rate": 0.0001327475765480214,
+      "loss": 1.2353,
+      "step": 5716
+    },
+    {
+      "epoch": 0.4120954371801341,
+      "grad_norm": 0.124472224039625,
+      "learning_rate": 0.00013272551587634605,
+      "loss": 1.3571,
+      "step": 5717
+    },
+    {
+      "epoch": 0.412167519642471,
+      "grad_norm": 0.1281840552954541,
+      "learning_rate": 0.0001327034534208373,
+      "loss": 1.2677,
+      "step": 5718
+    },
+    {
+      "epoch": 0.4122396021048079,
+      "grad_norm": 0.11329001736276889,
+      "learning_rate": 0.00013268138918269776,
+      "loss": 1.2335,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4123116845671448,
+      "grad_norm": 0.11642943591020229,
+      "learning_rate": 0.00013265932316313013,
+      "loss": 1.2443,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4123837670294817,
+      "grad_norm": 0.10705950202844583,
+      "learning_rate": 0.00013263725536333723,
+      "loss": 1.2931,
+      "step": 5721
+    },
+    {
+      "epoch": 0.4124558494918186,
+      "grad_norm": 0.10596976152183164,
+      "learning_rate": 0.0001326151857845219,
+      "loss": 1.2661,
+      "step": 5722
+    },
+    {
+      "epoch": 0.41252793195415555,
+      "grad_norm": 0.11616574234535154,
+      "learning_rate": 0.00013259311442788716,
+      "loss": 1.2783,
+      "step": 5723
+    },
+    {
+      "epoch": 0.41260001441649247,
+      "grad_norm": 0.1569605186693033,
+      "learning_rate": 0.00013257104129463614,
+      "loss": 1.2692,
+      "step": 5724
+    },
+    {
+      "epoch": 0.4126720968788294,
+      "grad_norm": 0.1190730203545132,
+      "learning_rate": 0.00013254896638597194,
+      "loss": 1.3042,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4127441793411663,
+      "grad_norm": 0.2140898320795892,
+      "learning_rate": 0.00013252688970309792,
+      "loss": 1.2345,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4128162618035032,
+      "grad_norm": 0.11379913756709181,
+      "learning_rate": 0.00013250481124721736,
+      "loss": 1.2539,
+      "step": 5727
+    },
+    {
+      "epoch": 0.41288834426584015,
+      "grad_norm": 0.11709807172101969,
+      "learning_rate": 0.00013248273101953387,
+      "loss": 1.2226,
+      "step": 5728
+    },
+    {
+      "epoch": 0.412960426728177,
+      "grad_norm": 0.11637493781625313,
+      "learning_rate": 0.00013246064902125085,
+      "loss": 1.3205,
+      "step": 5729
+    },
+    {
+      "epoch": 0.41303250919051393,
+      "grad_norm": 0.10374111723985312,
+      "learning_rate": 0.00013243856525357213,
+      "loss": 1.3509,
+      "step": 5730
+    },
+    {
+      "epoch": 0.41310459165285085,
+      "grad_norm": 0.10814065459120716,
+      "learning_rate": 0.00013241647971770135,
+      "loss": 1.29,
+      "step": 5731
+    },
+    {
+      "epoch": 0.41317667411518777,
+      "grad_norm": 0.11258059102785575,
+      "learning_rate": 0.00013239439241484246,
+      "loss": 1.3869,
+      "step": 5732
+    },
+    {
+      "epoch": 0.4132487565775247,
+      "grad_norm": 0.1044766688331684,
+      "learning_rate": 0.00013237230334619935,
+      "loss": 1.3134,
+      "step": 5733
+    },
+    {
+      "epoch": 0.4133208390398616,
+      "grad_norm": 0.12756888163021196,
+      "learning_rate": 0.0001323502125129761,
+      "loss": 1.3696,
+      "step": 5734
+    },
+    {
+      "epoch": 0.41339292150219853,
+      "grad_norm": 0.10054656785411854,
+      "learning_rate": 0.00013232811991637685,
+      "loss": 1.1691,
+      "step": 5735
+    },
+    {
+      "epoch": 0.41346500396453545,
+      "grad_norm": 0.11148892938127969,
+      "learning_rate": 0.00013230602555760587,
+      "loss": 1.3741,
+      "step": 5736
+    },
+    {
+      "epoch": 0.41353708642687237,
+      "grad_norm": 0.11278775855622056,
+      "learning_rate": 0.00013228392943786748,
+      "loss": 1.2846,
+      "step": 5737
+    },
+    {
+      "epoch": 0.41360916888920923,
+      "grad_norm": 0.10030062522877578,
+      "learning_rate": 0.00013226183155836608,
+      "loss": 1.2758,
+      "step": 5738
+    },
+    {
+      "epoch": 0.41368125135154615,
+      "grad_norm": 0.1098357960728305,
+      "learning_rate": 0.00013223973192030624,
+      "loss": 1.2857,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4137533338138831,
+      "grad_norm": 0.11653637587256298,
+      "learning_rate": 0.00013221763052489262,
+      "loss": 1.293,
+      "step": 5740
+    },
+    {
+      "epoch": 0.41382541627622,
+      "grad_norm": 0.11371480477021054,
+      "learning_rate": 0.0001321955273733299,
+      "loss": 1.2134,
+      "step": 5741
+    },
+    {
+      "epoch": 0.4138974987385569,
+      "grad_norm": 0.10979330966709969,
+      "learning_rate": 0.0001321734224668229,
+      "loss": 1.2229,
+      "step": 5742
+    },
+    {
+      "epoch": 0.41396958120089383,
+      "grad_norm": 0.11370099392691241,
+      "learning_rate": 0.00013215131580657653,
+      "loss": 1.2825,
+      "step": 5743
+    },
+    {
+      "epoch": 0.41404166366323075,
+      "grad_norm": 0.11254628814476295,
+      "learning_rate": 0.00013212920739379583,
+      "loss": 1.293,
+      "step": 5744
+    },
+    {
+      "epoch": 0.41411374612556767,
+      "grad_norm": 0.10831339698838317,
+      "learning_rate": 0.0001321070972296859,
+      "loss": 1.3218,
+      "step": 5745
+    },
+    {
+      "epoch": 0.41418582858790454,
+      "grad_norm": 0.11477512419623954,
+      "learning_rate": 0.00013208498531545193,
+      "loss": 1.2891,
+      "step": 5746
+    },
+    {
+      "epoch": 0.41425791105024146,
+      "grad_norm": 0.11548331039029583,
+      "learning_rate": 0.0001320628716522992,
+      "loss": 1.3178,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4143299935125784,
+      "grad_norm": 0.12070579217375459,
+      "learning_rate": 0.00013204075624143315,
+      "loss": 1.3098,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4144020759749153,
+      "grad_norm": 0.15614379391625516,
+      "learning_rate": 0.00013201863908405921,
+      "loss": 1.3405,
+      "step": 5749
+    },
+    {
+      "epoch": 0.4144741584372522,
+      "grad_norm": 0.1157058821515502,
+      "learning_rate": 0.00013199652018138305,
+      "loss": 1.3379,
+      "step": 5750
+    },
+    {
+      "epoch": 0.41454624089958914,
+      "grad_norm": 0.10292561629740869,
+      "learning_rate": 0.0001319743995346102,
+      "loss": 1.2631,
+      "step": 5751
+    },
+    {
+      "epoch": 0.41461832336192606,
+      "grad_norm": 0.11375077839785128,
+      "learning_rate": 0.0001319522771449466,
+      "loss": 1.3156,
+      "step": 5752
+    },
+    {
+      "epoch": 0.414690405824263,
+      "grad_norm": 0.1071203816138007,
+      "learning_rate": 0.000131930153013598,
+      "loss": 1.3131,
+      "step": 5753
+    },
+    {
+      "epoch": 0.4147624882865999,
+      "grad_norm": 0.1294971290765176,
+      "learning_rate": 0.00013190802714177044,
+      "loss": 1.3538,
+      "step": 5754
+    },
+    {
+      "epoch": 0.41483457074893676,
+      "grad_norm": 0.11264996806075384,
+      "learning_rate": 0.0001318858995306699,
+      "loss": 1.2519,
+      "step": 5755
+    },
+    {
+      "epoch": 0.4149066532112737,
+      "grad_norm": 0.12234996614199385,
+      "learning_rate": 0.00013186377018150263,
+      "loss": 1.3605,
+      "step": 5756
+    },
+    {
+      "epoch": 0.4149787356736106,
+      "grad_norm": 0.11423327423157063,
+      "learning_rate": 0.00013184163909547478,
+      "loss": 1.2944,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4150508181359475,
+      "grad_norm": 0.18715231678759536,
+      "learning_rate": 0.00013181950627379275,
+      "loss": 1.3195,
+      "step": 5758
+    },
+    {
+      "epoch": 0.41512290059828444,
+      "grad_norm": 0.11124244950786112,
+      "learning_rate": 0.00013179737171766298,
+      "loss": 1.3089,
+      "step": 5759
+    },
+    {
+      "epoch": 0.41519498306062136,
+      "grad_norm": 0.14922516244303155,
+      "learning_rate": 0.00013177523542829195,
+      "loss": 1.2605,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4152670655229583,
+      "grad_norm": 0.1101061599766577,
+      "learning_rate": 0.00013175309740688635,
+      "loss": 1.242,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4153391479852952,
+      "grad_norm": 0.11788276111689312,
+      "learning_rate": 0.00013173095765465286,
+      "loss": 1.3913,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4154112304476321,
+      "grad_norm": 0.1122174238913854,
+      "learning_rate": 0.00013170881617279828,
+      "loss": 1.3416,
+      "step": 5763
+    },
+    {
+      "epoch": 0.415483312909969,
+      "grad_norm": 0.13740005168327665,
+      "learning_rate": 0.0001316866729625296,
+      "loss": 1.3223,
+      "step": 5764
+    },
+    {
+      "epoch": 0.4155553953723059,
+      "grad_norm": 0.11983923328947645,
+      "learning_rate": 0.0001316645280250537,
+      "loss": 1.2949,
+      "step": 5765
+    },
+    {
+      "epoch": 0.4156274778346428,
+      "grad_norm": 0.155017525359119,
+      "learning_rate": 0.00013164238136157777,
+      "loss": 1.314,
+      "step": 5766
+    },
+    {
+      "epoch": 0.41569956029697974,
+      "grad_norm": 0.10762441924205723,
+      "learning_rate": 0.00013162023297330898,
+      "loss": 1.3406,
+      "step": 5767
+    },
+    {
+      "epoch": 0.41577164275931666,
+      "grad_norm": 0.11398164812829764,
+      "learning_rate": 0.00013159808286145462,
+      "loss": 1.3621,
+      "step": 5768
+    },
+    {
+      "epoch": 0.4158437252216536,
+      "grad_norm": 0.10251603163381236,
+      "learning_rate": 0.00013157593102722205,
+      "loss": 1.3084,
+      "step": 5769
+    },
+    {
+      "epoch": 0.4159158076839905,
+      "grad_norm": 0.11741698041864027,
+      "learning_rate": 0.00013155377747181877,
+      "loss": 1.426,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4159878901463274,
+      "grad_norm": 0.11370211447302637,
+      "learning_rate": 0.0001315316221964523,
+      "loss": 1.319,
+      "step": 5771
+    },
+    {
+      "epoch": 0.4160599726086643,
+      "grad_norm": 0.09855929856167671,
+      "learning_rate": 0.00013150946520233033,
+      "loss": 1.2814,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4161320550710012,
+      "grad_norm": 0.1221957182987909,
+      "learning_rate": 0.00013148730649066063,
+      "loss": 1.3552,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4162041375333381,
+      "grad_norm": 0.11789000313965839,
+      "learning_rate": 0.00013146514606265103,
+      "loss": 1.4112,
+      "step": 5774
+    },
+    {
+      "epoch": 0.41627621999567505,
+      "grad_norm": 0.12103076357965763,
+      "learning_rate": 0.00013144298391950949,
+      "loss": 1.3872,
+      "step": 5775
+    },
+    {
+      "epoch": 0.41634830245801197,
+      "grad_norm": 0.11652355704587519,
+      "learning_rate": 0.00013142082006244402,
+      "loss": 1.3243,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4164203849203489,
+      "grad_norm": 0.11099973221504232,
+      "learning_rate": 0.00013139865449266272,
+      "loss": 1.3189,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4164924673826858,
+      "grad_norm": 0.10906793957665359,
+      "learning_rate": 0.0001313764872113739,
+      "loss": 1.3767,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4165645498450227,
+      "grad_norm": 0.10520475817019434,
+      "learning_rate": 0.00013135431821978578,
+      "loss": 1.3738,
+      "step": 5779
+    },
+    {
+      "epoch": 0.41663663230735964,
+      "grad_norm": 0.12633137391628285,
+      "learning_rate": 0.00013133214751910686,
+      "loss": 1.2844,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4167087147696965,
+      "grad_norm": 0.11383805582224164,
+      "learning_rate": 0.00013130997511054559,
+      "loss": 1.2351,
+      "step": 5781
+    },
+    {
+      "epoch": 0.41678079723203343,
+      "grad_norm": 0.11142929532580026,
+      "learning_rate": 0.00013128780099531056,
+      "loss": 1.2734,
+      "step": 5782
+    },
+    {
+      "epoch": 0.41685287969437035,
+      "grad_norm": 0.1221908184004263,
+      "learning_rate": 0.00013126562517461044,
+      "loss": 1.363,
+      "step": 5783
+    },
+    {
+      "epoch": 0.41692496215670727,
+      "grad_norm": 0.12165021882885346,
+      "learning_rate": 0.00013124344764965407,
+      "loss": 1.2442,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4169970446190442,
+      "grad_norm": 0.10160642309308507,
+      "learning_rate": 0.00013122126842165027,
+      "loss": 1.2459,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4170691270813811,
+      "grad_norm": 0.1194326860923622,
+      "learning_rate": 0.00013119908749180802,
+      "loss": 1.2941,
+      "step": 5786
+    },
+    {
+      "epoch": 0.41714120954371803,
+      "grad_norm": 0.11295709807067898,
+      "learning_rate": 0.00013117690486133646,
+      "loss": 1.2344,
+      "step": 5787
+    },
+    {
+      "epoch": 0.41721329200605495,
+      "grad_norm": 0.11710558230813294,
+      "learning_rate": 0.00013115472053144457,
+      "loss": 1.3373,
+      "step": 5788
+    },
+    {
+      "epoch": 0.41728537446839187,
+      "grad_norm": 0.12308191782372366,
+      "learning_rate": 0.00013113253450334176,
+      "loss": 1.2818,
+      "step": 5789
+    },
+    {
+      "epoch": 0.41735745693072873,
+      "grad_norm": 0.10085624745431937,
+      "learning_rate": 0.00013111034677823727,
+      "loss": 1.2626,
+      "step": 5790
+    },
+    {
+      "epoch": 0.41742953939306565,
+      "grad_norm": 0.11799548910390327,
+      "learning_rate": 0.00013108815735734058,
+      "loss": 1.2158,
+      "step": 5791
+    },
+    {
+      "epoch": 0.41750162185540257,
+      "grad_norm": 0.11842423104826943,
+      "learning_rate": 0.00013106596624186116,
+      "loss": 1.3126,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4175737043177395,
+      "grad_norm": 0.10763707075893512,
+      "learning_rate": 0.00013104377343300868,
+      "loss": 1.3083,
+      "step": 5793
+    },
+    {
+      "epoch": 0.4176457867800764,
+      "grad_norm": 0.11803709323989009,
+      "learning_rate": 0.0001310215789319928,
+      "loss": 1.2854,
+      "step": 5794
+    },
+    {
+      "epoch": 0.41771786924241333,
+      "grad_norm": 0.11561034493802463,
+      "learning_rate": 0.00013099938274002337,
+      "loss": 1.2409,
+      "step": 5795
+    },
+    {
+      "epoch": 0.41778995170475025,
+      "grad_norm": 0.12041002524230612,
+      "learning_rate": 0.00013097718485831022,
+      "loss": 1.4041,
+      "step": 5796
+    },
+    {
+      "epoch": 0.41786203416708717,
+      "grad_norm": 0.13306528328632655,
+      "learning_rate": 0.00013095498528806337,
+      "loss": 1.2806,
+      "step": 5797
+    },
+    {
+      "epoch": 0.41793411662942404,
+      "grad_norm": 0.12111016556685925,
+      "learning_rate": 0.00013093278403049293,
+      "loss": 1.2991,
+      "step": 5798
+    },
+    {
+      "epoch": 0.41800619909176095,
+      "grad_norm": 0.12678938409734886,
+      "learning_rate": 0.000130910581086809,
+      "loss": 1.3521,
+      "step": 5799
+    },
+    {
+      "epoch": 0.4180782815540979,
+      "grad_norm": 0.10186451634064025,
+      "learning_rate": 0.00013088837645822187,
+      "loss": 1.2498,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4181503640164348,
+      "grad_norm": 0.12059713587842788,
+      "learning_rate": 0.0001308661701459419,
+      "loss": 1.2344,
+      "step": 5801
+    },
+    {
+      "epoch": 0.4182224464787717,
+      "grad_norm": 0.31769294986730506,
+      "learning_rate": 0.00013084396215117948,
+      "loss": 1.3323,
+      "step": 5802
+    },
+    {
+      "epoch": 0.41829452894110863,
+      "grad_norm": 0.11060773350869568,
+      "learning_rate": 0.00013082175247514522,
+      "loss": 1.3706,
+      "step": 5803
+    },
+    {
+      "epoch": 0.41836661140344555,
+      "grad_norm": 0.11974476416181043,
+      "learning_rate": 0.00013079954111904968,
+      "loss": 1.3386,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4184386938657825,
+      "grad_norm": 0.11379437297646786,
+      "learning_rate": 0.00013077732808410367,
+      "loss": 1.2637,
+      "step": 5805
+    },
+    {
+      "epoch": 0.4185107763281194,
+      "grad_norm": 0.10394237205843383,
+      "learning_rate": 0.00013075511337151788,
+      "loss": 1.2816,
+      "step": 5806
+    },
+    {
+      "epoch": 0.41858285879045626,
+      "grad_norm": 0.10857818009139476,
+      "learning_rate": 0.0001307328969825033,
+      "loss": 1.253,
+      "step": 5807
+    },
+    {
+      "epoch": 0.4186549412527932,
+      "grad_norm": 0.11129218902499605,
+      "learning_rate": 0.00013071067891827086,
+      "loss": 1.2276,
+      "step": 5808
+    },
+    {
+      "epoch": 0.4187270237151301,
+      "grad_norm": 0.10044495026153469,
+      "learning_rate": 0.0001306884591800317,
+      "loss": 1.2855,
+      "step": 5809
+    },
+    {
+      "epoch": 0.418799106177467,
+      "grad_norm": 0.09638904860424145,
+      "learning_rate": 0.00013066623776899694,
+      "loss": 1.2599,
+      "step": 5810
+    },
+    {
+      "epoch": 0.41887118863980394,
+      "grad_norm": 0.10921158020442018,
+      "learning_rate": 0.00013064401468637792,
+      "loss": 1.2713,
+      "step": 5811
+    },
+    {
+      "epoch": 0.41894327110214086,
+      "grad_norm": 0.1057345047939826,
+      "learning_rate": 0.00013062178993338588,
+      "loss": 1.2648,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4190153535644778,
+      "grad_norm": 0.15785404356538402,
+      "learning_rate": 0.0001305995635112324,
+      "loss": 1.3137,
+      "step": 5813
+    },
+    {
+      "epoch": 0.4190874360268147,
+      "grad_norm": 0.10808185619938047,
+      "learning_rate": 0.00013057733542112896,
+      "loss": 1.2388,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4191595184891516,
+      "grad_norm": 0.12063243413042626,
+      "learning_rate": 0.00013055510566428716,
+      "loss": 1.2814,
+      "step": 5815
+    },
+    {
+      "epoch": 0.4192316009514885,
+      "grad_norm": 0.11012780517875904,
+      "learning_rate": 0.00013053287424191876,
+      "loss": 1.2854,
+      "step": 5816
+    },
+    {
+      "epoch": 0.4193036834138254,
+      "grad_norm": 0.14874599852866077,
+      "learning_rate": 0.00013051064115523553,
+      "loss": 1.334,
+      "step": 5817
+    },
+    {
+      "epoch": 0.4193757658761623,
+      "grad_norm": 0.11292175722380536,
+      "learning_rate": 0.00013048840640544944,
+      "loss": 1.3586,
+      "step": 5818
+    },
+    {
+      "epoch": 0.41944784833849924,
+      "grad_norm": 0.11311835459466472,
+      "learning_rate": 0.0001304661699937724,
+      "loss": 1.3684,
+      "step": 5819
+    },
+    {
+      "epoch": 0.41951993080083616,
+      "grad_norm": 0.10923944846665483,
+      "learning_rate": 0.00013044393192141658,
+      "loss": 1.3127,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4195920132631731,
+      "grad_norm": 0.11473208378729607,
+      "learning_rate": 0.00013042169218959408,
+      "loss": 1.3785,
+      "step": 5821
+    },
+    {
+      "epoch": 0.41966409572551,
+      "grad_norm": 0.1028403491304812,
+      "learning_rate": 0.00013039945079951722,
+      "loss": 1.3696,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4197361781878469,
+      "grad_norm": 0.1014265579633888,
+      "learning_rate": 0.00013037720775239834,
+      "loss": 1.3024,
+      "step": 5823
+    },
+    {
+      "epoch": 0.4198082606501838,
+      "grad_norm": 0.11026740824804053,
+      "learning_rate": 0.00013035496304944982,
+      "loss": 1.331,
+      "step": 5824
+    },
+    {
+      "epoch": 0.4198803431125207,
+      "grad_norm": 0.10480226424516784,
+      "learning_rate": 0.00013033271669188432,
+      "loss": 1.2934,
+      "step": 5825
+    },
+    {
+      "epoch": 0.4199524255748576,
+      "grad_norm": 0.25456077335473465,
+      "learning_rate": 0.00013031046868091435,
+      "loss": 1.2823,
+      "step": 5826
+    },
+    {
+      "epoch": 0.42002450803719454,
+      "grad_norm": 0.127490067132824,
+      "learning_rate": 0.0001302882190177527,
+      "loss": 1.2558,
+      "step": 5827
+    },
+    {
+      "epoch": 0.42009659049953146,
+      "grad_norm": 0.1170693919657135,
+      "learning_rate": 0.0001302659677036121,
+      "loss": 1.4018,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4201686729618684,
+      "grad_norm": 0.1326499289913845,
+      "learning_rate": 0.00013024371473970554,
+      "loss": 1.2489,
+      "step": 5829
+    },
+    {
+      "epoch": 0.4202407554242053,
+      "grad_norm": 0.23190052442813497,
+      "learning_rate": 0.00013022146012724593,
+      "loss": 1.4003,
+      "step": 5830
+    },
+    {
+      "epoch": 0.4203128378865422,
+      "grad_norm": 0.11465681935853253,
+      "learning_rate": 0.0001301992038674464,
+      "loss": 1.3223,
+      "step": 5831
+    },
+    {
+      "epoch": 0.42038492034887914,
+      "grad_norm": 0.10027535250065935,
+      "learning_rate": 0.00013017694596152005,
+      "loss": 1.2324,
+      "step": 5832
+    },
+    {
+      "epoch": 0.420457002811216,
+      "grad_norm": 0.10742291782988428,
+      "learning_rate": 0.00013015468641068022,
+      "loss": 1.3507,
+      "step": 5833
+    },
+    {
+      "epoch": 0.4205290852735529,
+      "grad_norm": 0.17558256257991758,
+      "learning_rate": 0.00013013242521614016,
+      "loss": 1.2802,
+      "step": 5834
+    },
+    {
+      "epoch": 0.42060116773588985,
+      "grad_norm": 0.11150127975450859,
+      "learning_rate": 0.00013011016237911342,
+      "loss": 1.2536,
+      "step": 5835
+    },
+    {
+      "epoch": 0.42067325019822677,
+      "grad_norm": 0.10635214024902168,
+      "learning_rate": 0.0001300878979008134,
+      "loss": 1.1931,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4207453326605637,
+      "grad_norm": 0.12083277982101578,
+      "learning_rate": 0.00013006563178245378,
+      "loss": 1.2972,
+      "step": 5837
+    },
+    {
+      "epoch": 0.4208174151229006,
+      "grad_norm": 0.11321232959006536,
+      "learning_rate": 0.00013004336402524827,
+      "loss": 1.3544,
+      "step": 5838
+    },
+    {
+      "epoch": 0.4208894975852375,
+      "grad_norm": 0.11502913124155117,
+      "learning_rate": 0.00013002109463041065,
+      "loss": 1.358,
+      "step": 5839
+    },
+    {
+      "epoch": 0.42096158004757445,
+      "grad_norm": 0.1723280728242107,
+      "learning_rate": 0.0001299988235991548,
+      "loss": 1.3078,
+      "step": 5840
+    },
+    {
+      "epoch": 0.42103366250991137,
+      "grad_norm": 0.10537671043726601,
+      "learning_rate": 0.00012997655093269466,
+      "loss": 1.3633,
+      "step": 5841
+    },
+    {
+      "epoch": 0.42110574497224823,
+      "grad_norm": 0.12327091635330037,
+      "learning_rate": 0.00012995427663224432,
+      "loss": 1.3083,
+      "step": 5842
+    },
+    {
+      "epoch": 0.42117782743458515,
+      "grad_norm": 0.220938417658764,
+      "learning_rate": 0.00012993200069901794,
+      "loss": 1.2797,
+      "step": 5843
+    },
+    {
+      "epoch": 0.42124990989692207,
+      "grad_norm": 0.11596788808012234,
+      "learning_rate": 0.00012990972313422973,
+      "loss": 1.1671,
+      "step": 5844
+    },
+    {
+      "epoch": 0.421321992359259,
+      "grad_norm": 0.11466929246667065,
+      "learning_rate": 0.00012988744393909403,
+      "loss": 1.2221,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4213940748215959,
+      "grad_norm": 0.10877505925517539,
+      "learning_rate": 0.00012986516311482525,
+      "loss": 1.3244,
+      "step": 5846
+    },
+    {
+      "epoch": 0.42146615728393283,
+      "grad_norm": 0.09850827171620755,
+      "learning_rate": 0.00012984288066263792,
+      "loss": 1.261,
+      "step": 5847
+    },
+    {
+      "epoch": 0.42153823974626975,
+      "grad_norm": 0.1009768629794085,
+      "learning_rate": 0.0001298205965837466,
+      "loss": 1.3214,
+      "step": 5848
+    },
+    {
+      "epoch": 0.42161032220860667,
+      "grad_norm": 0.11608980139588242,
+      "learning_rate": 0.00012979831087936598,
+      "loss": 1.4015,
+      "step": 5849
+    },
+    {
+      "epoch": 0.42168240467094353,
+      "grad_norm": 0.12382845890643349,
+      "learning_rate": 0.00012977602355071085,
+      "loss": 1.3882,
+      "step": 5850
+    },
+    {
+      "epoch": 0.42175448713328045,
+      "grad_norm": 0.15899527184725737,
+      "learning_rate": 0.00012975373459899603,
+      "loss": 1.3289,
+      "step": 5851
+    },
+    {
+      "epoch": 0.4218265695956174,
+      "grad_norm": 0.18146300055317333,
+      "learning_rate": 0.0001297314440254365,
+      "loss": 1.3099,
+      "step": 5852
+    },
+    {
+      "epoch": 0.4218986520579543,
+      "grad_norm": 0.11547981994583766,
+      "learning_rate": 0.00012970915183124727,
+      "loss": 1.2653,
+      "step": 5853
+    },
+    {
+      "epoch": 0.4219707345202912,
+      "grad_norm": 0.1271125565309649,
+      "learning_rate": 0.00012968685801764348,
+      "loss": 1.4272,
+      "step": 5854
+    },
+    {
+      "epoch": 0.42204281698262813,
+      "grad_norm": 0.1256133378897278,
+      "learning_rate": 0.00012966456258584036,
+      "loss": 1.3372,
+      "step": 5855
+    },
+    {
+      "epoch": 0.42211489944496505,
+      "grad_norm": 0.12134453699395899,
+      "learning_rate": 0.0001296422655370532,
+      "loss": 1.4028,
+      "step": 5856
+    },
+    {
+      "epoch": 0.422186981907302,
+      "grad_norm": 0.11303325662824436,
+      "learning_rate": 0.00012961996687249738,
+      "loss": 1.2423,
+      "step": 5857
+    },
+    {
+      "epoch": 0.4222590643696389,
+      "grad_norm": 0.11256265376279281,
+      "learning_rate": 0.00012959766659338836,
+      "loss": 1.3433,
+      "step": 5858
+    },
+    {
+      "epoch": 0.42233114683197576,
+      "grad_norm": 0.11572370586411072,
+      "learning_rate": 0.00012957536470094172,
+      "loss": 1.3874,
+      "step": 5859
+    },
+    {
+      "epoch": 0.4224032292943127,
+      "grad_norm": 0.12904072916645284,
+      "learning_rate": 0.00012955306119637316,
+      "loss": 1.3172,
+      "step": 5860
+    },
+    {
+      "epoch": 0.4224753117566496,
+      "grad_norm": 0.10627045432943456,
+      "learning_rate": 0.00012953075608089832,
+      "loss": 1.2957,
+      "step": 5861
+    },
+    {
+      "epoch": 0.4225473942189865,
+      "grad_norm": 0.11051937828390557,
+      "learning_rate": 0.00012950844935573312,
+      "loss": 1.2476,
+      "step": 5862
+    },
+    {
+      "epoch": 0.42261947668132344,
+      "grad_norm": 0.2805990729892988,
+      "learning_rate": 0.0001294861410220934,
+      "loss": 1.2212,
+      "step": 5863
+    },
+    {
+      "epoch": 0.42269155914366036,
+      "grad_norm": 0.12504017132282794,
+      "learning_rate": 0.00012946383108119525,
+      "loss": 1.2945,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4227636416059973,
+      "grad_norm": 0.11637587142796796,
+      "learning_rate": 0.00012944151953425464,
+      "loss": 1.2396,
+      "step": 5865
+    },
+    {
+      "epoch": 0.4228357240683342,
+      "grad_norm": 0.10537114054035275,
+      "learning_rate": 0.00012941920638248792,
+      "loss": 1.2619,
+      "step": 5866
+    },
+    {
+      "epoch": 0.4229078065306711,
+      "grad_norm": 0.10653747008694342,
+      "learning_rate": 0.0001293968916271112,
+      "loss": 1.3334,
+      "step": 5867
+    },
+    {
+      "epoch": 0.422979888993008,
+      "grad_norm": 0.11714658172203664,
+      "learning_rate": 0.00012937457526934092,
+      "loss": 1.2882,
+      "step": 5868
+    },
+    {
+      "epoch": 0.4230519714553449,
+      "grad_norm": 0.11406214283937859,
+      "learning_rate": 0.00012935225731039348,
+      "loss": 1.3175,
+      "step": 5869
+    },
+    {
+      "epoch": 0.4231240539176818,
+      "grad_norm": 0.11147341975742099,
+      "learning_rate": 0.00012932993775148543,
+      "loss": 1.228,
+      "step": 5870
+    },
+    {
+      "epoch": 0.42319613638001874,
+      "grad_norm": 0.13126414753802007,
+      "learning_rate": 0.0001293076165938334,
+      "loss": 1.2417,
+      "step": 5871
+    },
+    {
+      "epoch": 0.42326821884235566,
+      "grad_norm": 0.11357323079508128,
+      "learning_rate": 0.00012928529383865404,
+      "loss": 1.3527,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4233403013046926,
+      "grad_norm": 0.10103614837544177,
+      "learning_rate": 0.00012926296948716418,
+      "loss": 1.237,
+      "step": 5873
+    },
+    {
+      "epoch": 0.4234123837670295,
+      "grad_norm": 0.12135404069055637,
+      "learning_rate": 0.0001292406435405807,
+      "loss": 1.3032,
+      "step": 5874
+    },
+    {
+      "epoch": 0.4234844662293664,
+      "grad_norm": 0.10211723243385856,
+      "learning_rate": 0.00012921831600012053,
+      "loss": 1.4031,
+      "step": 5875
+    },
+    {
+      "epoch": 0.4235565486917033,
+      "grad_norm": 0.11529134401246115,
+      "learning_rate": 0.00012919598686700077,
+      "loss": 1.2472,
+      "step": 5876
+    },
+    {
+      "epoch": 0.4236286311540402,
+      "grad_norm": 0.11710682427022916,
+      "learning_rate": 0.0001291736561424385,
+      "loss": 1.303,
+      "step": 5877
+    },
+    {
+      "epoch": 0.4237007136163771,
+      "grad_norm": 0.09699993743914648,
+      "learning_rate": 0.000129151323827651,
+      "loss": 1.2172,
+      "step": 5878
+    },
+    {
+      "epoch": 0.42377279607871404,
+      "grad_norm": 0.12551525105250652,
+      "learning_rate": 0.00012912898992385556,
+      "loss": 1.2053,
+      "step": 5879
+    },
+    {
+      "epoch": 0.42384487854105096,
+      "grad_norm": 0.11567935819785866,
+      "learning_rate": 0.00012910665443226959,
+      "loss": 1.2796,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4239169610033879,
+      "grad_norm": 0.11175601733293496,
+      "learning_rate": 0.0001290843173541105,
+      "loss": 1.2174,
+      "step": 5881
+    },
+    {
+      "epoch": 0.4239890434657248,
+      "grad_norm": 0.12124051350811293,
+      "learning_rate": 0.00012906197869059596,
+      "loss": 1.2295,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4240611259280617,
+      "grad_norm": 0.11395202640508614,
+      "learning_rate": 0.00012903963844294357,
+      "loss": 1.3453,
+      "step": 5883
+    },
+    {
+      "epoch": 0.42413320839039864,
+      "grad_norm": 0.12006307609395785,
+      "learning_rate": 0.00012901729661237113,
+      "loss": 1.3717,
+      "step": 5884
+    },
+    {
+      "epoch": 0.4242052908527355,
+      "grad_norm": 0.17069796446166097,
+      "learning_rate": 0.00012899495320009636,
+      "loss": 1.3398,
+      "step": 5885
+    },
+    {
+      "epoch": 0.4242773733150724,
+      "grad_norm": 0.15632832349104125,
+      "learning_rate": 0.0001289726082073373,
+      "loss": 1.2891,
+      "step": 5886
+    },
+    {
+      "epoch": 0.42434945577740935,
+      "grad_norm": 0.12546624839898962,
+      "learning_rate": 0.00012895026163531185,
+      "loss": 1.2315,
+      "step": 5887
+    },
+    {
+      "epoch": 0.42442153823974627,
+      "grad_norm": 0.14834019229231427,
+      "learning_rate": 0.00012892791348523817,
+      "loss": 1.3956,
+      "step": 5888
+    },
+    {
+      "epoch": 0.4244936207020832,
+      "grad_norm": 0.12126263424167104,
+      "learning_rate": 0.00012890556375833444,
+      "loss": 1.2215,
+      "step": 5889
+    },
+    {
+      "epoch": 0.4245657031644201,
+      "grad_norm": 0.10315927841213106,
+      "learning_rate": 0.00012888321245581885,
+      "loss": 1.2784,
+      "step": 5890
+    },
+    {
+      "epoch": 0.424637785626757,
+      "grad_norm": 0.10742230810290158,
+      "learning_rate": 0.0001288608595789098,
+      "loss": 1.2537,
+      "step": 5891
+    },
+    {
+      "epoch": 0.42470986808909394,
+      "grad_norm": 0.12065020608863788,
+      "learning_rate": 0.0001288385051288257,
+      "loss": 1.2909,
+      "step": 5892
+    },
+    {
+      "epoch": 0.4247819505514308,
+      "grad_norm": 0.10915634593556547,
+      "learning_rate": 0.00012881614910678505,
+      "loss": 1.3678,
+      "step": 5893
+    },
+    {
+      "epoch": 0.42485403301376773,
+      "grad_norm": 0.11382995388896339,
+      "learning_rate": 0.0001287937915140065,
+      "loss": 1.2935,
+      "step": 5894
+    },
+    {
+      "epoch": 0.42492611547610465,
+      "grad_norm": 0.10069477482850021,
+      "learning_rate": 0.00012877143235170873,
+      "loss": 1.3218,
+      "step": 5895
+    },
+    {
+      "epoch": 0.42499819793844157,
+      "grad_norm": 0.12153711912311264,
+      "learning_rate": 0.00012874907162111046,
+      "loss": 1.2379,
+      "step": 5896
+    },
+    {
+      "epoch": 0.4250702804007785,
+      "grad_norm": 0.11254059830449073,
+      "learning_rate": 0.00012872670932343063,
+      "loss": 1.268,
+      "step": 5897
+    },
+    {
+      "epoch": 0.4251423628631154,
+      "grad_norm": 0.13703734786769434,
+      "learning_rate": 0.00012870434545988812,
+      "loss": 1.2733,
+      "step": 5898
+    },
+    {
+      "epoch": 0.42521444532545233,
+      "grad_norm": 0.11381214752427882,
+      "learning_rate": 0.000128681980031702,
+      "loss": 1.335,
+      "step": 5899
+    },
+    {
+      "epoch": 0.42528652778778925,
+      "grad_norm": 0.11455506519600654,
+      "learning_rate": 0.0001286596130400914,
+      "loss": 1.322,
+      "step": 5900
+    },
+    {
+      "epoch": 0.42535861025012617,
+      "grad_norm": 0.11800515342237063,
+      "learning_rate": 0.00012863724448627543,
+      "loss": 1.3052,
+      "step": 5901
+    },
+    {
+      "epoch": 0.42543069271246303,
+      "grad_norm": 0.11039951424009174,
+      "learning_rate": 0.0001286148743714735,
+      "loss": 1.2309,
+      "step": 5902
+    },
+    {
+      "epoch": 0.42550277517479995,
+      "grad_norm": 0.12205408023131892,
+      "learning_rate": 0.00012859250269690487,
+      "loss": 1.1516,
+      "step": 5903
+    },
+    {
+      "epoch": 0.42557485763713687,
+      "grad_norm": 0.11471860466035828,
+      "learning_rate": 0.00012857012946378908,
+      "loss": 1.2603,
+      "step": 5904
+    },
+    {
+      "epoch": 0.4256469400994738,
+      "grad_norm": 0.10794815554299066,
+      "learning_rate": 0.0001285477546733456,
+      "loss": 1.2944,
+      "step": 5905
+    },
+    {
+      "epoch": 0.4257190225618107,
+      "grad_norm": 0.12086152690495612,
+      "learning_rate": 0.00012852537832679415,
+      "loss": 1.3231,
+      "step": 5906
+    },
+    {
+      "epoch": 0.42579110502414763,
+      "grad_norm": 0.12126255995808202,
+      "learning_rate": 0.00012850300042535433,
+      "loss": 1.2484,
+      "step": 5907
+    },
+    {
+      "epoch": 0.42586318748648455,
+      "grad_norm": 0.1142396432549583,
+      "learning_rate": 0.00012848062097024605,
+      "loss": 1.2881,
+      "step": 5908
+    },
+    {
+      "epoch": 0.42593526994882147,
+      "grad_norm": 0.1315468772506774,
+      "learning_rate": 0.00012845823996268906,
+      "loss": 1.3244,
+      "step": 5909
+    },
+    {
+      "epoch": 0.4260073524111584,
+      "grad_norm": 0.13246926101587736,
+      "learning_rate": 0.00012843585740390348,
+      "loss": 1.2666,
+      "step": 5910
+    },
+    {
+      "epoch": 0.42607943487349526,
+      "grad_norm": 0.10510564248694575,
+      "learning_rate": 0.00012841347329510923,
+      "loss": 1.2954,
+      "step": 5911
+    },
+    {
+      "epoch": 0.4261515173358322,
+      "grad_norm": 0.104667597021422,
+      "learning_rate": 0.0001283910876375265,
+      "loss": 1.2858,
+      "step": 5912
+    },
+    {
+      "epoch": 0.4262235997981691,
+      "grad_norm": 0.11454108608884551,
+      "learning_rate": 0.00012836870043237547,
+      "loss": 1.3277,
+      "step": 5913
+    },
+    {
+      "epoch": 0.426295682260506,
+      "grad_norm": 0.11460490871892062,
+      "learning_rate": 0.00012834631168087652,
+      "loss": 1.2855,
+      "step": 5914
+    },
+    {
+      "epoch": 0.42636776472284293,
+      "grad_norm": 0.12125230431511966,
+      "learning_rate": 0.00012832392138424997,
+      "loss": 1.2806,
+      "step": 5915
+    },
+    {
+      "epoch": 0.42643984718517985,
+      "grad_norm": 0.1031694622435946,
+      "learning_rate": 0.0001283015295437163,
+      "loss": 1.3358,
+      "step": 5916
+    },
+    {
+      "epoch": 0.4265119296475168,
+      "grad_norm": 0.11778132725403467,
+      "learning_rate": 0.0001282791361604961,
+      "loss": 1.2875,
+      "step": 5917
+    },
+    {
+      "epoch": 0.4265840121098537,
+      "grad_norm": 0.10993557943560862,
+      "learning_rate": 0.00012825674123580997,
+      "loss": 1.3423,
+      "step": 5918
+    },
+    {
+      "epoch": 0.42665609457219056,
+      "grad_norm": 0.10824784370138037,
+      "learning_rate": 0.00012823434477087867,
+      "loss": 1.2823,
+      "step": 5919
+    },
+    {
+      "epoch": 0.4267281770345275,
+      "grad_norm": 0.11817029162189069,
+      "learning_rate": 0.00012821194676692295,
+      "loss": 1.3019,
+      "step": 5920
+    },
+    {
+      "epoch": 0.4268002594968644,
+      "grad_norm": 0.12084636576315963,
+      "learning_rate": 0.00012818954722516377,
+      "loss": 1.2282,
+      "step": 5921
+    },
+    {
+      "epoch": 0.4268723419592013,
+      "grad_norm": 0.12204778568509272,
+      "learning_rate": 0.00012816714614682207,
+      "loss": 1.2146,
+      "step": 5922
+    },
+    {
+      "epoch": 0.42694442442153824,
+      "grad_norm": 0.10323906553941516,
+      "learning_rate": 0.00012814474353311893,
+      "loss": 1.2948,
+      "step": 5923
+    },
+    {
+      "epoch": 0.42701650688387516,
+      "grad_norm": 0.10654066792107893,
+      "learning_rate": 0.00012812233938527543,
+      "loss": 1.1661,
+      "step": 5924
+    },
+    {
+      "epoch": 0.4270885893462121,
+      "grad_norm": 0.10558981866773345,
+      "learning_rate": 0.00012809993370451288,
+      "loss": 1.2425,
+      "step": 5925
+    },
+    {
+      "epoch": 0.427160671808549,
+      "grad_norm": 0.12063404032601874,
+      "learning_rate": 0.00012807752649205255,
+      "loss": 1.2798,
+      "step": 5926
+    },
+    {
+      "epoch": 0.4272327542708859,
+      "grad_norm": 0.11867919996332976,
+      "learning_rate": 0.00012805511774911584,
+      "loss": 1.3537,
+      "step": 5927
+    },
+    {
+      "epoch": 0.4273048367332228,
+      "grad_norm": 0.11646647271836985,
+      "learning_rate": 0.00012803270747692422,
+      "loss": 1.2838,
+      "step": 5928
+    },
+    {
+      "epoch": 0.4273769191955597,
+      "grad_norm": 0.1254385966521564,
+      "learning_rate": 0.00012801029567669926,
+      "loss": 1.2858,
+      "step": 5929
+    },
+    {
+      "epoch": 0.4274490016578966,
+      "grad_norm": 0.11573196001008917,
+      "learning_rate": 0.00012798788234966258,
+      "loss": 1.3285,
+      "step": 5930
+    },
+    {
+      "epoch": 0.42752108412023354,
+      "grad_norm": 0.15008203995917666,
+      "learning_rate": 0.00012796546749703594,
+      "loss": 1.3137,
+      "step": 5931
+    },
+    {
+      "epoch": 0.42759316658257046,
+      "grad_norm": 0.11995526362986839,
+      "learning_rate": 0.00012794305112004113,
+      "loss": 1.3005,
+      "step": 5932
+    },
+    {
+      "epoch": 0.4276652490449074,
+      "grad_norm": 0.10423626687707538,
+      "learning_rate": 0.00012792063321990008,
+      "loss": 1.2742,
+      "step": 5933
+    },
+    {
+      "epoch": 0.4277373315072443,
+      "grad_norm": 0.13044328954105547,
+      "learning_rate": 0.00012789821379783468,
+      "loss": 1.2152,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4278094139695812,
+      "grad_norm": 0.0991528294190453,
+      "learning_rate": 0.0001278757928550671,
+      "loss": 1.2415,
+      "step": 5935
+    },
+    {
+      "epoch": 0.42788149643191814,
+      "grad_norm": 0.10390158323400892,
+      "learning_rate": 0.00012785337039281937,
+      "loss": 1.3297,
+      "step": 5936
+    },
+    {
+      "epoch": 0.427953578894255,
+      "grad_norm": 0.11419249407126836,
+      "learning_rate": 0.00012783094641231383,
+      "loss": 1.2019,
+      "step": 5937
+    },
+    {
+      "epoch": 0.4280256613565919,
+      "grad_norm": 0.11000666769160385,
+      "learning_rate": 0.00012780852091477267,
+      "loss": 1.2843,
+      "step": 5938
+    },
+    {
+      "epoch": 0.42809774381892884,
+      "grad_norm": 0.12252237501572535,
+      "learning_rate": 0.0001277860939014184,
+      "loss": 1.3234,
+      "step": 5939
+    },
+    {
+      "epoch": 0.42816982628126576,
+      "grad_norm": 0.10488980741250442,
+      "learning_rate": 0.00012776366537347336,
+      "loss": 1.2514,
+      "step": 5940
+    },
+    {
+      "epoch": 0.4282419087436027,
+      "grad_norm": 0.1116943992588385,
+      "learning_rate": 0.00012774123533216023,
+      "loss": 1.2606,
+      "step": 5941
+    },
+    {
+      "epoch": 0.4283139912059396,
+      "grad_norm": 0.10325061239390473,
+      "learning_rate": 0.00012771880377870156,
+      "loss": 1.3234,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4283860736682765,
+      "grad_norm": 0.11901982823631839,
+      "learning_rate": 0.00012769637071432011,
+      "loss": 1.3294,
+      "step": 5943
+    },
+    {
+      "epoch": 0.42845815613061344,
+      "grad_norm": 0.11367259206274181,
+      "learning_rate": 0.0001276739361402387,
+      "loss": 1.2413,
+      "step": 5944
+    },
+    {
+      "epoch": 0.4285302385929503,
+      "grad_norm": 0.10749605401829262,
+      "learning_rate": 0.00012765150005768016,
+      "loss": 1.2698,
+      "step": 5945
+    },
+    {
+      "epoch": 0.4286023210552872,
+      "grad_norm": 0.10139773010041246,
+      "learning_rate": 0.00012762906246786753,
+      "loss": 1.2715,
+      "step": 5946
+    },
+    {
+      "epoch": 0.42867440351762415,
+      "grad_norm": 0.10162378460340356,
+      "learning_rate": 0.00012760662337202378,
+      "loss": 1.2495,
+      "step": 5947
+    },
+    {
+      "epoch": 0.42874648597996107,
+      "grad_norm": 0.1107625008712845,
+      "learning_rate": 0.0001275841827713721,
+      "loss": 1.1189,
+      "step": 5948
+    },
+    {
+      "epoch": 0.428818568442298,
+      "grad_norm": 0.13248805460454607,
+      "learning_rate": 0.0001275617406671357,
+      "loss": 1.296,
+      "step": 5949
+    },
+    {
+      "epoch": 0.4288906509046349,
+      "grad_norm": 0.12354817301566465,
+      "learning_rate": 0.00012753929706053782,
+      "loss": 1.3346,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4289627333669718,
+      "grad_norm": 0.11479157635959818,
+      "learning_rate": 0.0001275168519528019,
+      "loss": 1.3241,
+      "step": 5951
+    },
+    {
+      "epoch": 0.42903481582930875,
+      "grad_norm": 0.1130399341647562,
+      "learning_rate": 0.00012749440534515135,
+      "loss": 1.306,
+      "step": 5952
+    },
+    {
+      "epoch": 0.42910689829164567,
+      "grad_norm": 0.11291163558064936,
+      "learning_rate": 0.00012747195723880977,
+      "loss": 1.2302,
+      "step": 5953
+    },
+    {
+      "epoch": 0.42917898075398253,
+      "grad_norm": 0.10392884441210835,
+      "learning_rate": 0.00012744950763500073,
+      "loss": 1.2568,
+      "step": 5954
+    },
+    {
+      "epoch": 0.42925106321631945,
+      "grad_norm": 0.11263593238928882,
+      "learning_rate": 0.000127427056534948,
+      "loss": 1.3172,
+      "step": 5955
+    },
+    {
+      "epoch": 0.42932314567865637,
+      "grad_norm": 0.11955898459538036,
+      "learning_rate": 0.00012740460393987526,
+      "loss": 1.3762,
+      "step": 5956
+    },
+    {
+      "epoch": 0.4293952281409933,
+      "grad_norm": 0.11714622663763133,
+      "learning_rate": 0.00012738214985100646,
+      "loss": 1.3133,
+      "step": 5957
+    },
+    {
+      "epoch": 0.4294673106033302,
+      "grad_norm": 0.10469113127637003,
+      "learning_rate": 0.00012735969426956551,
+      "loss": 1.291,
+      "step": 5958
+    },
+    {
+      "epoch": 0.42953939306566713,
+      "grad_norm": 0.10928817293166507,
+      "learning_rate": 0.00012733723719677648,
+      "loss": 1.3512,
+      "step": 5959
+    },
+    {
+      "epoch": 0.42961147552800405,
+      "grad_norm": 0.09712562444412318,
+      "learning_rate": 0.00012731477863386342,
+      "loss": 1.3135,
+      "step": 5960
+    },
+    {
+      "epoch": 0.42968355799034097,
+      "grad_norm": 0.10617983694043975,
+      "learning_rate": 0.00012729231858205062,
+      "loss": 1.212,
+      "step": 5961
+    },
+    {
+      "epoch": 0.4297556404526779,
+      "grad_norm": 0.11194187969076658,
+      "learning_rate": 0.00012726985704256222,
+      "loss": 1.2225,
+      "step": 5962
+    },
+    {
+      "epoch": 0.42982772291501475,
+      "grad_norm": 0.10575754502410442,
+      "learning_rate": 0.00012724739401662272,
+      "loss": 1.2438,
+      "step": 5963
+    },
+    {
+      "epoch": 0.4298998053773517,
+      "grad_norm": 0.1576761243441324,
+      "learning_rate": 0.00012722492950545644,
+      "loss": 1.2064,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4299718878396886,
+      "grad_norm": 0.11411990582589052,
+      "learning_rate": 0.00012720246351028794,
+      "loss": 1.2908,
+      "step": 5965
+    },
+    {
+      "epoch": 0.4300439703020255,
+      "grad_norm": 0.12433308805302791,
+      "learning_rate": 0.00012717999603234183,
+      "loss": 1.3453,
+      "step": 5966
+    },
+    {
+      "epoch": 0.43011605276436243,
+      "grad_norm": 0.11925420772753433,
+      "learning_rate": 0.00012715752707284278,
+      "loss": 1.2293,
+      "step": 5967
+    },
+    {
+      "epoch": 0.43018813522669935,
+      "grad_norm": 0.11133682995920709,
+      "learning_rate": 0.00012713505663301556,
+      "loss": 1.2804,
+      "step": 5968
+    },
+    {
+      "epoch": 0.4302602176890363,
+      "grad_norm": 0.12161049241724763,
+      "learning_rate": 0.00012711258471408496,
+      "loss": 1.3471,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4303323001513732,
+      "grad_norm": 0.11294756006414503,
+      "learning_rate": 0.00012709011131727596,
+      "loss": 1.3286,
+      "step": 5970
+    },
+    {
+      "epoch": 0.43040438261371006,
+      "grad_norm": 0.3078478388891434,
+      "learning_rate": 0.00012706763644381355,
+      "loss": 1.259,
+      "step": 5971
+    },
+    {
+      "epoch": 0.430476465076047,
+      "grad_norm": 0.10703078828257706,
+      "learning_rate": 0.00012704516009492278,
+      "loss": 1.2813,
+      "step": 5972
+    },
+    {
+      "epoch": 0.4305485475383839,
+      "grad_norm": 0.11089357282905786,
+      "learning_rate": 0.00012702268227182884,
+      "loss": 1.18,
+      "step": 5973
+    },
+    {
+      "epoch": 0.4306206300007208,
+      "grad_norm": 0.10941257122714299,
+      "learning_rate": 0.00012700020297575694,
+      "loss": 1.3231,
+      "step": 5974
+    },
+    {
+      "epoch": 0.43069271246305774,
+      "grad_norm": 0.11566904790063512,
+      "learning_rate": 0.00012697772220793244,
+      "loss": 1.2803,
+      "step": 5975
+    },
+    {
+      "epoch": 0.43076479492539466,
+      "grad_norm": 0.1086805152109447,
+      "learning_rate": 0.00012695523996958074,
+      "loss": 1.2469,
+      "step": 5976
+    },
+    {
+      "epoch": 0.4308368773877316,
+      "grad_norm": 0.1448621937948636,
+      "learning_rate": 0.00012693275626192733,
+      "loss": 1.281,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4309089598500685,
+      "grad_norm": 0.11672557607619412,
+      "learning_rate": 0.00012691027108619772,
+      "loss": 1.3074,
+      "step": 5978
+    },
+    {
+      "epoch": 0.4309810423124054,
+      "grad_norm": 0.107194411233828,
+      "learning_rate": 0.00012688778444361765,
+      "loss": 1.3564,
+      "step": 5979
+    },
+    {
+      "epoch": 0.4310531247747423,
+      "grad_norm": 0.10646855050639889,
+      "learning_rate": 0.0001268652963354127,
+      "loss": 1.3132,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4311252072370792,
+      "grad_norm": 0.11981262468689367,
+      "learning_rate": 0.00012684280676280882,
+      "loss": 1.2194,
+      "step": 5981
+    },
+    {
+      "epoch": 0.4311972896994161,
+      "grad_norm": 0.11331751029534436,
+      "learning_rate": 0.00012682031572703177,
+      "loss": 1.2281,
+      "step": 5982
+    },
+    {
+      "epoch": 0.43126937216175304,
+      "grad_norm": 0.10589917128660203,
+      "learning_rate": 0.0001267978232293076,
+      "loss": 1.1997,
+      "step": 5983
+    },
+    {
+      "epoch": 0.43134145462408996,
+      "grad_norm": 0.12031956643353998,
+      "learning_rate": 0.00012677532927086234,
+      "loss": 1.2781,
+      "step": 5984
+    },
+    {
+      "epoch": 0.4314135370864269,
+      "grad_norm": 0.11411255279598043,
+      "learning_rate": 0.00012675283385292212,
+      "loss": 1.3089,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4314856195487638,
+      "grad_norm": 0.11475137735871216,
+      "learning_rate": 0.00012673033697671303,
+      "loss": 1.2844,
+      "step": 5986
+    },
+    {
+      "epoch": 0.4315577020111007,
+      "grad_norm": 0.1106754087695093,
+      "learning_rate": 0.00012670783864346153,
+      "loss": 1.3024,
+      "step": 5987
+    },
+    {
+      "epoch": 0.43162978447343764,
+      "grad_norm": 0.11541776461167863,
+      "learning_rate": 0.00012668533885439384,
+      "loss": 1.2853,
+      "step": 5988
+    },
+    {
+      "epoch": 0.4317018669357745,
+      "grad_norm": 0.1078006725038903,
+      "learning_rate": 0.00012666283761073648,
+      "loss": 1.2359,
+      "step": 5989
+    },
+    {
+      "epoch": 0.4317739493981114,
+      "grad_norm": 0.12225236439139667,
+      "learning_rate": 0.0001266403349137159,
+      "loss": 1.248,
+      "step": 5990
+    },
+    {
+      "epoch": 0.43184603186044834,
+      "grad_norm": 0.10971593040105473,
+      "learning_rate": 0.00012661783076455875,
+      "loss": 1.3237,
+      "step": 5991
+    },
+    {
+      "epoch": 0.43191811432278526,
+      "grad_norm": 0.106510758373042,
+      "learning_rate": 0.00012659532516449168,
+      "loss": 1.2322,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4319901967851222,
+      "grad_norm": 0.11474016676179333,
+      "learning_rate": 0.0001265728181147415,
+      "loss": 1.2833,
+      "step": 5993
+    },
+    {
+      "epoch": 0.4320622792474591,
+      "grad_norm": 0.3274741208215998,
+      "learning_rate": 0.00012655030961653498,
+      "loss": 1.2616,
+      "step": 5994
+    },
+    {
+      "epoch": 0.432134361709796,
+      "grad_norm": 0.10571762199246779,
+      "learning_rate": 0.00012652779967109904,
+      "loss": 1.3227,
+      "step": 5995
+    },
+    {
+      "epoch": 0.43220644417213294,
+      "grad_norm": 0.11354820827513522,
+      "learning_rate": 0.00012650528827966073,
+      "loss": 1.3052,
+      "step": 5996
+    },
+    {
+      "epoch": 0.4322785266344698,
+      "grad_norm": 0.11055411972995331,
+      "learning_rate": 0.00012648277544344706,
+      "loss": 1.334,
+      "step": 5997
+    },
+    {
+      "epoch": 0.4323506090968067,
+      "grad_norm": 0.13034009164614763,
+      "learning_rate": 0.00012646026116368522,
+      "loss": 1.2309,
+      "step": 5998
+    },
+    {
+      "epoch": 0.43242269155914365,
+      "grad_norm": 0.10881092158582081,
+      "learning_rate": 0.0001264377454416024,
+      "loss": 1.3036,
+      "step": 5999
+    },
+    {
+      "epoch": 0.43249477402148057,
+      "grad_norm": 0.12130969427539966,
+      "learning_rate": 0.00012641522827842598,
+      "loss": 1.2837,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4325668564838175,
+      "grad_norm": 0.1177225132951214,
+      "learning_rate": 0.00012639270967538328,
+      "loss": 1.3959,
+      "step": 6001
+    },
+    {
+      "epoch": 0.4326389389461544,
+      "grad_norm": 0.13577113321252643,
+      "learning_rate": 0.00012637018963370178,
+      "loss": 1.2818,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4327110214084913,
+      "grad_norm": 0.1143886651087867,
+      "learning_rate": 0.00012634766815460902,
+      "loss": 1.3085,
+      "step": 6003
+    },
+    {
+      "epoch": 0.43278310387082825,
+      "grad_norm": 0.12003043825991133,
+      "learning_rate": 0.00012632514523933265,
+      "loss": 1.2635,
+      "step": 6004
+    },
+    {
+      "epoch": 0.43285518633316517,
+      "grad_norm": 0.12069908214555626,
+      "learning_rate": 0.00012630262088910037,
+      "loss": 1.2771,
+      "step": 6005
+    },
+    {
+      "epoch": 0.43292726879550203,
+      "grad_norm": 0.11880659713555444,
+      "learning_rate": 0.00012628009510513993,
+      "loss": 1.2839,
+      "step": 6006
+    },
+    {
+      "epoch": 0.43299935125783895,
+      "grad_norm": 0.1262081695182104,
+      "learning_rate": 0.00012625756788867918,
+      "loss": 1.2612,
+      "step": 6007
+    },
+    {
+      "epoch": 0.43307143372017587,
+      "grad_norm": 0.11170343726975708,
+      "learning_rate": 0.0001262350392409461,
+      "loss": 1.2865,
+      "step": 6008
+    },
+    {
+      "epoch": 0.4331435161825128,
+      "grad_norm": 0.10250472551892477,
+      "learning_rate": 0.00012621250916316864,
+      "loss": 1.1783,
+      "step": 6009
+    },
+    {
+      "epoch": 0.4332155986448497,
+      "grad_norm": 0.12080968985783086,
+      "learning_rate": 0.00012618997765657497,
+      "loss": 1.3063,
+      "step": 6010
+    },
+    {
+      "epoch": 0.43328768110718663,
+      "grad_norm": 0.3019183718649359,
+      "learning_rate": 0.00012616744472239316,
+      "loss": 1.2707,
+      "step": 6011
+    },
+    {
+      "epoch": 0.43335976356952355,
+      "grad_norm": 0.11129190735257444,
+      "learning_rate": 0.00012614491036185157,
+      "loss": 1.3425,
+      "step": 6012
+    },
+    {
+      "epoch": 0.43343184603186047,
+      "grad_norm": 0.12160621026145742,
+      "learning_rate": 0.00012612237457617842,
+      "loss": 1.3316,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4335039284941974,
+      "grad_norm": 0.11399957881485813,
+      "learning_rate": 0.0001260998373666022,
+      "loss": 1.2171,
+      "step": 6014
+    },
+    {
+      "epoch": 0.43357601095653425,
+      "grad_norm": 0.12232407350661996,
+      "learning_rate": 0.0001260772987343513,
+      "loss": 1.2017,
+      "step": 6015
+    },
+    {
+      "epoch": 0.43364809341887117,
+      "grad_norm": 0.11756648196860203,
+      "learning_rate": 0.00012605475868065436,
+      "loss": 1.3495,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4337201758812081,
+      "grad_norm": 0.11764908525919299,
+      "learning_rate": 0.00012603221720673997,
+      "loss": 1.2528,
+      "step": 6017
+    },
+    {
+      "epoch": 0.433792258343545,
+      "grad_norm": 0.11441203557156991,
+      "learning_rate": 0.00012600967431383687,
+      "loss": 1.3631,
+      "step": 6018
+    },
+    {
+      "epoch": 0.43386434080588193,
+      "grad_norm": 0.11886612547080508,
+      "learning_rate": 0.0001259871300031738,
+      "loss": 1.3752,
+      "step": 6019
+    },
+    {
+      "epoch": 0.43393642326821885,
+      "grad_norm": 0.14534340094917095,
+      "learning_rate": 0.00012596458427597968,
+      "loss": 1.3362,
+      "step": 6020
+    },
+    {
+      "epoch": 0.43400850573055577,
+      "grad_norm": 0.11106019026905768,
+      "learning_rate": 0.00012594203713348342,
+      "loss": 1.3163,
+      "step": 6021
+    },
+    {
+      "epoch": 0.4340805881928927,
+      "grad_norm": 0.113520840793546,
+      "learning_rate": 0.00012591948857691406,
+      "loss": 1.273,
+      "step": 6022
+    },
+    {
+      "epoch": 0.43415267065522956,
+      "grad_norm": 0.14548240041877186,
+      "learning_rate": 0.00012589693860750066,
+      "loss": 1.3464,
+      "step": 6023
+    },
+    {
+      "epoch": 0.4342247531175665,
+      "grad_norm": 0.10513068624172843,
+      "learning_rate": 0.00012587438722647248,
+      "loss": 1.2869,
+      "step": 6024
+    },
+    {
+      "epoch": 0.4342968355799034,
+      "grad_norm": 0.11787256470068318,
+      "learning_rate": 0.00012585183443505869,
+      "loss": 1.2866,
+      "step": 6025
+    },
+    {
+      "epoch": 0.4343689180422403,
+      "grad_norm": 0.10943635804561624,
+      "learning_rate": 0.00012582928023448863,
+      "loss": 1.2529,
+      "step": 6026
+    },
+    {
+      "epoch": 0.43444100050457723,
+      "grad_norm": 0.12116569335568374,
+      "learning_rate": 0.00012580672462599173,
+      "loss": 1.2792,
+      "step": 6027
+    },
+    {
+      "epoch": 0.43451308296691415,
+      "grad_norm": 0.1167625676731035,
+      "learning_rate": 0.0001257841676107975,
+      "loss": 1.2812,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4345851654292511,
+      "grad_norm": 0.11759235643775234,
+      "learning_rate": 0.0001257616091901354,
+      "loss": 1.2741,
+      "step": 6029
+    },
+    {
+      "epoch": 0.434657247891588,
+      "grad_norm": 0.11552904740499924,
+      "learning_rate": 0.00012573904936523516,
+      "loss": 1.2357,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4347293303539249,
+      "grad_norm": 0.1153494135926155,
+      "learning_rate": 0.00012571648813732646,
+      "loss": 1.1996,
+      "step": 6031
+    },
+    {
+      "epoch": 0.4348014128162618,
+      "grad_norm": 0.12398861323491774,
+      "learning_rate": 0.00012569392550763913,
+      "loss": 1.2495,
+      "step": 6032
+    },
+    {
+      "epoch": 0.4348734952785987,
+      "grad_norm": 0.115514233205805,
+      "learning_rate": 0.00012567136147740294,
+      "loss": 1.3979,
+      "step": 6033
+    },
+    {
+      "epoch": 0.4349455777409356,
+      "grad_norm": 0.12330589201223284,
+      "learning_rate": 0.0001256487960478479,
+      "loss": 1.2991,
+      "step": 6034
+    },
+    {
+      "epoch": 0.43501766020327254,
+      "grad_norm": 0.11696498080067984,
+      "learning_rate": 0.000125626229220204,
+      "loss": 1.3527,
+      "step": 6035
+    },
+    {
+      "epoch": 0.43508974266560946,
+      "grad_norm": 0.11802709560950991,
+      "learning_rate": 0.0001256036609957014,
+      "loss": 1.2324,
+      "step": 6036
+    },
+    {
+      "epoch": 0.4351618251279464,
+      "grad_norm": 0.11301904806975523,
+      "learning_rate": 0.00012558109137557016,
+      "loss": 1.2674,
+      "step": 6037
+    },
+    {
+      "epoch": 0.4352339075902833,
+      "grad_norm": 0.10888619433580962,
+      "learning_rate": 0.00012555852036104064,
+      "loss": 1.3086,
+      "step": 6038
+    },
+    {
+      "epoch": 0.4353059900526202,
+      "grad_norm": 0.1072630920399545,
+      "learning_rate": 0.00012553594795334307,
+      "loss": 1.2682,
+      "step": 6039
+    },
+    {
+      "epoch": 0.43537807251495714,
+      "grad_norm": 0.12300147146733284,
+      "learning_rate": 0.0001255133741537079,
+      "loss": 1.2106,
+      "step": 6040
+    },
+    {
+      "epoch": 0.435450154977294,
+      "grad_norm": 0.1517211088319215,
+      "learning_rate": 0.00012549079896336557,
+      "loss": 1.2865,
+      "step": 6041
+    },
+    {
+      "epoch": 0.4355222374396309,
+      "grad_norm": 0.11689985926761938,
+      "learning_rate": 0.00012546822238354666,
+      "loss": 1.2534,
+      "step": 6042
+    },
+    {
+      "epoch": 0.43559431990196784,
+      "grad_norm": 0.1227655739145898,
+      "learning_rate": 0.00012544564441548182,
+      "loss": 1.3136,
+      "step": 6043
+    },
+    {
+      "epoch": 0.43566640236430476,
+      "grad_norm": 0.1293989550188248,
+      "learning_rate": 0.00012542306506040167,
+      "loss": 1.2672,
+      "step": 6044
+    },
+    {
+      "epoch": 0.4357384848266417,
+      "grad_norm": 0.1549229375219695,
+      "learning_rate": 0.00012540048431953707,
+      "loss": 1.2793,
+      "step": 6045
+    },
+    {
+      "epoch": 0.4358105672889786,
+      "grad_norm": 0.11546904309780359,
+      "learning_rate": 0.0001253779021941188,
+      "loss": 1.201,
+      "step": 6046
+    },
+    {
+      "epoch": 0.4358826497513155,
+      "grad_norm": 0.12616199716863435,
+      "learning_rate": 0.00012535531868537785,
+      "loss": 1.3066,
+      "step": 6047
+    },
+    {
+      "epoch": 0.43595473221365244,
+      "grad_norm": 0.11759778732017785,
+      "learning_rate": 0.00012533273379454517,
+      "loss": 1.208,
+      "step": 6048
+    },
+    {
+      "epoch": 0.4360268146759893,
+      "grad_norm": 0.11396219151211187,
+      "learning_rate": 0.00012531014752285188,
+      "loss": 1.2464,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4360988971383262,
+      "grad_norm": 0.12074673627968806,
+      "learning_rate": 0.00012528755987152912,
+      "loss": 1.2859,
+      "step": 6050
+    },
+    {
+      "epoch": 0.43617097960066314,
+      "grad_norm": 0.11285105248443873,
+      "learning_rate": 0.0001252649708418081,
+      "loss": 1.2843,
+      "step": 6051
+    },
+    {
+      "epoch": 0.43624306206300006,
+      "grad_norm": 0.1249545078057542,
+      "learning_rate": 0.00012524238043492016,
+      "loss": 1.3856,
+      "step": 6052
+    },
+    {
+      "epoch": 0.436315144525337,
+      "grad_norm": 0.12555119541409873,
+      "learning_rate": 0.00012521978865209668,
+      "loss": 1.2641,
+      "step": 6053
+    },
+    {
+      "epoch": 0.4363872269876739,
+      "grad_norm": 0.12071504862664831,
+      "learning_rate": 0.0001251971954945691,
+      "loss": 1.3711,
+      "step": 6054
+    },
+    {
+      "epoch": 0.4364593094500108,
+      "grad_norm": 0.10470133037504746,
+      "learning_rate": 0.00012517460096356887,
+      "loss": 1.3802,
+      "step": 6055
+    },
+    {
+      "epoch": 0.43653139191234774,
+      "grad_norm": 0.12435989149440656,
+      "learning_rate": 0.00012515200506032774,
+      "loss": 1.3112,
+      "step": 6056
+    },
+    {
+      "epoch": 0.43660347437468466,
+      "grad_norm": 0.11599279613641474,
+      "learning_rate": 0.00012512940778607726,
+      "loss": 1.2721,
+      "step": 6057
+    },
+    {
+      "epoch": 0.43667555683702153,
+      "grad_norm": 0.13688431215515548,
+      "learning_rate": 0.0001251068091420493,
+      "loss": 1.2734,
+      "step": 6058
+    },
+    {
+      "epoch": 0.43674763929935845,
+      "grad_norm": 0.12723619890369212,
+      "learning_rate": 0.00012508420912947562,
+      "loss": 1.3002,
+      "step": 6059
+    },
+    {
+      "epoch": 0.43681972176169537,
+      "grad_norm": 0.16296857734465245,
+      "learning_rate": 0.00012506160774958814,
+      "loss": 1.235,
+      "step": 6060
+    },
+    {
+      "epoch": 0.4368918042240323,
+      "grad_norm": 0.1102715304209742,
+      "learning_rate": 0.00012503900500361877,
+      "loss": 1.2419,
+      "step": 6061
+    },
+    {
+      "epoch": 0.4369638866863692,
+      "grad_norm": 0.11029009837246491,
+      "learning_rate": 0.00012501640089279968,
+      "loss": 1.2593,
+      "step": 6062
+    },
+    {
+      "epoch": 0.4370359691487061,
+      "grad_norm": 0.11154288569765503,
+      "learning_rate": 0.0001249937954183629,
+      "loss": 1.2968,
+      "step": 6063
+    },
+    {
+      "epoch": 0.43710805161104305,
+      "grad_norm": 0.10697611816642338,
+      "learning_rate": 0.00012497118858154072,
+      "loss": 1.3194,
+      "step": 6064
+    },
+    {
+      "epoch": 0.43718013407337997,
+      "grad_norm": 0.11227467520972563,
+      "learning_rate": 0.00012494858038356531,
+      "loss": 1.2936,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4372522165357169,
+      "grad_norm": 0.11128857234312853,
+      "learning_rate": 0.0001249259708256691,
+      "loss": 1.2066,
+      "step": 6066
+    },
+    {
+      "epoch": 0.43732429899805375,
+      "grad_norm": 0.11445574392642889,
+      "learning_rate": 0.00012490335990908452,
+      "loss": 1.2735,
+      "step": 6067
+    },
+    {
+      "epoch": 0.43739638146039067,
+      "grad_norm": 0.11176694310483756,
+      "learning_rate": 0.00012488074763504397,
+      "loss": 1.3772,
+      "step": 6068
+    },
+    {
+      "epoch": 0.4374684639227276,
+      "grad_norm": 0.10579239215878351,
+      "learning_rate": 0.0001248581340047801,
+      "loss": 1.2262,
+      "step": 6069
+    },
+    {
+      "epoch": 0.4375405463850645,
+      "grad_norm": 0.10409787949270378,
+      "learning_rate": 0.00012483551901952558,
+      "loss": 1.2604,
+      "step": 6070
+    },
+    {
+      "epoch": 0.43761262884740143,
+      "grad_norm": 0.10970819830783977,
+      "learning_rate": 0.00012481290268051304,
+      "loss": 1.285,
+      "step": 6071
+    },
+    {
+      "epoch": 0.43768471130973835,
+      "grad_norm": 0.12216659723111035,
+      "learning_rate": 0.00012479028498897535,
+      "loss": 1.2363,
+      "step": 6072
+    },
+    {
+      "epoch": 0.43775679377207527,
+      "grad_norm": 0.11240004115645864,
+      "learning_rate": 0.00012476766594614534,
+      "loss": 1.3515,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4378288762344122,
+      "grad_norm": 0.11562957582973675,
+      "learning_rate": 0.00012474504555325594,
+      "loss": 1.3117,
+      "step": 6074
+    },
+    {
+      "epoch": 0.43790095869674905,
+      "grad_norm": 0.11814056547257085,
+      "learning_rate": 0.0001247224238115402,
+      "loss": 1.327,
+      "step": 6075
+    },
+    {
+      "epoch": 0.437973041159086,
+      "grad_norm": 0.11148471235570584,
+      "learning_rate": 0.00012469980072223118,
+      "loss": 1.3155,
+      "step": 6076
+    },
+    {
+      "epoch": 0.4380451236214229,
+      "grad_norm": 0.1253987568642912,
+      "learning_rate": 0.00012467717628656205,
+      "loss": 1.3321,
+      "step": 6077
+    },
+    {
+      "epoch": 0.4381172060837598,
+      "grad_norm": 0.10073558018700048,
+      "learning_rate": 0.00012465455050576606,
+      "loss": 1.2874,
+      "step": 6078
+    },
+    {
+      "epoch": 0.43818928854609673,
+      "grad_norm": 0.1025648594759843,
+      "learning_rate": 0.00012463192338107648,
+      "loss": 1.2077,
+      "step": 6079
+    },
+    {
+      "epoch": 0.43826137100843365,
+      "grad_norm": 0.10686611468133184,
+      "learning_rate": 0.00012460929491372674,
+      "loss": 1.3553,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4383334534707706,
+      "grad_norm": 0.10986061099993256,
+      "learning_rate": 0.00012458666510495022,
+      "loss": 1.2728,
+      "step": 6081
+    },
+    {
+      "epoch": 0.4384055359331075,
+      "grad_norm": 0.1189148647464115,
+      "learning_rate": 0.00012456403395598052,
+      "loss": 1.2315,
+      "step": 6082
+    },
+    {
+      "epoch": 0.4384776183954444,
+      "grad_norm": 0.12680242357553617,
+      "learning_rate": 0.00012454140146805125,
+      "loss": 1.1678,
+      "step": 6083
+    },
+    {
+      "epoch": 0.4385497008577813,
+      "grad_norm": 0.10509839760677844,
+      "learning_rate": 0.000124518767642396,
+      "loss": 1.2868,
+      "step": 6084
+    },
+    {
+      "epoch": 0.4386217833201182,
+      "grad_norm": 0.2201555546608077,
+      "learning_rate": 0.0001244961324802486,
+      "loss": 1.2682,
+      "step": 6085
+    },
+    {
+      "epoch": 0.4386938657824551,
+      "grad_norm": 0.10587289947500111,
+      "learning_rate": 0.00012447349598284277,
+      "loss": 1.2331,
+      "step": 6086
+    },
+    {
+      "epoch": 0.43876594824479204,
+      "grad_norm": 0.1130914905651986,
+      "learning_rate": 0.00012445085815141252,
+      "loss": 1.2046,
+      "step": 6087
+    },
+    {
+      "epoch": 0.43883803070712896,
+      "grad_norm": 0.11111753207264405,
+      "learning_rate": 0.00012442821898719174,
+      "loss": 1.325,
+      "step": 6088
+    },
+    {
+      "epoch": 0.4389101131694659,
+      "grad_norm": 0.12204989954179231,
+      "learning_rate": 0.00012440557849141448,
+      "loss": 1.2836,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4389821956318028,
+      "grad_norm": 0.12619992008063952,
+      "learning_rate": 0.00012438293666531484,
+      "loss": 1.2974,
+      "step": 6090
+    },
+    {
+      "epoch": 0.4390542780941397,
+      "grad_norm": 0.11483082643423838,
+      "learning_rate": 0.00012436029351012705,
+      "loss": 1.4021,
+      "step": 6091
+    },
+    {
+      "epoch": 0.43912636055647664,
+      "grad_norm": 0.11802560552673791,
+      "learning_rate": 0.0001243376490270853,
+      "loss": 1.295,
+      "step": 6092
+    },
+    {
+      "epoch": 0.4391984430188135,
+      "grad_norm": 0.10812444616428626,
+      "learning_rate": 0.00012431500321742395,
+      "loss": 1.334,
+      "step": 6093
+    },
+    {
+      "epoch": 0.4392705254811504,
+      "grad_norm": 0.11264576805227115,
+      "learning_rate": 0.0001242923560823774,
+      "loss": 1.3211,
+      "step": 6094
+    },
+    {
+      "epoch": 0.43934260794348734,
+      "grad_norm": 0.11108318635645215,
+      "learning_rate": 0.00012426970762318014,
+      "loss": 1.3563,
+      "step": 6095
+    },
+    {
+      "epoch": 0.43941469040582426,
+      "grad_norm": 0.11219503318083526,
+      "learning_rate": 0.00012424705784106664,
+      "loss": 1.3606,
+      "step": 6096
+    },
+    {
+      "epoch": 0.4394867728681612,
+      "grad_norm": 0.11665316127356277,
+      "learning_rate": 0.00012422440673727164,
+      "loss": 1.3862,
+      "step": 6097
+    },
+    {
+      "epoch": 0.4395588553304981,
+      "grad_norm": 0.10932987643097195,
+      "learning_rate": 0.0001242017543130297,
+      "loss": 1.2891,
+      "step": 6098
+    },
+    {
+      "epoch": 0.439630937792835,
+      "grad_norm": 0.11958892962052757,
+      "learning_rate": 0.00012417910056957565,
+      "loss": 1.3433,
+      "step": 6099
+    },
+    {
+      "epoch": 0.43970302025517194,
+      "grad_norm": 0.11336757793279394,
+      "learning_rate": 0.00012415644550814431,
+      "loss": 1.2849,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4397751027175088,
+      "grad_norm": 0.11624575269682721,
+      "learning_rate": 0.00012413378912997058,
+      "loss": 1.2296,
+      "step": 6101
+    },
+    {
+      "epoch": 0.4398471851798457,
+      "grad_norm": 0.1441044646829946,
+      "learning_rate": 0.00012411113143628944,
+      "loss": 1.3082,
+      "step": 6102
+    },
+    {
+      "epoch": 0.43991926764218264,
+      "grad_norm": 0.12763933703499017,
+      "learning_rate": 0.0001240884724283359,
+      "loss": 1.2589,
+      "step": 6103
+    },
+    {
+      "epoch": 0.43999135010451956,
+      "grad_norm": 0.1172929186848597,
+      "learning_rate": 0.00012406581210734513,
+      "loss": 1.2566,
+      "step": 6104
+    },
+    {
+      "epoch": 0.4400634325668565,
+      "grad_norm": 0.11389086582188714,
+      "learning_rate": 0.00012404315047455234,
+      "loss": 1.2806,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4401355150291934,
+      "grad_norm": 0.11216017251418797,
+      "learning_rate": 0.0001240204875311927,
+      "loss": 1.2978,
+      "step": 6106
+    },
+    {
+      "epoch": 0.4402075974915303,
+      "grad_norm": 0.10913159109278624,
+      "learning_rate": 0.00012399782327850162,
+      "loss": 1.1913,
+      "step": 6107
+    },
+    {
+      "epoch": 0.44027967995386724,
+      "grad_norm": 0.10965471187280228,
+      "learning_rate": 0.00012397515771771448,
+      "loss": 1.3435,
+      "step": 6108
+    },
+    {
+      "epoch": 0.44035176241620416,
+      "grad_norm": 0.11538132570324143,
+      "learning_rate": 0.00012395249085006676,
+      "loss": 1.2797,
+      "step": 6109
+    },
+    {
+      "epoch": 0.440423844878541,
+      "grad_norm": 0.12352849599173399,
+      "learning_rate": 0.00012392982267679398,
+      "loss": 1.2217,
+      "step": 6110
+    },
+    {
+      "epoch": 0.44049592734087795,
+      "grad_norm": 0.11217834636230206,
+      "learning_rate": 0.00012390715319913182,
+      "loss": 1.3615,
+      "step": 6111
+    },
+    {
+      "epoch": 0.44056800980321487,
+      "grad_norm": 0.11496936878385994,
+      "learning_rate": 0.0001238844824183159,
+      "loss": 1.3327,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4406400922655518,
+      "grad_norm": 0.11671948139587257,
+      "learning_rate": 0.00012386181033558204,
+      "loss": 1.219,
+      "step": 6113
+    },
+    {
+      "epoch": 0.4407121747278887,
+      "grad_norm": 0.11620757294103688,
+      "learning_rate": 0.000123839136952166,
+      "loss": 1.2772,
+      "step": 6114
+    },
+    {
+      "epoch": 0.4407842571902256,
+      "grad_norm": 0.12264066706220123,
+      "learning_rate": 0.00012381646226930376,
+      "loss": 1.2889,
+      "step": 6115
+    },
+    {
+      "epoch": 0.44085633965256255,
+      "grad_norm": 0.11900173373780774,
+      "learning_rate": 0.00012379378628823125,
+      "loss": 1.2509,
+      "step": 6116
+    },
+    {
+      "epoch": 0.44092842211489947,
+      "grad_norm": 0.11766343676032522,
+      "learning_rate": 0.0001237711090101845,
+      "loss": 1.3118,
+      "step": 6117
+    },
+    {
+      "epoch": 0.4410005045772364,
+      "grad_norm": 0.11935256607309158,
+      "learning_rate": 0.00012374843043639965,
+      "loss": 1.3533,
+      "step": 6118
+    },
+    {
+      "epoch": 0.44107258703957325,
+      "grad_norm": 0.11459213272644242,
+      "learning_rate": 0.00012372575056811294,
+      "loss": 1.3047,
+      "step": 6119
+    },
+    {
+      "epoch": 0.44114466950191017,
+      "grad_norm": 0.12368810507791098,
+      "learning_rate": 0.00012370306940656051,
+      "loss": 1.2916,
+      "step": 6120
+    },
+    {
+      "epoch": 0.4412167519642471,
+      "grad_norm": 0.10781255270451219,
+      "learning_rate": 0.00012368038695297876,
+      "loss": 1.2499,
+      "step": 6121
+    },
+    {
+      "epoch": 0.441288834426584,
+      "grad_norm": 0.13557001524985232,
+      "learning_rate": 0.00012365770320860406,
+      "loss": 1.2715,
+      "step": 6122
+    },
+    {
+      "epoch": 0.44136091688892093,
+      "grad_norm": 0.10438511543506705,
+      "learning_rate": 0.0001236350181746729,
+      "loss": 1.3173,
+      "step": 6123
+    },
+    {
+      "epoch": 0.44143299935125785,
+      "grad_norm": 0.11937657080183532,
+      "learning_rate": 0.0001236123318524218,
+      "loss": 1.3355,
+      "step": 6124
+    },
+    {
+      "epoch": 0.44150508181359477,
+      "grad_norm": 0.11571564725724427,
+      "learning_rate": 0.00012358964424308738,
+      "loss": 1.3169,
+      "step": 6125
+    },
+    {
+      "epoch": 0.4415771642759317,
+      "grad_norm": 0.10589081549156819,
+      "learning_rate": 0.0001235669553479063,
+      "loss": 1.2848,
+      "step": 6126
+    },
+    {
+      "epoch": 0.44164924673826855,
+      "grad_norm": 0.1247922190162885,
+      "learning_rate": 0.00012354426516811534,
+      "loss": 1.2791,
+      "step": 6127
+    },
+    {
+      "epoch": 0.4417213292006055,
+      "grad_norm": 0.10635673045573944,
+      "learning_rate": 0.00012352157370495125,
+      "loss": 1.2258,
+      "step": 6128
+    },
+    {
+      "epoch": 0.4417934116629424,
+      "grad_norm": 0.11337527541113464,
+      "learning_rate": 0.00012349888095965097,
+      "loss": 1.1656,
+      "step": 6129
+    },
+    {
+      "epoch": 0.4418654941252793,
+      "grad_norm": 0.11709616720912704,
+      "learning_rate": 0.0001234761869334515,
+      "loss": 1.3158,
+      "step": 6130
+    },
+    {
+      "epoch": 0.44193757658761623,
+      "grad_norm": 0.11209568339661248,
+      "learning_rate": 0.00012345349162758978,
+      "loss": 1.2387,
+      "step": 6131
+    },
+    {
+      "epoch": 0.44200965904995315,
+      "grad_norm": 0.19421867677050417,
+      "learning_rate": 0.00012343079504330295,
+      "loss": 1.3243,
+      "step": 6132
+    },
+    {
+      "epoch": 0.44208174151229007,
+      "grad_norm": 0.20039573349795312,
+      "learning_rate": 0.00012340809718182819,
+      "loss": 1.2701,
+      "step": 6133
+    },
+    {
+      "epoch": 0.442153823974627,
+      "grad_norm": 0.1153604073857216,
+      "learning_rate": 0.00012338539804440267,
+      "loss": 1.261,
+      "step": 6134
+    },
+    {
+      "epoch": 0.4422259064369639,
+      "grad_norm": 0.10842655685281861,
+      "learning_rate": 0.00012336269763226382,
+      "loss": 1.2637,
+      "step": 6135
+    },
+    {
+      "epoch": 0.4422979888993008,
+      "grad_norm": 0.10892914988477416,
+      "learning_rate": 0.00012333999594664886,
+      "loss": 1.2778,
+      "step": 6136
+    },
+    {
+      "epoch": 0.4423700713616377,
+      "grad_norm": 0.11627308641531679,
+      "learning_rate": 0.0001233172929887954,
+      "loss": 1.3198,
+      "step": 6137
+    },
+    {
+      "epoch": 0.4424421538239746,
+      "grad_norm": 0.11155827286113511,
+      "learning_rate": 0.0001232945887599408,
+      "loss": 1.3257,
+      "step": 6138
+    },
+    {
+      "epoch": 0.44251423628631154,
+      "grad_norm": 0.11537085931099843,
+      "learning_rate": 0.00012327188326132274,
+      "loss": 1.2862,
+      "step": 6139
+    },
+    {
+      "epoch": 0.44258631874864846,
+      "grad_norm": 0.11259428037415513,
+      "learning_rate": 0.0001232491764941788,
+      "loss": 1.2319,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4426584012109854,
+      "grad_norm": 0.10668355655793729,
+      "learning_rate": 0.00012322646845974684,
+      "loss": 1.2734,
+      "step": 6141
+    },
+    {
+      "epoch": 0.4427304836733223,
+      "grad_norm": 0.10939646978993468,
+      "learning_rate": 0.0001232037591592645,
+      "loss": 1.2858,
+      "step": 6142
+    },
+    {
+      "epoch": 0.4428025661356592,
+      "grad_norm": 0.12410195140199048,
+      "learning_rate": 0.00012318104859396973,
+      "loss": 1.2584,
+      "step": 6143
+    },
+    {
+      "epoch": 0.44287464859799613,
+      "grad_norm": 0.11845068683530852,
+      "learning_rate": 0.0001231583367651004,
+      "loss": 1.2881,
+      "step": 6144
+    },
+    {
+      "epoch": 0.442946731060333,
+      "grad_norm": 0.11377432250809363,
+      "learning_rate": 0.00012313562367389453,
+      "loss": 1.3056,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4430188135226699,
+      "grad_norm": 0.24260730616142437,
+      "learning_rate": 0.0001231129093215902,
+      "loss": 1.2085,
+      "step": 6146
+    },
+    {
+      "epoch": 0.44309089598500684,
+      "grad_norm": 0.10053055430571453,
+      "learning_rate": 0.00012309019370942555,
+      "loss": 1.2449,
+      "step": 6147
+    },
+    {
+      "epoch": 0.44316297844734376,
+      "grad_norm": 0.1064719995200622,
+      "learning_rate": 0.00012306747683863878,
+      "loss": 1.3351,
+      "step": 6148
+    },
+    {
+      "epoch": 0.4432350609096807,
+      "grad_norm": 0.40079505076091365,
+      "learning_rate": 0.00012304475871046812,
+      "loss": 1.3754,
+      "step": 6149
+    },
+    {
+      "epoch": 0.4433071433720176,
+      "grad_norm": 0.11631160718543326,
+      "learning_rate": 0.00012302203932615198,
+      "loss": 1.3051,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4433792258343545,
+      "grad_norm": 0.11032011891666405,
+      "learning_rate": 0.0001229993186869287,
+      "loss": 1.2485,
+      "step": 6151
+    },
+    {
+      "epoch": 0.44345130829669144,
+      "grad_norm": 0.1158137852875097,
+      "learning_rate": 0.00012297659679403683,
+      "loss": 1.3662,
+      "step": 6152
+    },
+    {
+      "epoch": 0.4435233907590283,
+      "grad_norm": 0.11676046719362458,
+      "learning_rate": 0.00012295387364871487,
+      "loss": 1.3153,
+      "step": 6153
+    },
+    {
+      "epoch": 0.4435954732213652,
+      "grad_norm": 0.12402122812734985,
+      "learning_rate": 0.00012293114925220145,
+      "loss": 1.323,
+      "step": 6154
+    },
+    {
+      "epoch": 0.44366755568370214,
+      "grad_norm": 0.10356838858319893,
+      "learning_rate": 0.00012290842360573523,
+      "loss": 1.2237,
+      "step": 6155
+    },
+    {
+      "epoch": 0.44373963814603906,
+      "grad_norm": 0.12591010024044288,
+      "learning_rate": 0.00012288569671055498,
+      "loss": 1.2633,
+      "step": 6156
+    },
+    {
+      "epoch": 0.443811720608376,
+      "grad_norm": 0.12489782039409159,
+      "learning_rate": 0.00012286296856789953,
+      "loss": 1.3078,
+      "step": 6157
+    },
+    {
+      "epoch": 0.4438838030707129,
+      "grad_norm": 0.11893771496951584,
+      "learning_rate": 0.00012284023917900775,
+      "loss": 1.2958,
+      "step": 6158
+    },
+    {
+      "epoch": 0.4439558855330498,
+      "grad_norm": 0.10814407111752877,
+      "learning_rate": 0.0001228175085451186,
+      "loss": 1.2753,
+      "step": 6159
+    },
+    {
+      "epoch": 0.44402796799538674,
+      "grad_norm": 0.11171073398920568,
+      "learning_rate": 0.0001227947766674711,
+      "loss": 1.3338,
+      "step": 6160
+    },
+    {
+      "epoch": 0.44410005045772366,
+      "grad_norm": 0.1110254460156328,
+      "learning_rate": 0.00012277204354730435,
+      "loss": 1.3551,
+      "step": 6161
+    },
+    {
+      "epoch": 0.4441721329200605,
+      "grad_norm": 0.11184821841464981,
+      "learning_rate": 0.00012274930918585749,
+      "loss": 1.2965,
+      "step": 6162
+    },
+    {
+      "epoch": 0.44424421538239744,
+      "grad_norm": 0.12111825452799499,
+      "learning_rate": 0.00012272657358436978,
+      "loss": 1.2812,
+      "step": 6163
+    },
+    {
+      "epoch": 0.44431629784473436,
+      "grad_norm": 0.11295193239834368,
+      "learning_rate": 0.0001227038367440805,
+      "loss": 1.3299,
+      "step": 6164
+    },
+    {
+      "epoch": 0.4443883803070713,
+      "grad_norm": 0.09707470646093042,
+      "learning_rate": 0.00012268109866622897,
+      "loss": 1.2111,
+      "step": 6165
+    },
+    {
+      "epoch": 0.4444604627694082,
+      "grad_norm": 0.11600411273185034,
+      "learning_rate": 0.00012265835935205467,
+      "loss": 1.2349,
+      "step": 6166
+    },
+    {
+      "epoch": 0.4445325452317451,
+      "grad_norm": 0.1133192419486963,
+      "learning_rate": 0.00012263561880279707,
+      "loss": 1.1631,
+      "step": 6167
+    },
+    {
+      "epoch": 0.44460462769408204,
+      "grad_norm": 0.1780357225117862,
+      "learning_rate": 0.00012261287701969574,
+      "loss": 1.2115,
+      "step": 6168
+    },
+    {
+      "epoch": 0.44467671015641896,
+      "grad_norm": 0.10246042418179444,
+      "learning_rate": 0.0001225901340039903,
+      "loss": 1.3566,
+      "step": 6169
+    },
+    {
+      "epoch": 0.4447487926187559,
+      "grad_norm": 0.1154150053118643,
+      "learning_rate": 0.0001225673897569205,
+      "loss": 1.3049,
+      "step": 6170
+    },
+    {
+      "epoch": 0.44482087508109275,
+      "grad_norm": 0.11280426959896243,
+      "learning_rate": 0.00012254464427972602,
+      "loss": 1.2833,
+      "step": 6171
+    },
+    {
+      "epoch": 0.44489295754342967,
+      "grad_norm": 0.11145951234746702,
+      "learning_rate": 0.00012252189757364678,
+      "loss": 1.2423,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4449650400057666,
+      "grad_norm": 0.14381507618070655,
+      "learning_rate": 0.0001224991496399226,
+      "loss": 1.3326,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4450371224681035,
+      "grad_norm": 0.11111792630294133,
+      "learning_rate": 0.0001224764004797935,
+      "loss": 1.3065,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4451092049304404,
+      "grad_norm": 0.1094546101980326,
+      "learning_rate": 0.0001224536500944995,
+      "loss": 1.3346,
+      "step": 6175
+    },
+    {
+      "epoch": 0.44518128739277735,
+      "grad_norm": 0.11920693101759212,
+      "learning_rate": 0.00012243089848528067,
+      "loss": 1.2787,
+      "step": 6176
+    },
+    {
+      "epoch": 0.44525336985511427,
+      "grad_norm": 0.11642799342103612,
+      "learning_rate": 0.00012240814565337723,
+      "loss": 1.2135,
+      "step": 6177
+    },
+    {
+      "epoch": 0.4453254523174512,
+      "grad_norm": 0.12315348868084296,
+      "learning_rate": 0.00012238539160002934,
+      "loss": 1.3308,
+      "step": 6178
+    },
+    {
+      "epoch": 0.44539753477978805,
+      "grad_norm": 0.11398484821734185,
+      "learning_rate": 0.00012236263632647737,
+      "loss": 1.2684,
+      "step": 6179
+    },
+    {
+      "epoch": 0.44546961724212497,
+      "grad_norm": 0.1167539952543268,
+      "learning_rate": 0.00012233987983396166,
+      "loss": 1.2943,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4455416997044619,
+      "grad_norm": 0.12315584595160435,
+      "learning_rate": 0.00012231712212372263,
+      "loss": 1.3452,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4456137821667988,
+      "grad_norm": 0.1060710286274634,
+      "learning_rate": 0.0001222943631970008,
+      "loss": 1.3259,
+      "step": 6182
+    },
+    {
+      "epoch": 0.44568586462913573,
+      "grad_norm": 0.11976884988259269,
+      "learning_rate": 0.0001222716030550367,
+      "loss": 1.2502,
+      "step": 6183
+    },
+    {
+      "epoch": 0.44575794709147265,
+      "grad_norm": 0.11550865860061085,
+      "learning_rate": 0.00012224884169907105,
+      "loss": 1.325,
+      "step": 6184
+    },
+    {
+      "epoch": 0.44583002955380957,
+      "grad_norm": 0.11184530643874704,
+      "learning_rate": 0.0001222260791303444,
+      "loss": 1.2225,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4459021120161465,
+      "grad_norm": 0.341874287549379,
+      "learning_rate": 0.0001222033153500977,
+      "loss": 1.3414,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4459741944784834,
+      "grad_norm": 0.11183294795211877,
+      "learning_rate": 0.00012218055035957155,
+      "loss": 1.3543,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4460462769408203,
+      "grad_norm": 0.11798965032535438,
+      "learning_rate": 0.00012215778416000707,
+      "loss": 1.2231,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4461183594031572,
+      "grad_norm": 0.11783780395218775,
+      "learning_rate": 0.00012213501675264508,
+      "loss": 1.3677,
+      "step": 6189
+    },
+    {
+      "epoch": 0.4461904418654941,
+      "grad_norm": 0.10979576167472976,
+      "learning_rate": 0.00012211224813872667,
+      "loss": 1.3008,
+      "step": 6190
+    },
+    {
+      "epoch": 0.44626252432783103,
+      "grad_norm": 0.11566090398481109,
+      "learning_rate": 0.00012208947831949288,
+      "loss": 1.2583,
+      "step": 6191
+    },
+    {
+      "epoch": 0.44633460679016795,
+      "grad_norm": 0.14898298645270414,
+      "learning_rate": 0.00012206670729618497,
+      "loss": 1.3316,
+      "step": 6192
+    },
+    {
+      "epoch": 0.4464066892525049,
+      "grad_norm": 0.18806865890030522,
+      "learning_rate": 0.00012204393507004403,
+      "loss": 1.2444,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4464787717148418,
+      "grad_norm": 0.13235715846909407,
+      "learning_rate": 0.00012202116164231145,
+      "loss": 1.2904,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4465508541771787,
+      "grad_norm": 0.1109614961508708,
+      "learning_rate": 0.00012199838701422856,
+      "loss": 1.3235,
+      "step": 6195
+    },
+    {
+      "epoch": 0.44662293663951563,
+      "grad_norm": 0.10455562071853947,
+      "learning_rate": 0.00012197561118703678,
+      "loss": 1.2797,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4466950191018525,
+      "grad_norm": 0.11608513357209674,
+      "learning_rate": 0.00012195283416197756,
+      "loss": 1.3316,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4467671015641894,
+      "grad_norm": 0.12891197728446196,
+      "learning_rate": 0.0001219300559402925,
+      "loss": 1.2145,
+      "step": 6198
+    },
+    {
+      "epoch": 0.44683918402652634,
+      "grad_norm": 0.11430332662537411,
+      "learning_rate": 0.00012190727652322321,
+      "loss": 1.2434,
+      "step": 6199
+    },
+    {
+      "epoch": 0.44691126648886326,
+      "grad_norm": 0.11501861108787079,
+      "learning_rate": 0.00012188449591201134,
+      "loss": 1.3638,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4469833489512002,
+      "grad_norm": 0.1221832493658122,
+      "learning_rate": 0.00012186171410789866,
+      "loss": 1.2891,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4470554314135371,
+      "grad_norm": 0.1045432489213537,
+      "learning_rate": 0.00012183893111212698,
+      "loss": 1.3162,
+      "step": 6202
+    },
+    {
+      "epoch": 0.447127513875874,
+      "grad_norm": 0.10349770427949422,
+      "learning_rate": 0.00012181614692593819,
+      "loss": 1.2493,
+      "step": 6203
+    },
+    {
+      "epoch": 0.44719959633821094,
+      "grad_norm": 0.11041387235148034,
+      "learning_rate": 0.00012179336155057422,
+      "loss": 1.3657,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4472716788005478,
+      "grad_norm": 0.1232271253869697,
+      "learning_rate": 0.00012177057498727706,
+      "loss": 1.2909,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4473437612628847,
+      "grad_norm": 0.10618367920275866,
+      "learning_rate": 0.00012174778723728884,
+      "loss": 1.3722,
+      "step": 6206
+    },
+    {
+      "epoch": 0.44741584372522164,
+      "grad_norm": 0.10761370763495524,
+      "learning_rate": 0.0001217249983018516,
+      "loss": 1.2935,
+      "step": 6207
+    },
+    {
+      "epoch": 0.44748792618755856,
+      "grad_norm": 0.1286738661411588,
+      "learning_rate": 0.00012170220818220763,
+      "loss": 1.2907,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4475600086498955,
+      "grad_norm": 0.11852642654910528,
+      "learning_rate": 0.00012167941687959914,
+      "loss": 1.311,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4476320911122324,
+      "grad_norm": 0.1420834114238846,
+      "learning_rate": 0.00012165662439526851,
+      "loss": 1.2973,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4477041735745693,
+      "grad_norm": 0.10559051165224981,
+      "learning_rate": 0.00012163383073045806,
+      "loss": 1.2214,
+      "step": 6211
+    },
+    {
+      "epoch": 0.44777625603690624,
+      "grad_norm": 0.12194787987533838,
+      "learning_rate": 0.00012161103588641034,
+      "loss": 1.31,
+      "step": 6212
+    },
+    {
+      "epoch": 0.44784833849924316,
+      "grad_norm": 0.11733603793115455,
+      "learning_rate": 0.00012158823986436778,
+      "loss": 1.2799,
+      "step": 6213
+    },
+    {
+      "epoch": 0.44792042096158,
+      "grad_norm": 0.18218797991883354,
+      "learning_rate": 0.00012156544266557305,
+      "loss": 1.2867,
+      "step": 6214
+    },
+    {
+      "epoch": 0.44799250342391694,
+      "grad_norm": 0.12581554642875592,
+      "learning_rate": 0.00012154264429126874,
+      "loss": 1.2605,
+      "step": 6215
+    },
+    {
+      "epoch": 0.44806458588625386,
+      "grad_norm": 0.11365739298989132,
+      "learning_rate": 0.00012151984474269761,
+      "loss": 1.3297,
+      "step": 6216
+    },
+    {
+      "epoch": 0.4481366683485908,
+      "grad_norm": 0.10470788069641399,
+      "learning_rate": 0.00012149704402110243,
+      "loss": 1.2482,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4482087508109277,
+      "grad_norm": 0.11264383425761285,
+      "learning_rate": 0.000121474242127726,
+      "loss": 1.27,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4482808332732646,
+      "grad_norm": 0.11384371111589146,
+      "learning_rate": 0.0001214514390638113,
+      "loss": 1.2667,
+      "step": 6219
+    },
+    {
+      "epoch": 0.44835291573560154,
+      "grad_norm": 0.11681866589149074,
+      "learning_rate": 0.00012142863483060123,
+      "loss": 1.1896,
+      "step": 6220
+    },
+    {
+      "epoch": 0.44842499819793846,
+      "grad_norm": 0.10926717671698677,
+      "learning_rate": 0.00012140582942933884,
+      "loss": 1.4085,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4484970806602754,
+      "grad_norm": 0.10616496128978532,
+      "learning_rate": 0.00012138302286126728,
+      "loss": 1.2747,
+      "step": 6222
+    },
+    {
+      "epoch": 0.44856916312261225,
+      "grad_norm": 0.13262816375450445,
+      "learning_rate": 0.00012136021512762965,
+      "loss": 1.3056,
+      "step": 6223
+    },
+    {
+      "epoch": 0.44864124558494917,
+      "grad_norm": 0.117100638617904,
+      "learning_rate": 0.00012133740622966923,
+      "loss": 1.1967,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4487133280472861,
+      "grad_norm": 0.1147065062223367,
+      "learning_rate": 0.00012131459616862926,
+      "loss": 1.23,
+      "step": 6225
+    },
+    {
+      "epoch": 0.448785410509623,
+      "grad_norm": 0.11568639736628643,
+      "learning_rate": 0.0001212917849457531,
+      "loss": 1.3234,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4488574929719599,
+      "grad_norm": 0.13247612327103375,
+      "learning_rate": 0.0001212689725622842,
+      "loss": 1.3116,
+      "step": 6227
+    },
+    {
+      "epoch": 0.44892957543429685,
+      "grad_norm": 0.11387268653044266,
+      "learning_rate": 0.00012124615901946601,
+      "loss": 1.2694,
+      "step": 6228
+    },
+    {
+      "epoch": 0.44900165789663377,
+      "grad_norm": 0.11167309843478021,
+      "learning_rate": 0.0001212233443185421,
+      "loss": 1.2395,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4490737403589707,
+      "grad_norm": 0.1244998784907486,
+      "learning_rate": 0.00012120052846075602,
+      "loss": 1.1932,
+      "step": 6230
+    },
+    {
+      "epoch": 0.44914582282130755,
+      "grad_norm": 0.10916254431177005,
+      "learning_rate": 0.0001211777114473515,
+      "loss": 1.2154,
+      "step": 6231
+    },
+    {
+      "epoch": 0.44921790528364447,
+      "grad_norm": 0.11541729519184904,
+      "learning_rate": 0.00012115489327957223,
+      "loss": 1.2778,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4492899877459814,
+      "grad_norm": 0.12380617953492777,
+      "learning_rate": 0.00012113207395866204,
+      "loss": 1.3211,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4493620702083183,
+      "grad_norm": 0.11218909847712409,
+      "learning_rate": 0.00012110925348586475,
+      "loss": 1.2115,
+      "step": 6234
+    },
+    {
+      "epoch": 0.44943415267065523,
+      "grad_norm": 0.11266011625883773,
+      "learning_rate": 0.00012108643186242429,
+      "loss": 1.2039,
+      "step": 6235
+    },
+    {
+      "epoch": 0.44950623513299215,
+      "grad_norm": 0.15031926838331888,
+      "learning_rate": 0.00012106360908958465,
+      "loss": 1.2851,
+      "step": 6236
+    },
+    {
+      "epoch": 0.44957831759532907,
+      "grad_norm": 0.10673010813504204,
+      "learning_rate": 0.0001210407851685899,
+      "loss": 1.2655,
+      "step": 6237
+    },
+    {
+      "epoch": 0.449650400057666,
+      "grad_norm": 0.10744163837515433,
+      "learning_rate": 0.00012101796010068409,
+      "loss": 1.3533,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4497224825200029,
+      "grad_norm": 0.11574807967264422,
+      "learning_rate": 0.00012099513388711146,
+      "loss": 1.2911,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4497945649823398,
+      "grad_norm": 0.10983007254343234,
+      "learning_rate": 0.00012097230652911617,
+      "loss": 1.3503,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4498666474446767,
+      "grad_norm": 0.13293757609146012,
+      "learning_rate": 0.00012094947802794258,
+      "loss": 1.3258,
+      "step": 6241
+    },
+    {
+      "epoch": 0.4499387299070136,
+      "grad_norm": 0.10981305685556575,
+      "learning_rate": 0.00012092664838483499,
+      "loss": 1.1912,
+      "step": 6242
+    },
+    {
+      "epoch": 0.45001081236935053,
+      "grad_norm": 0.1053496326339946,
+      "learning_rate": 0.00012090381760103786,
+      "loss": 1.2925,
+      "step": 6243
+    },
+    {
+      "epoch": 0.45008289483168745,
+      "grad_norm": 0.11502301927458566,
+      "learning_rate": 0.00012088098567779564,
+      "loss": 1.3275,
+      "step": 6244
+    },
+    {
+      "epoch": 0.45015497729402437,
+      "grad_norm": 0.16346180651170986,
+      "learning_rate": 0.00012085815261635293,
+      "loss": 1.2943,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4502270597563613,
+      "grad_norm": 0.14426737879283735,
+      "learning_rate": 0.00012083531841795425,
+      "loss": 1.2602,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4502991422186982,
+      "grad_norm": 0.11966354760977718,
+      "learning_rate": 0.00012081248308384438,
+      "loss": 1.2621,
+      "step": 6247
+    },
+    {
+      "epoch": 0.45037122468103513,
+      "grad_norm": 0.10644514833740698,
+      "learning_rate": 0.00012078964661526795,
+      "loss": 1.2403,
+      "step": 6248
+    },
+    {
+      "epoch": 0.450443307143372,
+      "grad_norm": 0.1143506294373547,
+      "learning_rate": 0.00012076680901346981,
+      "loss": 1.307,
+      "step": 6249
+    },
+    {
+      "epoch": 0.4505153896057089,
+      "grad_norm": 0.1124918937282338,
+      "learning_rate": 0.00012074397027969477,
+      "loss": 1.3259,
+      "step": 6250
+    },
+    {
+      "epoch": 0.45058747206804584,
+      "grad_norm": 0.10841951036148743,
+      "learning_rate": 0.0001207211304151878,
+      "loss": 1.2527,
+      "step": 6251
+    },
+    {
+      "epoch": 0.45065955453038276,
+      "grad_norm": 0.11820066415643023,
+      "learning_rate": 0.00012069828942119381,
+      "loss": 1.3629,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4507316369927197,
+      "grad_norm": 0.11720999503360523,
+      "learning_rate": 0.0001206754472989579,
+      "loss": 1.3623,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4508037194550566,
+      "grad_norm": 0.112127801905998,
+      "learning_rate": 0.00012065260404972514,
+      "loss": 1.2493,
+      "step": 6254
+    },
+    {
+      "epoch": 0.4508758019173935,
+      "grad_norm": 0.10316092088375668,
+      "learning_rate": 0.00012062975967474069,
+      "loss": 1.3238,
+      "step": 6255
+    },
+    {
+      "epoch": 0.45094788437973043,
+      "grad_norm": 0.12904361107423995,
+      "learning_rate": 0.00012060691417524976,
+      "loss": 1.2639,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4510199668420673,
+      "grad_norm": 0.10030206878318672,
+      "learning_rate": 0.00012058406755249769,
+      "loss": 1.2379,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4510920493044042,
+      "grad_norm": 0.10331193132449466,
+      "learning_rate": 0.00012056121980772977,
+      "loss": 1.2802,
+      "step": 6258
+    },
+    {
+      "epoch": 0.45116413176674114,
+      "grad_norm": 0.11508599560687456,
+      "learning_rate": 0.00012053837094219142,
+      "loss": 1.3882,
+      "step": 6259
+    },
+    {
+      "epoch": 0.45123621422907806,
+      "grad_norm": 0.10853276321230915,
+      "learning_rate": 0.00012051552095712812,
+      "loss": 1.3354,
+      "step": 6260
+    },
+    {
+      "epoch": 0.451308296691415,
+      "grad_norm": 0.11335350390419428,
+      "learning_rate": 0.00012049266985378541,
+      "loss": 1.2414,
+      "step": 6261
+    },
+    {
+      "epoch": 0.4513803791537519,
+      "grad_norm": 0.11420167591898431,
+      "learning_rate": 0.00012046981763340882,
+      "loss": 1.2707,
+      "step": 6262
+    },
+    {
+      "epoch": 0.4514524616160888,
+      "grad_norm": 0.11187802028241932,
+      "learning_rate": 0.00012044696429724407,
+      "loss": 1.2827,
+      "step": 6263
+    },
+    {
+      "epoch": 0.45152454407842574,
+      "grad_norm": 0.1244284296620371,
+      "learning_rate": 0.0001204241098465368,
+      "loss": 1.2373,
+      "step": 6264
+    },
+    {
+      "epoch": 0.45159662654076266,
+      "grad_norm": 0.10774649229499247,
+      "learning_rate": 0.00012040125428253288,
+      "loss": 1.2744,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4516687090030995,
+      "grad_norm": 0.11660834219173045,
+      "learning_rate": 0.00012037839760647805,
+      "loss": 1.3412,
+      "step": 6266
+    },
+    {
+      "epoch": 0.45174079146543644,
+      "grad_norm": 0.11385963065991304,
+      "learning_rate": 0.00012035553981961827,
+      "loss": 1.442,
+      "step": 6267
+    },
+    {
+      "epoch": 0.45181287392777336,
+      "grad_norm": 0.11147570730827226,
+      "learning_rate": 0.00012033268092319943,
+      "loss": 1.2048,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4518849563901103,
+      "grad_norm": 0.12038261067237646,
+      "learning_rate": 0.0001203098209184676,
+      "loss": 1.2162,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4519570388524472,
+      "grad_norm": 0.12065611425869718,
+      "learning_rate": 0.00012028695980666882,
+      "loss": 1.3109,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4520291213147841,
+      "grad_norm": 0.11159741375154589,
+      "learning_rate": 0.00012026409758904927,
+      "loss": 1.3725,
+      "step": 6271
+    },
+    {
+      "epoch": 0.45210120377712104,
+      "grad_norm": 0.10074942809513684,
+      "learning_rate": 0.00012024123426685508,
+      "loss": 1.3784,
+      "step": 6272
+    },
+    {
+      "epoch": 0.45217328623945796,
+      "grad_norm": 0.116351813277158,
+      "learning_rate": 0.00012021836984133255,
+      "loss": 1.3491,
+      "step": 6273
+    },
+    {
+      "epoch": 0.4522453687017949,
+      "grad_norm": 0.13552263305471476,
+      "learning_rate": 0.000120195504313728,
+      "loss": 1.2634,
+      "step": 6274
+    },
+    {
+      "epoch": 0.45231745116413175,
+      "grad_norm": 0.11456870928857249,
+      "learning_rate": 0.00012017263768528775,
+      "loss": 1.2874,
+      "step": 6275
+    },
+    {
+      "epoch": 0.45238953362646867,
+      "grad_norm": 0.12016115905409067,
+      "learning_rate": 0.0001201497699572583,
+      "loss": 1.2156,
+      "step": 6276
+    },
+    {
+      "epoch": 0.4524616160888056,
+      "grad_norm": 0.1300937825537205,
+      "learning_rate": 0.00012012690113088613,
+      "loss": 1.3277,
+      "step": 6277
+    },
+    {
+      "epoch": 0.4525336985511425,
+      "grad_norm": 0.1180030921668264,
+      "learning_rate": 0.00012010403120741778,
+      "loss": 1.2472,
+      "step": 6278
+    },
+    {
+      "epoch": 0.4526057810134794,
+      "grad_norm": 0.1058777521661178,
+      "learning_rate": 0.00012008116018809988,
+      "loss": 1.2213,
+      "step": 6279
+    },
+    {
+      "epoch": 0.45267786347581634,
+      "grad_norm": 0.11319434094752877,
+      "learning_rate": 0.00012005828807417909,
+      "loss": 1.2151,
+      "step": 6280
+    },
+    {
+      "epoch": 0.45274994593815326,
+      "grad_norm": 0.11644029027535932,
+      "learning_rate": 0.00012003541486690216,
+      "loss": 1.3678,
+      "step": 6281
+    },
+    {
+      "epoch": 0.4528220284004902,
+      "grad_norm": 0.135666380671819,
+      "learning_rate": 0.00012001254056751587,
+      "loss": 1.2896,
+      "step": 6282
+    },
+    {
+      "epoch": 0.45289411086282705,
+      "grad_norm": 0.20214530300726827,
+      "learning_rate": 0.0001199896651772671,
+      "loss": 1.2344,
+      "step": 6283
+    },
+    {
+      "epoch": 0.45296619332516397,
+      "grad_norm": 0.11204125679866801,
+      "learning_rate": 0.0001199667886974027,
+      "loss": 1.266,
+      "step": 6284
+    },
+    {
+      "epoch": 0.4530382757875009,
+      "grad_norm": 0.10730956432062984,
+      "learning_rate": 0.00011994391112916977,
+      "loss": 1.308,
+      "step": 6285
+    },
+    {
+      "epoch": 0.4531103582498378,
+      "grad_norm": 0.11057996009326553,
+      "learning_rate": 0.00011992103247381516,
+      "loss": 1.4075,
+      "step": 6286
+    },
+    {
+      "epoch": 0.45318244071217473,
+      "grad_norm": 0.10551223028085217,
+      "learning_rate": 0.00011989815273258614,
+      "loss": 1.2512,
+      "step": 6287
+    },
+    {
+      "epoch": 0.45325452317451165,
+      "grad_norm": 0.10957704368618436,
+      "learning_rate": 0.00011987527190672973,
+      "loss": 1.3169,
+      "step": 6288
+    },
+    {
+      "epoch": 0.45332660563684857,
+      "grad_norm": 0.10794660248916897,
+      "learning_rate": 0.00011985238999749325,
+      "loss": 1.2039,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4533986880991855,
+      "grad_norm": 0.108125671185429,
+      "learning_rate": 0.00011982950700612383,
+      "loss": 1.2544,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4534707705615224,
+      "grad_norm": 0.10588193369631008,
+      "learning_rate": 0.00011980662293386896,
+      "loss": 1.2041,
+      "step": 6291
+    },
+    {
+      "epoch": 0.45354285302385927,
+      "grad_norm": 0.10962343663060811,
+      "learning_rate": 0.0001197837377819759,
+      "loss": 1.2581,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4536149354861962,
+      "grad_norm": 0.1620004878972137,
+      "learning_rate": 0.00011976085155169215,
+      "loss": 1.3899,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4536870179485331,
+      "grad_norm": 0.11175016668310074,
+      "learning_rate": 0.0001197379642442652,
+      "loss": 1.2496,
+      "step": 6294
+    },
+    {
+      "epoch": 0.45375910041087003,
+      "grad_norm": 0.112587139657002,
+      "learning_rate": 0.00011971507586094261,
+      "loss": 1.2319,
+      "step": 6295
+    },
+    {
+      "epoch": 0.45383118287320695,
+      "grad_norm": 0.10660754953521301,
+      "learning_rate": 0.00011969218640297204,
+      "loss": 1.3464,
+      "step": 6296
+    },
+    {
+      "epoch": 0.45390326533554387,
+      "grad_norm": 0.11751766310548097,
+      "learning_rate": 0.00011966929587160111,
+      "loss": 1.1587,
+      "step": 6297
+    },
+    {
+      "epoch": 0.4539753477978808,
+      "grad_norm": 0.10264304653698171,
+      "learning_rate": 0.0001196464042680776,
+      "loss": 1.2603,
+      "step": 6298
+    },
+    {
+      "epoch": 0.4540474302602177,
+      "grad_norm": 0.23250924050289382,
+      "learning_rate": 0.00011962351159364931,
+      "loss": 1.3191,
+      "step": 6299
+    },
+    {
+      "epoch": 0.45411951272255463,
+      "grad_norm": 0.10827713738357429,
+      "learning_rate": 0.00011960061784956404,
+      "loss": 1.3625,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4541915951848915,
+      "grad_norm": 0.10636047225513107,
+      "learning_rate": 0.00011957772303706978,
+      "loss": 1.251,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4542636776472284,
+      "grad_norm": 0.09900988383542932,
+      "learning_rate": 0.00011955482715741446,
+      "loss": 1.2786,
+      "step": 6302
+    },
+    {
+      "epoch": 0.45433576010956533,
+      "grad_norm": 0.12160024031538758,
+      "learning_rate": 0.00011953193021184613,
+      "loss": 1.1897,
+      "step": 6303
+    },
+    {
+      "epoch": 0.45440784257190225,
+      "grad_norm": 0.10344296510105436,
+      "learning_rate": 0.00011950903220161285,
+      "loss": 1.2314,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4544799250342392,
+      "grad_norm": 0.1360584863558019,
+      "learning_rate": 0.00011948613312796277,
+      "loss": 1.2704,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4545520074965761,
+      "grad_norm": 0.11818766342709913,
+      "learning_rate": 0.00011946323299214412,
+      "loss": 1.3209,
+      "step": 6306
+    },
+    {
+      "epoch": 0.454624089958913,
+      "grad_norm": 0.12092876786320089,
+      "learning_rate": 0.00011944033179540515,
+      "loss": 1.3418,
+      "step": 6307
+    },
+    {
+      "epoch": 0.45469617242124993,
+      "grad_norm": 0.10167502719453904,
+      "learning_rate": 0.00011941742953899417,
+      "loss": 1.3258,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4547682548835868,
+      "grad_norm": 0.09876293001679952,
+      "learning_rate": 0.00011939452622415959,
+      "loss": 1.2799,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4548403373459237,
+      "grad_norm": 0.1137847981604281,
+      "learning_rate": 0.0001193716218521498,
+      "loss": 1.3153,
+      "step": 6310
+    },
+    {
+      "epoch": 0.45491241980826064,
+      "grad_norm": 0.12232478883554639,
+      "learning_rate": 0.00011934871642421333,
+      "loss": 1.2908,
+      "step": 6311
+    },
+    {
+      "epoch": 0.45498450227059756,
+      "grad_norm": 0.11896772128323718,
+      "learning_rate": 0.0001193258099415987,
+      "loss": 1.3539,
+      "step": 6312
+    },
+    {
+      "epoch": 0.4550565847329345,
+      "grad_norm": 0.10627655576066967,
+      "learning_rate": 0.00011930290240555456,
+      "loss": 1.2587,
+      "step": 6313
+    },
+    {
+      "epoch": 0.4551286671952714,
+      "grad_norm": 0.12075731878155815,
+      "learning_rate": 0.0001192799938173295,
+      "loss": 1.2779,
+      "step": 6314
+    },
+    {
+      "epoch": 0.4552007496576083,
+      "grad_norm": 0.1096785418244709,
+      "learning_rate": 0.00011925708417817234,
+      "loss": 1.2782,
+      "step": 6315
+    },
+    {
+      "epoch": 0.45527283211994524,
+      "grad_norm": 0.0998924985653576,
+      "learning_rate": 0.00011923417348933183,
+      "loss": 1.234,
+      "step": 6316
+    },
+    {
+      "epoch": 0.45534491458228216,
+      "grad_norm": 0.12126539808425317,
+      "learning_rate": 0.00011921126175205675,
+      "loss": 1.3479,
+      "step": 6317
+    },
+    {
+      "epoch": 0.455416997044619,
+      "grad_norm": 0.1125978209338988,
+      "learning_rate": 0.00011918834896759606,
+      "loss": 1.4337,
+      "step": 6318
+    },
+    {
+      "epoch": 0.45548907950695594,
+      "grad_norm": 0.1089319145279081,
+      "learning_rate": 0.00011916543513719864,
+      "loss": 1.2337,
+      "step": 6319
+    },
+    {
+      "epoch": 0.45556116196929286,
+      "grad_norm": 0.11718650252265311,
+      "learning_rate": 0.00011914252026211361,
+      "loss": 1.3513,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4556332444316298,
+      "grad_norm": 0.12882137094455498,
+      "learning_rate": 0.00011911960434358991,
+      "loss": 1.2971,
+      "step": 6321
+    },
+    {
+      "epoch": 0.4557053268939667,
+      "grad_norm": 0.1258027878930256,
+      "learning_rate": 0.00011909668738287679,
+      "loss": 1.4175,
+      "step": 6322
+    },
+    {
+      "epoch": 0.4557774093563036,
+      "grad_norm": 0.12135326314994901,
+      "learning_rate": 0.0001190737693812233,
+      "loss": 1.3245,
+      "step": 6323
+    },
+    {
+      "epoch": 0.45584949181864054,
+      "grad_norm": 0.12154203485602146,
+      "learning_rate": 0.0001190508503398788,
+      "loss": 1.2761,
+      "step": 6324
+    },
+    {
+      "epoch": 0.45592157428097746,
+      "grad_norm": 0.11041929101192825,
+      "learning_rate": 0.00011902793026009247,
+      "loss": 1.2998,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4559936567433143,
+      "grad_norm": 0.12585164933638743,
+      "learning_rate": 0.00011900500914311376,
+      "loss": 1.2713,
+      "step": 6326
+    },
+    {
+      "epoch": 0.45606573920565124,
+      "grad_norm": 0.11083753234775774,
+      "learning_rate": 0.000118982086990192,
+      "loss": 1.282,
+      "step": 6327
+    },
+    {
+      "epoch": 0.45613782166798816,
+      "grad_norm": 0.10204749702286364,
+      "learning_rate": 0.00011895916380257668,
+      "loss": 1.2922,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4562099041303251,
+      "grad_norm": 0.10173866185533695,
+      "learning_rate": 0.00011893623958151732,
+      "loss": 1.3409,
+      "step": 6329
+    },
+    {
+      "epoch": 0.456281986592662,
+      "grad_norm": 0.10826188401608597,
+      "learning_rate": 0.00011891331432826351,
+      "loss": 1.2465,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4563540690549989,
+      "grad_norm": 0.11810052983269038,
+      "learning_rate": 0.00011889038804406485,
+      "loss": 1.2668,
+      "step": 6331
+    },
+    {
+      "epoch": 0.45642615151733584,
+      "grad_norm": 0.11107800662230059,
+      "learning_rate": 0.00011886746073017105,
+      "loss": 1.316,
+      "step": 6332
+    },
+    {
+      "epoch": 0.45649823397967276,
+      "grad_norm": 0.11712267473233848,
+      "learning_rate": 0.00011884453238783185,
+      "loss": 1.2537,
+      "step": 6333
+    },
+    {
+      "epoch": 0.4565703164420097,
+      "grad_norm": 0.2935700307771365,
+      "learning_rate": 0.00011882160301829703,
+      "loss": 1.2126,
+      "step": 6334
+    },
+    {
+      "epoch": 0.45664239890434655,
+      "grad_norm": 0.14273255485706624,
+      "learning_rate": 0.00011879867262281648,
+      "loss": 1.2624,
+      "step": 6335
+    },
+    {
+      "epoch": 0.45671448136668347,
+      "grad_norm": 0.11431166056447542,
+      "learning_rate": 0.00011877574120264008,
+      "loss": 1.2465,
+      "step": 6336
+    },
+    {
+      "epoch": 0.4567865638290204,
+      "grad_norm": 0.11142533360112059,
+      "learning_rate": 0.00011875280875901782,
+      "loss": 1.2923,
+      "step": 6337
+    },
+    {
+      "epoch": 0.4568586462913573,
+      "grad_norm": 0.1159379312551039,
+      "learning_rate": 0.00011872987529319974,
+      "loss": 1.3752,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4569307287536942,
+      "grad_norm": 0.11106184230281022,
+      "learning_rate": 0.00011870694080643585,
+      "loss": 1.2434,
+      "step": 6339
+    },
+    {
+      "epoch": 0.45700281121603115,
+      "grad_norm": 0.11999579472835126,
+      "learning_rate": 0.00011868400529997636,
+      "loss": 1.2183,
+      "step": 6340
+    },
+    {
+      "epoch": 0.45707489367836807,
+      "grad_norm": 0.12081117364863095,
+      "learning_rate": 0.0001186610687750714,
+      "loss": 1.2384,
+      "step": 6341
+    },
+    {
+      "epoch": 0.457146976140705,
+      "grad_norm": 0.11993622639313149,
+      "learning_rate": 0.00011863813123297128,
+      "loss": 1.3192,
+      "step": 6342
+    },
+    {
+      "epoch": 0.4572190586030419,
+      "grad_norm": 0.11678336210568285,
+      "learning_rate": 0.00011861519267492623,
+      "loss": 1.1642,
+      "step": 6343
+    },
+    {
+      "epoch": 0.45729114106537877,
+      "grad_norm": 0.11522399516645977,
+      "learning_rate": 0.00011859225310218666,
+      "loss": 1.3312,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4573632235277157,
+      "grad_norm": 0.121030797751944,
+      "learning_rate": 0.00011856931251600297,
+      "loss": 1.2982,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4574353059900526,
+      "grad_norm": 0.10095669511775239,
+      "learning_rate": 0.00011854637091762563,
+      "loss": 1.2721,
+      "step": 6346
+    },
+    {
+      "epoch": 0.45750738845238953,
+      "grad_norm": 0.11652173249350654,
+      "learning_rate": 0.00011852342830830511,
+      "loss": 1.2794,
+      "step": 6347
+    },
+    {
+      "epoch": 0.45757947091472645,
+      "grad_norm": 0.12646796089424744,
+      "learning_rate": 0.00011850048468929209,
+      "loss": 1.3405,
+      "step": 6348
+    },
+    {
+      "epoch": 0.45765155337706337,
+      "grad_norm": 0.11805797733400293,
+      "learning_rate": 0.00011847754006183713,
+      "loss": 1.2928,
+      "step": 6349
+    },
+    {
+      "epoch": 0.4577236358394003,
+      "grad_norm": 0.11344494597837224,
+      "learning_rate": 0.00011845459442719092,
+      "loss": 1.2148,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4577957183017372,
+      "grad_norm": 0.12413885205485269,
+      "learning_rate": 0.0001184316477866042,
+      "loss": 1.2624,
+      "step": 6351
+    },
+    {
+      "epoch": 0.4578678007640741,
+      "grad_norm": 0.1096288420048504,
+      "learning_rate": 0.00011840870014132782,
+      "loss": 1.285,
+      "step": 6352
+    },
+    {
+      "epoch": 0.457939883226411,
+      "grad_norm": 0.11539673621227144,
+      "learning_rate": 0.00011838575149261255,
+      "loss": 1.264,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4580119656887479,
+      "grad_norm": 0.12675309563267986,
+      "learning_rate": 0.00011836280184170938,
+      "loss": 1.292,
+      "step": 6354
+    },
+    {
+      "epoch": 0.45808404815108483,
+      "grad_norm": 0.11267634688552802,
+      "learning_rate": 0.00011833985118986919,
+      "loss": 1.3475,
+      "step": 6355
+    },
+    {
+      "epoch": 0.45815613061342175,
+      "grad_norm": 0.12746554999140675,
+      "learning_rate": 0.00011831689953834306,
+      "loss": 1.2663,
+      "step": 6356
+    },
+    {
+      "epoch": 0.4582282130757587,
+      "grad_norm": 0.14402609523041981,
+      "learning_rate": 0.00011829394688838205,
+      "loss": 1.1895,
+      "step": 6357
+    },
+    {
+      "epoch": 0.4583002955380956,
+      "grad_norm": 0.11661132521569158,
+      "learning_rate": 0.00011827099324123725,
+      "loss": 1.3593,
+      "step": 6358
+    },
+    {
+      "epoch": 0.4583723780004325,
+      "grad_norm": 0.10924170034164744,
+      "learning_rate": 0.00011824803859815988,
+      "loss": 1.2514,
+      "step": 6359
+    },
+    {
+      "epoch": 0.45844446046276943,
+      "grad_norm": 0.10972364921999637,
+      "learning_rate": 0.00011822508296040116,
+      "loss": 1.3529,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4585165429251063,
+      "grad_norm": 0.10751943035807042,
+      "learning_rate": 0.00011820212632921233,
+      "loss": 1.2364,
+      "step": 6361
+    },
+    {
+      "epoch": 0.4585886253874432,
+      "grad_norm": 0.10927991045756887,
+      "learning_rate": 0.00011817916870584482,
+      "loss": 1.3141,
+      "step": 6362
+    },
+    {
+      "epoch": 0.45866070784978014,
+      "grad_norm": 0.11419690269455746,
+      "learning_rate": 0.00011815621009154994,
+      "loss": 1.2906,
+      "step": 6363
+    },
+    {
+      "epoch": 0.45873279031211706,
+      "grad_norm": 0.12497875899148445,
+      "learning_rate": 0.00011813325048757922,
+      "loss": 1.2757,
+      "step": 6364
+    },
+    {
+      "epoch": 0.458804872774454,
+      "grad_norm": 0.12262561654682452,
+      "learning_rate": 0.00011811028989518409,
+      "loss": 1.2427,
+      "step": 6365
+    },
+    {
+      "epoch": 0.4588769552367909,
+      "grad_norm": 0.10258864118632048,
+      "learning_rate": 0.00011808732831561617,
+      "loss": 1.2969,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4589490376991278,
+      "grad_norm": 0.11107848274369195,
+      "learning_rate": 0.000118064365750127,
+      "loss": 1.1987,
+      "step": 6367
+    },
+    {
+      "epoch": 0.45902112016146474,
+      "grad_norm": 0.11624976971703624,
+      "learning_rate": 0.00011804140219996835,
+      "loss": 1.2167,
+      "step": 6368
+    },
+    {
+      "epoch": 0.45909320262380166,
+      "grad_norm": 0.10198319556425335,
+      "learning_rate": 0.00011801843766639182,
+      "loss": 1.2219,
+      "step": 6369
+    },
+    {
+      "epoch": 0.4591652850861385,
+      "grad_norm": 0.12008465251424066,
+      "learning_rate": 0.00011799547215064928,
+      "loss": 1.42,
+      "step": 6370
+    },
+    {
+      "epoch": 0.45923736754847544,
+      "grad_norm": 0.11673851822573282,
+      "learning_rate": 0.00011797250565399252,
+      "loss": 1.4312,
+      "step": 6371
+    },
+    {
+      "epoch": 0.45930945001081236,
+      "grad_norm": 0.1268255888565791,
+      "learning_rate": 0.00011794953817767338,
+      "loss": 1.2639,
+      "step": 6372
+    },
+    {
+      "epoch": 0.4593815324731493,
+      "grad_norm": 0.11271288085460293,
+      "learning_rate": 0.00011792656972294385,
+      "loss": 1.2341,
+      "step": 6373
+    },
+    {
+      "epoch": 0.4594536149354862,
+      "grad_norm": 0.11065960322968779,
+      "learning_rate": 0.00011790360029105588,
+      "loss": 1.279,
+      "step": 6374
+    },
+    {
+      "epoch": 0.4595256973978231,
+      "grad_norm": 0.14375698429603526,
+      "learning_rate": 0.00011788062988326155,
+      "loss": 1.3252,
+      "step": 6375
+    },
+    {
+      "epoch": 0.45959777986016004,
+      "grad_norm": 0.22930718720627696,
+      "learning_rate": 0.0001178576585008129,
+      "loss": 1.3183,
+      "step": 6376
+    },
+    {
+      "epoch": 0.45966986232249696,
+      "grad_norm": 0.1161007884510719,
+      "learning_rate": 0.00011783468614496209,
+      "loss": 1.2729,
+      "step": 6377
+    },
+    {
+      "epoch": 0.4597419447848338,
+      "grad_norm": 0.1193892699870104,
+      "learning_rate": 0.00011781171281696135,
+      "loss": 1.3319,
+      "step": 6378
+    },
+    {
+      "epoch": 0.45981402724717074,
+      "grad_norm": 0.11620418724763544,
+      "learning_rate": 0.00011778873851806294,
+      "loss": 1.328,
+      "step": 6379
+    },
+    {
+      "epoch": 0.45988610970950766,
+      "grad_norm": 0.10897419544525498,
+      "learning_rate": 0.0001177657632495191,
+      "loss": 1.3396,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4599581921718446,
+      "grad_norm": 0.10758148573351571,
+      "learning_rate": 0.00011774278701258223,
+      "loss": 1.2557,
+      "step": 6381
+    },
+    {
+      "epoch": 0.4600302746341815,
+      "grad_norm": 0.11632123535146997,
+      "learning_rate": 0.00011771980980850475,
+      "loss": 1.4534,
+      "step": 6382
+    },
+    {
+      "epoch": 0.4601023570965184,
+      "grad_norm": 0.10411263937552012,
+      "learning_rate": 0.0001176968316385391,
+      "loss": 1.3049,
+      "step": 6383
+    },
+    {
+      "epoch": 0.46017443955885534,
+      "grad_norm": 0.24609620097484364,
+      "learning_rate": 0.00011767385250393781,
+      "loss": 1.2984,
+      "step": 6384
+    },
+    {
+      "epoch": 0.46024652202119226,
+      "grad_norm": 0.11653009029840669,
+      "learning_rate": 0.00011765087240595345,
+      "loss": 1.3582,
+      "step": 6385
+    },
+    {
+      "epoch": 0.4603186044835292,
+      "grad_norm": 0.11396082357405397,
+      "learning_rate": 0.00011762789134583863,
+      "loss": 1.321,
+      "step": 6386
+    },
+    {
+      "epoch": 0.46039068694586605,
+      "grad_norm": 0.1527995572032769,
+      "learning_rate": 0.00011760490932484602,
+      "loss": 1.2784,
+      "step": 6387
+    },
+    {
+      "epoch": 0.46046276940820297,
+      "grad_norm": 0.12158524369578265,
+      "learning_rate": 0.00011758192634422837,
+      "loss": 1.2868,
+      "step": 6388
+    },
+    {
+      "epoch": 0.4605348518705399,
+      "grad_norm": 0.11547010936127863,
+      "learning_rate": 0.00011755894240523844,
+      "loss": 1.3018,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4606069343328768,
+      "grad_norm": 0.12091803590579307,
+      "learning_rate": 0.00011753595750912905,
+      "loss": 1.2581,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4606790167952137,
+      "grad_norm": 0.12153316738024923,
+      "learning_rate": 0.00011751297165715309,
+      "loss": 1.3046,
+      "step": 6391
+    },
+    {
+      "epoch": 0.46075109925755064,
+      "grad_norm": 0.10109789684196466,
+      "learning_rate": 0.0001174899848505635,
+      "loss": 1.2153,
+      "step": 6392
+    },
+    {
+      "epoch": 0.46082318171988756,
+      "grad_norm": 0.12027638858201402,
+      "learning_rate": 0.00011746699709061331,
+      "loss": 1.3439,
+      "step": 6393
+    },
+    {
+      "epoch": 0.4608952641822245,
+      "grad_norm": 0.11262954244878595,
+      "learning_rate": 0.00011744400837855545,
+      "loss": 1.3042,
+      "step": 6394
+    },
+    {
+      "epoch": 0.4609673466445614,
+      "grad_norm": 0.11708309876263591,
+      "learning_rate": 0.00011742101871564313,
+      "loss": 1.3243,
+      "step": 6395
+    },
+    {
+      "epoch": 0.46103942910689827,
+      "grad_norm": 0.10313446098595672,
+      "learning_rate": 0.00011739802810312939,
+      "loss": 1.2838,
+      "step": 6396
+    },
+    {
+      "epoch": 0.4611115115692352,
+      "grad_norm": 0.12332463343722518,
+      "learning_rate": 0.00011737503654226751,
+      "loss": 1.3534,
+      "step": 6397
+    },
+    {
+      "epoch": 0.4611835940315721,
+      "grad_norm": 0.134614751117181,
+      "learning_rate": 0.00011735204403431066,
+      "loss": 1.2422,
+      "step": 6398
+    },
+    {
+      "epoch": 0.46125567649390903,
+      "grad_norm": 0.1234225680019613,
+      "learning_rate": 0.0001173290505805122,
+      "loss": 1.3061,
+      "step": 6399
+    },
+    {
+      "epoch": 0.46132775895624595,
+      "grad_norm": 0.12707362452267273,
+      "learning_rate": 0.00011730605618212544,
+      "loss": 1.2076,
+      "step": 6400
+    },
+    {
+      "epoch": 0.46139984141858287,
+      "grad_norm": 0.13162406723439937,
+      "learning_rate": 0.00011728306084040381,
+      "loss": 1.2557,
+      "step": 6401
+    },
+    {
+      "epoch": 0.4614719238809198,
+      "grad_norm": 0.1262229670436941,
+      "learning_rate": 0.00011726006455660072,
+      "loss": 1.3149,
+      "step": 6402
+    },
+    {
+      "epoch": 0.4615440063432567,
+      "grad_norm": 0.11821489290535485,
+      "learning_rate": 0.00011723706733196973,
+      "loss": 1.2341,
+      "step": 6403
+    },
+    {
+      "epoch": 0.46161608880559357,
+      "grad_norm": 0.1165121628156832,
+      "learning_rate": 0.00011721406916776434,
+      "loss": 1.2743,
+      "step": 6404
+    },
+    {
+      "epoch": 0.4616881712679305,
+      "grad_norm": 0.12152394988027565,
+      "learning_rate": 0.0001171910700652382,
+      "loss": 1.2464,
+      "step": 6405
+    },
+    {
+      "epoch": 0.4617602537302674,
+      "grad_norm": 0.11631721740992997,
+      "learning_rate": 0.00011716807002564491,
+      "loss": 1.3039,
+      "step": 6406
+    },
+    {
+      "epoch": 0.46183233619260433,
+      "grad_norm": 0.11614809878125501,
+      "learning_rate": 0.00011714506905023826,
+      "loss": 1.3091,
+      "step": 6407
+    },
+    {
+      "epoch": 0.46190441865494125,
+      "grad_norm": 0.10953083955287068,
+      "learning_rate": 0.00011712206714027192,
+      "loss": 1.302,
+      "step": 6408
+    },
+    {
+      "epoch": 0.46197650111727817,
+      "grad_norm": 0.11759845227758693,
+      "learning_rate": 0.00011709906429699977,
+      "loss": 1.2843,
+      "step": 6409
+    },
+    {
+      "epoch": 0.4620485835796151,
+      "grad_norm": 0.11557290284520418,
+      "learning_rate": 0.00011707606052167563,
+      "loss": 1.3341,
+      "step": 6410
+    },
+    {
+      "epoch": 0.462120666041952,
+      "grad_norm": 0.12175912327976177,
+      "learning_rate": 0.00011705305581555344,
+      "loss": 1.3549,
+      "step": 6411
+    },
+    {
+      "epoch": 0.46219274850428893,
+      "grad_norm": 0.12176008556319953,
+      "learning_rate": 0.00011703005017988712,
+      "loss": 1.3004,
+      "step": 6412
+    },
+    {
+      "epoch": 0.4622648309666258,
+      "grad_norm": 0.1119792085211442,
+      "learning_rate": 0.00011700704361593073,
+      "loss": 1.2669,
+      "step": 6413
+    },
+    {
+      "epoch": 0.4623369134289627,
+      "grad_norm": 0.13243837087233506,
+      "learning_rate": 0.00011698403612493832,
+      "loss": 1.3384,
+      "step": 6414
+    },
+    {
+      "epoch": 0.46240899589129963,
+      "grad_norm": 0.10857528782747108,
+      "learning_rate": 0.00011696102770816399,
+      "loss": 1.2365,
+      "step": 6415
+    },
+    {
+      "epoch": 0.46248107835363655,
+      "grad_norm": 0.10878508985155999,
+      "learning_rate": 0.00011693801836686189,
+      "loss": 1.3299,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4625531608159735,
+      "grad_norm": 0.11416528080089615,
+      "learning_rate": 0.0001169150081022863,
+      "loss": 1.2468,
+      "step": 6417
+    },
+    {
+      "epoch": 0.4626252432783104,
+      "grad_norm": 0.2639050609090086,
+      "learning_rate": 0.0001168919969156914,
+      "loss": 1.3429,
+      "step": 6418
+    },
+    {
+      "epoch": 0.4626973257406473,
+      "grad_norm": 0.1261378580334504,
+      "learning_rate": 0.0001168689848083316,
+      "loss": 1.2487,
+      "step": 6419
+    },
+    {
+      "epoch": 0.46276940820298423,
+      "grad_norm": 0.10713307308215285,
+      "learning_rate": 0.00011684597178146115,
+      "loss": 1.3172,
+      "step": 6420
+    },
+    {
+      "epoch": 0.46284149066532115,
+      "grad_norm": 0.11373113495921679,
+      "learning_rate": 0.00011682295783633457,
+      "loss": 1.2997,
+      "step": 6421
+    },
+    {
+      "epoch": 0.462913573127658,
+      "grad_norm": 0.15731958627458148,
+      "learning_rate": 0.00011679994297420627,
+      "loss": 1.3302,
+      "step": 6422
+    },
+    {
+      "epoch": 0.46298565558999494,
+      "grad_norm": 0.11385562282070424,
+      "learning_rate": 0.0001167769271963308,
+      "loss": 1.2805,
+      "step": 6423
+    },
+    {
+      "epoch": 0.46305773805233186,
+      "grad_norm": 0.10989266139759926,
+      "learning_rate": 0.00011675391050396267,
+      "loss": 1.3218,
+      "step": 6424
+    },
+    {
+      "epoch": 0.4631298205146688,
+      "grad_norm": 0.15357616119270998,
+      "learning_rate": 0.00011673089289835657,
+      "loss": 1.3051,
+      "step": 6425
+    },
+    {
+      "epoch": 0.4632019029770057,
+      "grad_norm": 0.12886191757162022,
+      "learning_rate": 0.00011670787438076711,
+      "loss": 1.3445,
+      "step": 6426
+    },
+    {
+      "epoch": 0.4632739854393426,
+      "grad_norm": 0.12113350110293364,
+      "learning_rate": 0.00011668485495244903,
+      "loss": 1.3366,
+      "step": 6427
+    },
+    {
+      "epoch": 0.46334606790167954,
+      "grad_norm": 0.11844883544069147,
+      "learning_rate": 0.00011666183461465709,
+      "loss": 1.1698,
+      "step": 6428
+    },
+    {
+      "epoch": 0.46341815036401646,
+      "grad_norm": 0.1148805963488104,
+      "learning_rate": 0.00011663881336864611,
+      "loss": 1.2461,
+      "step": 6429
+    },
+    {
+      "epoch": 0.4634902328263533,
+      "grad_norm": 0.1123960828199724,
+      "learning_rate": 0.00011661579121567092,
+      "loss": 1.3278,
+      "step": 6430
+    },
+    {
+      "epoch": 0.46356231528869024,
+      "grad_norm": 0.11370452361120477,
+      "learning_rate": 0.00011659276815698648,
+      "loss": 1.2931,
+      "step": 6431
+    },
+    {
+      "epoch": 0.46363439775102716,
+      "grad_norm": 0.10973410912948611,
+      "learning_rate": 0.00011656974419384771,
+      "loss": 1.2922,
+      "step": 6432
+    },
+    {
+      "epoch": 0.4637064802133641,
+      "grad_norm": 0.1238693319547951,
+      "learning_rate": 0.00011654671932750969,
+      "loss": 1.3648,
+      "step": 6433
+    },
+    {
+      "epoch": 0.463778562675701,
+      "grad_norm": 0.11453589516634798,
+      "learning_rate": 0.00011652369355922739,
+      "loss": 1.3243,
+      "step": 6434
+    },
+    {
+      "epoch": 0.4638506451380379,
+      "grad_norm": 0.1329808201947811,
+      "learning_rate": 0.00011650066689025599,
+      "loss": 1.2404,
+      "step": 6435
+    },
+    {
+      "epoch": 0.46392272760037484,
+      "grad_norm": 0.11170961262342344,
+      "learning_rate": 0.0001164776393218506,
+      "loss": 1.3035,
+      "step": 6436
+    },
+    {
+      "epoch": 0.46399481006271176,
+      "grad_norm": 0.12806298929505883,
+      "learning_rate": 0.0001164546108552665,
+      "loss": 1.2879,
+      "step": 6437
+    },
+    {
+      "epoch": 0.4640668925250487,
+      "grad_norm": 0.26218331966780917,
+      "learning_rate": 0.00011643158149175886,
+      "loss": 1.2315,
+      "step": 6438
+    },
+    {
+      "epoch": 0.46413897498738554,
+      "grad_norm": 0.11763240009974474,
+      "learning_rate": 0.00011640855123258306,
+      "loss": 1.2448,
+      "step": 6439
+    },
+    {
+      "epoch": 0.46421105744972246,
+      "grad_norm": 0.2370265182201456,
+      "learning_rate": 0.00011638552007899437,
+      "loss": 1.2345,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4642831399120594,
+      "grad_norm": 0.11601862030081322,
+      "learning_rate": 0.00011636248803224832,
+      "loss": 1.2961,
+      "step": 6441
+    },
+    {
+      "epoch": 0.4643552223743963,
+      "grad_norm": 0.11764539859209766,
+      "learning_rate": 0.00011633945509360023,
+      "loss": 1.3405,
+      "step": 6442
+    },
+    {
+      "epoch": 0.4644273048367332,
+      "grad_norm": 0.12104054828431378,
+      "learning_rate": 0.00011631642126430568,
+      "loss": 1.2978,
+      "step": 6443
+    },
+    {
+      "epoch": 0.46449938729907014,
+      "grad_norm": 0.11670010220567054,
+      "learning_rate": 0.00011629338654562018,
+      "loss": 1.286,
+      "step": 6444
+    },
+    {
+      "epoch": 0.46457146976140706,
+      "grad_norm": 0.11447121181475796,
+      "learning_rate": 0.00011627035093879939,
+      "loss": 1.2732,
+      "step": 6445
+    },
+    {
+      "epoch": 0.464643552223744,
+      "grad_norm": 0.12398313827639287,
+      "learning_rate": 0.0001162473144450989,
+      "loss": 1.3964,
+      "step": 6446
+    },
+    {
+      "epoch": 0.4647156346860809,
+      "grad_norm": 0.12012681133245492,
+      "learning_rate": 0.00011622427706577441,
+      "loss": 1.3306,
+      "step": 6447
+    },
+    {
+      "epoch": 0.46478771714841777,
+      "grad_norm": 0.11501191051682401,
+      "learning_rate": 0.00011620123880208167,
+      "loss": 1.2807,
+      "step": 6448
+    },
+    {
+      "epoch": 0.4648597996107547,
+      "grad_norm": 0.12032741782807024,
+      "learning_rate": 0.0001161781996552765,
+      "loss": 1.3929,
+      "step": 6449
+    },
+    {
+      "epoch": 0.4649318820730916,
+      "grad_norm": 0.10889371086667705,
+      "learning_rate": 0.00011615515962661469,
+      "loss": 1.3067,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4650039645354285,
+      "grad_norm": 0.11652214068758367,
+      "learning_rate": 0.00011613211871735214,
+      "loss": 1.1413,
+      "step": 6451
+    },
+    {
+      "epoch": 0.46507604699776545,
+      "grad_norm": 0.11480954771152317,
+      "learning_rate": 0.00011610907692874482,
+      "loss": 1.2993,
+      "step": 6452
+    },
+    {
+      "epoch": 0.46514812946010237,
+      "grad_norm": 0.1082511676058477,
+      "learning_rate": 0.0001160860342620487,
+      "loss": 1.3179,
+      "step": 6453
+    },
+    {
+      "epoch": 0.4652202119224393,
+      "grad_norm": 0.10777397633916425,
+      "learning_rate": 0.00011606299071851978,
+      "loss": 1.295,
+      "step": 6454
+    },
+    {
+      "epoch": 0.4652922943847762,
+      "grad_norm": 0.11374364458617493,
+      "learning_rate": 0.00011603994629941417,
+      "loss": 1.3512,
+      "step": 6455
+    },
+    {
+      "epoch": 0.46536437684711307,
+      "grad_norm": 0.11530603420209237,
+      "learning_rate": 0.00011601690100598801,
+      "loss": 1.3791,
+      "step": 6456
+    },
+    {
+      "epoch": 0.46543645930945,
+      "grad_norm": 0.11879921963103994,
+      "learning_rate": 0.00011599385483949743,
+      "loss": 1.3011,
+      "step": 6457
+    },
+    {
+      "epoch": 0.4655085417717869,
+      "grad_norm": 0.1120374936760742,
+      "learning_rate": 0.00011597080780119868,
+      "loss": 1.2684,
+      "step": 6458
+    },
+    {
+      "epoch": 0.46558062423412383,
+      "grad_norm": 0.10613128190519387,
+      "learning_rate": 0.00011594775989234803,
+      "loss": 1.299,
+      "step": 6459
+    },
+    {
+      "epoch": 0.46565270669646075,
+      "grad_norm": 0.11579179196387689,
+      "learning_rate": 0.00011592471111420178,
+      "loss": 1.274,
+      "step": 6460
+    },
+    {
+      "epoch": 0.46572478915879767,
+      "grad_norm": 0.10637021862392047,
+      "learning_rate": 0.00011590166146801633,
+      "loss": 1.2331,
+      "step": 6461
+    },
+    {
+      "epoch": 0.4657968716211346,
+      "grad_norm": 0.25614186492200675,
+      "learning_rate": 0.00011587861095504805,
+      "loss": 1.2836,
+      "step": 6462
+    },
+    {
+      "epoch": 0.4658689540834715,
+      "grad_norm": 0.11698921875621124,
+      "learning_rate": 0.00011585555957655342,
+      "loss": 1.2605,
+      "step": 6463
+    },
+    {
+      "epoch": 0.46594103654580843,
+      "grad_norm": 0.11631706274148662,
+      "learning_rate": 0.00011583250733378896,
+      "loss": 1.2609,
+      "step": 6464
+    },
+    {
+      "epoch": 0.4660131190081453,
+      "grad_norm": 0.11360766432266374,
+      "learning_rate": 0.00011580945422801119,
+      "loss": 1.2843,
+      "step": 6465
+    },
+    {
+      "epoch": 0.4660852014704822,
+      "grad_norm": 0.11148632082206832,
+      "learning_rate": 0.00011578640026047676,
+      "loss": 1.3101,
+      "step": 6466
+    },
+    {
+      "epoch": 0.46615728393281913,
+      "grad_norm": 0.117467587146344,
+      "learning_rate": 0.00011576334543244225,
+      "loss": 1.2894,
+      "step": 6467
+    },
+    {
+      "epoch": 0.46622936639515605,
+      "grad_norm": 0.12006262721208269,
+      "learning_rate": 0.00011574028974516444,
+      "loss": 1.3721,
+      "step": 6468
+    },
+    {
+      "epoch": 0.466301448857493,
+      "grad_norm": 0.12387097234849258,
+      "learning_rate": 0.00011571723319989996,
+      "loss": 1.2267,
+      "step": 6469
+    },
+    {
+      "epoch": 0.4663735313198299,
+      "grad_norm": 0.11714886378544177,
+      "learning_rate": 0.00011569417579790571,
+      "loss": 1.2822,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4664456137821668,
+      "grad_norm": 0.12867238655438837,
+      "learning_rate": 0.00011567111754043844,
+      "loss": 1.2488,
+      "step": 6471
+    },
+    {
+      "epoch": 0.46651769624450373,
+      "grad_norm": 0.1070697991915659,
+      "learning_rate": 0.00011564805842875511,
+      "loss": 1.4446,
+      "step": 6472
+    },
+    {
+      "epoch": 0.46658977870684065,
+      "grad_norm": 0.12318796165478549,
+      "learning_rate": 0.00011562499846411254,
+      "loss": 1.16,
+      "step": 6473
+    },
+    {
+      "epoch": 0.4666618611691775,
+      "grad_norm": 0.11793604635497831,
+      "learning_rate": 0.00011560193764776783,
+      "loss": 1.3839,
+      "step": 6474
+    },
+    {
+      "epoch": 0.46673394363151444,
+      "grad_norm": 0.11885810704041225,
+      "learning_rate": 0.0001155788759809779,
+      "loss": 1.2758,
+      "step": 6475
+    },
+    {
+      "epoch": 0.46680602609385136,
+      "grad_norm": 0.133105256736894,
+      "learning_rate": 0.00011555581346499988,
+      "loss": 1.2713,
+      "step": 6476
+    },
+    {
+      "epoch": 0.4668781085561883,
+      "grad_norm": 0.10415842149036643,
+      "learning_rate": 0.00011553275010109084,
+      "loss": 1.2457,
+      "step": 6477
+    },
+    {
+      "epoch": 0.4669501910185252,
+      "grad_norm": 0.11923045112300627,
+      "learning_rate": 0.00011550968589050799,
+      "loss": 1.2656,
+      "step": 6478
+    },
+    {
+      "epoch": 0.4670222734808621,
+      "grad_norm": 0.11614138879233368,
+      "learning_rate": 0.00011548662083450848,
+      "loss": 1.2897,
+      "step": 6479
+    },
+    {
+      "epoch": 0.46709435594319904,
+      "grad_norm": 0.10302723242895209,
+      "learning_rate": 0.00011546355493434959,
+      "loss": 1.2533,
+      "step": 6480
+    },
+    {
+      "epoch": 0.46716643840553596,
+      "grad_norm": 0.13108803560555476,
+      "learning_rate": 0.00011544048819128862,
+      "loss": 1.3559,
+      "step": 6481
+    },
+    {
+      "epoch": 0.4672385208678728,
+      "grad_norm": 0.11046974295000782,
+      "learning_rate": 0.0001154174206065829,
+      "loss": 1.2406,
+      "step": 6482
+    },
+    {
+      "epoch": 0.46731060333020974,
+      "grad_norm": 0.11849043251252794,
+      "learning_rate": 0.00011539435218148984,
+      "loss": 1.2463,
+      "step": 6483
+    },
+    {
+      "epoch": 0.46738268579254666,
+      "grad_norm": 0.12115786994275228,
+      "learning_rate": 0.00011537128291726685,
+      "loss": 1.2933,
+      "step": 6484
+    },
+    {
+      "epoch": 0.4674547682548836,
+      "grad_norm": 0.11426861402527307,
+      "learning_rate": 0.00011534821281517143,
+      "loss": 1.1762,
+      "step": 6485
+    },
+    {
+      "epoch": 0.4675268507172205,
+      "grad_norm": 0.11676749784747349,
+      "learning_rate": 0.00011532514187646111,
+      "loss": 1.2355,
+      "step": 6486
+    },
+    {
+      "epoch": 0.4675989331795574,
+      "grad_norm": 0.12184119955472573,
+      "learning_rate": 0.00011530207010239344,
+      "loss": 1.3615,
+      "step": 6487
+    },
+    {
+      "epoch": 0.46767101564189434,
+      "grad_norm": 0.1147051225162204,
+      "learning_rate": 0.00011527899749422608,
+      "loss": 1.2744,
+      "step": 6488
+    },
+    {
+      "epoch": 0.46774309810423126,
+      "grad_norm": 0.21254942356414397,
+      "learning_rate": 0.00011525592405321666,
+      "loss": 1.2776,
+      "step": 6489
+    },
+    {
+      "epoch": 0.4678151805665682,
+      "grad_norm": 0.11026891716373907,
+      "learning_rate": 0.0001152328497806229,
+      "loss": 1.2844,
+      "step": 6490
+    },
+    {
+      "epoch": 0.46788726302890504,
+      "grad_norm": 0.11462443923700213,
+      "learning_rate": 0.00011520977467770251,
+      "loss": 1.3556,
+      "step": 6491
+    },
+    {
+      "epoch": 0.46795934549124196,
+      "grad_norm": 0.13337245083684368,
+      "learning_rate": 0.0001151866987457134,
+      "loss": 1.2067,
+      "step": 6492
+    },
+    {
+      "epoch": 0.4680314279535789,
+      "grad_norm": 0.12988253897197355,
+      "learning_rate": 0.00011516362198591326,
+      "loss": 1.2242,
+      "step": 6493
+    },
+    {
+      "epoch": 0.4681035104159158,
+      "grad_norm": 0.1128865130469948,
+      "learning_rate": 0.00011514054439956013,
+      "loss": 1.2333,
+      "step": 6494
+    },
+    {
+      "epoch": 0.4681755928782527,
+      "grad_norm": 0.12759042649591074,
+      "learning_rate": 0.00011511746598791182,
+      "loss": 1.3066,
+      "step": 6495
+    },
+    {
+      "epoch": 0.46824767534058964,
+      "grad_norm": 0.10367320602784094,
+      "learning_rate": 0.00011509438675222643,
+      "loss": 1.2459,
+      "step": 6496
+    },
+    {
+      "epoch": 0.46831975780292656,
+      "grad_norm": 0.11524528375129693,
+      "learning_rate": 0.00011507130669376187,
+      "loss": 1.2401,
+      "step": 6497
+    },
+    {
+      "epoch": 0.4683918402652635,
+      "grad_norm": 0.11169479355145666,
+      "learning_rate": 0.00011504822581377633,
+      "loss": 1.2448,
+      "step": 6498
+    },
+    {
+      "epoch": 0.4684639227276004,
+      "grad_norm": 0.11536937777397653,
+      "learning_rate": 0.00011502514411352778,
+      "loss": 1.2857,
+      "step": 6499
+    },
+    {
+      "epoch": 0.46853600518993727,
+      "grad_norm": 0.1224672297164719,
+      "learning_rate": 0.00011500206159427453,
+      "loss": 1.3165,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4686080876522742,
+      "grad_norm": 0.11321953053758567,
+      "learning_rate": 0.00011497897825727469,
+      "loss": 1.3714,
+      "step": 6501
+    },
+    {
+      "epoch": 0.4686801701146111,
+      "grad_norm": 0.12301764999735171,
+      "learning_rate": 0.0001149558941037865,
+      "loss": 1.3122,
+      "step": 6502
+    },
+    {
+      "epoch": 0.468752252576948,
+      "grad_norm": 0.12992848033132073,
+      "learning_rate": 0.0001149328091350683,
+      "loss": 1.23,
+      "step": 6503
+    },
+    {
+      "epoch": 0.46882433503928495,
+      "grad_norm": 0.12151311835291106,
+      "learning_rate": 0.00011490972335237842,
+      "loss": 1.2181,
+      "step": 6504
+    },
+    {
+      "epoch": 0.46889641750162186,
+      "grad_norm": 0.11630533076298913,
+      "learning_rate": 0.0001148866367569752,
+      "loss": 1.2182,
+      "step": 6505
+    },
+    {
+      "epoch": 0.4689684999639588,
+      "grad_norm": 0.1745933011838296,
+      "learning_rate": 0.00011486354935011711,
+      "loss": 1.3819,
+      "step": 6506
+    },
+    {
+      "epoch": 0.4690405824262957,
+      "grad_norm": 0.10448151070040798,
+      "learning_rate": 0.00011484046113306262,
+      "loss": 1.2289,
+      "step": 6507
+    },
+    {
+      "epoch": 0.46911266488863257,
+      "grad_norm": 0.11180837505123203,
+      "learning_rate": 0.00011481737210707022,
+      "loss": 1.292,
+      "step": 6508
+    },
+    {
+      "epoch": 0.4691847473509695,
+      "grad_norm": 0.12096053985157339,
+      "learning_rate": 0.0001147942822733985,
+      "loss": 1.2313,
+      "step": 6509
+    },
+    {
+      "epoch": 0.4692568298133064,
+      "grad_norm": 0.12280530585853097,
+      "learning_rate": 0.000114771191633306,
+      "loss": 1.3137,
+      "step": 6510
+    },
+    {
+      "epoch": 0.46932891227564333,
+      "grad_norm": 0.194888914728498,
+      "learning_rate": 0.00011474810018805144,
+      "loss": 1.3825,
+      "step": 6511
+    },
+    {
+      "epoch": 0.46940099473798025,
+      "grad_norm": 0.1121433223184387,
+      "learning_rate": 0.00011472500793889352,
+      "loss": 1.3011,
+      "step": 6512
+    },
+    {
+      "epoch": 0.46947307720031717,
+      "grad_norm": 0.11043633356545385,
+      "learning_rate": 0.00011470191488709086,
+      "loss": 1.3107,
+      "step": 6513
+    },
+    {
+      "epoch": 0.4695451596626541,
+      "grad_norm": 0.10092081719010644,
+      "learning_rate": 0.00011467882103390235,
+      "loss": 1.3478,
+      "step": 6514
+    },
+    {
+      "epoch": 0.469617242124991,
+      "grad_norm": 0.13234525737803948,
+      "learning_rate": 0.00011465572638058675,
+      "loss": 1.2602,
+      "step": 6515
+    },
+    {
+      "epoch": 0.4696893245873279,
+      "grad_norm": 0.11331110368562208,
+      "learning_rate": 0.000114632630928403,
+      "loss": 1.2758,
+      "step": 6516
+    },
+    {
+      "epoch": 0.4697614070496648,
+      "grad_norm": 0.14956523324953358,
+      "learning_rate": 0.00011460953467860989,
+      "loss": 1.2113,
+      "step": 6517
+    },
+    {
+      "epoch": 0.4698334895120017,
+      "grad_norm": 0.10983482124623807,
+      "learning_rate": 0.00011458643763246648,
+      "loss": 1.3667,
+      "step": 6518
+    },
+    {
+      "epoch": 0.46990557197433863,
+      "grad_norm": 0.1047811574513841,
+      "learning_rate": 0.0001145633397912317,
+      "loss": 1.2165,
+      "step": 6519
+    },
+    {
+      "epoch": 0.46997765443667555,
+      "grad_norm": 0.1289689771075662,
+      "learning_rate": 0.00011454024115616464,
+      "loss": 1.3075,
+      "step": 6520
+    },
+    {
+      "epoch": 0.47004973689901247,
+      "grad_norm": 0.11966054017565217,
+      "learning_rate": 0.00011451714172852433,
+      "loss": 1.1376,
+      "step": 6521
+    },
+    {
+      "epoch": 0.4701218193613494,
+      "grad_norm": 0.12687376919314175,
+      "learning_rate": 0.00011449404150956994,
+      "loss": 1.2363,
+      "step": 6522
+    },
+    {
+      "epoch": 0.4701939018236863,
+      "grad_norm": 0.11901943068886853,
+      "learning_rate": 0.00011447094050056064,
+      "loss": 1.2168,
+      "step": 6523
+    },
+    {
+      "epoch": 0.47026598428602323,
+      "grad_norm": 0.1165900146184186,
+      "learning_rate": 0.0001144478387027556,
+      "loss": 1.3747,
+      "step": 6524
+    },
+    {
+      "epoch": 0.47033806674836015,
+      "grad_norm": 0.12860593746851254,
+      "learning_rate": 0.0001144247361174141,
+      "loss": 1.4003,
+      "step": 6525
+    },
+    {
+      "epoch": 0.470410149210697,
+      "grad_norm": 0.10893417363451731,
+      "learning_rate": 0.00011440163274579541,
+      "loss": 1.393,
+      "step": 6526
+    },
+    {
+      "epoch": 0.47048223167303393,
+      "grad_norm": 0.12634965151183436,
+      "learning_rate": 0.00011437852858915892,
+      "loss": 1.3428,
+      "step": 6527
+    },
+    {
+      "epoch": 0.47055431413537085,
+      "grad_norm": 0.10999801118281113,
+      "learning_rate": 0.00011435542364876398,
+      "loss": 1.2037,
+      "step": 6528
+    },
+    {
+      "epoch": 0.4706263965977078,
+      "grad_norm": 0.1211521913920346,
+      "learning_rate": 0.00011433231792587003,
+      "loss": 1.2994,
+      "step": 6529
+    },
+    {
+      "epoch": 0.4706984790600447,
+      "grad_norm": 0.11134549022926137,
+      "learning_rate": 0.00011430921142173652,
+      "loss": 1.2514,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4707705615223816,
+      "grad_norm": 0.12313189346920926,
+      "learning_rate": 0.000114286104137623,
+      "loss": 1.2479,
+      "step": 6531
+    },
+    {
+      "epoch": 0.47084264398471853,
+      "grad_norm": 0.11450068613706148,
+      "learning_rate": 0.00011426299607478896,
+      "loss": 1.3756,
+      "step": 6532
+    },
+    {
+      "epoch": 0.47091472644705545,
+      "grad_norm": 0.11757337785464327,
+      "learning_rate": 0.00011423988723449406,
+      "loss": 1.3017,
+      "step": 6533
+    },
+    {
+      "epoch": 0.4709868089093923,
+      "grad_norm": 0.10800446457792932,
+      "learning_rate": 0.0001142167776179979,
+      "loss": 1.3339,
+      "step": 6534
+    },
+    {
+      "epoch": 0.47105889137172924,
+      "grad_norm": 0.11317361681170651,
+      "learning_rate": 0.00011419366722656017,
+      "loss": 1.3639,
+      "step": 6535
+    },
+    {
+      "epoch": 0.47113097383406616,
+      "grad_norm": 0.11328999405449011,
+      "learning_rate": 0.0001141705560614406,
+      "loss": 1.2811,
+      "step": 6536
+    },
+    {
+      "epoch": 0.4712030562964031,
+      "grad_norm": 0.12082718069284965,
+      "learning_rate": 0.00011414744412389897,
+      "loss": 1.1984,
+      "step": 6537
+    },
+    {
+      "epoch": 0.47127513875874,
+      "grad_norm": 0.12041917072227684,
+      "learning_rate": 0.00011412433141519504,
+      "loss": 1.233,
+      "step": 6538
+    },
+    {
+      "epoch": 0.4713472212210769,
+      "grad_norm": 0.11182157729503696,
+      "learning_rate": 0.0001141012179365887,
+      "loss": 1.3072,
+      "step": 6539
+    },
+    {
+      "epoch": 0.47141930368341384,
+      "grad_norm": 0.10904243805845829,
+      "learning_rate": 0.00011407810368933984,
+      "loss": 1.2787,
+      "step": 6540
+    },
+    {
+      "epoch": 0.47149138614575076,
+      "grad_norm": 0.16383957360039397,
+      "learning_rate": 0.00011405498867470839,
+      "loss": 1.3705,
+      "step": 6541
+    },
+    {
+      "epoch": 0.4715634686080877,
+      "grad_norm": 0.10261966658983347,
+      "learning_rate": 0.0001140318728939543,
+      "loss": 1.3521,
+      "step": 6542
+    },
+    {
+      "epoch": 0.47163555107042454,
+      "grad_norm": 0.11349918564706585,
+      "learning_rate": 0.00011400875634833758,
+      "loss": 1.3151,
+      "step": 6543
+    },
+    {
+      "epoch": 0.47170763353276146,
+      "grad_norm": 0.11895140479423519,
+      "learning_rate": 0.00011398563903911835,
+      "loss": 1.1529,
+      "step": 6544
+    },
+    {
+      "epoch": 0.4717797159950984,
+      "grad_norm": 0.10870967045779348,
+      "learning_rate": 0.00011396252096755668,
+      "loss": 1.3509,
+      "step": 6545
+    },
+    {
+      "epoch": 0.4718517984574353,
+      "grad_norm": 0.11516973059113121,
+      "learning_rate": 0.00011393940213491267,
+      "loss": 1.2714,
+      "step": 6546
+    },
+    {
+      "epoch": 0.4719238809197722,
+      "grad_norm": 0.13688416851118243,
+      "learning_rate": 0.00011391628254244658,
+      "loss": 1.2476,
+      "step": 6547
+    },
+    {
+      "epoch": 0.47199596338210914,
+      "grad_norm": 0.10797684096960992,
+      "learning_rate": 0.00011389316219141854,
+      "loss": 1.2688,
+      "step": 6548
+    },
+    {
+      "epoch": 0.47206804584444606,
+      "grad_norm": 0.1166062157780419,
+      "learning_rate": 0.00011387004108308892,
+      "loss": 1.3206,
+      "step": 6549
+    },
+    {
+      "epoch": 0.472140128306783,
+      "grad_norm": 0.10474031348484616,
+      "learning_rate": 0.00011384691921871793,
+      "loss": 1.244,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4722122107691199,
+      "grad_norm": 0.1186324263543942,
+      "learning_rate": 0.00011382379659956601,
+      "loss": 1.2325,
+      "step": 6551
+    },
+    {
+      "epoch": 0.47228429323145676,
+      "grad_norm": 0.11443542488333544,
+      "learning_rate": 0.00011380067322689345,
+      "loss": 1.3083,
+      "step": 6552
+    },
+    {
+      "epoch": 0.4723563756937937,
+      "grad_norm": 0.11068671045805405,
+      "learning_rate": 0.00011377754910196079,
+      "loss": 1.2458,
+      "step": 6553
+    },
+    {
+      "epoch": 0.4724284581561306,
+      "grad_norm": 0.12342324586330647,
+      "learning_rate": 0.0001137544242260284,
+      "loss": 1.1612,
+      "step": 6554
+    },
+    {
+      "epoch": 0.4725005406184675,
+      "grad_norm": 0.11778925712078572,
+      "learning_rate": 0.0001137312986003569,
+      "loss": 1.3606,
+      "step": 6555
+    },
+    {
+      "epoch": 0.47257262308080444,
+      "grad_norm": 0.12246550483540086,
+      "learning_rate": 0.00011370817222620673,
+      "loss": 1.476,
+      "step": 6556
+    },
+    {
+      "epoch": 0.47264470554314136,
+      "grad_norm": 0.14800772588692657,
+      "learning_rate": 0.00011368504510483854,
+      "loss": 1.2736,
+      "step": 6557
+    },
+    {
+      "epoch": 0.4727167880054783,
+      "grad_norm": 0.1182207116609803,
+      "learning_rate": 0.00011366191723751299,
+      "loss": 1.2977,
+      "step": 6558
+    },
+    {
+      "epoch": 0.4727888704678152,
+      "grad_norm": 0.11753972818662609,
+      "learning_rate": 0.00011363878862549072,
+      "loss": 1.346,
+      "step": 6559
+    },
+    {
+      "epoch": 0.47286095293015207,
+      "grad_norm": 0.10506243661487923,
+      "learning_rate": 0.00011361565927003243,
+      "loss": 1.2153,
+      "step": 6560
+    },
+    {
+      "epoch": 0.472933035392489,
+      "grad_norm": 0.11757972125264783,
+      "learning_rate": 0.00011359252917239893,
+      "loss": 1.3197,
+      "step": 6561
+    },
+    {
+      "epoch": 0.4730051178548259,
+      "grad_norm": 0.27115650379746675,
+      "learning_rate": 0.00011356939833385097,
+      "loss": 1.3993,
+      "step": 6562
+    },
+    {
+      "epoch": 0.4730772003171628,
+      "grad_norm": 0.10344240274293441,
+      "learning_rate": 0.00011354626675564942,
+      "loss": 1.2805,
+      "step": 6563
+    },
+    {
+      "epoch": 0.47314928277949975,
+      "grad_norm": 0.12198326054154794,
+      "learning_rate": 0.00011352313443905511,
+      "loss": 1.1753,
+      "step": 6564
+    },
+    {
+      "epoch": 0.47322136524183667,
+      "grad_norm": 0.17696461857471327,
+      "learning_rate": 0.00011350000138532902,
+      "loss": 1.2858,
+      "step": 6565
+    },
+    {
+      "epoch": 0.4732934477041736,
+      "grad_norm": 0.11732827093401406,
+      "learning_rate": 0.00011347686759573206,
+      "loss": 1.2558,
+      "step": 6566
+    },
+    {
+      "epoch": 0.4733655301665105,
+      "grad_norm": 0.11013284479901297,
+      "learning_rate": 0.00011345373307152529,
+      "loss": 1.3405,
+      "step": 6567
+    },
+    {
+      "epoch": 0.4734376126288474,
+      "grad_norm": 0.12189561256123123,
+      "learning_rate": 0.00011343059781396965,
+      "loss": 1.2719,
+      "step": 6568
+    },
+    {
+      "epoch": 0.4735096950911843,
+      "grad_norm": 0.11830360625352956,
+      "learning_rate": 0.00011340746182432633,
+      "loss": 1.179,
+      "step": 6569
+    },
+    {
+      "epoch": 0.4735817775535212,
+      "grad_norm": 0.10334526992268529,
+      "learning_rate": 0.00011338432510385635,
+      "loss": 1.2952,
+      "step": 6570
+    },
+    {
+      "epoch": 0.47365386001585813,
+      "grad_norm": 0.12047428059922938,
+      "learning_rate": 0.00011336118765382095,
+      "loss": 1.2416,
+      "step": 6571
+    },
+    {
+      "epoch": 0.47372594247819505,
+      "grad_norm": 0.10686079318706426,
+      "learning_rate": 0.00011333804947548125,
+      "loss": 1.3959,
+      "step": 6572
+    },
+    {
+      "epoch": 0.47379802494053197,
+      "grad_norm": 0.13220550287177124,
+      "learning_rate": 0.00011331491057009858,
+      "loss": 1.2652,
+      "step": 6573
+    },
+    {
+      "epoch": 0.4738701074028689,
+      "grad_norm": 0.10932425775978673,
+      "learning_rate": 0.00011329177093893411,
+      "loss": 1.3342,
+      "step": 6574
+    },
+    {
+      "epoch": 0.4739421898652058,
+      "grad_norm": 0.09875134187110385,
+      "learning_rate": 0.00011326863058324926,
+      "loss": 1.2983,
+      "step": 6575
+    },
+    {
+      "epoch": 0.47401427232754273,
+      "grad_norm": 0.10408255018532686,
+      "learning_rate": 0.00011324548950430529,
+      "loss": 1.3267,
+      "step": 6576
+    },
+    {
+      "epoch": 0.47408635478987965,
+      "grad_norm": 0.11967568000829236,
+      "learning_rate": 0.00011322234770336368,
+      "loss": 1.2574,
+      "step": 6577
+    },
+    {
+      "epoch": 0.4741584372522165,
+      "grad_norm": 0.1248377646903826,
+      "learning_rate": 0.00011319920518168584,
+      "loss": 1.3362,
+      "step": 6578
+    },
+    {
+      "epoch": 0.47423051971455343,
+      "grad_norm": 0.0965246962659956,
+      "learning_rate": 0.0001131760619405332,
+      "loss": 1.2671,
+      "step": 6579
+    },
+    {
+      "epoch": 0.47430260217689035,
+      "grad_norm": 0.1163008287832273,
+      "learning_rate": 0.00011315291798116732,
+      "loss": 1.3243,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4743746846392273,
+      "grad_norm": 0.11881620099228277,
+      "learning_rate": 0.00011312977330484975,
+      "loss": 1.227,
+      "step": 6581
+    },
+    {
+      "epoch": 0.4744467671015642,
+      "grad_norm": 0.11950082967663751,
+      "learning_rate": 0.00011310662791284205,
+      "loss": 1.3255,
+      "step": 6582
+    },
+    {
+      "epoch": 0.4745188495639011,
+      "grad_norm": 0.10576207305416321,
+      "learning_rate": 0.00011308348180640592,
+      "loss": 1.1724,
+      "step": 6583
+    },
+    {
+      "epoch": 0.47459093202623803,
+      "grad_norm": 0.11712675475846852,
+      "learning_rate": 0.00011306033498680292,
+      "loss": 1.3035,
+      "step": 6584
+    },
+    {
+      "epoch": 0.47466301448857495,
+      "grad_norm": 0.1182049430260817,
+      "learning_rate": 0.00011303718745529485,
+      "loss": 1.2877,
+      "step": 6585
+    },
+    {
+      "epoch": 0.4747350969509118,
+      "grad_norm": 0.12440828054509638,
+      "learning_rate": 0.00011301403921314345,
+      "loss": 1.1839,
+      "step": 6586
+    },
+    {
+      "epoch": 0.47480717941324874,
+      "grad_norm": 0.11117357567330838,
+      "learning_rate": 0.00011299089026161047,
+      "loss": 1.3857,
+      "step": 6587
+    },
+    {
+      "epoch": 0.47487926187558566,
+      "grad_norm": 0.10374122347495388,
+      "learning_rate": 0.00011296774060195776,
+      "loss": 1.3274,
+      "step": 6588
+    },
+    {
+      "epoch": 0.4749513443379226,
+      "grad_norm": 0.11688446770682642,
+      "learning_rate": 0.00011294459023544717,
+      "loss": 1.2694,
+      "step": 6589
+    },
+    {
+      "epoch": 0.4750234268002595,
+      "grad_norm": 0.10827484616251197,
+      "learning_rate": 0.00011292143916334059,
+      "loss": 1.1774,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4750955092625964,
+      "grad_norm": 0.18034505991330255,
+      "learning_rate": 0.0001128982873869,
+      "loss": 1.216,
+      "step": 6591
+    },
+    {
+      "epoch": 0.47516759172493334,
+      "grad_norm": 0.12578519402201488,
+      "learning_rate": 0.00011287513490738732,
+      "loss": 1.2912,
+      "step": 6592
+    },
+    {
+      "epoch": 0.47523967418727026,
+      "grad_norm": 0.1163398331625733,
+      "learning_rate": 0.00011285198172606465,
+      "loss": 1.368,
+      "step": 6593
+    },
+    {
+      "epoch": 0.4753117566496072,
+      "grad_norm": 0.1273386784887031,
+      "learning_rate": 0.00011282882784419398,
+      "loss": 1.1705,
+      "step": 6594
+    },
+    {
+      "epoch": 0.47538383911194404,
+      "grad_norm": 0.1257212798254381,
+      "learning_rate": 0.00011280567326303742,
+      "loss": 1.247,
+      "step": 6595
+    },
+    {
+      "epoch": 0.47545592157428096,
+      "grad_norm": 0.11203578541786073,
+      "learning_rate": 0.00011278251798385708,
+      "loss": 1.3038,
+      "step": 6596
+    },
+    {
+      "epoch": 0.4755280040366179,
+      "grad_norm": 0.1121422298044936,
+      "learning_rate": 0.00011275936200791522,
+      "loss": 1.3258,
+      "step": 6597
+    },
+    {
+      "epoch": 0.4756000864989548,
+      "grad_norm": 0.10864852782113668,
+      "learning_rate": 0.00011273620533647391,
+      "loss": 1.3078,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4756721689612917,
+      "grad_norm": 0.1083443081385807,
+      "learning_rate": 0.00011271304797079551,
+      "loss": 1.3296,
+      "step": 6599
+    },
+    {
+      "epoch": 0.47574425142362864,
+      "grad_norm": 0.109738644569416,
+      "learning_rate": 0.00011268988991214225,
+      "loss": 1.31,
+      "step": 6600
+    },
+    {
+      "epoch": 0.47581633388596556,
+      "grad_norm": 0.11234902890203449,
+      "learning_rate": 0.00011266673116177646,
+      "loss": 1.2841,
+      "step": 6601
+    },
+    {
+      "epoch": 0.4758884163483025,
+      "grad_norm": 0.12275103040120826,
+      "learning_rate": 0.00011264357172096049,
+      "loss": 1.2798,
+      "step": 6602
+    },
+    {
+      "epoch": 0.4759604988106394,
+      "grad_norm": 0.1110044926712897,
+      "learning_rate": 0.00011262041159095678,
+      "loss": 1.2563,
+      "step": 6603
+    },
+    {
+      "epoch": 0.47603258127297626,
+      "grad_norm": 0.12065193066476597,
+      "learning_rate": 0.00011259725077302772,
+      "loss": 1.3392,
+      "step": 6604
+    },
+    {
+      "epoch": 0.4761046637353132,
+      "grad_norm": 0.11531521209395344,
+      "learning_rate": 0.00011257408926843578,
+      "loss": 1.2412,
+      "step": 6605
+    },
+    {
+      "epoch": 0.4761767461976501,
+      "grad_norm": 0.11195891123047894,
+      "learning_rate": 0.0001125509270784435,
+      "loss": 1.2664,
+      "step": 6606
+    },
+    {
+      "epoch": 0.476248828659987,
+      "grad_norm": 0.12948325395374222,
+      "learning_rate": 0.00011252776420431338,
+      "loss": 1.3045,
+      "step": 6607
+    },
+    {
+      "epoch": 0.47632091112232394,
+      "grad_norm": 0.1112960270535804,
+      "learning_rate": 0.00011250460064730807,
+      "loss": 1.3289,
+      "step": 6608
+    },
+    {
+      "epoch": 0.47639299358466086,
+      "grad_norm": 0.13389925390511814,
+      "learning_rate": 0.00011248143640869016,
+      "loss": 1.2797,
+      "step": 6609
+    },
+    {
+      "epoch": 0.4764650760469978,
+      "grad_norm": 0.11712112736944384,
+      "learning_rate": 0.00011245827148972228,
+      "loss": 1.3449,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4765371585093347,
+      "grad_norm": 0.12134331096273968,
+      "learning_rate": 0.00011243510589166714,
+      "loss": 1.2674,
+      "step": 6611
+    },
+    {
+      "epoch": 0.47660924097167157,
+      "grad_norm": 0.11957900987967593,
+      "learning_rate": 0.00011241193961578751,
+      "loss": 1.3546,
+      "step": 6612
+    },
+    {
+      "epoch": 0.4766813234340085,
+      "grad_norm": 0.1179160332972117,
+      "learning_rate": 0.00011238877266334611,
+      "loss": 1.3356,
+      "step": 6613
+    },
+    {
+      "epoch": 0.4767534058963454,
+      "grad_norm": 0.12067802778141214,
+      "learning_rate": 0.00011236560503560578,
+      "loss": 1.2679,
+      "step": 6614
+    },
+    {
+      "epoch": 0.4768254883586823,
+      "grad_norm": 0.10886711660947127,
+      "learning_rate": 0.00011234243673382936,
+      "loss": 1.2706,
+      "step": 6615
+    },
+    {
+      "epoch": 0.47689757082101925,
+      "grad_norm": 0.11338147862767747,
+      "learning_rate": 0.0001123192677592797,
+      "loss": 1.2405,
+      "step": 6616
+    },
+    {
+      "epoch": 0.47696965328335617,
+      "grad_norm": 0.10885522540495282,
+      "learning_rate": 0.00011229609811321972,
+      "loss": 1.3059,
+      "step": 6617
+    },
+    {
+      "epoch": 0.4770417357456931,
+      "grad_norm": 0.12582600534213642,
+      "learning_rate": 0.00011227292779691243,
+      "loss": 1.2986,
+      "step": 6618
+    },
+    {
+      "epoch": 0.47711381820803,
+      "grad_norm": 0.1036537119329675,
+      "learning_rate": 0.00011224975681162075,
+      "loss": 1.3045,
+      "step": 6619
+    },
+    {
+      "epoch": 0.4771859006703669,
+      "grad_norm": 0.11736672640123194,
+      "learning_rate": 0.00011222658515860773,
+      "loss": 1.3444,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4772579831327038,
+      "grad_norm": 0.11663349104505674,
+      "learning_rate": 0.00011220341283913645,
+      "loss": 1.3032,
+      "step": 6621
+    },
+    {
+      "epoch": 0.4773300655950407,
+      "grad_norm": 0.12556631222041992,
+      "learning_rate": 0.00011218023985447001,
+      "loss": 1.1968,
+      "step": 6622
+    },
+    {
+      "epoch": 0.47740214805737763,
+      "grad_norm": 0.10788635757702193,
+      "learning_rate": 0.00011215706620587149,
+      "loss": 1.2894,
+      "step": 6623
+    },
+    {
+      "epoch": 0.47747423051971455,
+      "grad_norm": 0.12613641791443872,
+      "learning_rate": 0.00011213389189460414,
+      "loss": 1.1995,
+      "step": 6624
+    },
+    {
+      "epoch": 0.47754631298205147,
+      "grad_norm": 0.111005109229988,
+      "learning_rate": 0.00011211071692193112,
+      "loss": 1.3606,
+      "step": 6625
+    },
+    {
+      "epoch": 0.4776183954443884,
+      "grad_norm": 0.11694161981152197,
+      "learning_rate": 0.00011208754128911569,
+      "loss": 1.2711,
+      "step": 6626
+    },
+    {
+      "epoch": 0.4776904779067253,
+      "grad_norm": 0.1109555534952357,
+      "learning_rate": 0.00011206436499742108,
+      "loss": 1.3531,
+      "step": 6627
+    },
+    {
+      "epoch": 0.47776256036906223,
+      "grad_norm": 0.1351998920579512,
+      "learning_rate": 0.0001120411880481107,
+      "loss": 1.3529,
+      "step": 6628
+    },
+    {
+      "epoch": 0.47783464283139915,
+      "grad_norm": 0.10917545928201229,
+      "learning_rate": 0.0001120180104424478,
+      "loss": 1.256,
+      "step": 6629
+    },
+    {
+      "epoch": 0.477906725293736,
+      "grad_norm": 0.10886711356928544,
+      "learning_rate": 0.00011199483218169587,
+      "loss": 1.2438,
+      "step": 6630
+    },
+    {
+      "epoch": 0.47797880775607293,
+      "grad_norm": 0.1183360996781391,
+      "learning_rate": 0.0001119716532671182,
+      "loss": 1.1675,
+      "step": 6631
+    },
+    {
+      "epoch": 0.47805089021840985,
+      "grad_norm": 0.12107607466334962,
+      "learning_rate": 0.0001119484736999784,
+      "loss": 1.362,
+      "step": 6632
+    },
+    {
+      "epoch": 0.47812297268074677,
+      "grad_norm": 0.23750754591051745,
+      "learning_rate": 0.00011192529348153986,
+      "loss": 1.3102,
+      "step": 6633
+    },
+    {
+      "epoch": 0.4781950551430837,
+      "grad_norm": 0.11871993588596784,
+      "learning_rate": 0.00011190211261306613,
+      "loss": 1.3069,
+      "step": 6634
+    },
+    {
+      "epoch": 0.4782671376054206,
+      "grad_norm": 0.11434830442132919,
+      "learning_rate": 0.0001118789310958208,
+      "loss": 1.2555,
+      "step": 6635
+    },
+    {
+      "epoch": 0.47833922006775753,
+      "grad_norm": 0.1210354819061694,
+      "learning_rate": 0.00011185574893106745,
+      "loss": 1.3267,
+      "step": 6636
+    },
+    {
+      "epoch": 0.47841130253009445,
+      "grad_norm": 0.12183612826222211,
+      "learning_rate": 0.00011183256612006971,
+      "loss": 1.3518,
+      "step": 6637
+    },
+    {
+      "epoch": 0.4784833849924313,
+      "grad_norm": 0.19650677482369586,
+      "learning_rate": 0.00011180938266409126,
+      "loss": 1.2391,
+      "step": 6638
+    },
+    {
+      "epoch": 0.47855546745476824,
+      "grad_norm": 0.12172210203615237,
+      "learning_rate": 0.0001117861985643958,
+      "loss": 1.2154,
+      "step": 6639
+    },
+    {
+      "epoch": 0.47862754991710516,
+      "grad_norm": 0.11652331880781351,
+      "learning_rate": 0.00011176301382224709,
+      "loss": 1.3394,
+      "step": 6640
+    },
+    {
+      "epoch": 0.4786996323794421,
+      "grad_norm": 0.11340329680798249,
+      "learning_rate": 0.00011173982843890889,
+      "loss": 1.2409,
+      "step": 6641
+    },
+    {
+      "epoch": 0.478771714841779,
+      "grad_norm": 0.1136949173802656,
+      "learning_rate": 0.00011171664241564498,
+      "loss": 1.251,
+      "step": 6642
+    },
+    {
+      "epoch": 0.4788437973041159,
+      "grad_norm": 0.33169212073237336,
+      "learning_rate": 0.00011169345575371927,
+      "loss": 1.4021,
+      "step": 6643
+    },
+    {
+      "epoch": 0.47891587976645283,
+      "grad_norm": 0.10868257998916443,
+      "learning_rate": 0.00011167026845439563,
+      "loss": 1.3355,
+      "step": 6644
+    },
+    {
+      "epoch": 0.47898796222878975,
+      "grad_norm": 0.11694931877809758,
+      "learning_rate": 0.00011164708051893791,
+      "loss": 1.3049,
+      "step": 6645
+    },
+    {
+      "epoch": 0.4790600446911267,
+      "grad_norm": 0.11172984879265925,
+      "learning_rate": 0.00011162389194861016,
+      "loss": 1.3237,
+      "step": 6646
+    },
+    {
+      "epoch": 0.47913212715346354,
+      "grad_norm": 0.11265164375904296,
+      "learning_rate": 0.00011160070274467624,
+      "loss": 1.2592,
+      "step": 6647
+    },
+    {
+      "epoch": 0.47920420961580046,
+      "grad_norm": 0.11515832194864066,
+      "learning_rate": 0.00011157751290840029,
+      "loss": 1.3413,
+      "step": 6648
+    },
+    {
+      "epoch": 0.4792762920781374,
+      "grad_norm": 0.12454084124651096,
+      "learning_rate": 0.00011155432244104627,
+      "loss": 1.2384,
+      "step": 6649
+    },
+    {
+      "epoch": 0.4793483745404743,
+      "grad_norm": 0.10638823681562361,
+      "learning_rate": 0.00011153113134387836,
+      "loss": 1.2761,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4794204570028112,
+      "grad_norm": 0.11242456056492704,
+      "learning_rate": 0.00011150793961816057,
+      "loss": 1.2733,
+      "step": 6651
+    },
+    {
+      "epoch": 0.47949253946514814,
+      "grad_norm": 0.10790236068284649,
+      "learning_rate": 0.00011148474726515716,
+      "loss": 1.3215,
+      "step": 6652
+    },
+    {
+      "epoch": 0.47956462192748506,
+      "grad_norm": 0.1244745566904934,
+      "learning_rate": 0.00011146155428613225,
+      "loss": 1.265,
+      "step": 6653
+    },
+    {
+      "epoch": 0.479636704389822,
+      "grad_norm": 0.12066803018680086,
+      "learning_rate": 0.00011143836068235014,
+      "loss": 1.2784,
+      "step": 6654
+    },
+    {
+      "epoch": 0.4797087868521589,
+      "grad_norm": 0.2539535350581273,
+      "learning_rate": 0.00011141516645507499,
+      "loss": 1.2309,
+      "step": 6655
+    },
+    {
+      "epoch": 0.47978086931449576,
+      "grad_norm": 0.11301349618126927,
+      "learning_rate": 0.00011139197160557118,
+      "loss": 1.3172,
+      "step": 6656
+    },
+    {
+      "epoch": 0.4798529517768327,
+      "grad_norm": 0.12230411461472779,
+      "learning_rate": 0.00011136877613510299,
+      "loss": 1.3001,
+      "step": 6657
+    },
+    {
+      "epoch": 0.4799250342391696,
+      "grad_norm": 0.11863509220294853,
+      "learning_rate": 0.00011134558004493478,
+      "loss": 1.3353,
+      "step": 6658
+    },
+    {
+      "epoch": 0.4799971167015065,
+      "grad_norm": 0.11667179837061116,
+      "learning_rate": 0.00011132238333633098,
+      "loss": 1.3894,
+      "step": 6659
+    },
+    {
+      "epoch": 0.48006919916384344,
+      "grad_norm": 0.12472518353045266,
+      "learning_rate": 0.00011129918601055598,
+      "loss": 1.3326,
+      "step": 6660
+    },
+    {
+      "epoch": 0.48014128162618036,
+      "grad_norm": 0.11283843192854598,
+      "learning_rate": 0.00011127598806887427,
+      "loss": 1.3967,
+      "step": 6661
+    },
+    {
+      "epoch": 0.4802133640885173,
+      "grad_norm": 0.11085444594578024,
+      "learning_rate": 0.00011125278951255032,
+      "loss": 1.2797,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4802854465508542,
+      "grad_norm": 0.11500773994743514,
+      "learning_rate": 0.00011122959034284868,
+      "loss": 1.2058,
+      "step": 6663
+    },
+    {
+      "epoch": 0.48035752901319106,
+      "grad_norm": 0.11220857232410424,
+      "learning_rate": 0.00011120639056103388,
+      "loss": 1.2065,
+      "step": 6664
+    },
+    {
+      "epoch": 0.480429611475528,
+      "grad_norm": 0.1299768071144842,
+      "learning_rate": 0.00011118319016837058,
+      "loss": 1.3065,
+      "step": 6665
+    },
+    {
+      "epoch": 0.4805016939378649,
+      "grad_norm": 0.10808860270550105,
+      "learning_rate": 0.00011115998916612336,
+      "loss": 1.2717,
+      "step": 6666
+    },
+    {
+      "epoch": 0.4805737764002018,
+      "grad_norm": 0.11294604125014757,
+      "learning_rate": 0.00011113678755555685,
+      "loss": 1.2716,
+      "step": 6667
+    },
+    {
+      "epoch": 0.48064585886253874,
+      "grad_norm": 0.11419254252275499,
+      "learning_rate": 0.00011111358533793583,
+      "loss": 1.3432,
+      "step": 6668
+    },
+    {
+      "epoch": 0.48071794132487566,
+      "grad_norm": 0.11454440086373131,
+      "learning_rate": 0.00011109038251452495,
+      "loss": 1.2786,
+      "step": 6669
+    },
+    {
+      "epoch": 0.4807900237872126,
+      "grad_norm": 0.24908818784123699,
+      "learning_rate": 0.00011106717908658904,
+      "loss": 1.3305,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4808621062495495,
+      "grad_norm": 0.1252115011005337,
+      "learning_rate": 0.00011104397505539282,
+      "loss": 1.3101,
+      "step": 6671
+    },
+    {
+      "epoch": 0.4809341887118864,
+      "grad_norm": 0.12544846134194712,
+      "learning_rate": 0.0001110207704222012,
+      "loss": 1.2933,
+      "step": 6672
+    },
+    {
+      "epoch": 0.4810062711742233,
+      "grad_norm": 0.11622821423295845,
+      "learning_rate": 0.00011099756518827896,
+      "loss": 1.3101,
+      "step": 6673
+    },
+    {
+      "epoch": 0.4810783536365602,
+      "grad_norm": 0.11467307615614815,
+      "learning_rate": 0.00011097435935489107,
+      "loss": 1.333,
+      "step": 6674
+    },
+    {
+      "epoch": 0.4811504360988971,
+      "grad_norm": 0.1110005359055924,
+      "learning_rate": 0.00011095115292330236,
+      "loss": 1.2569,
+      "step": 6675
+    },
+    {
+      "epoch": 0.48122251856123405,
+      "grad_norm": 0.12518833260331,
+      "learning_rate": 0.00011092794589477791,
+      "loss": 1.2717,
+      "step": 6676
+    },
+    {
+      "epoch": 0.48129460102357097,
+      "grad_norm": 0.110104384149153,
+      "learning_rate": 0.00011090473827058261,
+      "loss": 1.3099,
+      "step": 6677
+    },
+    {
+      "epoch": 0.4813666834859079,
+      "grad_norm": 0.11751088183246654,
+      "learning_rate": 0.0001108815300519815,
+      "loss": 1.2518,
+      "step": 6678
+    },
+    {
+      "epoch": 0.4814387659482448,
+      "grad_norm": 0.10788385574615697,
+      "learning_rate": 0.00011085832124023966,
+      "loss": 1.2491,
+      "step": 6679
+    },
+    {
+      "epoch": 0.4815108484105817,
+      "grad_norm": 0.11823916585685167,
+      "learning_rate": 0.0001108351118366222,
+      "loss": 1.2959,
+      "step": 6680
+    },
+    {
+      "epoch": 0.48158293087291865,
+      "grad_norm": 0.117497073438747,
+      "learning_rate": 0.00011081190184239419,
+      "loss": 1.2902,
+      "step": 6681
+    },
+    {
+      "epoch": 0.4816550133352555,
+      "grad_norm": 0.10771517660434016,
+      "learning_rate": 0.00011078869125882081,
+      "loss": 1.304,
+      "step": 6682
+    },
+    {
+      "epoch": 0.48172709579759243,
+      "grad_norm": 0.11754296872265683,
+      "learning_rate": 0.00011076548008716723,
+      "loss": 1.3157,
+      "step": 6683
+    },
+    {
+      "epoch": 0.48179917825992935,
+      "grad_norm": 0.10711045334908918,
+      "learning_rate": 0.00011074226832869867,
+      "loss": 1.2053,
+      "step": 6684
+    },
+    {
+      "epoch": 0.48187126072226627,
+      "grad_norm": 0.10964257732164819,
+      "learning_rate": 0.0001107190559846804,
+      "loss": 1.2688,
+      "step": 6685
+    },
+    {
+      "epoch": 0.4819433431846032,
+      "grad_norm": 0.12927616129272768,
+      "learning_rate": 0.00011069584305637768,
+      "loss": 1.2053,
+      "step": 6686
+    },
+    {
+      "epoch": 0.4820154256469401,
+      "grad_norm": 0.10285068642492298,
+      "learning_rate": 0.00011067262954505582,
+      "loss": 1.3398,
+      "step": 6687
+    },
+    {
+      "epoch": 0.48208750810927703,
+      "grad_norm": 0.11684011732196121,
+      "learning_rate": 0.00011064941545198018,
+      "loss": 1.2664,
+      "step": 6688
+    },
+    {
+      "epoch": 0.48215959057161395,
+      "grad_norm": 0.11780944062527147,
+      "learning_rate": 0.0001106262007784161,
+      "loss": 1.3371,
+      "step": 6689
+    },
+    {
+      "epoch": 0.4822316730339508,
+      "grad_norm": 0.11442499764092631,
+      "learning_rate": 0.00011060298552562905,
+      "loss": 1.302,
+      "step": 6690
+    },
+    {
+      "epoch": 0.48230375549628773,
+      "grad_norm": 0.13565389799336577,
+      "learning_rate": 0.00011057976969488439,
+      "loss": 1.4222,
+      "step": 6691
+    },
+    {
+      "epoch": 0.48237583795862465,
+      "grad_norm": 0.1301610051546298,
+      "learning_rate": 0.00011055655328744767,
+      "loss": 1.2404,
+      "step": 6692
+    },
+    {
+      "epoch": 0.4824479204209616,
+      "grad_norm": 0.1303023659060739,
+      "learning_rate": 0.00011053333630458434,
+      "loss": 1.2995,
+      "step": 6693
+    },
+    {
+      "epoch": 0.4825200028832985,
+      "grad_norm": 0.10902530567045365,
+      "learning_rate": 0.00011051011874755995,
+      "loss": 1.2088,
+      "step": 6694
+    },
+    {
+      "epoch": 0.4825920853456354,
+      "grad_norm": 0.11357089601795133,
+      "learning_rate": 0.00011048690061764005,
+      "loss": 1.2523,
+      "step": 6695
+    },
+    {
+      "epoch": 0.48266416780797233,
+      "grad_norm": 0.11757939259341749,
+      "learning_rate": 0.00011046368191609025,
+      "loss": 1.2669,
+      "step": 6696
+    },
+    {
+      "epoch": 0.48273625027030925,
+      "grad_norm": 0.16144160948730854,
+      "learning_rate": 0.00011044046264417618,
+      "loss": 1.2861,
+      "step": 6697
+    },
+    {
+      "epoch": 0.4828083327326462,
+      "grad_norm": 0.11205162203737991,
+      "learning_rate": 0.0001104172428031635,
+      "loss": 1.2851,
+      "step": 6698
+    },
+    {
+      "epoch": 0.48288041519498304,
+      "grad_norm": 0.11949470013075704,
+      "learning_rate": 0.00011039402239431791,
+      "loss": 1.275,
+      "step": 6699
+    },
+    {
+      "epoch": 0.48295249765731996,
+      "grad_norm": 0.1142610605096984,
+      "learning_rate": 0.00011037080141890505,
+      "loss": 1.3177,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4830245801196569,
+      "grad_norm": 0.11179364205275656,
+      "learning_rate": 0.0001103475798781908,
+      "loss": 1.194,
+      "step": 6701
+    },
+    {
+      "epoch": 0.4830966625819938,
+      "grad_norm": 0.1186492850313083,
+      "learning_rate": 0.0001103243577734408,
+      "loss": 1.3286,
+      "step": 6702
+    },
+    {
+      "epoch": 0.4831687450443307,
+      "grad_norm": 0.1043609385740912,
+      "learning_rate": 0.00011030113510592101,
+      "loss": 1.2547,
+      "step": 6703
+    },
+    {
+      "epoch": 0.48324082750666764,
+      "grad_norm": 0.11571111743344233,
+      "learning_rate": 0.00011027791187689715,
+      "loss": 1.3118,
+      "step": 6704
+    },
+    {
+      "epoch": 0.48331290996900456,
+      "grad_norm": 0.11117556459823182,
+      "learning_rate": 0.00011025468808763519,
+      "loss": 1.378,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4833849924313415,
+      "grad_norm": 0.1070941460992657,
+      "learning_rate": 0.00011023146373940095,
+      "loss": 1.2578,
+      "step": 6706
+    },
+    {
+      "epoch": 0.4834570748936784,
+      "grad_norm": 0.12163832949384037,
+      "learning_rate": 0.00011020823883346043,
+      "loss": 1.3055,
+      "step": 6707
+    },
+    {
+      "epoch": 0.48352915735601526,
+      "grad_norm": 0.10620911073536292,
+      "learning_rate": 0.00011018501337107954,
+      "loss": 1.3233,
+      "step": 6708
+    },
+    {
+      "epoch": 0.4836012398183522,
+      "grad_norm": 0.18541007831453554,
+      "learning_rate": 0.00011016178735352434,
+      "loss": 1.241,
+      "step": 6709
+    },
+    {
+      "epoch": 0.4836733222806891,
+      "grad_norm": 0.11775412113493963,
+      "learning_rate": 0.0001101385607820608,
+      "loss": 1.2453,
+      "step": 6710
+    },
+    {
+      "epoch": 0.483745404743026,
+      "grad_norm": 0.20657295352949362,
+      "learning_rate": 0.00011011533365795498,
+      "loss": 1.2087,
+      "step": 6711
+    },
+    {
+      "epoch": 0.48381748720536294,
+      "grad_norm": 0.11690180662248317,
+      "learning_rate": 0.00011009210598247302,
+      "loss": 1.2674,
+      "step": 6712
+    },
+    {
+      "epoch": 0.48388956966769986,
+      "grad_norm": 0.11958336030747393,
+      "learning_rate": 0.00011006887775688095,
+      "loss": 1.2536,
+      "step": 6713
+    },
+    {
+      "epoch": 0.4839616521300368,
+      "grad_norm": 0.11794650583199211,
+      "learning_rate": 0.00011004564898244498,
+      "loss": 1.3733,
+      "step": 6714
+    },
+    {
+      "epoch": 0.4840337345923737,
+      "grad_norm": 0.10963774109255373,
+      "learning_rate": 0.00011002241966043128,
+      "loss": 1.2876,
+      "step": 6715
+    },
+    {
+      "epoch": 0.48410581705471056,
+      "grad_norm": 0.11982493843287786,
+      "learning_rate": 0.00010999918979210604,
+      "loss": 1.2829,
+      "step": 6716
+    },
+    {
+      "epoch": 0.4841778995170475,
+      "grad_norm": 0.10969184409396843,
+      "learning_rate": 0.0001099759593787355,
+      "loss": 1.2611,
+      "step": 6717
+    },
+    {
+      "epoch": 0.4842499819793844,
+      "grad_norm": 0.11639842948676624,
+      "learning_rate": 0.00010995272842158593,
+      "loss": 1.3564,
+      "step": 6718
+    },
+    {
+      "epoch": 0.4843220644417213,
+      "grad_norm": 0.11818817824670831,
+      "learning_rate": 0.00010992949692192362,
+      "loss": 1.2086,
+      "step": 6719
+    },
+    {
+      "epoch": 0.48439414690405824,
+      "grad_norm": 0.11772707632839953,
+      "learning_rate": 0.0001099062648810149,
+      "loss": 1.3377,
+      "step": 6720
+    },
+    {
+      "epoch": 0.48446622936639516,
+      "grad_norm": 0.12226402557724682,
+      "learning_rate": 0.00010988303230012613,
+      "loss": 1.3409,
+      "step": 6721
+    },
+    {
+      "epoch": 0.4845383118287321,
+      "grad_norm": 0.1346169695724381,
+      "learning_rate": 0.00010985979918052364,
+      "loss": 1.3115,
+      "step": 6722
+    },
+    {
+      "epoch": 0.484610394291069,
+      "grad_norm": 0.11866881964120574,
+      "learning_rate": 0.00010983656552347394,
+      "loss": 1.2273,
+      "step": 6723
+    },
+    {
+      "epoch": 0.4846824767534059,
+      "grad_norm": 0.1179204118053401,
+      "learning_rate": 0.00010981333133024337,
+      "loss": 1.3097,
+      "step": 6724
+    },
+    {
+      "epoch": 0.4847545592157428,
+      "grad_norm": 0.11211113886421821,
+      "learning_rate": 0.00010979009660209849,
+      "loss": 1.2765,
+      "step": 6725
+    },
+    {
+      "epoch": 0.4848266416780797,
+      "grad_norm": 0.11509953432009283,
+      "learning_rate": 0.00010976686134030574,
+      "loss": 1.3225,
+      "step": 6726
+    },
+    {
+      "epoch": 0.4848987241404166,
+      "grad_norm": 0.11745896851248416,
+      "learning_rate": 0.0001097436255461317,
+      "loss": 1.3181,
+      "step": 6727
+    },
+    {
+      "epoch": 0.48497080660275355,
+      "grad_norm": 0.10847494603200637,
+      "learning_rate": 0.00010972038922084283,
+      "loss": 1.3094,
+      "step": 6728
+    },
+    {
+      "epoch": 0.48504288906509047,
+      "grad_norm": 0.10788281287222388,
+      "learning_rate": 0.00010969715236570586,
+      "loss": 1.2645,
+      "step": 6729
+    },
+    {
+      "epoch": 0.4851149715274274,
+      "grad_norm": 0.108379233292937,
+      "learning_rate": 0.00010967391498198729,
+      "loss": 1.2881,
+      "step": 6730
+    },
+    {
+      "epoch": 0.4851870539897643,
+      "grad_norm": 0.11573867881557215,
+      "learning_rate": 0.00010965067707095382,
+      "loss": 1.2507,
+      "step": 6731
+    },
+    {
+      "epoch": 0.4852591364521012,
+      "grad_norm": 0.11345686751423673,
+      "learning_rate": 0.0001096274386338721,
+      "loss": 1.1911,
+      "step": 6732
+    },
+    {
+      "epoch": 0.4853312189144381,
+      "grad_norm": 0.11816596287665855,
+      "learning_rate": 0.00010960419967200888,
+      "loss": 1.2617,
+      "step": 6733
+    },
+    {
+      "epoch": 0.485403301376775,
+      "grad_norm": 0.12255067840151984,
+      "learning_rate": 0.00010958096018663083,
+      "loss": 1.3459,
+      "step": 6734
+    },
+    {
+      "epoch": 0.48547538383911193,
+      "grad_norm": 0.1148482255986348,
+      "learning_rate": 0.00010955772017900474,
+      "loss": 1.2805,
+      "step": 6735
+    },
+    {
+      "epoch": 0.48554746630144885,
+      "grad_norm": 0.10526656971440132,
+      "learning_rate": 0.00010953447965039737,
+      "loss": 1.2898,
+      "step": 6736
+    },
+    {
+      "epoch": 0.48561954876378577,
+      "grad_norm": 0.128754216655674,
+      "learning_rate": 0.0001095112386020756,
+      "loss": 1.3039,
+      "step": 6737
+    },
+    {
+      "epoch": 0.4856916312261227,
+      "grad_norm": 0.11941666583423508,
+      "learning_rate": 0.00010948799703530621,
+      "loss": 1.2935,
+      "step": 6738
+    },
+    {
+      "epoch": 0.4857637136884596,
+      "grad_norm": 0.12198453049799741,
+      "learning_rate": 0.0001094647549513561,
+      "loss": 1.4503,
+      "step": 6739
+    },
+    {
+      "epoch": 0.48583579615079653,
+      "grad_norm": 0.12296313491205539,
+      "learning_rate": 0.0001094415123514922,
+      "loss": 1.4106,
+      "step": 6740
+    },
+    {
+      "epoch": 0.48590787861313345,
+      "grad_norm": 0.13756675216523614,
+      "learning_rate": 0.00010941826923698143,
+      "loss": 1.3422,
+      "step": 6741
+    },
+    {
+      "epoch": 0.4859799610754703,
+      "grad_norm": 0.11028618369384587,
+      "learning_rate": 0.00010939502560909065,
+      "loss": 1.2947,
+      "step": 6742
+    },
+    {
+      "epoch": 0.48605204353780723,
+      "grad_norm": 0.12061037880781181,
+      "learning_rate": 0.000109371781469087,
+      "loss": 1.2243,
+      "step": 6743
+    },
+    {
+      "epoch": 0.48612412600014415,
+      "grad_norm": 0.14768591952596508,
+      "learning_rate": 0.00010934853681823736,
+      "loss": 1.2357,
+      "step": 6744
+    },
+    {
+      "epoch": 0.48619620846248107,
+      "grad_norm": 0.10677490637862676,
+      "learning_rate": 0.00010932529165780887,
+      "loss": 1.2738,
+      "step": 6745
+    },
+    {
+      "epoch": 0.486268290924818,
+      "grad_norm": 0.1117442150070134,
+      "learning_rate": 0.0001093020459890685,
+      "loss": 1.3279,
+      "step": 6746
+    },
+    {
+      "epoch": 0.4863403733871549,
+      "grad_norm": 0.11478700226388606,
+      "learning_rate": 0.00010927879981328348,
+      "loss": 1.2911,
+      "step": 6747
+    },
+    {
+      "epoch": 0.48641245584949183,
+      "grad_norm": 0.11531100066298801,
+      "learning_rate": 0.0001092555531317208,
+      "loss": 1.312,
+      "step": 6748
+    },
+    {
+      "epoch": 0.48648453831182875,
+      "grad_norm": 0.12228790864837116,
+      "learning_rate": 0.00010923230594564771,
+      "loss": 1.2544,
+      "step": 6749
+    },
+    {
+      "epoch": 0.48655662077416567,
+      "grad_norm": 0.1234759506509443,
+      "learning_rate": 0.00010920905825633132,
+      "loss": 1.2638,
+      "step": 6750
+    },
+    {
+      "epoch": 0.48662870323650254,
+      "grad_norm": 0.1126290181373236,
+      "learning_rate": 0.0001091858100650389,
+      "loss": 1.2539,
+      "step": 6751
+    },
+    {
+      "epoch": 0.48670078569883946,
+      "grad_norm": 0.11359668195468578,
+      "learning_rate": 0.00010916256137303764,
+      "loss": 1.3091,
+      "step": 6752
+    },
+    {
+      "epoch": 0.4867728681611764,
+      "grad_norm": 0.11143017374124674,
+      "learning_rate": 0.00010913931218159481,
+      "loss": 1.2677,
+      "step": 6753
+    },
+    {
+      "epoch": 0.4868449506235133,
+      "grad_norm": 0.12524006725832637,
+      "learning_rate": 0.00010911606249197773,
+      "loss": 1.2216,
+      "step": 6754
+    },
+    {
+      "epoch": 0.4869170330858502,
+      "grad_norm": 0.10234069677833989,
+      "learning_rate": 0.00010909281230545365,
+      "loss": 1.2075,
+      "step": 6755
+    },
+    {
+      "epoch": 0.48698911554818713,
+      "grad_norm": 0.11823059901679571,
+      "learning_rate": 0.00010906956162328996,
+      "loss": 1.2293,
+      "step": 6756
+    },
+    {
+      "epoch": 0.48706119801052405,
+      "grad_norm": 0.1149588026132385,
+      "learning_rate": 0.00010904631044675404,
+      "loss": 1.2716,
+      "step": 6757
+    },
+    {
+      "epoch": 0.487133280472861,
+      "grad_norm": 0.10914749711507159,
+      "learning_rate": 0.00010902305877711325,
+      "loss": 1.3383,
+      "step": 6758
+    },
+    {
+      "epoch": 0.48720536293519784,
+      "grad_norm": 0.10814980500586929,
+      "learning_rate": 0.00010899980661563505,
+      "loss": 1.3476,
+      "step": 6759
+    },
+    {
+      "epoch": 0.48727744539753476,
+      "grad_norm": 0.2444905431432667,
+      "learning_rate": 0.00010897655396358688,
+      "loss": 1.2196,
+      "step": 6760
+    },
+    {
+      "epoch": 0.4873495278598717,
+      "grad_norm": 0.15515953627525284,
+      "learning_rate": 0.0001089533008222362,
+      "loss": 1.2621,
+      "step": 6761
+    },
+    {
+      "epoch": 0.4874216103222086,
+      "grad_norm": 0.1015593127820306,
+      "learning_rate": 0.00010893004719285054,
+      "loss": 1.2573,
+      "step": 6762
+    },
+    {
+      "epoch": 0.4874936927845455,
+      "grad_norm": 0.11710132167017417,
+      "learning_rate": 0.0001089067930766974,
+      "loss": 1.2924,
+      "step": 6763
+    },
+    {
+      "epoch": 0.48756577524688244,
+      "grad_norm": 0.11678077176022073,
+      "learning_rate": 0.00010888353847504436,
+      "loss": 1.2871,
+      "step": 6764
+    },
+    {
+      "epoch": 0.48763785770921936,
+      "grad_norm": 0.10704141756961287,
+      "learning_rate": 0.00010886028338915899,
+      "loss": 1.2769,
+      "step": 6765
+    },
+    {
+      "epoch": 0.4877099401715563,
+      "grad_norm": 0.13465563188496532,
+      "learning_rate": 0.00010883702782030894,
+      "loss": 1.2598,
+      "step": 6766
+    },
+    {
+      "epoch": 0.4877820226338932,
+      "grad_norm": 0.10844341584540582,
+      "learning_rate": 0.00010881377176976178,
+      "loss": 1.2367,
+      "step": 6767
+    },
+    {
+      "epoch": 0.48785410509623006,
+      "grad_norm": 0.12294406269364307,
+      "learning_rate": 0.00010879051523878522,
+      "loss": 1.216,
+      "step": 6768
+    },
+    {
+      "epoch": 0.487926187558567,
+      "grad_norm": 0.11932305089527712,
+      "learning_rate": 0.00010876725822864695,
+      "loss": 1.3331,
+      "step": 6769
+    },
+    {
+      "epoch": 0.4879982700209039,
+      "grad_norm": 0.12367495854758476,
+      "learning_rate": 0.00010874400074061466,
+      "loss": 1.2758,
+      "step": 6770
+    },
+    {
+      "epoch": 0.4880703524832408,
+      "grad_norm": 0.11948570860133968,
+      "learning_rate": 0.0001087207427759561,
+      "loss": 1.2092,
+      "step": 6771
+    },
+    {
+      "epoch": 0.48814243494557774,
+      "grad_norm": 0.10919702541213241,
+      "learning_rate": 0.00010869748433593905,
+      "loss": 1.2671,
+      "step": 6772
+    },
+    {
+      "epoch": 0.48821451740791466,
+      "grad_norm": 0.11581071775711974,
+      "learning_rate": 0.0001086742254218313,
+      "loss": 1.2457,
+      "step": 6773
+    },
+    {
+      "epoch": 0.4882865998702516,
+      "grad_norm": 0.12630293520632627,
+      "learning_rate": 0.00010865096603490066,
+      "loss": 1.2659,
+      "step": 6774
+    },
+    {
+      "epoch": 0.4883586823325885,
+      "grad_norm": 0.11528899984407981,
+      "learning_rate": 0.00010862770617641495,
+      "loss": 1.1927,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4884307647949254,
+      "grad_norm": 0.12155869757620753,
+      "learning_rate": 0.00010860444584764211,
+      "loss": 1.2506,
+      "step": 6776
+    },
+    {
+      "epoch": 0.4885028472572623,
+      "grad_norm": 0.10778585586156092,
+      "learning_rate": 0.00010858118504984995,
+      "loss": 1.2124,
+      "step": 6777
+    },
+    {
+      "epoch": 0.4885749297195992,
+      "grad_norm": 0.12277825427860099,
+      "learning_rate": 0.0001085579237843065,
+      "loss": 1.2996,
+      "step": 6778
+    },
+    {
+      "epoch": 0.4886470121819361,
+      "grad_norm": 0.10596066028674372,
+      "learning_rate": 0.00010853466205227955,
+      "loss": 1.1841,
+      "step": 6779
+    },
+    {
+      "epoch": 0.48871909464427304,
+      "grad_norm": 0.13691651404832295,
+      "learning_rate": 0.00010851139985503724,
+      "loss": 1.3186,
+      "step": 6780
+    },
+    {
+      "epoch": 0.48879117710660996,
+      "grad_norm": 0.10303587005178277,
+      "learning_rate": 0.00010848813719384745,
+      "loss": 1.2868,
+      "step": 6781
+    },
+    {
+      "epoch": 0.4888632595689469,
+      "grad_norm": 0.1289208058221435,
+      "learning_rate": 0.0001084648740699783,
+      "loss": 1.2199,
+      "step": 6782
+    },
+    {
+      "epoch": 0.4889353420312838,
+      "grad_norm": 0.1145896005841746,
+      "learning_rate": 0.00010844161048469772,
+      "loss": 1.2821,
+      "step": 6783
+    },
+    {
+      "epoch": 0.4890074244936207,
+      "grad_norm": 0.16840108133552367,
+      "learning_rate": 0.00010841834643927389,
+      "loss": 1.2347,
+      "step": 6784
+    },
+    {
+      "epoch": 0.4890795069559576,
+      "grad_norm": 0.1260523961088543,
+      "learning_rate": 0.00010839508193497485,
+      "loss": 1.3472,
+      "step": 6785
+    },
+    {
+      "epoch": 0.4891515894182945,
+      "grad_norm": 0.16172651235233707,
+      "learning_rate": 0.00010837181697306875,
+      "loss": 1.3206,
+      "step": 6786
+    },
+    {
+      "epoch": 0.4892236718806314,
+      "grad_norm": 0.1092991212338904,
+      "learning_rate": 0.00010834855155482373,
+      "loss": 1.2728,
+      "step": 6787
+    },
+    {
+      "epoch": 0.48929575434296835,
+      "grad_norm": 0.1041826453859726,
+      "learning_rate": 0.00010832528568150795,
+      "loss": 1.2412,
+      "step": 6788
+    },
+    {
+      "epoch": 0.48936783680530527,
+      "grad_norm": 0.11998241648878576,
+      "learning_rate": 0.00010830201935438961,
+      "loss": 1.2384,
+      "step": 6789
+    },
+    {
+      "epoch": 0.4894399192676422,
+      "grad_norm": 0.20445150416230753,
+      "learning_rate": 0.00010827875257473697,
+      "loss": 1.1935,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4895120017299791,
+      "grad_norm": 0.1111254498071482,
+      "learning_rate": 0.00010825548534381824,
+      "loss": 1.3116,
+      "step": 6791
+    },
+    {
+      "epoch": 0.489584084192316,
+      "grad_norm": 0.11217687394354275,
+      "learning_rate": 0.0001082322176629017,
+      "loss": 1.3061,
+      "step": 6792
+    },
+    {
+      "epoch": 0.48965616665465295,
+      "grad_norm": 0.10587699927092915,
+      "learning_rate": 0.00010820894953325564,
+      "loss": 1.2538,
+      "step": 6793
+    },
+    {
+      "epoch": 0.4897282491169898,
+      "grad_norm": 0.12295878588538568,
+      "learning_rate": 0.00010818568095614841,
+      "loss": 1.3453,
+      "step": 6794
+    },
+    {
+      "epoch": 0.48980033157932673,
+      "grad_norm": 0.11114117857929862,
+      "learning_rate": 0.00010816241193284833,
+      "loss": 1.2425,
+      "step": 6795
+    },
+    {
+      "epoch": 0.48987241404166365,
+      "grad_norm": 0.11800271918326383,
+      "learning_rate": 0.00010813914246462377,
+      "loss": 1.3728,
+      "step": 6796
+    },
+    {
+      "epoch": 0.48994449650400057,
+      "grad_norm": 0.10573707739934553,
+      "learning_rate": 0.00010811587255274313,
+      "loss": 1.2897,
+      "step": 6797
+    },
+    {
+      "epoch": 0.4900165789663375,
+      "grad_norm": 0.11851933021176159,
+      "learning_rate": 0.00010809260219847485,
+      "loss": 1.2826,
+      "step": 6798
+    },
+    {
+      "epoch": 0.4900886614286744,
+      "grad_norm": 0.13123707481657415,
+      "learning_rate": 0.00010806933140308733,
+      "loss": 1.3026,
+      "step": 6799
+    },
+    {
+      "epoch": 0.49016074389101133,
+      "grad_norm": 0.12605712592314458,
+      "learning_rate": 0.00010804606016784908,
+      "loss": 1.3648,
+      "step": 6800
+    },
+    {
+      "epoch": 0.49023282635334825,
+      "grad_norm": 0.19418223649896546,
+      "learning_rate": 0.00010802278849402854,
+      "loss": 1.2039,
+      "step": 6801
+    },
+    {
+      "epoch": 0.49030490881568517,
+      "grad_norm": 0.11449773999669557,
+      "learning_rate": 0.00010799951638289427,
+      "loss": 1.3986,
+      "step": 6802
+    },
+    {
+      "epoch": 0.49037699127802203,
+      "grad_norm": 0.13249216044266246,
+      "learning_rate": 0.00010797624383571476,
+      "loss": 1.2504,
+      "step": 6803
+    },
+    {
+      "epoch": 0.49044907374035895,
+      "grad_norm": 0.11684864516886379,
+      "learning_rate": 0.00010795297085375866,
+      "loss": 1.2762,
+      "step": 6804
+    },
+    {
+      "epoch": 0.4905211562026959,
+      "grad_norm": 0.11269600948503311,
+      "learning_rate": 0.00010792969743829444,
+      "loss": 1.3103,
+      "step": 6805
+    },
+    {
+      "epoch": 0.4905932386650328,
+      "grad_norm": 0.13076698778318294,
+      "learning_rate": 0.00010790642359059085,
+      "loss": 1.2942,
+      "step": 6806
+    },
+    {
+      "epoch": 0.4906653211273697,
+      "grad_norm": 0.11897524513996839,
+      "learning_rate": 0.00010788314931191637,
+      "loss": 1.315,
+      "step": 6807
+    },
+    {
+      "epoch": 0.49073740358970663,
+      "grad_norm": 0.11545025320203194,
+      "learning_rate": 0.00010785987460353975,
+      "loss": 1.315,
+      "step": 6808
+    },
+    {
+      "epoch": 0.49080948605204355,
+      "grad_norm": 0.11694700555448548,
+      "learning_rate": 0.00010783659946672964,
+      "loss": 1.2101,
+      "step": 6809
+    },
+    {
+      "epoch": 0.4908815685143805,
+      "grad_norm": 0.11303514988276292,
+      "learning_rate": 0.00010781332390275476,
+      "loss": 1.2687,
+      "step": 6810
+    },
+    {
+      "epoch": 0.49095365097671734,
+      "grad_norm": 0.10971820025771965,
+      "learning_rate": 0.00010779004791288383,
+      "loss": 1.2658,
+      "step": 6811
+    },
+    {
+      "epoch": 0.49102573343905426,
+      "grad_norm": 0.12634982181147728,
+      "learning_rate": 0.00010776677149838559,
+      "loss": 1.3211,
+      "step": 6812
+    },
+    {
+      "epoch": 0.4910978159013912,
+      "grad_norm": 0.1254575286851772,
+      "learning_rate": 0.00010774349466052882,
+      "loss": 1.2471,
+      "step": 6813
+    },
+    {
+      "epoch": 0.4911698983637281,
+      "grad_norm": 0.10991600320228413,
+      "learning_rate": 0.00010772021740058231,
+      "loss": 1.3257,
+      "step": 6814
+    },
+    {
+      "epoch": 0.491241980826065,
+      "grad_norm": 0.11678631899097877,
+      "learning_rate": 0.0001076969397198149,
+      "loss": 1.2696,
+      "step": 6815
+    },
+    {
+      "epoch": 0.49131406328840194,
+      "grad_norm": 0.11007096127760886,
+      "learning_rate": 0.00010767366161949542,
+      "loss": 1.2331,
+      "step": 6816
+    },
+    {
+      "epoch": 0.49138614575073886,
+      "grad_norm": 0.10683332748964164,
+      "learning_rate": 0.0001076503831008927,
+      "loss": 1.3255,
+      "step": 6817
+    },
+    {
+      "epoch": 0.4914582282130758,
+      "grad_norm": 0.11542381687014025,
+      "learning_rate": 0.00010762710416527571,
+      "loss": 1.3107,
+      "step": 6818
+    },
+    {
+      "epoch": 0.4915303106754127,
+      "grad_norm": 0.13711088642103195,
+      "learning_rate": 0.00010760382481391325,
+      "loss": 1.2297,
+      "step": 6819
+    },
+    {
+      "epoch": 0.49160239313774956,
+      "grad_norm": 0.11367402329570657,
+      "learning_rate": 0.00010758054504807435,
+      "loss": 1.3513,
+      "step": 6820
+    },
+    {
+      "epoch": 0.4916744756000865,
+      "grad_norm": 0.10481323379486192,
+      "learning_rate": 0.0001075572648690279,
+      "loss": 1.1772,
+      "step": 6821
+    },
+    {
+      "epoch": 0.4917465580624234,
+      "grad_norm": 0.10603385491778457,
+      "learning_rate": 0.00010753398427804293,
+      "loss": 1.2754,
+      "step": 6822
+    },
+    {
+      "epoch": 0.4918186405247603,
+      "grad_norm": 0.11115073448550458,
+      "learning_rate": 0.0001075107032763884,
+      "loss": 1.2809,
+      "step": 6823
+    },
+    {
+      "epoch": 0.49189072298709724,
+      "grad_norm": 0.1180878674836699,
+      "learning_rate": 0.00010748742186533338,
+      "loss": 1.2425,
+      "step": 6824
+    },
+    {
+      "epoch": 0.49196280544943416,
+      "grad_norm": 0.11035177287595509,
+      "learning_rate": 0.00010746414004614683,
+      "loss": 1.4545,
+      "step": 6825
+    },
+    {
+      "epoch": 0.4920348879117711,
+      "grad_norm": 0.12618769756503598,
+      "learning_rate": 0.00010744085782009792,
+      "loss": 1.1661,
+      "step": 6826
+    },
+    {
+      "epoch": 0.492106970374108,
+      "grad_norm": 0.18087337673828593,
+      "learning_rate": 0.00010741757518845565,
+      "loss": 1.3043,
+      "step": 6827
+    },
+    {
+      "epoch": 0.4921790528364449,
+      "grad_norm": 0.11708722394575866,
+      "learning_rate": 0.00010739429215248922,
+      "loss": 1.2373,
+      "step": 6828
+    },
+    {
+      "epoch": 0.4922511352987818,
+      "grad_norm": 0.10141647280121105,
+      "learning_rate": 0.0001073710087134677,
+      "loss": 1.2884,
+      "step": 6829
+    },
+    {
+      "epoch": 0.4923232177611187,
+      "grad_norm": 0.10969457388007585,
+      "learning_rate": 0.00010734772487266026,
+      "loss": 1.3647,
+      "step": 6830
+    },
+    {
+      "epoch": 0.4923953002234556,
+      "grad_norm": 0.11378573158175849,
+      "learning_rate": 0.00010732444063133607,
+      "loss": 1.2607,
+      "step": 6831
+    },
+    {
+      "epoch": 0.49246738268579254,
+      "grad_norm": 0.11885416175276062,
+      "learning_rate": 0.00010730115599076434,
+      "loss": 1.2543,
+      "step": 6832
+    },
+    {
+      "epoch": 0.49253946514812946,
+      "grad_norm": 0.10321282629844597,
+      "learning_rate": 0.00010727787095221429,
+      "loss": 1.2779,
+      "step": 6833
+    },
+    {
+      "epoch": 0.4926115476104664,
+      "grad_norm": 0.11806315164889812,
+      "learning_rate": 0.00010725458551695515,
+      "loss": 1.2282,
+      "step": 6834
+    },
+    {
+      "epoch": 0.4926836300728033,
+      "grad_norm": 0.11725491739862538,
+      "learning_rate": 0.00010723129968625623,
+      "loss": 1.2452,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4927557125351402,
+      "grad_norm": 0.11429542476093173,
+      "learning_rate": 0.00010720801346138676,
+      "loss": 1.2398,
+      "step": 6836
+    },
+    {
+      "epoch": 0.4928277949974771,
+      "grad_norm": 0.12310585928371179,
+      "learning_rate": 0.00010718472684361609,
+      "loss": 1.2628,
+      "step": 6837
+    },
+    {
+      "epoch": 0.492899877459814,
+      "grad_norm": 0.11287961212558892,
+      "learning_rate": 0.00010716143983421352,
+      "loss": 1.3367,
+      "step": 6838
+    },
+    {
+      "epoch": 0.4929719599221509,
+      "grad_norm": 0.2800499960259994,
+      "learning_rate": 0.00010713815243444841,
+      "loss": 1.2824,
+      "step": 6839
+    },
+    {
+      "epoch": 0.49304404238448785,
+      "grad_norm": 0.11381867075657004,
+      "learning_rate": 0.00010711486464559013,
+      "loss": 1.3446,
+      "step": 6840
+    },
+    {
+      "epoch": 0.49311612484682477,
+      "grad_norm": 0.1252115890266457,
+      "learning_rate": 0.00010709157646890809,
+      "loss": 1.2895,
+      "step": 6841
+    },
+    {
+      "epoch": 0.4931882073091617,
+      "grad_norm": 0.12265315683020213,
+      "learning_rate": 0.00010706828790567167,
+      "loss": 1.2675,
+      "step": 6842
+    },
+    {
+      "epoch": 0.4932602897714986,
+      "grad_norm": 0.11142364266529224,
+      "learning_rate": 0.00010704499895715032,
+      "loss": 1.3427,
+      "step": 6843
+    },
+    {
+      "epoch": 0.4933323722338355,
+      "grad_norm": 0.11045587060172599,
+      "learning_rate": 0.00010702170962461354,
+      "loss": 1.2035,
+      "step": 6844
+    },
+    {
+      "epoch": 0.49340445469617245,
+      "grad_norm": 0.12871661965114023,
+      "learning_rate": 0.00010699841990933073,
+      "loss": 1.3662,
+      "step": 6845
+    },
+    {
+      "epoch": 0.4934765371585093,
+      "grad_norm": 0.11582282340683116,
+      "learning_rate": 0.00010697512981257143,
+      "loss": 1.2886,
+      "step": 6846
+    },
+    {
+      "epoch": 0.49354861962084623,
+      "grad_norm": 0.11987268169283884,
+      "learning_rate": 0.00010695183933560515,
+      "loss": 1.399,
+      "step": 6847
+    },
+    {
+      "epoch": 0.49362070208318315,
+      "grad_norm": 0.12591689803531442,
+      "learning_rate": 0.00010692854847970147,
+      "loss": 1.2618,
+      "step": 6848
+    },
+    {
+      "epoch": 0.49369278454552007,
+      "grad_norm": 0.11361364471251292,
+      "learning_rate": 0.00010690525724612987,
+      "loss": 1.3458,
+      "step": 6849
+    },
+    {
+      "epoch": 0.493764867007857,
+      "grad_norm": 0.12045347988047232,
+      "learning_rate": 0.00010688196563616,
+      "loss": 1.1959,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4938369494701939,
+      "grad_norm": 0.12522734995092094,
+      "learning_rate": 0.00010685867365106143,
+      "loss": 1.2013,
+      "step": 6851
+    },
+    {
+      "epoch": 0.49390903193253083,
+      "grad_norm": 0.11550724624821057,
+      "learning_rate": 0.00010683538129210377,
+      "loss": 1.2569,
+      "step": 6852
+    },
+    {
+      "epoch": 0.49398111439486775,
+      "grad_norm": 0.1176304730643414,
+      "learning_rate": 0.00010681208856055671,
+      "loss": 1.3448,
+      "step": 6853
+    },
+    {
+      "epoch": 0.49405319685720467,
+      "grad_norm": 0.11263218052337659,
+      "learning_rate": 0.00010678879545768985,
+      "loss": 1.2507,
+      "step": 6854
+    },
+    {
+      "epoch": 0.49412527931954153,
+      "grad_norm": 0.10790991024426275,
+      "learning_rate": 0.00010676550198477293,
+      "loss": 1.2673,
+      "step": 6855
+    },
+    {
+      "epoch": 0.49419736178187845,
+      "grad_norm": 0.12430009481599463,
+      "learning_rate": 0.00010674220814307559,
+      "loss": 1.3238,
+      "step": 6856
+    },
+    {
+      "epoch": 0.4942694442442154,
+      "grad_norm": 0.1204146616425452,
+      "learning_rate": 0.00010671891393386764,
+      "loss": 1.2926,
+      "step": 6857
+    },
+    {
+      "epoch": 0.4943415267065523,
+      "grad_norm": 0.1258695062428468,
+      "learning_rate": 0.00010669561935841871,
+      "loss": 1.3561,
+      "step": 6858
+    },
+    {
+      "epoch": 0.4944136091688892,
+      "grad_norm": 0.11238291281488844,
+      "learning_rate": 0.00010667232441799867,
+      "loss": 1.2107,
+      "step": 6859
+    },
+    {
+      "epoch": 0.49448569163122613,
+      "grad_norm": 0.14064999180265675,
+      "learning_rate": 0.00010664902911387722,
+      "loss": 1.3533,
+      "step": 6860
+    },
+    {
+      "epoch": 0.49455777409356305,
+      "grad_norm": 0.11895364712753051,
+      "learning_rate": 0.00010662573344732425,
+      "loss": 1.2795,
+      "step": 6861
+    },
+    {
+      "epoch": 0.49462985655589997,
+      "grad_norm": 0.12076537674603803,
+      "learning_rate": 0.00010660243741960952,
+      "loss": 1.3036,
+      "step": 6862
+    },
+    {
+      "epoch": 0.49470193901823684,
+      "grad_norm": 0.11776780163590178,
+      "learning_rate": 0.00010657914103200286,
+      "loss": 1.2484,
+      "step": 6863
+    },
+    {
+      "epoch": 0.49477402148057376,
+      "grad_norm": 0.10551161721084648,
+      "learning_rate": 0.00010655584428577415,
+      "loss": 1.2013,
+      "step": 6864
+    },
+    {
+      "epoch": 0.4948461039429107,
+      "grad_norm": 0.11967204879822464,
+      "learning_rate": 0.0001065325471821933,
+      "loss": 1.3351,
+      "step": 6865
+    },
+    {
+      "epoch": 0.4949181864052476,
+      "grad_norm": 0.120952019126291,
+      "learning_rate": 0.00010650924972253016,
+      "loss": 1.3757,
+      "step": 6866
+    },
+    {
+      "epoch": 0.4949902688675845,
+      "grad_norm": 0.12248146095661473,
+      "learning_rate": 0.00010648595190805471,
+      "loss": 1.3135,
+      "step": 6867
+    },
+    {
+      "epoch": 0.49506235132992144,
+      "grad_norm": 0.30856788175316036,
+      "learning_rate": 0.00010646265374003683,
+      "loss": 1.2429,
+      "step": 6868
+    },
+    {
+      "epoch": 0.49513443379225835,
+      "grad_norm": 0.1107488511889901,
+      "learning_rate": 0.00010643935521974651,
+      "loss": 1.247,
+      "step": 6869
+    },
+    {
+      "epoch": 0.4952065162545953,
+      "grad_norm": 0.11230032222510844,
+      "learning_rate": 0.00010641605634845373,
+      "loss": 1.3098,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4952785987169322,
+      "grad_norm": 0.12296773195699903,
+      "learning_rate": 0.00010639275712742847,
+      "loss": 1.2731,
+      "step": 6871
+    },
+    {
+      "epoch": 0.49535068117926906,
+      "grad_norm": 0.1362369429106335,
+      "learning_rate": 0.00010636945755794074,
+      "loss": 1.274,
+      "step": 6872
+    },
+    {
+      "epoch": 0.495422763641606,
+      "grad_norm": 0.1116051281755247,
+      "learning_rate": 0.00010634615764126063,
+      "loss": 1.341,
+      "step": 6873
+    },
+    {
+      "epoch": 0.4954948461039429,
+      "grad_norm": 0.12115866490862838,
+      "learning_rate": 0.00010632285737865813,
+      "loss": 1.2296,
+      "step": 6874
+    },
+    {
+      "epoch": 0.4955669285662798,
+      "grad_norm": 0.12297757842131547,
+      "learning_rate": 0.00010629955677140335,
+      "loss": 1.2688,
+      "step": 6875
+    },
+    {
+      "epoch": 0.49563901102861674,
+      "grad_norm": 0.12261724918529951,
+      "learning_rate": 0.00010627625582076634,
+      "loss": 1.2978,
+      "step": 6876
+    },
+    {
+      "epoch": 0.49571109349095366,
+      "grad_norm": 0.10074885610599724,
+      "learning_rate": 0.0001062529545280173,
+      "loss": 1.3096,
+      "step": 6877
+    },
+    {
+      "epoch": 0.4957831759532906,
+      "grad_norm": 0.10610290534379146,
+      "learning_rate": 0.00010622965289442622,
+      "loss": 1.2813,
+      "step": 6878
+    },
+    {
+      "epoch": 0.4958552584156275,
+      "grad_norm": 0.12306865868572389,
+      "learning_rate": 0.00010620635092126338,
+      "loss": 1.2581,
+      "step": 6879
+    },
+    {
+      "epoch": 0.4959273408779644,
+      "grad_norm": 0.12978136745900473,
+      "learning_rate": 0.00010618304860979884,
+      "loss": 1.2678,
+      "step": 6880
+    },
+    {
+      "epoch": 0.4959994233403013,
+      "grad_norm": 0.10541171978355798,
+      "learning_rate": 0.0001061597459613029,
+      "loss": 1.1707,
+      "step": 6881
+    },
+    {
+      "epoch": 0.4960715058026382,
+      "grad_norm": 0.11763545969135332,
+      "learning_rate": 0.00010613644297704563,
+      "loss": 1.3027,
+      "step": 6882
+    },
+    {
+      "epoch": 0.4961435882649751,
+      "grad_norm": 0.11481127081276359,
+      "learning_rate": 0.00010611313965829738,
+      "loss": 1.1956,
+      "step": 6883
+    },
+    {
+      "epoch": 0.49621567072731204,
+      "grad_norm": 0.11557470499234863,
+      "learning_rate": 0.00010608983600632831,
+      "loss": 1.3319,
+      "step": 6884
+    },
+    {
+      "epoch": 0.49628775318964896,
+      "grad_norm": 0.12787888892549099,
+      "learning_rate": 0.00010606653202240871,
+      "loss": 1.2284,
+      "step": 6885
+    },
+    {
+      "epoch": 0.4963598356519859,
+      "grad_norm": 0.10788142162439854,
+      "learning_rate": 0.00010604322770780881,
+      "loss": 1.3532,
+      "step": 6886
+    },
+    {
+      "epoch": 0.4964319181143228,
+      "grad_norm": 0.14491325966446725,
+      "learning_rate": 0.00010601992306379896,
+      "loss": 1.3467,
+      "step": 6887
+    },
+    {
+      "epoch": 0.4965040005766597,
+      "grad_norm": 0.11346495829047057,
+      "learning_rate": 0.00010599661809164943,
+      "loss": 1.3584,
+      "step": 6888
+    },
+    {
+      "epoch": 0.4965760830389966,
+      "grad_norm": 0.12012578940009846,
+      "learning_rate": 0.00010597331279263057,
+      "loss": 1.3486,
+      "step": 6889
+    },
+    {
+      "epoch": 0.4966481655013335,
+      "grad_norm": 0.10841375321913939,
+      "learning_rate": 0.00010595000716801274,
+      "loss": 1.2377,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4967202479636704,
+      "grad_norm": 0.13229651008354618,
+      "learning_rate": 0.00010592670121906625,
+      "loss": 1.2585,
+      "step": 6891
+    },
+    {
+      "epoch": 0.49679233042600734,
+      "grad_norm": 0.16916953789401679,
+      "learning_rate": 0.00010590339494706157,
+      "loss": 1.1375,
+      "step": 6892
+    },
+    {
+      "epoch": 0.49686441288834426,
+      "grad_norm": 0.1213162320307477,
+      "learning_rate": 0.00010588008835326902,
+      "loss": 1.3626,
+      "step": 6893
+    },
+    {
+      "epoch": 0.4969364953506812,
+      "grad_norm": 0.17087668735696548,
+      "learning_rate": 0.00010585678143895906,
+      "loss": 1.2144,
+      "step": 6894
+    },
+    {
+      "epoch": 0.4970085778130181,
+      "grad_norm": 0.1437471742573615,
+      "learning_rate": 0.00010583347420540213,
+      "loss": 1.2918,
+      "step": 6895
+    },
+    {
+      "epoch": 0.497080660275355,
+      "grad_norm": 0.10682779047675818,
+      "learning_rate": 0.00010581016665386862,
+      "loss": 1.2631,
+      "step": 6896
+    },
+    {
+      "epoch": 0.49715274273769194,
+      "grad_norm": 0.10593275967346906,
+      "learning_rate": 0.00010578685878562912,
+      "loss": 1.2985,
+      "step": 6897
+    },
+    {
+      "epoch": 0.4972248252000288,
+      "grad_norm": 0.11160574758252655,
+      "learning_rate": 0.00010576355060195397,
+      "loss": 1.3167,
+      "step": 6898
+    },
+    {
+      "epoch": 0.49729690766236573,
+      "grad_norm": 0.12664028551111028,
+      "learning_rate": 0.00010574024210411382,
+      "loss": 1.3303,
+      "step": 6899
+    },
+    {
+      "epoch": 0.49736899012470265,
+      "grad_norm": 0.13366445754227305,
+      "learning_rate": 0.00010571693329337906,
+      "loss": 1.4026,
+      "step": 6900
+    },
+    {
+      "epoch": 0.49744107258703957,
+      "grad_norm": 0.11601253239219124,
+      "learning_rate": 0.00010569362417102035,
+      "loss": 1.2733,
+      "step": 6901
+    },
+    {
+      "epoch": 0.4975131550493765,
+      "grad_norm": 0.11631289303423299,
+      "learning_rate": 0.00010567031473830813,
+      "loss": 1.2753,
+      "step": 6902
+    },
+    {
+      "epoch": 0.4975852375117134,
+      "grad_norm": 0.12110585403991796,
+      "learning_rate": 0.00010564700499651308,
+      "loss": 1.21,
+      "step": 6903
+    },
+    {
+      "epoch": 0.4976573199740503,
+      "grad_norm": 0.12365706098051729,
+      "learning_rate": 0.00010562369494690569,
+      "loss": 1.2964,
+      "step": 6904
+    },
+    {
+      "epoch": 0.49772940243638725,
+      "grad_norm": 0.10182225207138314,
+      "learning_rate": 0.00010560038459075665,
+      "loss": 1.1918,
+      "step": 6905
+    },
+    {
+      "epoch": 0.49780148489872417,
+      "grad_norm": 0.12479040235841656,
+      "learning_rate": 0.00010557707392933654,
+      "loss": 1.3306,
+      "step": 6906
+    },
+    {
+      "epoch": 0.49787356736106103,
+      "grad_norm": 0.1260761956075266,
+      "learning_rate": 0.00010555376296391599,
+      "loss": 1.2964,
+      "step": 6907
+    },
+    {
+      "epoch": 0.49794564982339795,
+      "grad_norm": 0.11706457558289177,
+      "learning_rate": 0.0001055304516957657,
+      "loss": 1.3377,
+      "step": 6908
+    },
+    {
+      "epoch": 0.49801773228573487,
+      "grad_norm": 0.10883328229377537,
+      "learning_rate": 0.00010550714012615628,
+      "loss": 1.3826,
+      "step": 6909
+    },
+    {
+      "epoch": 0.4980898147480718,
+      "grad_norm": 0.11068094341851113,
+      "learning_rate": 0.00010548382825635847,
+      "loss": 1.3528,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4981618972104087,
+      "grad_norm": 0.11910339854880814,
+      "learning_rate": 0.00010546051608764298,
+      "loss": 1.4055,
+      "step": 6911
+    },
+    {
+      "epoch": 0.49823397967274563,
+      "grad_norm": 0.1063155403666711,
+      "learning_rate": 0.0001054372036212805,
+      "loss": 1.3028,
+      "step": 6912
+    },
+    {
+      "epoch": 0.49830606213508255,
+      "grad_norm": 0.12136599938667825,
+      "learning_rate": 0.00010541389085854176,
+      "loss": 1.2881,
+      "step": 6913
+    },
+    {
+      "epoch": 0.49837814459741947,
+      "grad_norm": 0.11199850056931408,
+      "learning_rate": 0.00010539057780069757,
+      "loss": 1.2311,
+      "step": 6914
+    },
+    {
+      "epoch": 0.49845022705975633,
+      "grad_norm": 0.12492471437260602,
+      "learning_rate": 0.00010536726444901863,
+      "loss": 1.3806,
+      "step": 6915
+    },
+    {
+      "epoch": 0.49852230952209325,
+      "grad_norm": 0.1180757792459564,
+      "learning_rate": 0.00010534395080477577,
+      "loss": 1.2291,
+      "step": 6916
+    },
+    {
+      "epoch": 0.4985943919844302,
+      "grad_norm": 0.121991321462135,
+      "learning_rate": 0.00010532063686923979,
+      "loss": 1.2546,
+      "step": 6917
+    },
+    {
+      "epoch": 0.4986664744467671,
+      "grad_norm": 0.11214869690138675,
+      "learning_rate": 0.00010529732264368151,
+      "loss": 1.2326,
+      "step": 6918
+    },
+    {
+      "epoch": 0.498738556909104,
+      "grad_norm": 0.14496401783964127,
+      "learning_rate": 0.00010527400812937174,
+      "loss": 1.2457,
+      "step": 6919
+    },
+    {
+      "epoch": 0.49881063937144093,
+      "grad_norm": 0.16777101183244772,
+      "learning_rate": 0.00010525069332758135,
+      "loss": 1.3523,
+      "step": 6920
+    },
+    {
+      "epoch": 0.49888272183377785,
+      "grad_norm": 0.11504645697892978,
+      "learning_rate": 0.0001052273782395812,
+      "loss": 1.2729,
+      "step": 6921
+    },
+    {
+      "epoch": 0.4989548042961148,
+      "grad_norm": 0.11371270325848473,
+      "learning_rate": 0.00010520406286664215,
+      "loss": 1.2603,
+      "step": 6922
+    },
+    {
+      "epoch": 0.4990268867584517,
+      "grad_norm": 0.1252545509333843,
+      "learning_rate": 0.00010518074721003514,
+      "loss": 1.3751,
+      "step": 6923
+    },
+    {
+      "epoch": 0.49909896922078856,
+      "grad_norm": 0.14795807425223312,
+      "learning_rate": 0.00010515743127103106,
+      "loss": 1.2495,
+      "step": 6924
+    },
+    {
+      "epoch": 0.4991710516831255,
+      "grad_norm": 0.13334549542590668,
+      "learning_rate": 0.00010513411505090083,
+      "loss": 1.3085,
+      "step": 6925
+    },
+    {
+      "epoch": 0.4992431341454624,
+      "grad_norm": 0.12562689981435787,
+      "learning_rate": 0.00010511079855091539,
+      "loss": 1.3388,
+      "step": 6926
+    },
+    {
+      "epoch": 0.4993152166077993,
+      "grad_norm": 0.11454085097055873,
+      "learning_rate": 0.00010508748177234574,
+      "loss": 1.351,
+      "step": 6927
+    },
+    {
+      "epoch": 0.49938729907013624,
+      "grad_norm": 0.10611206996795167,
+      "learning_rate": 0.00010506416471646278,
+      "loss": 1.2126,
+      "step": 6928
+    },
+    {
+      "epoch": 0.49945938153247316,
+      "grad_norm": 0.10774466362184408,
+      "learning_rate": 0.00010504084738453755,
+      "loss": 1.2579,
+      "step": 6929
+    },
+    {
+      "epoch": 0.4995314639948101,
+      "grad_norm": 0.10773684842386143,
+      "learning_rate": 0.00010501752977784105,
+      "loss": 1.3422,
+      "step": 6930
+    },
+    {
+      "epoch": 0.499603546457147,
+      "grad_norm": 0.10857165935727281,
+      "learning_rate": 0.00010499421189764427,
+      "loss": 1.2499,
+      "step": 6931
+    },
+    {
+      "epoch": 0.4996756289194839,
+      "grad_norm": 0.11471966861522961,
+      "learning_rate": 0.00010497089374521829,
+      "loss": 1.2446,
+      "step": 6932
+    },
+    {
+      "epoch": 0.4997477113818208,
+      "grad_norm": 0.11661136523697468,
+      "learning_rate": 0.0001049475753218341,
+      "loss": 1.2764,
+      "step": 6933
+    },
+    {
+      "epoch": 0.4998197938441577,
+      "grad_norm": 0.11874244887956131,
+      "learning_rate": 0.00010492425662876282,
+      "loss": 1.3055,
+      "step": 6934
+    },
+    {
+      "epoch": 0.4998918763064946,
+      "grad_norm": 0.1057634477233999,
+      "learning_rate": 0.00010490093766727546,
+      "loss": 1.181,
+      "step": 6935
+    },
+    {
+      "epoch": 0.49996395876883154,
+      "grad_norm": 0.12497602079017478,
+      "learning_rate": 0.00010487761843864321,
+      "loss": 1.2072,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5000360412311684,
+      "grad_norm": 0.12795733411701035,
+      "learning_rate": 0.00010485429894413707,
+      "loss": 1.3311,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5001081236935053,
+      "grad_norm": 0.10778050654762039,
+      "learning_rate": 0.00010483097918502824,
+      "loss": 1.3146,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5001802061558422,
+      "grad_norm": 0.1157629374831069,
+      "learning_rate": 0.00010480765916258781,
+      "loss": 1.253,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5002522886181792,
+      "grad_norm": 0.1193445448171705,
+      "learning_rate": 0.00010478433887808696,
+      "loss": 1.2317,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5003243710805161,
+      "grad_norm": 0.11053271910930955,
+      "learning_rate": 0.00010476101833279681,
+      "loss": 1.1575,
+      "step": 6941
+    },
+    {
+      "epoch": 0.500396453542853,
+      "grad_norm": 0.1158403741938708,
+      "learning_rate": 0.00010473769752798859,
+      "loss": 1.3175,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5004685360051899,
+      "grad_norm": 0.10429903451398084,
+      "learning_rate": 0.00010471437646493347,
+      "loss": 1.3294,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5005406184675268,
+      "grad_norm": 0.21738332583854253,
+      "learning_rate": 0.00010469105514490265,
+      "loss": 1.2295,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5006127009298638,
+      "grad_norm": 0.10604131511871959,
+      "learning_rate": 0.00010466773356916737,
+      "loss": 1.3317,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5006847833922007,
+      "grad_norm": 0.11070453350746867,
+      "learning_rate": 0.00010464441173899884,
+      "loss": 1.2139,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5007568658545376,
+      "grad_norm": 0.1157911970974042,
+      "learning_rate": 0.0001046210896556683,
+      "loss": 1.2704,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5008289483168745,
+      "grad_norm": 0.45187643252869647,
+      "learning_rate": 0.00010459776732044706,
+      "loss": 1.2043,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5009010307792114,
+      "grad_norm": 0.11254036766562509,
+      "learning_rate": 0.00010457444473460635,
+      "loss": 1.2543,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5009731132415484,
+      "grad_norm": 0.11909861914075419,
+      "learning_rate": 0.00010455112189941752,
+      "loss": 1.3413,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5010451957038853,
+      "grad_norm": 0.12806471578098585,
+      "learning_rate": 0.00010452779881615177,
+      "loss": 1.2436,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5011172781662222,
+      "grad_norm": 0.10956416917503267,
+      "learning_rate": 0.00010450447548608054,
+      "loss": 1.3122,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5011893606285591,
+      "grad_norm": 0.10728670024470575,
+      "learning_rate": 0.00010448115191047505,
+      "loss": 1.2457,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5012614430908959,
+      "grad_norm": 0.117551005193128,
+      "learning_rate": 0.00010445782809060671,
+      "loss": 1.3251,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5013335255532329,
+      "grad_norm": 0.10561948950354993,
+      "learning_rate": 0.00010443450402774684,
+      "loss": 1.2942,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5014056080155698,
+      "grad_norm": 0.10442584980562998,
+      "learning_rate": 0.00010441117972316688,
+      "loss": 1.3628,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5014776904779067,
+      "grad_norm": 0.11742962701118431,
+      "learning_rate": 0.00010438785517813809,
+      "loss": 1.2508,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5015497729402436,
+      "grad_norm": 0.11249601606805194,
+      "learning_rate": 0.00010436453039393199,
+      "loss": 1.2105,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5016218554025805,
+      "grad_norm": 0.14495667889429326,
+      "learning_rate": 0.0001043412053718199,
+      "loss": 1.3089,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5016939378649174,
+      "grad_norm": 0.11296283939789438,
+      "learning_rate": 0.00010431788011307333,
+      "loss": 1.2603,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5017660203272544,
+      "grad_norm": 0.3212738174490349,
+      "learning_rate": 0.00010429455461896363,
+      "loss": 1.277,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5018381027895913,
+      "grad_norm": 0.10567542953820852,
+      "learning_rate": 0.00010427122889076228,
+      "loss": 1.3008,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5019101852519282,
+      "grad_norm": 0.15122028443145444,
+      "learning_rate": 0.00010424790292974076,
+      "loss": 1.1993,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5019822677142651,
+      "grad_norm": 0.11655119270503408,
+      "learning_rate": 0.00010422457673717052,
+      "loss": 1.2266,
+      "step": 6964
+    },
+    {
+      "epoch": 0.502054350176602,
+      "grad_norm": 0.13267127788869168,
+      "learning_rate": 0.00010420125031432304,
+      "loss": 1.3622,
+      "step": 6965
+    },
+    {
+      "epoch": 0.502126432638939,
+      "grad_norm": 0.11414000132070531,
+      "learning_rate": 0.00010417792366246986,
+      "loss": 1.3517,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5021985151012759,
+      "grad_norm": 0.12079610031581506,
+      "learning_rate": 0.00010415459678288244,
+      "loss": 1.154,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5022705975636128,
+      "grad_norm": 0.11142861226885566,
+      "learning_rate": 0.00010413126967683233,
+      "loss": 1.3521,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5023426800259497,
+      "grad_norm": 0.11180542187212855,
+      "learning_rate": 0.00010410794234559107,
+      "loss": 1.3368,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5024147624882866,
+      "grad_norm": 0.1117856544673813,
+      "learning_rate": 0.00010408461479043021,
+      "loss": 1.2831,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5024868449506236,
+      "grad_norm": 0.11878634679379872,
+      "learning_rate": 0.00010406128701262128,
+      "loss": 1.4089,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5025589274129604,
+      "grad_norm": 0.12895956679673884,
+      "learning_rate": 0.00010403795901343592,
+      "loss": 1.1959,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5026310098752973,
+      "grad_norm": 0.12287259597437529,
+      "learning_rate": 0.00010401463079414562,
+      "loss": 1.3408,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5027030923376342,
+      "grad_norm": 0.1669144284896794,
+      "learning_rate": 0.00010399130235602206,
+      "loss": 1.1912,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5027751747999711,
+      "grad_norm": 0.12319136142680005,
+      "learning_rate": 0.00010396797370033678,
+      "loss": 1.2555,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5028472572623081,
+      "grad_norm": 0.10949442758226624,
+      "learning_rate": 0.00010394464482836146,
+      "loss": 1.2385,
+      "step": 6976
+    },
+    {
+      "epoch": 0.502919339724645,
+      "grad_norm": 0.1269273339285458,
+      "learning_rate": 0.00010392131574136767,
+      "loss": 1.2024,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5029914221869819,
+      "grad_norm": 0.13045458599531365,
+      "learning_rate": 0.00010389798644062714,
+      "loss": 1.2317,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5030635046493188,
+      "grad_norm": 0.10617506174701635,
+      "learning_rate": 0.00010387465692741142,
+      "loss": 1.2862,
+      "step": 6979
+    },
+    {
+      "epoch": 0.5031355871116557,
+      "grad_norm": 0.18083997395279244,
+      "learning_rate": 0.00010385132720299231,
+      "loss": 1.3271,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5032076695739927,
+      "grad_norm": 0.13687415244669107,
+      "learning_rate": 0.00010382799726864135,
+      "loss": 1.2436,
+      "step": 6981
+    },
+    {
+      "epoch": 0.5032797520363296,
+      "grad_norm": 0.10669982956016602,
+      "learning_rate": 0.00010380466712563033,
+      "loss": 1.2033,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5033518344986665,
+      "grad_norm": 0.1409296676026083,
+      "learning_rate": 0.0001037813367752309,
+      "loss": 1.2501,
+      "step": 6983
+    },
+    {
+      "epoch": 0.5034239169610034,
+      "grad_norm": 0.12120034772741203,
+      "learning_rate": 0.00010375800621871478,
+      "loss": 1.2785,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5034959994233403,
+      "grad_norm": 0.11198150114036606,
+      "learning_rate": 0.00010373467545735371,
+      "loss": 1.2992,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5035680818856773,
+      "grad_norm": 0.11834900656504306,
+      "learning_rate": 0.00010371134449241941,
+      "loss": 1.1989,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5036401643480142,
+      "grad_norm": 0.11611235895235254,
+      "learning_rate": 0.00010368801332518365,
+      "loss": 1.3188,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5037122468103511,
+      "grad_norm": 0.12432896498794097,
+      "learning_rate": 0.00010366468195691814,
+      "loss": 1.2797,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5037843292726879,
+      "grad_norm": 0.12197522109275467,
+      "learning_rate": 0.0001036413503888947,
+      "loss": 1.2386,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5038564117350248,
+      "grad_norm": 0.20666758562617563,
+      "learning_rate": 0.0001036180186223851,
+      "loss": 1.2695,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5039284941973617,
+      "grad_norm": 0.11013892985333235,
+      "learning_rate": 0.00010359468665866108,
+      "loss": 1.2654,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5040005766596987,
+      "grad_norm": 0.11097601815313837,
+      "learning_rate": 0.00010357135449899449,
+      "loss": 1.2605,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5040726591220356,
+      "grad_norm": 0.12193783381995935,
+      "learning_rate": 0.00010354802214465715,
+      "loss": 1.3979,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5041447415843725,
+      "grad_norm": 0.10707030727407998,
+      "learning_rate": 0.00010352468959692083,
+      "loss": 1.2741,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5042168240467094,
+      "grad_norm": 0.10687913282270942,
+      "learning_rate": 0.00010350135685705742,
+      "loss": 1.219,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5042889065090463,
+      "grad_norm": 0.12345405174808748,
+      "learning_rate": 0.00010347802392633874,
+      "loss": 1.3272,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5043609889713833,
+      "grad_norm": 0.1255012192858554,
+      "learning_rate": 0.00010345469080603663,
+      "loss": 1.1901,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5044330714337202,
+      "grad_norm": 0.1053038867981158,
+      "learning_rate": 0.00010343135749742296,
+      "loss": 1.2595,
+      "step": 6998
+    },
+    {
+      "epoch": 0.5045051538960571,
+      "grad_norm": 0.1189012588622437,
+      "learning_rate": 0.00010340802400176962,
+      "loss": 1.2259,
+      "step": 6999
+    },
+    {
+      "epoch": 0.504577236358394,
+      "grad_norm": 0.11684965197895306,
+      "learning_rate": 0.00010338469032034845,
+      "loss": 1.2767,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5046493188207309,
+      "grad_norm": 0.11439774207548804,
+      "learning_rate": 0.00010336135645443142,
+      "loss": 1.3205,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5047214012830679,
+      "grad_norm": 0.18527831983049076,
+      "learning_rate": 0.00010333802240529038,
+      "loss": 1.3196,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5047934837454048,
+      "grad_norm": 0.10944244236382766,
+      "learning_rate": 0.00010331468817419725,
+      "loss": 1.2382,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5048655662077417,
+      "grad_norm": 0.10515857764160709,
+      "learning_rate": 0.00010329135376242396,
+      "loss": 1.4424,
+      "step": 7004
+    },
+    {
+      "epoch": 0.5049376486700786,
+      "grad_norm": 0.12378928966502582,
+      "learning_rate": 0.00010326801917124248,
+      "loss": 1.2264,
+      "step": 7005
+    },
+    {
+      "epoch": 0.5050097311324154,
+      "grad_norm": 0.10860782856856664,
+      "learning_rate": 0.00010324468440192464,
+      "loss": 1.2902,
+      "step": 7006
+    },
+    {
+      "epoch": 0.5050818135947523,
+      "grad_norm": 0.11231096761948563,
+      "learning_rate": 0.00010322134945574253,
+      "loss": 1.304,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5051538960570893,
+      "grad_norm": 0.10338694424015128,
+      "learning_rate": 0.00010319801433396802,
+      "loss": 1.3331,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5052259785194262,
+      "grad_norm": 0.1335348822559058,
+      "learning_rate": 0.00010317467903787317,
+      "loss": 1.2348,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5052980609817631,
+      "grad_norm": 0.12313362323528995,
+      "learning_rate": 0.00010315134356872986,
+      "loss": 1.2657,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5053701434441,
+      "grad_norm": 0.10965701046684272,
+      "learning_rate": 0.00010312800792781016,
+      "loss": 1.2295,
+      "step": 7011
+    },
+    {
+      "epoch": 0.505442225906437,
+      "grad_norm": 0.1131393506953316,
+      "learning_rate": 0.00010310467211638603,
+      "loss": 1.2892,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5055143083687739,
+      "grad_norm": 0.12264790745718121,
+      "learning_rate": 0.00010308133613572952,
+      "loss": 1.1712,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5055863908311108,
+      "grad_norm": 0.12438389874048691,
+      "learning_rate": 0.0001030579999871126,
+      "loss": 1.3633,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5056584732934477,
+      "grad_norm": 0.12299103515704593,
+      "learning_rate": 0.00010303466367180735,
+      "loss": 1.3679,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5057305557557846,
+      "grad_norm": 0.13052398880422758,
+      "learning_rate": 0.00010301132719108576,
+      "loss": 1.3171,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5058026382181215,
+      "grad_norm": 0.11439471594034265,
+      "learning_rate": 0.0001029879905462199,
+      "loss": 1.344,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5058747206804585,
+      "grad_norm": 0.1086903681850103,
+      "learning_rate": 0.0001029646537384818,
+      "loss": 1.2586,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5059468031427954,
+      "grad_norm": 0.1940453953716343,
+      "learning_rate": 0.00010294131676914358,
+      "loss": 1.2736,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5060188856051323,
+      "grad_norm": 0.10012727633664513,
+      "learning_rate": 0.00010291797963947729,
+      "loss": 1.2998,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5060909680674692,
+      "grad_norm": 0.10872842425595666,
+      "learning_rate": 0.00010289464235075499,
+      "loss": 1.2338,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5061630505298061,
+      "grad_norm": 0.11267727332706863,
+      "learning_rate": 0.00010287130490424877,
+      "loss": 1.2033,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5062351329921431,
+      "grad_norm": 0.11395706373007196,
+      "learning_rate": 0.00010284796730123076,
+      "loss": 1.352,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5063072154544799,
+      "grad_norm": 0.11843298662977926,
+      "learning_rate": 0.00010282462954297305,
+      "loss": 1.2729,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5063792979168168,
+      "grad_norm": 0.12348909566353686,
+      "learning_rate": 0.00010280129163074778,
+      "loss": 1.3408,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5064513803791537,
+      "grad_norm": 0.11625031654054746,
+      "learning_rate": 0.00010277795356582702,
+      "loss": 1.2338,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5065234628414906,
+      "grad_norm": 0.11500499509704035,
+      "learning_rate": 0.00010275461534948298,
+      "loss": 1.3062,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5065955453038276,
+      "grad_norm": 0.17571402776546083,
+      "learning_rate": 0.00010273127698298772,
+      "loss": 1.1924,
+      "step": 7028
+    },
+    {
+      "epoch": 0.5066676277661645,
+      "grad_norm": 0.11836956817522856,
+      "learning_rate": 0.00010270793846761347,
+      "loss": 1.3234,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5067397102285014,
+      "grad_norm": 0.11340406473735079,
+      "learning_rate": 0.0001026845998046323,
+      "loss": 1.2127,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5068117926908383,
+      "grad_norm": 0.11384954806449944,
+      "learning_rate": 0.00010266126099531648,
+      "loss": 1.26,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5068838751531752,
+      "grad_norm": 0.13063168586775573,
+      "learning_rate": 0.00010263792204093807,
+      "loss": 1.2691,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5069559576155122,
+      "grad_norm": 0.11534704566592087,
+      "learning_rate": 0.00010261458294276935,
+      "loss": 1.3722,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5070280400778491,
+      "grad_norm": 0.11949005061736985,
+      "learning_rate": 0.00010259124370208243,
+      "loss": 1.3583,
+      "step": 7034
+    },
+    {
+      "epoch": 0.507100122540186,
+      "grad_norm": 0.12116280457044948,
+      "learning_rate": 0.0001025679043201496,
+      "loss": 1.2911,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5071722050025229,
+      "grad_norm": 0.11434477456174348,
+      "learning_rate": 0.00010254456479824296,
+      "loss": 1.3152,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5072442874648598,
+      "grad_norm": 0.11980512368442159,
+      "learning_rate": 0.0001025212251376348,
+      "loss": 1.2102,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5073163699271968,
+      "grad_norm": 0.11321216446066355,
+      "learning_rate": 0.00010249788533959732,
+      "loss": 1.2035,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5073884523895337,
+      "grad_norm": 0.12968432206203206,
+      "learning_rate": 0.00010247454540540273,
+      "loss": 1.3067,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5074605348518706,
+      "grad_norm": 0.1238481025173741,
+      "learning_rate": 0.00010245120533632329,
+      "loss": 1.2949,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5075326173142074,
+      "grad_norm": 0.10234043758401795,
+      "learning_rate": 0.00010242786513363121,
+      "loss": 1.2155,
+      "step": 7041
+    },
+    {
+      "epoch": 0.5076046997765443,
+      "grad_norm": 0.10660251531375739,
+      "learning_rate": 0.00010240452479859876,
+      "loss": 1.282,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5076767822388812,
+      "grad_norm": 0.10215016869733715,
+      "learning_rate": 0.00010238118433249822,
+      "loss": 1.2898,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5077488647012182,
+      "grad_norm": 0.11610152329131827,
+      "learning_rate": 0.00010235784373660182,
+      "loss": 1.2666,
+      "step": 7044
+    },
+    {
+      "epoch": 0.5078209471635551,
+      "grad_norm": 0.11149258196168135,
+      "learning_rate": 0.00010233450301218183,
+      "loss": 1.334,
+      "step": 7045
+    },
+    {
+      "epoch": 0.507893029625892,
+      "grad_norm": 0.12843242501085153,
+      "learning_rate": 0.00010231116216051055,
+      "loss": 1.2815,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5079651120882289,
+      "grad_norm": 0.10860019498953792,
+      "learning_rate": 0.00010228782118286028,
+      "loss": 1.3012,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5080371945505658,
+      "grad_norm": 0.14432266164134358,
+      "learning_rate": 0.00010226448008050325,
+      "loss": 1.2399,
+      "step": 7048
+    },
+    {
+      "epoch": 0.5081092770129028,
+      "grad_norm": 0.10987996810541616,
+      "learning_rate": 0.00010224113885471183,
+      "loss": 1.3097,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5081813594752397,
+      "grad_norm": 0.10673568223320415,
+      "learning_rate": 0.00010221779750675824,
+      "loss": 1.3217,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5082534419375766,
+      "grad_norm": 0.12173110775256318,
+      "learning_rate": 0.00010219445603791491,
+      "loss": 1.245,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5083255243999135,
+      "grad_norm": 0.24303579239859635,
+      "learning_rate": 0.00010217111444945405,
+      "loss": 1.1892,
+      "step": 7052
+    },
+    {
+      "epoch": 0.5083976068622504,
+      "grad_norm": 0.11118011838096162,
+      "learning_rate": 0.00010214777274264806,
+      "loss": 1.2825,
+      "step": 7053
+    },
+    {
+      "epoch": 0.5084696893245874,
+      "grad_norm": 0.11821571784318938,
+      "learning_rate": 0.0001021244309187692,
+      "loss": 1.3168,
+      "step": 7054
+    },
+    {
+      "epoch": 0.5085417717869243,
+      "grad_norm": 0.13146786880173733,
+      "learning_rate": 0.0001021010889790899,
+      "loss": 1.2941,
+      "step": 7055
+    },
+    {
+      "epoch": 0.5086138542492612,
+      "grad_norm": 0.12188735318915522,
+      "learning_rate": 0.00010207774692488238,
+      "loss": 1.2555,
+      "step": 7056
+    },
+    {
+      "epoch": 0.5086859367115981,
+      "grad_norm": 0.11049965037600919,
+      "learning_rate": 0.00010205440475741915,
+      "loss": 1.3968,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5087580191739349,
+      "grad_norm": 0.10952335485364043,
+      "learning_rate": 0.00010203106247797243,
+      "loss": 1.1981,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5088301016362718,
+      "grad_norm": 0.11272537100810133,
+      "learning_rate": 0.00010200772008781465,
+      "loss": 1.3891,
+      "step": 7059
+    },
+    {
+      "epoch": 0.5089021840986088,
+      "grad_norm": 0.3398371180736666,
+      "learning_rate": 0.00010198437758821817,
+      "loss": 1.2988,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5089742665609457,
+      "grad_norm": 0.10772010992416078,
+      "learning_rate": 0.00010196103498045535,
+      "loss": 1.2066,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5090463490232826,
+      "grad_norm": 0.11485504393156523,
+      "learning_rate": 0.00010193769226579857,
+      "loss": 1.288,
+      "step": 7062
+    },
+    {
+      "epoch": 0.5091184314856195,
+      "grad_norm": 0.11311272059326001,
+      "learning_rate": 0.00010191434944552022,
+      "loss": 1.3162,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5091905139479564,
+      "grad_norm": 0.12724829346885888,
+      "learning_rate": 0.00010189100652089274,
+      "loss": 1.3268,
+      "step": 7064
+    },
+    {
+      "epoch": 0.5092625964102934,
+      "grad_norm": 0.10659964901553032,
+      "learning_rate": 0.00010186766349318846,
+      "loss": 1.3028,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5093346788726303,
+      "grad_norm": 0.13830652445364297,
+      "learning_rate": 0.00010184432036367981,
+      "loss": 1.2699,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5094067613349672,
+      "grad_norm": 0.1099804446465307,
+      "learning_rate": 0.0001018209771336392,
+      "loss": 1.3081,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5094788437973041,
+      "grad_norm": 0.2002250034992289,
+      "learning_rate": 0.00010179763380433904,
+      "loss": 1.3682,
+      "step": 7068
+    },
+    {
+      "epoch": 0.509550926259641,
+      "grad_norm": 0.10581648982776251,
+      "learning_rate": 0.00010177429037705177,
+      "loss": 1.3218,
+      "step": 7069
+    },
+    {
+      "epoch": 0.509623008721978,
+      "grad_norm": 0.11243308082267349,
+      "learning_rate": 0.0001017509468530498,
+      "loss": 1.3306,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5096950911843149,
+      "grad_norm": 0.10638017943221456,
+      "learning_rate": 0.00010172760323360555,
+      "loss": 1.2495,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5097671736466518,
+      "grad_norm": 0.11269091367366199,
+      "learning_rate": 0.00010170425951999148,
+      "loss": 1.2649,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5098392561089887,
+      "grad_norm": 0.12322922100405526,
+      "learning_rate": 0.00010168091571348002,
+      "loss": 1.3521,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5099113385713256,
+      "grad_norm": 0.11516243510726772,
+      "learning_rate": 0.00010165757181534359,
+      "loss": 1.335,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5099834210336626,
+      "grad_norm": 0.11646459986738819,
+      "learning_rate": 0.0001016342278268547,
+      "loss": 1.3025,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5100555034959994,
+      "grad_norm": 0.1120665178464888,
+      "learning_rate": 0.00010161088374928576,
+      "loss": 1.258,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5101275859583363,
+      "grad_norm": 0.11827599713644568,
+      "learning_rate": 0.00010158753958390923,
+      "loss": 1.1935,
+      "step": 7077
+    },
+    {
+      "epoch": 0.5101996684206732,
+      "grad_norm": 0.11423085898857405,
+      "learning_rate": 0.00010156419533199761,
+      "loss": 1.2602,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5102717508830101,
+      "grad_norm": 0.1179884186428919,
+      "learning_rate": 0.00010154085099482335,
+      "loss": 1.2486,
+      "step": 7079
+    },
+    {
+      "epoch": 0.510343833345347,
+      "grad_norm": 0.1179421188054214,
+      "learning_rate": 0.00010151750657365895,
+      "loss": 1.345,
+      "step": 7080
+    },
+    {
+      "epoch": 0.510415915807684,
+      "grad_norm": 0.11354284556559702,
+      "learning_rate": 0.0001014941620697768,
+      "loss": 1.3086,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5104879982700209,
+      "grad_norm": 0.13094599182934544,
+      "learning_rate": 0.00010147081748444951,
+      "loss": 1.3002,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5105600807323578,
+      "grad_norm": 0.12036786772796915,
+      "learning_rate": 0.00010144747281894946,
+      "loss": 1.3186,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5106321631946947,
+      "grad_norm": 0.09972007885835515,
+      "learning_rate": 0.00010142412807454924,
+      "loss": 1.2143,
+      "step": 7084
+    },
+    {
+      "epoch": 0.5107042456570317,
+      "grad_norm": 0.11555788649033709,
+      "learning_rate": 0.00010140078325252127,
+      "loss": 1.2786,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5107763281193686,
+      "grad_norm": 0.1125270273639237,
+      "learning_rate": 0.0001013774383541381,
+      "loss": 1.1685,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5108484105817055,
+      "grad_norm": 0.1205865063322386,
+      "learning_rate": 0.00010135409338067219,
+      "loss": 1.3529,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5109204930440424,
+      "grad_norm": 0.11132340374979316,
+      "learning_rate": 0.00010133074833339611,
+      "loss": 1.4143,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5109925755063793,
+      "grad_norm": 0.11658384836455722,
+      "learning_rate": 0.0001013074032135823,
+      "loss": 1.3645,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5110646579687163,
+      "grad_norm": 0.1148340479609266,
+      "learning_rate": 0.00010128405802250336,
+      "loss": 1.3787,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5111367404310532,
+      "grad_norm": 0.10784131735261002,
+      "learning_rate": 0.00010126071276143176,
+      "loss": 1.3204,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5112088228933901,
+      "grad_norm": 0.11249822809550535,
+      "learning_rate": 0.00010123736743164002,
+      "loss": 1.2249,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5112809053557269,
+      "grad_norm": 0.13152641736820136,
+      "learning_rate": 0.0001012140220344007,
+      "loss": 1.2954,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5113529878180638,
+      "grad_norm": 0.11514662451841444,
+      "learning_rate": 0.00010119067657098631,
+      "loss": 1.2706,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5114250702804007,
+      "grad_norm": 0.11113819282550233,
+      "learning_rate": 0.0001011673310426694,
+      "loss": 1.2733,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5114971527427377,
+      "grad_norm": 0.11575718573063988,
+      "learning_rate": 0.0001011439854507225,
+      "loss": 1.3144,
+      "step": 7096
+    },
+    {
+      "epoch": 0.5115692352050746,
+      "grad_norm": 0.11094155205030591,
+      "learning_rate": 0.00010112063979641815,
+      "loss": 1.2601,
+      "step": 7097
+    },
+    {
+      "epoch": 0.5116413176674115,
+      "grad_norm": 0.1178017773740219,
+      "learning_rate": 0.00010109729408102891,
+      "loss": 1.3717,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5117134001297484,
+      "grad_norm": 0.10540724497308113,
+      "learning_rate": 0.00010107394830582732,
+      "loss": 1.2692,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5117854825920853,
+      "grad_norm": 0.11565431856677566,
+      "learning_rate": 0.00010105060247208595,
+      "loss": 1.3529,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5118575650544223,
+      "grad_norm": 0.27045704040874236,
+      "learning_rate": 0.00010102725658107737,
+      "loss": 1.3883,
+      "step": 7101
+    },
+    {
+      "epoch": 0.5119296475167592,
+      "grad_norm": 0.10726972432460037,
+      "learning_rate": 0.00010100391063407411,
+      "loss": 1.2022,
+      "step": 7102
+    },
+    {
+      "epoch": 0.5120017299790961,
+      "grad_norm": 0.11414371332155908,
+      "learning_rate": 0.00010098056463234869,
+      "loss": 1.2106,
+      "step": 7103
+    },
+    {
+      "epoch": 0.512073812441433,
+      "grad_norm": 0.11548387149202304,
+      "learning_rate": 0.00010095721857717378,
+      "loss": 1.2739,
+      "step": 7104
+    },
+    {
+      "epoch": 0.5121458949037699,
+      "grad_norm": 0.11809455530634978,
+      "learning_rate": 0.00010093387246982184,
+      "loss": 1.2285,
+      "step": 7105
+    },
+    {
+      "epoch": 0.5122179773661069,
+      "grad_norm": 0.10485638224365453,
+      "learning_rate": 0.00010091052631156555,
+      "loss": 1.3384,
+      "step": 7106
+    },
+    {
+      "epoch": 0.5122900598284438,
+      "grad_norm": 0.11303068501109216,
+      "learning_rate": 0.00010088718010367739,
+      "loss": 1.2735,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5123621422907807,
+      "grad_norm": 0.12885999368400453,
+      "learning_rate": 0.00010086383384743001,
+      "loss": 1.3451,
+      "step": 7108
+    },
+    {
+      "epoch": 0.5124342247531176,
+      "grad_norm": 0.1011109717728009,
+      "learning_rate": 0.00010084048754409592,
+      "loss": 1.3226,
+      "step": 7109
+    },
+    {
+      "epoch": 0.5125063072154544,
+      "grad_norm": 0.11338885835814896,
+      "learning_rate": 0.00010081714119494779,
+      "loss": 1.3371,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5125783896777913,
+      "grad_norm": 0.11987788337003753,
+      "learning_rate": 0.0001007937948012581,
+      "loss": 1.3126,
+      "step": 7111
+    },
+    {
+      "epoch": 0.5126504721401283,
+      "grad_norm": 0.1237783481175856,
+      "learning_rate": 0.00010077044836429954,
+      "loss": 1.2966,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5127225546024652,
+      "grad_norm": 0.11374828902877025,
+      "learning_rate": 0.00010074710188534462,
+      "loss": 1.3541,
+      "step": 7113
+    },
+    {
+      "epoch": 0.5127946370648021,
+      "grad_norm": 0.10799391415773524,
+      "learning_rate": 0.00010072375536566599,
+      "loss": 1.3585,
+      "step": 7114
+    },
+    {
+      "epoch": 0.512866719527139,
+      "grad_norm": 0.10673750259818049,
+      "learning_rate": 0.0001007004088065362,
+      "loss": 1.1935,
+      "step": 7115
+    },
+    {
+      "epoch": 0.512938801989476,
+      "grad_norm": 0.12480703336746679,
+      "learning_rate": 0.00010067706220922785,
+      "loss": 1.2702,
+      "step": 7116
+    },
+    {
+      "epoch": 0.5130108844518129,
+      "grad_norm": 0.11907788465984848,
+      "learning_rate": 0.00010065371557501358,
+      "loss": 1.2713,
+      "step": 7117
+    },
+    {
+      "epoch": 0.5130829669141498,
+      "grad_norm": 0.13155636199626664,
+      "learning_rate": 0.00010063036890516593,
+      "loss": 1.2813,
+      "step": 7118
+    },
+    {
+      "epoch": 0.5131550493764867,
+      "grad_norm": 0.1071252613204054,
+      "learning_rate": 0.00010060702220095756,
+      "loss": 1.2155,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5132271318388236,
+      "grad_norm": 0.12945122161481346,
+      "learning_rate": 0.00010058367546366105,
+      "loss": 1.2727,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5132992143011605,
+      "grad_norm": 0.11689002085848324,
+      "learning_rate": 0.000100560328694549,
+      "loss": 1.2561,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5133712967634975,
+      "grad_norm": 0.20498219147219104,
+      "learning_rate": 0.00010053698189489402,
+      "loss": 1.2674,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5134433792258344,
+      "grad_norm": 0.11331375921383362,
+      "learning_rate": 0.00010051363506596872,
+      "loss": 1.3201,
+      "step": 7123
+    },
+    {
+      "epoch": 0.5135154616881713,
+      "grad_norm": 0.12031063852596065,
+      "learning_rate": 0.00010049028820904574,
+      "loss": 1.3129,
+      "step": 7124
+    },
+    {
+      "epoch": 0.5135875441505082,
+      "grad_norm": 0.10979728688722737,
+      "learning_rate": 0.00010046694132539762,
+      "loss": 1.1386,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5136596266128451,
+      "grad_norm": 0.12018275359136113,
+      "learning_rate": 0.00010044359441629707,
+      "loss": 1.3135,
+      "step": 7126
+    },
+    {
+      "epoch": 0.5137317090751821,
+      "grad_norm": 0.1023466136646112,
+      "learning_rate": 0.00010042024748301661,
+      "loss": 1.342,
+      "step": 7127
+    },
+    {
+      "epoch": 0.5138037915375189,
+      "grad_norm": 0.11560915581833245,
+      "learning_rate": 0.00010039690052682895,
+      "loss": 1.2789,
+      "step": 7128
+    },
+    {
+      "epoch": 0.5138758739998558,
+      "grad_norm": 0.10830590472364983,
+      "learning_rate": 0.00010037355354900661,
+      "loss": 1.2286,
+      "step": 7129
+    },
+    {
+      "epoch": 0.5139479564621927,
+      "grad_norm": 0.1356163930946624,
+      "learning_rate": 0.0001003502065508223,
+      "loss": 1.3135,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5140200389245296,
+      "grad_norm": 0.1145914567400856,
+      "learning_rate": 0.00010032685953354857,
+      "loss": 1.3089,
+      "step": 7131
+    },
+    {
+      "epoch": 0.5140921213868666,
+      "grad_norm": 0.11889180897672226,
+      "learning_rate": 0.0001003035124984581,
+      "loss": 1.327,
+      "step": 7132
+    },
+    {
+      "epoch": 0.5141642038492035,
+      "grad_norm": 0.11043095322688636,
+      "learning_rate": 0.00010028016544682346,
+      "loss": 1.282,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5142362863115404,
+      "grad_norm": 0.12114549854856871,
+      "learning_rate": 0.00010025681837991731,
+      "loss": 1.281,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5143083687738773,
+      "grad_norm": 0.11437881195008924,
+      "learning_rate": 0.00010023347129901226,
+      "loss": 1.3043,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5143804512362142,
+      "grad_norm": 0.3118873871479386,
+      "learning_rate": 0.00010021012420538094,
+      "loss": 1.3691,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5144525336985512,
+      "grad_norm": 0.11714218750677677,
+      "learning_rate": 0.00010018677710029596,
+      "loss": 1.3486,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5145246161608881,
+      "grad_norm": 0.12368166082934172,
+      "learning_rate": 0.00010016342998502997,
+      "loss": 1.2675,
+      "step": 7138
+    },
+    {
+      "epoch": 0.514596698623225,
+      "grad_norm": 0.12590363299731078,
+      "learning_rate": 0.00010014008286085556,
+      "loss": 1.3458,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5146687810855619,
+      "grad_norm": 0.11409311473265477,
+      "learning_rate": 0.00010011673572904541,
+      "loss": 1.243,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5147408635478988,
+      "grad_norm": 0.10696401025683092,
+      "learning_rate": 0.00010009338859087209,
+      "loss": 1.3104,
+      "step": 7141
+    },
+    {
+      "epoch": 0.5148129460102358,
+      "grad_norm": 0.1154605926667675,
+      "learning_rate": 0.00010007004144760828,
+      "loss": 1.1947,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5148850284725727,
+      "grad_norm": 0.10906765768543772,
+      "learning_rate": 0.00010004669430052656,
+      "loss": 1.3103,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5149571109349096,
+      "grad_norm": 0.11397224360971865,
+      "learning_rate": 0.00010002334715089961,
+      "loss": 1.3015,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5150291933972464,
+      "grad_norm": 0.30039647860450247,
+      "learning_rate": 0.0001,
+      "loss": 1.2568,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5151012758595833,
+      "grad_norm": 0.12323845668560508,
+      "learning_rate": 9.997665284910042e-05,
+      "loss": 1.2781,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5151733583219202,
+      "grad_norm": 0.11380923673597977,
+      "learning_rate": 9.995330569947348e-05,
+      "loss": 1.2815,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5152454407842572,
+      "grad_norm": 0.1485372880150334,
+      "learning_rate": 9.992995855239174e-05,
+      "loss": 1.2418,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5153175232465941,
+      "grad_norm": 0.12025744397204861,
+      "learning_rate": 9.990661140912793e-05,
+      "loss": 1.2236,
+      "step": 7149
+    },
+    {
+      "epoch": 0.515389605708931,
+      "grad_norm": 0.11641003752000491,
+      "learning_rate": 9.988326427095463e-05,
+      "loss": 1.3439,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5154616881712679,
+      "grad_norm": 0.11486834857238017,
+      "learning_rate": 9.985991713914448e-05,
+      "loss": 1.2307,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5155337706336048,
+      "grad_norm": 0.11910449085647644,
+      "learning_rate": 9.983657001497005e-05,
+      "loss": 1.241,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5156058530959418,
+      "grad_norm": 0.11462343819251723,
+      "learning_rate": 9.981322289970407e-05,
+      "loss": 1.2418,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5156779355582787,
+      "grad_norm": 0.1180408778178796,
+      "learning_rate": 9.978987579461908e-05,
+      "loss": 1.2965,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5157500180206156,
+      "grad_norm": 0.12332040190131259,
+      "learning_rate": 9.976652870098777e-05,
+      "loss": 1.3014,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5158221004829525,
+      "grad_norm": 0.10990217861461092,
+      "learning_rate": 9.974318162008273e-05,
+      "loss": 1.3284,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5158941829452894,
+      "grad_norm": 0.13674743934717018,
+      "learning_rate": 9.971983455317655e-05,
+      "loss": 1.3957,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5159662654076264,
+      "grad_norm": 0.12248959868393366,
+      "learning_rate": 9.969648750154192e-05,
+      "loss": 1.3123,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5160383478699633,
+      "grad_norm": 0.109846590168548,
+      "learning_rate": 9.967314046645145e-05,
+      "loss": 1.3479,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5161104303323002,
+      "grad_norm": 0.11543020582158676,
+      "learning_rate": 9.964979344917775e-05,
+      "loss": 1.3432,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5161825127946371,
+      "grad_norm": 0.11781093032360426,
+      "learning_rate": 9.96264464509934e-05,
+      "loss": 1.2213,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5162545952569739,
+      "grad_norm": 0.10487386198787729,
+      "learning_rate": 9.960309947317106e-05,
+      "loss": 1.2997,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5163266777193108,
+      "grad_norm": 0.1336708464838374,
+      "learning_rate": 9.957975251698341e-05,
+      "loss": 1.2605,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5163987601816478,
+      "grad_norm": 0.14031902247997202,
+      "learning_rate": 9.955640558370297e-05,
+      "loss": 1.156,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5164708426439847,
+      "grad_norm": 0.10388763373115396,
+      "learning_rate": 9.953305867460238e-05,
+      "loss": 1.1475,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5165429251063216,
+      "grad_norm": 0.12433886760589709,
+      "learning_rate": 9.950971179095429e-05,
+      "loss": 1.2128,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5166150075686585,
+      "grad_norm": 0.12446056867707815,
+      "learning_rate": 9.94863649340313e-05,
+      "loss": 1.2463,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5166870900309954,
+      "grad_norm": 0.10719976748893395,
+      "learning_rate": 9.9463018105106e-05,
+      "loss": 1.2322,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5167591724933324,
+      "grad_norm": 0.11837427012840653,
+      "learning_rate": 9.943967130545105e-05,
+      "loss": 1.3259,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5168312549556693,
+      "grad_norm": 0.1102087515297322,
+      "learning_rate": 9.941632453633896e-05,
+      "loss": 1.4071,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5169033374180062,
+      "grad_norm": 0.11894011294705953,
+      "learning_rate": 9.939297779904246e-05,
+      "loss": 1.2701,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5169754198803431,
+      "grad_norm": 0.12287579971785105,
+      "learning_rate": 9.936963109483408e-05,
+      "loss": 1.2948,
+      "step": 7172
+    },
+    {
+      "epoch": 0.51704750234268,
+      "grad_norm": 0.11870841247990305,
+      "learning_rate": 9.934628442498649e-05,
+      "loss": 1.1972,
+      "step": 7173
+    },
+    {
+      "epoch": 0.517119584805017,
+      "grad_norm": 0.12348518954708558,
+      "learning_rate": 9.932293779077216e-05,
+      "loss": 1.3956,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5171916672673539,
+      "grad_norm": 0.161518634592365,
+      "learning_rate": 9.929959119346384e-05,
+      "loss": 1.2461,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5172637497296908,
+      "grad_norm": 0.12402867880783464,
+      "learning_rate": 9.927624463433405e-05,
+      "loss": 1.284,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5173358321920277,
+      "grad_norm": 0.11263108798141801,
+      "learning_rate": 9.925289811465542e-05,
+      "loss": 1.3169,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5174079146543646,
+      "grad_norm": 0.11171950308547363,
+      "learning_rate": 9.92295516357005e-05,
+      "loss": 1.2477,
+      "step": 7178
+    },
+    {
+      "epoch": 0.5174799971167015,
+      "grad_norm": 0.10863943467423663,
+      "learning_rate": 9.92062051987419e-05,
+      "loss": 1.2744,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5175520795790384,
+      "grad_norm": 0.1114944173446608,
+      "learning_rate": 9.918285880505223e-05,
+      "loss": 1.2765,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5176241620413753,
+      "grad_norm": 0.1123476127700827,
+      "learning_rate": 9.91595124559041e-05,
+      "loss": 1.4002,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5176962445037122,
+      "grad_norm": 0.1174116317428876,
+      "learning_rate": 9.913616615257004e-05,
+      "loss": 1.3287,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5177683269660491,
+      "grad_norm": 0.1062968761415639,
+      "learning_rate": 9.911281989632261e-05,
+      "loss": 1.211,
+      "step": 7183
+    },
+    {
+      "epoch": 0.517840409428386,
+      "grad_norm": 0.11119727386212755,
+      "learning_rate": 9.908947368843446e-05,
+      "loss": 1.2354,
+      "step": 7184
+    },
+    {
+      "epoch": 0.517912491890723,
+      "grad_norm": 0.11204635238205646,
+      "learning_rate": 9.906612753017818e-05,
+      "loss": 1.3022,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5179845743530599,
+      "grad_norm": 0.1140071849066326,
+      "learning_rate": 9.904278142282627e-05,
+      "loss": 1.2004,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5180566568153968,
+      "grad_norm": 0.11016047329679181,
+      "learning_rate": 9.901943536765131e-05,
+      "loss": 1.3439,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5181287392777337,
+      "grad_norm": 0.11201029965813118,
+      "learning_rate": 9.899608936592591e-05,
+      "loss": 1.1785,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5182008217400706,
+      "grad_norm": 0.11221909889853911,
+      "learning_rate": 9.897274341892267e-05,
+      "loss": 1.3104,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5182729042024076,
+      "grad_norm": 0.11397892814982645,
+      "learning_rate": 9.894939752791406e-05,
+      "loss": 1.3457,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5183449866647445,
+      "grad_norm": 0.1194041942698514,
+      "learning_rate": 9.892605169417271e-05,
+      "loss": 1.3266,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5184170691270814,
+      "grad_norm": 0.13670777044171262,
+      "learning_rate": 9.890270591897109e-05,
+      "loss": 1.2222,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5184891515894183,
+      "grad_norm": 0.12068395275694493,
+      "learning_rate": 9.887936020358187e-05,
+      "loss": 1.2797,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5185612340517552,
+      "grad_norm": 0.10621805758213121,
+      "learning_rate": 9.885601454927753e-05,
+      "loss": 1.2341,
+      "step": 7194
+    },
+    {
+      "epoch": 0.5186333165140922,
+      "grad_norm": 0.1204412724382485,
+      "learning_rate": 9.883266895733066e-05,
+      "loss": 1.3379,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5187053989764291,
+      "grad_norm": 0.11835536940012367,
+      "learning_rate": 9.880932342901368e-05,
+      "loss": 1.328,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5187774814387659,
+      "grad_norm": 0.13385744494234741,
+      "learning_rate": 9.878597796559933e-05,
+      "loss": 1.2717,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5188495639011028,
+      "grad_norm": 0.10287679451085262,
+      "learning_rate": 9.876263256836e-05,
+      "loss": 1.3825,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5189216463634397,
+      "grad_norm": 0.11750487152645837,
+      "learning_rate": 9.873928723856829e-05,
+      "loss": 1.2979,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5189937288257767,
+      "grad_norm": 0.12390589926732176,
+      "learning_rate": 9.871594197749668e-05,
+      "loss": 1.4209,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5190658112881136,
+      "grad_norm": 0.11557420030414903,
+      "learning_rate": 9.86925967864177e-05,
+      "loss": 1.1778,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5191378937504505,
+      "grad_norm": 0.1290865081386808,
+      "learning_rate": 9.866925166660391e-05,
+      "loss": 1.3051,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5192099762127874,
+      "grad_norm": 0.11137060045632191,
+      "learning_rate": 9.864590661932783e-05,
+      "loss": 1.2246,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5192820586751243,
+      "grad_norm": 0.10307572966494788,
+      "learning_rate": 9.862256164586193e-05,
+      "loss": 1.2184,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5193541411374613,
+      "grad_norm": 0.19131738017917865,
+      "learning_rate": 9.859921674747873e-05,
+      "loss": 1.3304,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5194262235997982,
+      "grad_norm": 0.11534448568074876,
+      "learning_rate": 9.857587192545077e-05,
+      "loss": 1.3204,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5194983060621351,
+      "grad_norm": 0.11031640994724057,
+      "learning_rate": 9.855252718105055e-05,
+      "loss": 1.2096,
+      "step": 7207
+    },
+    {
+      "epoch": 0.519570388524472,
+      "grad_norm": 0.11306148104736735,
+      "learning_rate": 9.852918251555052e-05,
+      "loss": 1.2339,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5196424709868089,
+      "grad_norm": 0.10441301065087104,
+      "learning_rate": 9.850583793022319e-05,
+      "loss": 1.2665,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5197145534491459,
+      "grad_norm": 0.11149643723179724,
+      "learning_rate": 9.848249342634108e-05,
+      "loss": 1.2467,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5197866359114828,
+      "grad_norm": 0.13053295337689536,
+      "learning_rate": 9.845914900517667e-05,
+      "loss": 1.3688,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5198587183738197,
+      "grad_norm": 0.11728371818107318,
+      "learning_rate": 9.843580466800241e-05,
+      "loss": 1.3079,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5199308008361566,
+      "grad_norm": 0.11566271573225374,
+      "learning_rate": 9.841246041609078e-05,
+      "loss": 1.3134,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5200028832984934,
+      "grad_norm": 0.11173916543797045,
+      "learning_rate": 9.838911625071424e-05,
+      "loss": 1.3313,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5200749657608303,
+      "grad_norm": 0.11956775400618803,
+      "learning_rate": 9.836577217314533e-05,
+      "loss": 1.2802,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5201470482231673,
+      "grad_norm": 0.11974440717611215,
+      "learning_rate": 9.834242818465642e-05,
+      "loss": 1.3169,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5202191306855042,
+      "grad_norm": 0.12320454819118733,
+      "learning_rate": 9.831908428652002e-05,
+      "loss": 1.3107,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5202912131478411,
+      "grad_norm": 0.12355041751617629,
+      "learning_rate": 9.829574048000852e-05,
+      "loss": 1.2898,
+      "step": 7218
+    },
+    {
+      "epoch": 0.520363295610178,
+      "grad_norm": 0.10923990840543599,
+      "learning_rate": 9.827239676639448e-05,
+      "loss": 1.3809,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5204353780725149,
+      "grad_norm": 0.11505483074493192,
+      "learning_rate": 9.824905314695023e-05,
+      "loss": 1.309,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5205074605348519,
+      "grad_norm": 0.1071517549399351,
+      "learning_rate": 9.822570962294826e-05,
+      "loss": 1.2386,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5205795429971888,
+      "grad_norm": 0.2826794816530416,
+      "learning_rate": 9.8202366195661e-05,
+      "loss": 1.2858,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5206516254595257,
+      "grad_norm": 0.11136920737862337,
+      "learning_rate": 9.817902286636082e-05,
+      "loss": 1.381,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5207237079218626,
+      "grad_norm": 0.1647083764001279,
+      "learning_rate": 9.815567963632022e-05,
+      "loss": 1.2525,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5207957903841995,
+      "grad_norm": 0.11198870189510103,
+      "learning_rate": 9.813233650681157e-05,
+      "loss": 1.3211,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5208678728465365,
+      "grad_norm": 0.11372496205238468,
+      "learning_rate": 9.810899347910732e-05,
+      "loss": 1.2686,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5209399553088734,
+      "grad_norm": 0.1189555235077661,
+      "learning_rate": 9.808565055447979e-05,
+      "loss": 1.2698,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5210120377712103,
+      "grad_norm": 0.10830131274950142,
+      "learning_rate": 9.806230773420145e-05,
+      "loss": 1.2784,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5210841202335472,
+      "grad_norm": 0.11326815014293853,
+      "learning_rate": 9.803896501954469e-05,
+      "loss": 1.1978,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5211562026958841,
+      "grad_norm": 0.11731294789531828,
+      "learning_rate": 9.801562241178188e-05,
+      "loss": 1.1936,
+      "step": 7230
+    },
+    {
+      "epoch": 0.521228285158221,
+      "grad_norm": 0.11028078133556968,
+      "learning_rate": 9.799227991218537e-05,
+      "loss": 1.3887,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5213003676205579,
+      "grad_norm": 0.1158329216835084,
+      "learning_rate": 9.796893752202758e-05,
+      "loss": 1.2926,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5213724500828948,
+      "grad_norm": 0.12272863511246122,
+      "learning_rate": 9.794559524258088e-05,
+      "loss": 1.3325,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5214445325452317,
+      "grad_norm": 0.10508591131750587,
+      "learning_rate": 9.792225307511763e-05,
+      "loss": 1.2895,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5215166150075686,
+      "grad_norm": 0.12035272220820702,
+      "learning_rate": 9.789891102091015e-05,
+      "loss": 1.2626,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5215886974699055,
+      "grad_norm": 0.11774276668107553,
+      "learning_rate": 9.78755690812308e-05,
+      "loss": 1.2454,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5216607799322425,
+      "grad_norm": 0.1412312765639459,
+      "learning_rate": 9.785222725735197e-05,
+      "loss": 1.1984,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5217328623945794,
+      "grad_norm": 0.11598339783349061,
+      "learning_rate": 9.782888555054598e-05,
+      "loss": 1.3256,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5218049448569163,
+      "grad_norm": 0.11719199924733839,
+      "learning_rate": 9.780554396208514e-05,
+      "loss": 1.3149,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5218770273192532,
+      "grad_norm": 0.11199213767718862,
+      "learning_rate": 9.778220249324174e-05,
+      "loss": 1.3353,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5219491097815901,
+      "grad_norm": 0.11456062474689108,
+      "learning_rate": 9.77588611452882e-05,
+      "loss": 1.294,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5220211922439271,
+      "grad_norm": 0.10811375598963548,
+      "learning_rate": 9.773551991949678e-05,
+      "loss": 1.2093,
+      "step": 7242
+    },
+    {
+      "epoch": 0.522093274706264,
+      "grad_norm": 0.13184467007266468,
+      "learning_rate": 9.771217881713977e-05,
+      "loss": 1.3067,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5221653571686009,
+      "grad_norm": 0.14112555127150017,
+      "learning_rate": 9.768883783948948e-05,
+      "loss": 1.325,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5222374396309378,
+      "grad_norm": 0.10490808679040044,
+      "learning_rate": 9.766549698781818e-05,
+      "loss": 1.3478,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5223095220932747,
+      "grad_norm": 0.11934461466729776,
+      "learning_rate": 9.764215626339822e-05,
+      "loss": 1.2264,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5223816045556117,
+      "grad_norm": 0.11601117174746298,
+      "learning_rate": 9.76188156675018e-05,
+      "loss": 1.2715,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5224536870179486,
+      "grad_norm": 0.12194799376226051,
+      "learning_rate": 9.759547520140128e-05,
+      "loss": 1.2999,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5225257694802854,
+      "grad_norm": 0.10383897161679227,
+      "learning_rate": 9.757213486636879e-05,
+      "loss": 1.2083,
+      "step": 7249
+    },
+    {
+      "epoch": 0.5225978519426223,
+      "grad_norm": 0.11977090319005426,
+      "learning_rate": 9.754879466367675e-05,
+      "loss": 1.3654,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5226699344049592,
+      "grad_norm": 0.10242736597436056,
+      "learning_rate": 9.752545459459729e-05,
+      "loss": 1.2565,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5227420168672962,
+      "grad_norm": 0.10550870535115335,
+      "learning_rate": 9.750211466040272e-05,
+      "loss": 1.2465,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5228140993296331,
+      "grad_norm": 0.1131124002890087,
+      "learning_rate": 9.747877486236518e-05,
+      "loss": 1.2359,
+      "step": 7253
+    },
+    {
+      "epoch": 0.52288618179197,
+      "grad_norm": 0.10233273542545084,
+      "learning_rate": 9.745543520175705e-05,
+      "loss": 1.2459,
+      "step": 7254
+    },
+    {
+      "epoch": 0.5229582642543069,
+      "grad_norm": 0.12599852952663707,
+      "learning_rate": 9.743209567985043e-05,
+      "loss": 1.309,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5230303467166438,
+      "grad_norm": 0.12947382205998756,
+      "learning_rate": 9.740875629791759e-05,
+      "loss": 1.1848,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5231024291789808,
+      "grad_norm": 0.10632837245688784,
+      "learning_rate": 9.738541705723069e-05,
+      "loss": 1.3157,
+      "step": 7257
+    },
+    {
+      "epoch": 0.5231745116413177,
+      "grad_norm": 0.11452869301271489,
+      "learning_rate": 9.736207795906194e-05,
+      "loss": 1.2633,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5232465941036546,
+      "grad_norm": 0.1246729229049783,
+      "learning_rate": 9.733873900468355e-05,
+      "loss": 1.3803,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5233186765659915,
+      "grad_norm": 0.14217158187270168,
+      "learning_rate": 9.731540019536772e-05,
+      "loss": 1.3089,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5233907590283284,
+      "grad_norm": 0.11365162472438861,
+      "learning_rate": 9.729206153238657e-05,
+      "loss": 1.2662,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5234628414906654,
+      "grad_norm": 0.12169220019740987,
+      "learning_rate": 9.726872301701227e-05,
+      "loss": 1.2577,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5235349239530023,
+      "grad_norm": 0.11847236814622256,
+      "learning_rate": 9.724538465051704e-05,
+      "loss": 1.1342,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5236070064153392,
+      "grad_norm": 0.11298895195281543,
+      "learning_rate": 9.722204643417299e-05,
+      "loss": 1.2605,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5236790888776761,
+      "grad_norm": 0.14575218353428307,
+      "learning_rate": 9.719870836925224e-05,
+      "loss": 1.2407,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5237511713400129,
+      "grad_norm": 0.11267975994983996,
+      "learning_rate": 9.717537045702698e-05,
+      "loss": 1.1969,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5238232538023498,
+      "grad_norm": 0.15375693214385186,
+      "learning_rate": 9.715203269876924e-05,
+      "loss": 1.2625,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5238953362646868,
+      "grad_norm": 0.11867174763467195,
+      "learning_rate": 9.712869509575124e-05,
+      "loss": 1.3166,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5239674187270237,
+      "grad_norm": 0.13486305110717484,
+      "learning_rate": 9.710535764924504e-05,
+      "loss": 1.2554,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5240395011893606,
+      "grad_norm": 0.1088809191365966,
+      "learning_rate": 9.708202036052276e-05,
+      "loss": 1.2693,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5241115836516975,
+      "grad_norm": 0.11970111745769267,
+      "learning_rate": 9.70586832308564e-05,
+      "loss": 1.2727,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5241836661140344,
+      "grad_norm": 0.12409730640583365,
+      "learning_rate": 9.70353462615182e-05,
+      "loss": 1.273,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5242557485763714,
+      "grad_norm": 0.12155895083981803,
+      "learning_rate": 9.701200945378012e-05,
+      "loss": 1.3408,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5243278310387083,
+      "grad_norm": 0.11370949260237409,
+      "learning_rate": 9.698867280891428e-05,
+      "loss": 1.3015,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5243999135010452,
+      "grad_norm": 0.11118339699783325,
+      "learning_rate": 9.696533632819266e-05,
+      "loss": 1.3108,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5244719959633821,
+      "grad_norm": 0.119960165437779,
+      "learning_rate": 9.69420000128874e-05,
+      "loss": 1.3322,
+      "step": 7276
+    },
+    {
+      "epoch": 0.524544078425719,
+      "grad_norm": 0.1903724215875718,
+      "learning_rate": 9.69186638642705e-05,
+      "loss": 1.4088,
+      "step": 7277
+    },
+    {
+      "epoch": 0.524616160888056,
+      "grad_norm": 0.1551594219226439,
+      "learning_rate": 9.689532788361399e-05,
+      "loss": 1.2746,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5246882433503929,
+      "grad_norm": 0.10547680092823697,
+      "learning_rate": 9.687199207218986e-05,
+      "loss": 1.2874,
+      "step": 7279
+    },
+    {
+      "epoch": 0.5247603258127298,
+      "grad_norm": 0.12993703285394276,
+      "learning_rate": 9.684865643127013e-05,
+      "loss": 1.2841,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5248324082750667,
+      "grad_norm": 0.11484813419613774,
+      "learning_rate": 9.682532096212685e-05,
+      "loss": 1.3059,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5249044907374036,
+      "grad_norm": 0.11259087120340065,
+      "learning_rate": 9.680198566603199e-05,
+      "loss": 1.2307,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5249765731997404,
+      "grad_norm": 0.12697461569623208,
+      "learning_rate": 9.677865054425752e-05,
+      "loss": 1.3207,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5250486556620774,
+      "grad_norm": 0.12441808542204309,
+      "learning_rate": 9.675531559807536e-05,
+      "loss": 1.3112,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5251207381244143,
+      "grad_norm": 0.11388604935818636,
+      "learning_rate": 9.673198082875756e-05,
+      "loss": 1.3285,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5251928205867512,
+      "grad_norm": 0.12947477953934897,
+      "learning_rate": 9.670864623757608e-05,
+      "loss": 1.2348,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5252649030490881,
+      "grad_norm": 0.11856148161942663,
+      "learning_rate": 9.668531182580278e-05,
+      "loss": 1.3106,
+      "step": 7287
+    },
+    {
+      "epoch": 0.525336985511425,
+      "grad_norm": 0.11532737889895187,
+      "learning_rate": 9.666197759470965e-05,
+      "loss": 1.2495,
+      "step": 7288
+    },
+    {
+      "epoch": 0.525409067973762,
+      "grad_norm": 0.10680704482249513,
+      "learning_rate": 9.663864354556858e-05,
+      "loss": 1.318,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5254811504360989,
+      "grad_norm": 0.10890683859191211,
+      "learning_rate": 9.661530967965156e-05,
+      "loss": 1.3272,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5255532328984358,
+      "grad_norm": 0.10685207968052232,
+      "learning_rate": 9.65919759982304e-05,
+      "loss": 1.2769,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5256253153607727,
+      "grad_norm": 0.12310380091279464,
+      "learning_rate": 9.656864250257706e-05,
+      "loss": 1.2863,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5256973978231096,
+      "grad_norm": 0.16512961769327578,
+      "learning_rate": 9.654530919396338e-05,
+      "loss": 1.2411,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5257694802854466,
+      "grad_norm": 0.11202916016647271,
+      "learning_rate": 9.652197607366129e-05,
+      "loss": 1.3603,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5258415627477835,
+      "grad_norm": 0.11081656092597453,
+      "learning_rate": 9.64986431429426e-05,
+      "loss": 1.3229,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5259136452101204,
+      "grad_norm": 0.11398962275007977,
+      "learning_rate": 9.647531040307918e-05,
+      "loss": 1.2721,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5259857276724573,
+      "grad_norm": 0.11089366766141522,
+      "learning_rate": 9.645197785534286e-05,
+      "loss": 1.2982,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5260578101347942,
+      "grad_norm": 0.10518388504252578,
+      "learning_rate": 9.642864550100552e-05,
+      "loss": 1.2994,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5261298925971312,
+      "grad_norm": 0.10850058779245937,
+      "learning_rate": 9.640531334133894e-05,
+      "loss": 1.2279,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5262019750594681,
+      "grad_norm": 0.1069179234330266,
+      "learning_rate": 9.638198137761494e-05,
+      "loss": 1.1756,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5262740575218049,
+      "grad_norm": 0.11882627208639167,
+      "learning_rate": 9.635864961110535e-05,
+      "loss": 1.3265,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5263461399841418,
+      "grad_norm": 0.11061399540112378,
+      "learning_rate": 9.633531804308187e-05,
+      "loss": 1.2346,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5264182224464787,
+      "grad_norm": 0.11388155701013879,
+      "learning_rate": 9.631198667481637e-05,
+      "loss": 1.2829,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5264903049088157,
+      "grad_norm": 0.12161632200906332,
+      "learning_rate": 9.62886555075806e-05,
+      "loss": 1.2581,
+      "step": 7304
+    },
+    {
+      "epoch": 0.5265623873711526,
+      "grad_norm": 0.1083432752591383,
+      "learning_rate": 9.626532454264633e-05,
+      "loss": 1.3434,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5266344698334895,
+      "grad_norm": 0.11887748789069416,
+      "learning_rate": 9.624199378128524e-05,
+      "loss": 1.2516,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5267065522958264,
+      "grad_norm": 0.12790896149751274,
+      "learning_rate": 9.621866322476911e-05,
+      "loss": 1.2626,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5267786347581633,
+      "grad_norm": 0.12944582815473146,
+      "learning_rate": 9.619533287436968e-05,
+      "loss": 1.2682,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5268507172205003,
+      "grad_norm": 0.11420009176087513,
+      "learning_rate": 9.617200273135866e-05,
+      "loss": 1.2355,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5269227996828372,
+      "grad_norm": 0.2106988643171106,
+      "learning_rate": 9.614867279700774e-05,
+      "loss": 1.3355,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5269948821451741,
+      "grad_norm": 0.11211769911618022,
+      "learning_rate": 9.612534307258856e-05,
+      "loss": 1.1876,
+      "step": 7311
+    },
+    {
+      "epoch": 0.527066964607511,
+      "grad_norm": 0.10961853715663278,
+      "learning_rate": 9.610201355937287e-05,
+      "loss": 1.2308,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5271390470698479,
+      "grad_norm": 0.11710104332232192,
+      "learning_rate": 9.607868425863235e-05,
+      "loss": 1.3204,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5272111295321849,
+      "grad_norm": 0.1122737831533485,
+      "learning_rate": 9.605535517163859e-05,
+      "loss": 1.2751,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5272832119945218,
+      "grad_norm": 0.11864038256537922,
+      "learning_rate": 9.603202629966323e-05,
+      "loss": 1.3243,
+      "step": 7315
+    },
+    {
+      "epoch": 0.5273552944568587,
+      "grad_norm": 0.11090848209555176,
+      "learning_rate": 9.600869764397798e-05,
+      "loss": 1.3261,
+      "step": 7316
+    },
+    {
+      "epoch": 0.5274273769191956,
+      "grad_norm": 0.10887228123082249,
+      "learning_rate": 9.598536920585442e-05,
+      "loss": 1.1879,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5274994593815324,
+      "grad_norm": 0.10608364246623174,
+      "learning_rate": 9.596204098656413e-05,
+      "loss": 1.2916,
+      "step": 7318
+    },
+    {
+      "epoch": 0.5275715418438693,
+      "grad_norm": 0.12607558693580617,
+      "learning_rate": 9.59387129873787e-05,
+      "loss": 1.3688,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5276436243062063,
+      "grad_norm": 0.10540564528596474,
+      "learning_rate": 9.59153852095698e-05,
+      "loss": 1.193,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5277157067685432,
+      "grad_norm": 0.11184984737746356,
+      "learning_rate": 9.589205765440895e-05,
+      "loss": 1.3127,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5277877892308801,
+      "grad_norm": 0.11610590716520178,
+      "learning_rate": 9.586873032316769e-05,
+      "loss": 1.3649,
+      "step": 7322
+    },
+    {
+      "epoch": 0.527859871693217,
+      "grad_norm": 0.11511965382278418,
+      "learning_rate": 9.58454032171176e-05,
+      "loss": 1.3496,
+      "step": 7323
+    },
+    {
+      "epoch": 0.5279319541555539,
+      "grad_norm": 0.10678780015925926,
+      "learning_rate": 9.582207633753015e-05,
+      "loss": 1.2833,
+      "step": 7324
+    },
+    {
+      "epoch": 0.5280040366178909,
+      "grad_norm": 0.1185562584673809,
+      "learning_rate": 9.579874968567697e-05,
+      "loss": 1.3012,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5280761190802278,
+      "grad_norm": 0.12311080145173445,
+      "learning_rate": 9.577542326282951e-05,
+      "loss": 1.2677,
+      "step": 7326
+    },
+    {
+      "epoch": 0.5281482015425647,
+      "grad_norm": 0.11465857156700891,
+      "learning_rate": 9.575209707025929e-05,
+      "loss": 1.2017,
+      "step": 7327
+    },
+    {
+      "epoch": 0.5282202840049016,
+      "grad_norm": 0.11896010168793375,
+      "learning_rate": 9.572877110923772e-05,
+      "loss": 1.2962,
+      "step": 7328
+    },
+    {
+      "epoch": 0.5282923664672385,
+      "grad_norm": 0.10520460525753542,
+      "learning_rate": 9.570544538103639e-05,
+      "loss": 1.3079,
+      "step": 7329
+    },
+    {
+      "epoch": 0.5283644489295755,
+      "grad_norm": 0.127414708680132,
+      "learning_rate": 9.56821198869267e-05,
+      "loss": 1.3314,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5284365313919124,
+      "grad_norm": 0.10954466112085269,
+      "learning_rate": 9.565879462818012e-05,
+      "loss": 1.2179,
+      "step": 7331
+    },
+    {
+      "epoch": 0.5285086138542493,
+      "grad_norm": 0.12220564314045829,
+      "learning_rate": 9.563546960606805e-05,
+      "loss": 1.2273,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5285806963165862,
+      "grad_norm": 0.1179611689688099,
+      "learning_rate": 9.561214482186192e-05,
+      "loss": 1.3524,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5286527787789231,
+      "grad_norm": 0.11830911855525388,
+      "learning_rate": 9.558882027683316e-05,
+      "loss": 1.2964,
+      "step": 7334
+    },
+    {
+      "epoch": 0.52872486124126,
+      "grad_norm": 0.1225483704917622,
+      "learning_rate": 9.556549597225319e-05,
+      "loss": 1.2834,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5287969437035969,
+      "grad_norm": 0.12511248857472101,
+      "learning_rate": 9.554217190939332e-05,
+      "loss": 1.2341,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5288690261659338,
+      "grad_norm": 0.10512178867725896,
+      "learning_rate": 9.551884808952496e-05,
+      "loss": 1.2724,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5289411086282707,
+      "grad_norm": 0.10387319652045754,
+      "learning_rate": 9.549552451391947e-05,
+      "loss": 1.2283,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5290131910906076,
+      "grad_norm": 0.10912535309507275,
+      "learning_rate": 9.547220118384824e-05,
+      "loss": 1.184,
+      "step": 7339
+    },
+    {
+      "epoch": 0.5290852735529445,
+      "grad_norm": 0.11763856962776381,
+      "learning_rate": 9.544887810058252e-05,
+      "loss": 1.2867,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5291573560152815,
+      "grad_norm": 0.12948547340898717,
+      "learning_rate": 9.542555526539364e-05,
+      "loss": 1.2131,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5292294384776184,
+      "grad_norm": 0.12140212009882448,
+      "learning_rate": 9.540223267955294e-05,
+      "loss": 1.3432,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5293015209399553,
+      "grad_norm": 0.11141216764702262,
+      "learning_rate": 9.53789103443317e-05,
+      "loss": 1.2886,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5293736034022922,
+      "grad_norm": 0.1533032030936138,
+      "learning_rate": 9.535558826100119e-05,
+      "loss": 1.293,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5294456858646291,
+      "grad_norm": 0.12643190074903796,
+      "learning_rate": 9.533226643083268e-05,
+      "loss": 1.4104,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5295177683269661,
+      "grad_norm": 0.11346731421859507,
+      "learning_rate": 9.530894485509735e-05,
+      "loss": 1.3133,
+      "step": 7346
+    },
+    {
+      "epoch": 0.529589850789303,
+      "grad_norm": 0.12483511680686454,
+      "learning_rate": 9.528562353506655e-05,
+      "loss": 1.3051,
+      "step": 7347
+    },
+    {
+      "epoch": 0.5296619332516399,
+      "grad_norm": 0.12651343021409167,
+      "learning_rate": 9.526230247201142e-05,
+      "loss": 1.244,
+      "step": 7348
+    },
+    {
+      "epoch": 0.5297340157139768,
+      "grad_norm": 0.11827675988682157,
+      "learning_rate": 9.523898166720323e-05,
+      "loss": 1.2236,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5298060981763137,
+      "grad_norm": 0.11096813217416847,
+      "learning_rate": 9.521566112191305e-05,
+      "loss": 1.3302,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5298781806386507,
+      "grad_norm": 0.1176090540083605,
+      "learning_rate": 9.519234083741221e-05,
+      "loss": 1.3268,
+      "step": 7351
+    },
+    {
+      "epoch": 0.5299502631009876,
+      "grad_norm": 0.12597813384083162,
+      "learning_rate": 9.516902081497177e-05,
+      "loss": 1.3431,
+      "step": 7352
+    },
+    {
+      "epoch": 0.5300223455633244,
+      "grad_norm": 0.10701312580419006,
+      "learning_rate": 9.514570105586296e-05,
+      "loss": 1.2309,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5300944280256613,
+      "grad_norm": 0.11915148027244207,
+      "learning_rate": 9.512238156135684e-05,
+      "loss": 1.2261,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5301665104879982,
+      "grad_norm": 0.12098448428323576,
+      "learning_rate": 9.509906233272453e-05,
+      "loss": 1.2759,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5302385929503352,
+      "grad_norm": 0.1181332157303745,
+      "learning_rate": 9.507574337123719e-05,
+      "loss": 1.2784,
+      "step": 7356
+    },
+    {
+      "epoch": 0.5303106754126721,
+      "grad_norm": 0.11502472927584168,
+      "learning_rate": 9.505242467816592e-05,
+      "loss": 1.2214,
+      "step": 7357
+    },
+    {
+      "epoch": 0.530382757875009,
+      "grad_norm": 0.11388683815172898,
+      "learning_rate": 9.502910625478175e-05,
+      "loss": 1.3066,
+      "step": 7358
+    },
+    {
+      "epoch": 0.5304548403373459,
+      "grad_norm": 0.11317942712711054,
+      "learning_rate": 9.500578810235573e-05,
+      "loss": 1.2085,
+      "step": 7359
+    },
+    {
+      "epoch": 0.5305269227996828,
+      "grad_norm": 0.11105431137014084,
+      "learning_rate": 9.498247022215896e-05,
+      "loss": 1.2934,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5305990052620198,
+      "grad_norm": 0.1285053341832049,
+      "learning_rate": 9.495915261546247e-05,
+      "loss": 1.3205,
+      "step": 7361
+    },
+    {
+      "epoch": 0.5306710877243567,
+      "grad_norm": 0.13257988243294133,
+      "learning_rate": 9.493583528353724e-05,
+      "loss": 1.2352,
+      "step": 7362
+    },
+    {
+      "epoch": 0.5307431701866936,
+      "grad_norm": 0.11506611656876747,
+      "learning_rate": 9.491251822765427e-05,
+      "loss": 1.3341,
+      "step": 7363
+    },
+    {
+      "epoch": 0.5308152526490305,
+      "grad_norm": 0.11903333116674662,
+      "learning_rate": 9.48892014490846e-05,
+      "loss": 1.2587,
+      "step": 7364
+    },
+    {
+      "epoch": 0.5308873351113674,
+      "grad_norm": 0.11830424672364288,
+      "learning_rate": 9.48658849490992e-05,
+      "loss": 1.2914,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5309594175737044,
+      "grad_norm": 0.12065948144208029,
+      "learning_rate": 9.484256872896896e-05,
+      "loss": 1.2613,
+      "step": 7366
+    },
+    {
+      "epoch": 0.5310315000360413,
+      "grad_norm": 0.10721587619439903,
+      "learning_rate": 9.481925278996488e-05,
+      "loss": 1.3414,
+      "step": 7367
+    },
+    {
+      "epoch": 0.5311035824983782,
+      "grad_norm": 0.1204456177958331,
+      "learning_rate": 9.479593713335784e-05,
+      "loss": 1.2989,
+      "step": 7368
+    },
+    {
+      "epoch": 0.5311756649607151,
+      "grad_norm": 0.11958430383184676,
+      "learning_rate": 9.477262176041883e-05,
+      "loss": 1.4231,
+      "step": 7369
+    },
+    {
+      "epoch": 0.5312477474230519,
+      "grad_norm": 0.11360982777296064,
+      "learning_rate": 9.474930667241868e-05,
+      "loss": 1.2757,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5313198298853888,
+      "grad_norm": 0.1168297219921061,
+      "learning_rate": 9.472599187062829e-05,
+      "loss": 1.278,
+      "step": 7371
+    },
+    {
+      "epoch": 0.5313919123477258,
+      "grad_norm": 0.11660766503209312,
+      "learning_rate": 9.47026773563185e-05,
+      "loss": 1.1834,
+      "step": 7372
+    },
+    {
+      "epoch": 0.5314639948100627,
+      "grad_norm": 0.11383422358206732,
+      "learning_rate": 9.467936313076022e-05,
+      "loss": 1.3421,
+      "step": 7373
+    },
+    {
+      "epoch": 0.5315360772723996,
+      "grad_norm": 0.11031215248384618,
+      "learning_rate": 9.465604919522425e-05,
+      "loss": 1.3639,
+      "step": 7374
+    },
+    {
+      "epoch": 0.5316081597347365,
+      "grad_norm": 0.1277583055510472,
+      "learning_rate": 9.46327355509814e-05,
+      "loss": 1.3148,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5316802421970734,
+      "grad_norm": 0.2959471021131825,
+      "learning_rate": 9.460942219930248e-05,
+      "loss": 1.2425,
+      "step": 7376
+    },
+    {
+      "epoch": 0.5317523246594104,
+      "grad_norm": 0.11307174403019675,
+      "learning_rate": 9.458610914145826e-05,
+      "loss": 1.235,
+      "step": 7377
+    },
+    {
+      "epoch": 0.5318244071217473,
+      "grad_norm": 0.11441784397019111,
+      "learning_rate": 9.456279637871953e-05,
+      "loss": 1.3919,
+      "step": 7378
+    },
+    {
+      "epoch": 0.5318964895840842,
+      "grad_norm": 0.12366426175034125,
+      "learning_rate": 9.453948391235706e-05,
+      "loss": 1.2062,
+      "step": 7379
+    },
+    {
+      "epoch": 0.5319685720464211,
+      "grad_norm": 0.11441266639392562,
+      "learning_rate": 9.451617174364157e-05,
+      "loss": 1.3157,
+      "step": 7380
+    },
+    {
+      "epoch": 0.532040654508758,
+      "grad_norm": 0.11102251513136403,
+      "learning_rate": 9.449285987384373e-05,
+      "loss": 1.293,
+      "step": 7381
+    },
+    {
+      "epoch": 0.532112736971095,
+      "grad_norm": 0.1129808038043404,
+      "learning_rate": 9.446954830423434e-05,
+      "loss": 1.2375,
+      "step": 7382
+    },
+    {
+      "epoch": 0.5321848194334319,
+      "grad_norm": 0.11732861583185526,
+      "learning_rate": 9.444623703608402e-05,
+      "loss": 1.334,
+      "step": 7383
+    },
+    {
+      "epoch": 0.5322569018957688,
+      "grad_norm": 0.12267122155558541,
+      "learning_rate": 9.442292607066351e-05,
+      "loss": 1.2743,
+      "step": 7384
+    },
+    {
+      "epoch": 0.5323289843581057,
+      "grad_norm": 0.12400218151450836,
+      "learning_rate": 9.439961540924337e-05,
+      "loss": 1.3368,
+      "step": 7385
+    },
+    {
+      "epoch": 0.5324010668204426,
+      "grad_norm": 0.1080094483069831,
+      "learning_rate": 9.437630505309432e-05,
+      "loss": 1.1941,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5324731492827794,
+      "grad_norm": 0.309526984695902,
+      "learning_rate": 9.435299500348695e-05,
+      "loss": 1.2563,
+      "step": 7387
+    },
+    {
+      "epoch": 0.5325452317451164,
+      "grad_norm": 0.11007670904778594,
+      "learning_rate": 9.43296852616919e-05,
+      "loss": 1.3192,
+      "step": 7388
+    },
+    {
+      "epoch": 0.5326173142074533,
+      "grad_norm": 0.11767283004366992,
+      "learning_rate": 9.43063758289797e-05,
+      "loss": 1.3021,
+      "step": 7389
+    },
+    {
+      "epoch": 0.5326893966697902,
+      "grad_norm": 0.10052249311117047,
+      "learning_rate": 9.428306670662093e-05,
+      "loss": 1.2762,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5327614791321271,
+      "grad_norm": 0.11100716845608054,
+      "learning_rate": 9.425975789588621e-05,
+      "loss": 1.2005,
+      "step": 7391
+    },
+    {
+      "epoch": 0.532833561594464,
+      "grad_norm": 0.12653754548044413,
+      "learning_rate": 9.423644939804605e-05,
+      "loss": 1.2325,
+      "step": 7392
+    },
+    {
+      "epoch": 0.532905644056801,
+      "grad_norm": 0.12006923364696608,
+      "learning_rate": 9.421314121437093e-05,
+      "loss": 1.2729,
+      "step": 7393
+    },
+    {
+      "epoch": 0.5329777265191379,
+      "grad_norm": 0.1221837950448006,
+      "learning_rate": 9.418983334613136e-05,
+      "loss": 1.316,
+      "step": 7394
+    },
+    {
+      "epoch": 0.5330498089814748,
+      "grad_norm": 0.11982496019180384,
+      "learning_rate": 9.416652579459789e-05,
+      "loss": 1.3523,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5331218914438117,
+      "grad_norm": 0.11344579662612883,
+      "learning_rate": 9.414321856104097e-05,
+      "loss": 1.2254,
+      "step": 7396
+    },
+    {
+      "epoch": 0.5331939739061486,
+      "grad_norm": 0.1188345019774624,
+      "learning_rate": 9.4119911646731e-05,
+      "loss": 1.2067,
+      "step": 7397
+    },
+    {
+      "epoch": 0.5332660563684856,
+      "grad_norm": 0.12078342795999134,
+      "learning_rate": 9.409660505293849e-05,
+      "loss": 1.2958,
+      "step": 7398
+    },
+    {
+      "epoch": 0.5333381388308225,
+      "grad_norm": 0.1297015096137038,
+      "learning_rate": 9.407329878093374e-05,
+      "loss": 1.3212,
+      "step": 7399
+    },
+    {
+      "epoch": 0.5334102212931594,
+      "grad_norm": 0.11130965021466861,
+      "learning_rate": 9.40499928319873e-05,
+      "loss": 1.2973,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5334823037554963,
+      "grad_norm": 0.11470268490912264,
+      "learning_rate": 9.402668720736946e-05,
+      "loss": 1.3177,
+      "step": 7401
+    },
+    {
+      "epoch": 0.5335543862178332,
+      "grad_norm": 0.12407228133535678,
+      "learning_rate": 9.400338190835062e-05,
+      "loss": 1.2936,
+      "step": 7402
+    },
+    {
+      "epoch": 0.5336264686801702,
+      "grad_norm": 0.10934204911343011,
+      "learning_rate": 9.398007693620105e-05,
+      "loss": 1.3893,
+      "step": 7403
+    },
+    {
+      "epoch": 0.5336985511425071,
+      "grad_norm": 0.10343400811507467,
+      "learning_rate": 9.395677229219121e-05,
+      "loss": 1.2967,
+      "step": 7404
+    },
+    {
+      "epoch": 0.5337706336048439,
+      "grad_norm": 0.1132971490405063,
+      "learning_rate": 9.393346797759133e-05,
+      "loss": 1.2044,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5338427160671808,
+      "grad_norm": 0.14743177900148022,
+      "learning_rate": 9.391016399367172e-05,
+      "loss": 1.1604,
+      "step": 7406
+    },
+    {
+      "epoch": 0.5339147985295177,
+      "grad_norm": 0.10863401644387757,
+      "learning_rate": 9.388686034170261e-05,
+      "loss": 1.3117,
+      "step": 7407
+    },
+    {
+      "epoch": 0.5339868809918547,
+      "grad_norm": 0.12056730397492661,
+      "learning_rate": 9.386355702295435e-05,
+      "loss": 1.2879,
+      "step": 7408
+    },
+    {
+      "epoch": 0.5340589634541916,
+      "grad_norm": 0.11730560594837597,
+      "learning_rate": 9.384025403869711e-05,
+      "loss": 1.3184,
+      "step": 7409
+    },
+    {
+      "epoch": 0.5341310459165285,
+      "grad_norm": 0.10758983539614599,
+      "learning_rate": 9.381695139020117e-05,
+      "loss": 1.2045,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5342031283788654,
+      "grad_norm": 0.11440394798596869,
+      "learning_rate": 9.379364907873667e-05,
+      "loss": 1.2219,
+      "step": 7411
+    },
+    {
+      "epoch": 0.5342752108412023,
+      "grad_norm": 0.12730378744554194,
+      "learning_rate": 9.377034710557378e-05,
+      "loss": 1.32,
+      "step": 7412
+    },
+    {
+      "epoch": 0.5343472933035393,
+      "grad_norm": 0.11788372753996447,
+      "learning_rate": 9.374704547198274e-05,
+      "loss": 1.2426,
+      "step": 7413
+    },
+    {
+      "epoch": 0.5344193757658762,
+      "grad_norm": 0.3852062189769867,
+      "learning_rate": 9.372374417923367e-05,
+      "loss": 1.2677,
+      "step": 7414
+    },
+    {
+      "epoch": 0.5344914582282131,
+      "grad_norm": 0.11653189919008958,
+      "learning_rate": 9.370044322859669e-05,
+      "loss": 1.3384,
+      "step": 7415
+    },
+    {
+      "epoch": 0.53456354069055,
+      "grad_norm": 0.11627642340418474,
+      "learning_rate": 9.367714262134188e-05,
+      "loss": 1.372,
+      "step": 7416
+    },
+    {
+      "epoch": 0.5346356231528869,
+      "grad_norm": 0.1175790940423925,
+      "learning_rate": 9.365384235873938e-05,
+      "loss": 1.1754,
+      "step": 7417
+    },
+    {
+      "epoch": 0.5347077056152238,
+      "grad_norm": 0.12497811932948198,
+      "learning_rate": 9.363054244205927e-05,
+      "loss": 1.2973,
+      "step": 7418
+    },
+    {
+      "epoch": 0.5347797880775608,
+      "grad_norm": 0.12627922902873676,
+      "learning_rate": 9.360724287257156e-05,
+      "loss": 1.3266,
+      "step": 7419
+    },
+    {
+      "epoch": 0.5348518705398977,
+      "grad_norm": 0.11768835736731456,
+      "learning_rate": 9.358394365154632e-05,
+      "loss": 1.3121,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5349239530022346,
+      "grad_norm": 0.11716827462304295,
+      "learning_rate": 9.356064478025349e-05,
+      "loss": 1.3207,
+      "step": 7421
+    },
+    {
+      "epoch": 0.5349960354645714,
+      "grad_norm": 0.12482530191261616,
+      "learning_rate": 9.35373462599632e-05,
+      "loss": 1.232,
+      "step": 7422
+    },
+    {
+      "epoch": 0.5350681179269083,
+      "grad_norm": 0.11461899247396398,
+      "learning_rate": 9.351404809194533e-05,
+      "loss": 1.2785,
+      "step": 7423
+    },
+    {
+      "epoch": 0.5351402003892453,
+      "grad_norm": 0.10858833887253082,
+      "learning_rate": 9.349075027746987e-05,
+      "loss": 1.2616,
+      "step": 7424
+    },
+    {
+      "epoch": 0.5352122828515822,
+      "grad_norm": 0.12112387776068197,
+      "learning_rate": 9.346745281780671e-05,
+      "loss": 1.3212,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5352843653139191,
+      "grad_norm": 0.11776692592060786,
+      "learning_rate": 9.344415571422586e-05,
+      "loss": 1.2812,
+      "step": 7426
+    },
+    {
+      "epoch": 0.535356447776256,
+      "grad_norm": 0.11393507386818869,
+      "learning_rate": 9.342085896799718e-05,
+      "loss": 1.2683,
+      "step": 7427
+    },
+    {
+      "epoch": 0.5354285302385929,
+      "grad_norm": 0.11366065749914481,
+      "learning_rate": 9.339756258039054e-05,
+      "loss": 1.368,
+      "step": 7428
+    },
+    {
+      "epoch": 0.5355006127009299,
+      "grad_norm": 0.10781377506055312,
+      "learning_rate": 9.33742665526758e-05,
+      "loss": 1.2672,
+      "step": 7429
+    },
+    {
+      "epoch": 0.5355726951632668,
+      "grad_norm": 0.1276493678721322,
+      "learning_rate": 9.335097088612276e-05,
+      "loss": 1.2714,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5356447776256037,
+      "grad_norm": 0.27884285133757936,
+      "learning_rate": 9.332767558200135e-05,
+      "loss": 1.2269,
+      "step": 7431
+    },
+    {
+      "epoch": 0.5357168600879406,
+      "grad_norm": 0.10349211775618855,
+      "learning_rate": 9.330438064158131e-05,
+      "loss": 1.2993,
+      "step": 7432
+    },
+    {
+      "epoch": 0.5357889425502775,
+      "grad_norm": 0.12363380252790919,
+      "learning_rate": 9.328108606613241e-05,
+      "loss": 1.3116,
+      "step": 7433
+    },
+    {
+      "epoch": 0.5358610250126145,
+      "grad_norm": 0.11911135292281909,
+      "learning_rate": 9.325779185692441e-05,
+      "loss": 1.2916,
+      "step": 7434
+    },
+    {
+      "epoch": 0.5359331074749514,
+      "grad_norm": 0.10558105141413665,
+      "learning_rate": 9.323449801522709e-05,
+      "loss": 1.2636,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5360051899372883,
+      "grad_norm": 0.11687447551338517,
+      "learning_rate": 9.321120454231019e-05,
+      "loss": 1.3125,
+      "step": 7436
+    },
+    {
+      "epoch": 0.5360772723996252,
+      "grad_norm": 0.1131427135177478,
+      "learning_rate": 9.318791143944333e-05,
+      "loss": 1.332,
+      "step": 7437
+    },
+    {
+      "epoch": 0.5361493548619621,
+      "grad_norm": 0.10508562815401726,
+      "learning_rate": 9.316461870789624e-05,
+      "loss": 1.2313,
+      "step": 7438
+    },
+    {
+      "epoch": 0.5362214373242989,
+      "grad_norm": 0.11092650453149393,
+      "learning_rate": 9.314132634893859e-05,
+      "loss": 1.2757,
+      "step": 7439
+    },
+    {
+      "epoch": 0.5362935197866359,
+      "grad_norm": 0.11453696435124658,
+      "learning_rate": 9.311803436384003e-05,
+      "loss": 1.2281,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5363656022489728,
+      "grad_norm": 0.1171822017769216,
+      "learning_rate": 9.309474275387015e-05,
+      "loss": 1.3191,
+      "step": 7441
+    },
+    {
+      "epoch": 0.5364376847113097,
+      "grad_norm": 0.11334516475799941,
+      "learning_rate": 9.307145152029857e-05,
+      "loss": 1.2316,
+      "step": 7442
+    },
+    {
+      "epoch": 0.5365097671736466,
+      "grad_norm": 0.1130892418140151,
+      "learning_rate": 9.304816066439484e-05,
+      "loss": 1.3228,
+      "step": 7443
+    },
+    {
+      "epoch": 0.5365818496359835,
+      "grad_norm": 0.1389770311472262,
+      "learning_rate": 9.30248701874286e-05,
+      "loss": 1.231,
+      "step": 7444
+    },
+    {
+      "epoch": 0.5366539320983205,
+      "grad_norm": 0.12105784296413,
+      "learning_rate": 9.30015800906693e-05,
+      "loss": 1.3653,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5367260145606574,
+      "grad_norm": 0.12880904912126165,
+      "learning_rate": 9.29782903753865e-05,
+      "loss": 1.2372,
+      "step": 7446
+    },
+    {
+      "epoch": 0.5367980970229943,
+      "grad_norm": 0.12297128936712325,
+      "learning_rate": 9.295500104284967e-05,
+      "loss": 1.184,
+      "step": 7447
+    },
+    {
+      "epoch": 0.5368701794853312,
+      "grad_norm": 0.11783950834168047,
+      "learning_rate": 9.293171209432836e-05,
+      "loss": 1.322,
+      "step": 7448
+    },
+    {
+      "epoch": 0.5369422619476681,
+      "grad_norm": 0.11571626117714288,
+      "learning_rate": 9.290842353109195e-05,
+      "loss": 1.3065,
+      "step": 7449
+    },
+    {
+      "epoch": 0.5370143444100051,
+      "grad_norm": 0.11727495513484885,
+      "learning_rate": 9.288513535440989e-05,
+      "loss": 1.3903,
+      "step": 7450
+    },
+    {
+      "epoch": 0.537086426872342,
+      "grad_norm": 0.1164750717060037,
+      "learning_rate": 9.286184756555164e-05,
+      "loss": 1.3313,
+      "step": 7451
+    },
+    {
+      "epoch": 0.5371585093346789,
+      "grad_norm": 0.12727860888567924,
+      "learning_rate": 9.283856016578651e-05,
+      "loss": 1.2011,
+      "step": 7452
+    },
+    {
+      "epoch": 0.5372305917970158,
+      "grad_norm": 0.11159854281228168,
+      "learning_rate": 9.281527315638394e-05,
+      "loss": 1.2035,
+      "step": 7453
+    },
+    {
+      "epoch": 0.5373026742593527,
+      "grad_norm": 0.11296379550750454,
+      "learning_rate": 9.279198653861327e-05,
+      "loss": 1.2585,
+      "step": 7454
+    },
+    {
+      "epoch": 0.5373747567216897,
+      "grad_norm": 0.1137005226734329,
+      "learning_rate": 9.276870031374382e-05,
+      "loss": 1.3127,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5374468391840266,
+      "grad_norm": 0.1235851541087,
+      "learning_rate": 9.274541448304486e-05,
+      "loss": 1.3098,
+      "step": 7456
+    },
+    {
+      "epoch": 0.5375189216463634,
+      "grad_norm": 0.11844503456750632,
+      "learning_rate": 9.272212904778575e-05,
+      "loss": 1.3192,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5375910041087003,
+      "grad_norm": 0.12132771898853575,
+      "learning_rate": 9.26988440092357e-05,
+      "loss": 1.3125,
+      "step": 7458
+    },
+    {
+      "epoch": 0.5376630865710372,
+      "grad_norm": 0.15510639624061862,
+      "learning_rate": 9.2675559368664e-05,
+      "loss": 1.2068,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5377351690333741,
+      "grad_norm": 0.12218832136662423,
+      "learning_rate": 9.265227512733978e-05,
+      "loss": 1.2882,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5378072514957111,
+      "grad_norm": 0.10667074879792221,
+      "learning_rate": 9.262899128653233e-05,
+      "loss": 1.2528,
+      "step": 7461
+    },
+    {
+      "epoch": 0.537879333958048,
+      "grad_norm": 0.1189448810526114,
+      "learning_rate": 9.26057078475108e-05,
+      "loss": 1.1702,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5379514164203849,
+      "grad_norm": 0.11017040808448045,
+      "learning_rate": 9.258242481154437e-05,
+      "loss": 1.2315,
+      "step": 7463
+    },
+    {
+      "epoch": 0.5380234988827218,
+      "grad_norm": 0.11280859610183984,
+      "learning_rate": 9.255914217990211e-05,
+      "loss": 1.2234,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5380955813450587,
+      "grad_norm": 0.10861186835328869,
+      "learning_rate": 9.253585995385317e-05,
+      "loss": 1.3633,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5381676638073957,
+      "grad_norm": 0.13062027408521393,
+      "learning_rate": 9.251257813466664e-05,
+      "loss": 1.3405,
+      "step": 7466
+    },
+    {
+      "epoch": 0.5382397462697326,
+      "grad_norm": 0.1474106757562423,
+      "learning_rate": 9.248929672361163e-05,
+      "loss": 1.2382,
+      "step": 7467
+    },
+    {
+      "epoch": 0.5383118287320695,
+      "grad_norm": 0.12055466685336334,
+      "learning_rate": 9.24660157219571e-05,
+      "loss": 1.2596,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5383839111944064,
+      "grad_norm": 0.11461472463312739,
+      "learning_rate": 9.244273513097211e-05,
+      "loss": 1.2924,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5384559936567433,
+      "grad_norm": 0.18942249609460438,
+      "learning_rate": 9.241945495192567e-05,
+      "loss": 1.277,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5385280761190803,
+      "grad_norm": 0.17745009231255068,
+      "learning_rate": 9.239617518608678e-05,
+      "loss": 1.2779,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5386001585814172,
+      "grad_norm": 0.127935642131505,
+      "learning_rate": 9.237289583472434e-05,
+      "loss": 1.2849,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5386722410437541,
+      "grad_norm": 0.10935652141776692,
+      "learning_rate": 9.234961689910734e-05,
+      "loss": 1.2222,
+      "step": 7473
+    },
+    {
+      "epoch": 0.5387443235060909,
+      "grad_norm": 0.1453977848821654,
+      "learning_rate": 9.232633838050459e-05,
+      "loss": 1.2661,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5388164059684278,
+      "grad_norm": 0.12197570049077276,
+      "learning_rate": 9.230306028018513e-05,
+      "loss": 1.2784,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5388884884307648,
+      "grad_norm": 0.11795962546623347,
+      "learning_rate": 9.227978259941771e-05,
+      "loss": 1.204,
+      "step": 7476
+    },
+    {
+      "epoch": 0.5389605708931017,
+      "grad_norm": 0.10983087866163357,
+      "learning_rate": 9.225650533947122e-05,
+      "loss": 1.1931,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5390326533554386,
+      "grad_norm": 0.12271669921146053,
+      "learning_rate": 9.223322850161441e-05,
+      "loss": 1.4238,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5391047358177755,
+      "grad_norm": 0.1156828112187912,
+      "learning_rate": 9.220995208711619e-05,
+      "loss": 1.3623,
+      "step": 7479
+    },
+    {
+      "epoch": 0.5391768182801124,
+      "grad_norm": 0.11883720382508955,
+      "learning_rate": 9.218667609724526e-05,
+      "loss": 1.1952,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5392489007424494,
+      "grad_norm": 0.1132850548523201,
+      "learning_rate": 9.216340053327039e-05,
+      "loss": 1.208,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5393209832047863,
+      "grad_norm": 0.11433178604267444,
+      "learning_rate": 9.214012539646025e-05,
+      "loss": 1.3515,
+      "step": 7482
+    },
+    {
+      "epoch": 0.5393930656671232,
+      "grad_norm": 0.19550170138194267,
+      "learning_rate": 9.211685068808364e-05,
+      "loss": 1.3188,
+      "step": 7483
+    },
+    {
+      "epoch": 0.5394651481294601,
+      "grad_norm": 0.11614179442343593,
+      "learning_rate": 9.209357640940919e-05,
+      "loss": 1.231,
+      "step": 7484
+    },
+    {
+      "epoch": 0.539537230591797,
+      "grad_norm": 0.10575420513587154,
+      "learning_rate": 9.207030256170557e-05,
+      "loss": 1.261,
+      "step": 7485
+    },
+    {
+      "epoch": 0.539609313054134,
+      "grad_norm": 0.11628861178441151,
+      "learning_rate": 9.204702914624138e-05,
+      "loss": 1.2763,
+      "step": 7486
+    },
+    {
+      "epoch": 0.5396813955164709,
+      "grad_norm": 0.13791121142829255,
+      "learning_rate": 9.202375616428522e-05,
+      "loss": 1.3442,
+      "step": 7487
+    },
+    {
+      "epoch": 0.5397534779788078,
+      "grad_norm": 0.105201117986868,
+      "learning_rate": 9.200048361710575e-05,
+      "loss": 1.1557,
+      "step": 7488
+    },
+    {
+      "epoch": 0.5398255604411447,
+      "grad_norm": 0.11680474095683747,
+      "learning_rate": 9.19772115059715e-05,
+      "loss": 1.3283,
+      "step": 7489
+    },
+    {
+      "epoch": 0.5398976429034816,
+      "grad_norm": 0.11954583262378388,
+      "learning_rate": 9.195393983215097e-05,
+      "loss": 1.4006,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5399697253658184,
+      "grad_norm": 0.11013109962826448,
+      "learning_rate": 9.193066859691268e-05,
+      "loss": 1.2178,
+      "step": 7491
+    },
+    {
+      "epoch": 0.5400418078281554,
+      "grad_norm": 0.11200780830926975,
+      "learning_rate": 9.190739780152516e-05,
+      "loss": 1.3004,
+      "step": 7492
+    },
+    {
+      "epoch": 0.5401138902904923,
+      "grad_norm": 0.10964692319918658,
+      "learning_rate": 9.18841274472569e-05,
+      "loss": 1.2619,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5401859727528292,
+      "grad_norm": 0.11054770309165553,
+      "learning_rate": 9.186085753537625e-05,
+      "loss": 1.1893,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5402580552151661,
+      "grad_norm": 0.11240178207892564,
+      "learning_rate": 9.183758806715172e-05,
+      "loss": 1.2308,
+      "step": 7495
+    },
+    {
+      "epoch": 0.540330137677503,
+      "grad_norm": 0.12137339017756846,
+      "learning_rate": 9.18143190438516e-05,
+      "loss": 1.299,
+      "step": 7496
+    },
+    {
+      "epoch": 0.54040222013984,
+      "grad_norm": 0.11041105303631017,
+      "learning_rate": 9.179105046674438e-05,
+      "loss": 1.1973,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5404743026021769,
+      "grad_norm": 0.12460945915420832,
+      "learning_rate": 9.176778233709833e-05,
+      "loss": 1.326,
+      "step": 7498
+    },
+    {
+      "epoch": 0.5405463850645138,
+      "grad_norm": 0.11097024968063395,
+      "learning_rate": 9.174451465618182e-05,
+      "loss": 1.2201,
+      "step": 7499
+    },
+    {
+      "epoch": 0.5406184675268507,
+      "grad_norm": 0.12111922603107267,
+      "learning_rate": 9.172124742526303e-05,
+      "loss": 1.3357,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5406905499891876,
+      "grad_norm": 0.12030245854975798,
+      "learning_rate": 9.169798064561041e-05,
+      "loss": 1.2897,
+      "step": 7501
+    },
+    {
+      "epoch": 0.5407626324515246,
+      "grad_norm": 0.11209714883460496,
+      "learning_rate": 9.167471431849207e-05,
+      "loss": 1.2041,
+      "step": 7502
+    },
+    {
+      "epoch": 0.5408347149138615,
+      "grad_norm": 0.11602224752094084,
+      "learning_rate": 9.165144844517632e-05,
+      "loss": 1.3257,
+      "step": 7503
+    },
+    {
+      "epoch": 0.5409067973761984,
+      "grad_norm": 0.10986225575232732,
+      "learning_rate": 9.162818302693124e-05,
+      "loss": 1.2354,
+      "step": 7504
+    },
+    {
+      "epoch": 0.5409788798385353,
+      "grad_norm": 0.10894007903472105,
+      "learning_rate": 9.160491806502516e-05,
+      "loss": 1.2705,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5410509623008722,
+      "grad_norm": 0.12067230968674919,
+      "learning_rate": 9.158165356072612e-05,
+      "loss": 1.2406,
+      "step": 7506
+    },
+    {
+      "epoch": 0.5411230447632092,
+      "grad_norm": 0.12675682679743966,
+      "learning_rate": 9.15583895153023e-05,
+      "loss": 1.2815,
+      "step": 7507
+    },
+    {
+      "epoch": 0.5411951272255461,
+      "grad_norm": 0.11392481670966305,
+      "learning_rate": 9.153512593002175e-05,
+      "loss": 1.2818,
+      "step": 7508
+    },
+    {
+      "epoch": 0.5412672096878829,
+      "grad_norm": 0.27030577514636533,
+      "learning_rate": 9.151186280615253e-05,
+      "loss": 1.3412,
+      "step": 7509
+    },
+    {
+      "epoch": 0.5413392921502198,
+      "grad_norm": 0.11139779782634601,
+      "learning_rate": 9.148860014496277e-05,
+      "loss": 1.2644,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5414113746125567,
+      "grad_norm": 0.11649779132960472,
+      "learning_rate": 9.146533794772046e-05,
+      "loss": 1.2588,
+      "step": 7511
+    },
+    {
+      "epoch": 0.5414834570748936,
+      "grad_norm": 0.13309104002339164,
+      "learning_rate": 9.144207621569356e-05,
+      "loss": 1.2514,
+      "step": 7512
+    },
+    {
+      "epoch": 0.5415555395372306,
+      "grad_norm": 0.11864326938191876,
+      "learning_rate": 9.141881495015004e-05,
+      "loss": 1.2428,
+      "step": 7513
+    },
+    {
+      "epoch": 0.5416276219995675,
+      "grad_norm": 0.1152355106516985,
+      "learning_rate": 9.13955541523579e-05,
+      "loss": 1.288,
+      "step": 7514
+    },
+    {
+      "epoch": 0.5416997044619044,
+      "grad_norm": 0.11731673772995199,
+      "learning_rate": 9.137229382358507e-05,
+      "loss": 1.3098,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5417717869242413,
+      "grad_norm": 0.11043662950395586,
+      "learning_rate": 9.134903396509938e-05,
+      "loss": 1.2883,
+      "step": 7516
+    },
+    {
+      "epoch": 0.5418438693865782,
+      "grad_norm": 0.13563140171267796,
+      "learning_rate": 9.132577457816873e-05,
+      "loss": 1.3106,
+      "step": 7517
+    },
+    {
+      "epoch": 0.5419159518489152,
+      "grad_norm": 0.12247614321488999,
+      "learning_rate": 9.130251566406095e-05,
+      "loss": 1.3629,
+      "step": 7518
+    },
+    {
+      "epoch": 0.5419880343112521,
+      "grad_norm": 0.10918272859573436,
+      "learning_rate": 9.127925722404391e-05,
+      "loss": 1.2809,
+      "step": 7519
+    },
+    {
+      "epoch": 0.542060116773589,
+      "grad_norm": 0.11530163767158645,
+      "learning_rate": 9.125599925938537e-05,
+      "loss": 1.3396,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5421321992359259,
+      "grad_norm": 0.11796924834231905,
+      "learning_rate": 9.123274177135308e-05,
+      "loss": 1.2292,
+      "step": 7521
+    },
+    {
+      "epoch": 0.5422042816982628,
+      "grad_norm": 0.29827127729719843,
+      "learning_rate": 9.120948476121479e-05,
+      "loss": 1.3082,
+      "step": 7522
+    },
+    {
+      "epoch": 0.5422763641605998,
+      "grad_norm": 0.10672773271641581,
+      "learning_rate": 9.118622823023823e-05,
+      "loss": 1.311,
+      "step": 7523
+    },
+    {
+      "epoch": 0.5423484466229367,
+      "grad_norm": 0.11168256156255803,
+      "learning_rate": 9.11629721796911e-05,
+      "loss": 1.3493,
+      "step": 7524
+    },
+    {
+      "epoch": 0.5424205290852736,
+      "grad_norm": 0.12070043416055005,
+      "learning_rate": 9.113971661084102e-05,
+      "loss": 1.2893,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5424926115476104,
+      "grad_norm": 0.1136621686557042,
+      "learning_rate": 9.111646152495565e-05,
+      "loss": 1.2742,
+      "step": 7526
+    },
+    {
+      "epoch": 0.5425646940099473,
+      "grad_norm": 0.11497600483422923,
+      "learning_rate": 9.109320692330263e-05,
+      "loss": 1.2388,
+      "step": 7527
+    },
+    {
+      "epoch": 0.5426367764722843,
+      "grad_norm": 0.2584434690675181,
+      "learning_rate": 9.10699528071495e-05,
+      "loss": 1.1888,
+      "step": 7528
+    },
+    {
+      "epoch": 0.5427088589346212,
+      "grad_norm": 0.10477270851938202,
+      "learning_rate": 9.104669917776383e-05,
+      "loss": 1.3076,
+      "step": 7529
+    },
+    {
+      "epoch": 0.5427809413969581,
+      "grad_norm": 0.11380772413972341,
+      "learning_rate": 9.102344603641318e-05,
+      "loss": 1.2844,
+      "step": 7530
+    },
+    {
+      "epoch": 0.542853023859295,
+      "grad_norm": 0.10636732924990382,
+      "learning_rate": 9.100019338436496e-05,
+      "loss": 1.202,
+      "step": 7531
+    },
+    {
+      "epoch": 0.5429251063216319,
+      "grad_norm": 0.10869003366972269,
+      "learning_rate": 9.097694122288676e-05,
+      "loss": 1.3326,
+      "step": 7532
+    },
+    {
+      "epoch": 0.5429971887839689,
+      "grad_norm": 0.10470567640627548,
+      "learning_rate": 9.095368955324599e-05,
+      "loss": 1.1932,
+      "step": 7533
+    },
+    {
+      "epoch": 0.5430692712463058,
+      "grad_norm": 0.11595243561219701,
+      "learning_rate": 9.093043837671008e-05,
+      "loss": 1.3226,
+      "step": 7534
+    },
+    {
+      "epoch": 0.5431413537086427,
+      "grad_norm": 0.11281791267817871,
+      "learning_rate": 9.090718769454636e-05,
+      "loss": 1.2896,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5432134361709796,
+      "grad_norm": 0.11931927466817613,
+      "learning_rate": 9.088393750802231e-05,
+      "loss": 1.3519,
+      "step": 7536
+    },
+    {
+      "epoch": 0.5432855186333165,
+      "grad_norm": 0.10797806402612889,
+      "learning_rate": 9.086068781840521e-05,
+      "loss": 1.2336,
+      "step": 7537
+    },
+    {
+      "epoch": 0.5433576010956535,
+      "grad_norm": 0.11342512771012053,
+      "learning_rate": 9.08374386269624e-05,
+      "loss": 1.3073,
+      "step": 7538
+    },
+    {
+      "epoch": 0.5434296835579904,
+      "grad_norm": 0.11193826988363084,
+      "learning_rate": 9.081418993496114e-05,
+      "loss": 1.2554,
+      "step": 7539
+    },
+    {
+      "epoch": 0.5435017660203273,
+      "grad_norm": 0.12160537471170904,
+      "learning_rate": 9.079094174366867e-05,
+      "loss": 1.2882,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5435738484826642,
+      "grad_norm": 0.11366576762490241,
+      "learning_rate": 9.07676940543523e-05,
+      "loss": 1.3584,
+      "step": 7541
+    },
+    {
+      "epoch": 0.5436459309450011,
+      "grad_norm": 0.10889197305128209,
+      "learning_rate": 9.074444686827923e-05,
+      "loss": 1.1782,
+      "step": 7542
+    },
+    {
+      "epoch": 0.5437180134073379,
+      "grad_norm": 0.10806452608122846,
+      "learning_rate": 9.072120018671657e-05,
+      "loss": 1.2893,
+      "step": 7543
+    },
+    {
+      "epoch": 0.5437900958696749,
+      "grad_norm": 0.11773753697906669,
+      "learning_rate": 9.06979540109315e-05,
+      "loss": 1.3529,
+      "step": 7544
+    },
+    {
+      "epoch": 0.5438621783320118,
+      "grad_norm": 0.1228142837771542,
+      "learning_rate": 9.067470834219116e-05,
+      "loss": 1.2769,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5439342607943487,
+      "grad_norm": 0.12588999157155295,
+      "learning_rate": 9.065146318176268e-05,
+      "loss": 1.1945,
+      "step": 7546
+    },
+    {
+      "epoch": 0.5440063432566856,
+      "grad_norm": 0.12101239902362552,
+      "learning_rate": 9.062821853091305e-05,
+      "loss": 1.3122,
+      "step": 7547
+    },
+    {
+      "epoch": 0.5440784257190225,
+      "grad_norm": 0.10553094910740025,
+      "learning_rate": 9.060497439090935e-05,
+      "loss": 1.3164,
+      "step": 7548
+    },
+    {
+      "epoch": 0.5441505081813595,
+      "grad_norm": 0.11817758142122947,
+      "learning_rate": 9.058173076301861e-05,
+      "loss": 1.3277,
+      "step": 7549
+    },
+    {
+      "epoch": 0.5442225906436964,
+      "grad_norm": 0.11467330337333792,
+      "learning_rate": 9.055848764850782e-05,
+      "loss": 1.2174,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5442946731060333,
+      "grad_norm": 0.11393983868045413,
+      "learning_rate": 9.05352450486439e-05,
+      "loss": 1.3039,
+      "step": 7551
+    },
+    {
+      "epoch": 0.5443667555683702,
+      "grad_norm": 0.11400195698312268,
+      "learning_rate": 9.051200296469382e-05,
+      "loss": 1.3689,
+      "step": 7552
+    },
+    {
+      "epoch": 0.5444388380307071,
+      "grad_norm": 0.13668238103875047,
+      "learning_rate": 9.048876139792441e-05,
+      "loss": 1.3171,
+      "step": 7553
+    },
+    {
+      "epoch": 0.5445109204930441,
+      "grad_norm": 0.11831531796396257,
+      "learning_rate": 9.046552034960263e-05,
+      "loss": 1.3058,
+      "step": 7554
+    },
+    {
+      "epoch": 0.544583002955381,
+      "grad_norm": 0.1286862205481974,
+      "learning_rate": 9.044227982099529e-05,
+      "loss": 1.175,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5446550854177179,
+      "grad_norm": 0.11114428835411638,
+      "learning_rate": 9.041903981336922e-05,
+      "loss": 1.3072,
+      "step": 7556
+    },
+    {
+      "epoch": 0.5447271678800548,
+      "grad_norm": 0.12456168663381026,
+      "learning_rate": 9.039580032799113e-05,
+      "loss": 1.2268,
+      "step": 7557
+    },
+    {
+      "epoch": 0.5447992503423917,
+      "grad_norm": 0.1281814365533053,
+      "learning_rate": 9.037256136612791e-05,
+      "loss": 1.2152,
+      "step": 7558
+    },
+    {
+      "epoch": 0.5448713328047287,
+      "grad_norm": 0.11954377382916251,
+      "learning_rate": 9.03493229290462e-05,
+      "loss": 1.3185,
+      "step": 7559
+    },
+    {
+      "epoch": 0.5449434152670656,
+      "grad_norm": 0.11148566098719168,
+      "learning_rate": 9.032608501801275e-05,
+      "loss": 1.1832,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5450154977294024,
+      "grad_norm": 0.10915343954448353,
+      "learning_rate": 9.030284763429419e-05,
+      "loss": 1.2989,
+      "step": 7561
+    },
+    {
+      "epoch": 0.5450875801917393,
+      "grad_norm": 0.1098884549079372,
+      "learning_rate": 9.027961077915717e-05,
+      "loss": 1.1952,
+      "step": 7562
+    },
+    {
+      "epoch": 0.5451596626540762,
+      "grad_norm": 0.12318393420685571,
+      "learning_rate": 9.025637445386834e-05,
+      "loss": 1.2463,
+      "step": 7563
+    },
+    {
+      "epoch": 0.5452317451164131,
+      "grad_norm": 0.1109263809867332,
+      "learning_rate": 9.02331386596943e-05,
+      "loss": 1.2624,
+      "step": 7564
+    },
+    {
+      "epoch": 0.5453038275787501,
+      "grad_norm": 0.10640272685379323,
+      "learning_rate": 9.020990339790155e-05,
+      "loss": 1.2469,
+      "step": 7565
+    },
+    {
+      "epoch": 0.545375910041087,
+      "grad_norm": 0.1260090087310851,
+      "learning_rate": 9.018666866975663e-05,
+      "loss": 1.3015,
+      "step": 7566
+    },
+    {
+      "epoch": 0.5454479925034239,
+      "grad_norm": 0.11497708062965693,
+      "learning_rate": 9.016343447652608e-05,
+      "loss": 1.2745,
+      "step": 7567
+    },
+    {
+      "epoch": 0.5455200749657608,
+      "grad_norm": 0.11357111598034937,
+      "learning_rate": 9.014020081947638e-05,
+      "loss": 1.2348,
+      "step": 7568
+    },
+    {
+      "epoch": 0.5455921574280977,
+      "grad_norm": 0.22051238528950076,
+      "learning_rate": 9.011696769987392e-05,
+      "loss": 1.3434,
+      "step": 7569
+    },
+    {
+      "epoch": 0.5456642398904347,
+      "grad_norm": 0.12145436196995273,
+      "learning_rate": 9.009373511898511e-05,
+      "loss": 1.2696,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5457363223527716,
+      "grad_norm": 0.24763196230914836,
+      "learning_rate": 9.007050307807639e-05,
+      "loss": 1.2455,
+      "step": 7571
+    },
+    {
+      "epoch": 0.5458084048151085,
+      "grad_norm": 0.11172933346664297,
+      "learning_rate": 9.00472715784141e-05,
+      "loss": 1.2669,
+      "step": 7572
+    },
+    {
+      "epoch": 0.5458804872774454,
+      "grad_norm": 0.1363365704918133,
+      "learning_rate": 9.002404062126452e-05,
+      "loss": 1.2621,
+      "step": 7573
+    },
+    {
+      "epoch": 0.5459525697397823,
+      "grad_norm": 0.11561570036120399,
+      "learning_rate": 9.0000810207894e-05,
+      "loss": 1.2924,
+      "step": 7574
+    },
+    {
+      "epoch": 0.5460246522021193,
+      "grad_norm": 0.12060262546011928,
+      "learning_rate": 8.997758033956871e-05,
+      "loss": 1.3243,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5460967346644562,
+      "grad_norm": 0.12099440955383589,
+      "learning_rate": 8.995435101755503e-05,
+      "loss": 1.2476,
+      "step": 7576
+    },
+    {
+      "epoch": 0.5461688171267931,
+      "grad_norm": 0.14017328422547784,
+      "learning_rate": 8.993112224311907e-05,
+      "loss": 1.373,
+      "step": 7577
+    },
+    {
+      "epoch": 0.5462408995891299,
+      "grad_norm": 0.12449590784220538,
+      "learning_rate": 8.990789401752703e-05,
+      "loss": 1.3816,
+      "step": 7578
+    },
+    {
+      "epoch": 0.5463129820514668,
+      "grad_norm": 0.20486048714745922,
+      "learning_rate": 8.988466634204501e-05,
+      "loss": 1.2807,
+      "step": 7579
+    },
+    {
+      "epoch": 0.5463850645138038,
+      "grad_norm": 0.11268537538771663,
+      "learning_rate": 8.986143921793923e-05,
+      "loss": 1.2886,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5464571469761407,
+      "grad_norm": 0.1235412063552395,
+      "learning_rate": 8.983821264647568e-05,
+      "loss": 1.329,
+      "step": 7581
+    },
+    {
+      "epoch": 0.5465292294384776,
+      "grad_norm": 0.11491151981579217,
+      "learning_rate": 8.981498662892048e-05,
+      "loss": 1.3381,
+      "step": 7582
+    },
+    {
+      "epoch": 0.5466013119008145,
+      "grad_norm": 0.12011890440139904,
+      "learning_rate": 8.97917611665396e-05,
+      "loss": 1.2094,
+      "step": 7583
+    },
+    {
+      "epoch": 0.5466733943631514,
+      "grad_norm": 0.34718535129585476,
+      "learning_rate": 8.976853626059905e-05,
+      "loss": 1.2883,
+      "step": 7584
+    },
+    {
+      "epoch": 0.5467454768254884,
+      "grad_norm": 0.10811108716111599,
+      "learning_rate": 8.974531191236482e-05,
+      "loss": 1.2436,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5468175592878253,
+      "grad_norm": 0.1122861476356958,
+      "learning_rate": 8.972208812310286e-05,
+      "loss": 1.2827,
+      "step": 7586
+    },
+    {
+      "epoch": 0.5468896417501622,
+      "grad_norm": 0.11297746173634698,
+      "learning_rate": 8.969886489407901e-05,
+      "loss": 1.306,
+      "step": 7587
+    },
+    {
+      "epoch": 0.5469617242124991,
+      "grad_norm": 0.12832828137809804,
+      "learning_rate": 8.967564222655918e-05,
+      "loss": 1.3277,
+      "step": 7588
+    },
+    {
+      "epoch": 0.547033806674836,
+      "grad_norm": 0.10622026674674709,
+      "learning_rate": 8.965242012180922e-05,
+      "loss": 1.1972,
+      "step": 7589
+    },
+    {
+      "epoch": 0.547105889137173,
+      "grad_norm": 0.13024821659199387,
+      "learning_rate": 8.962919858109496e-05,
+      "loss": 1.2623,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5471779715995099,
+      "grad_norm": 0.11649643637636801,
+      "learning_rate": 8.960597760568215e-05,
+      "loss": 1.2145,
+      "step": 7591
+    },
+    {
+      "epoch": 0.5472500540618468,
+      "grad_norm": 0.12439036431162076,
+      "learning_rate": 8.95827571968365e-05,
+      "loss": 1.345,
+      "step": 7592
+    },
+    {
+      "epoch": 0.5473221365241837,
+      "grad_norm": 0.1225117021950515,
+      "learning_rate": 8.955953735582382e-05,
+      "loss": 1.3638,
+      "step": 7593
+    },
+    {
+      "epoch": 0.5473942189865206,
+      "grad_norm": 0.21166060544341253,
+      "learning_rate": 8.953631808390976e-05,
+      "loss": 1.2192,
+      "step": 7594
+    },
+    {
+      "epoch": 0.5474663014488574,
+      "grad_norm": 0.11972810630892496,
+      "learning_rate": 8.951309938235998e-05,
+      "loss": 1.3914,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5475383839111944,
+      "grad_norm": 0.1163936069413252,
+      "learning_rate": 8.948988125244009e-05,
+      "loss": 1.3841,
+      "step": 7596
+    },
+    {
+      "epoch": 0.5476104663735313,
+      "grad_norm": 0.10315381956501381,
+      "learning_rate": 8.946666369541567e-05,
+      "loss": 1.3924,
+      "step": 7597
+    },
+    {
+      "epoch": 0.5476825488358682,
+      "grad_norm": 0.11010956011429399,
+      "learning_rate": 8.944344671255236e-05,
+      "loss": 1.1706,
+      "step": 7598
+    },
+    {
+      "epoch": 0.5477546312982051,
+      "grad_norm": 0.10701279778592335,
+      "learning_rate": 8.942023030511562e-05,
+      "loss": 1.3432,
+      "step": 7599
+    },
+    {
+      "epoch": 0.547826713760542,
+      "grad_norm": 0.15415132690046865,
+      "learning_rate": 8.939701447437097e-05,
+      "loss": 1.2442,
+      "step": 7600
+    },
+    {
+      "epoch": 0.547898796222879,
+      "grad_norm": 0.11839146280243672,
+      "learning_rate": 8.937379922158389e-05,
+      "loss": 1.3287,
+      "step": 7601
+    },
+    {
+      "epoch": 0.5479708786852159,
+      "grad_norm": 0.12487649380733835,
+      "learning_rate": 8.935058454801984e-05,
+      "loss": 1.2915,
+      "step": 7602
+    },
+    {
+      "epoch": 0.5480429611475528,
+      "grad_norm": 0.28863873376198756,
+      "learning_rate": 8.932737045494421e-05,
+      "loss": 1.2739,
+      "step": 7603
+    },
+    {
+      "epoch": 0.5481150436098897,
+      "grad_norm": 0.23829200961239105,
+      "learning_rate": 8.930415694362235e-05,
+      "loss": 1.2597,
+      "step": 7604
+    },
+    {
+      "epoch": 0.5481871260722266,
+      "grad_norm": 0.11368330821853831,
+      "learning_rate": 8.928094401531964e-05,
+      "loss": 1.3288,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5482592085345636,
+      "grad_norm": 0.11650974984692379,
+      "learning_rate": 8.925773167130133e-05,
+      "loss": 1.289,
+      "step": 7606
+    },
+    {
+      "epoch": 0.5483312909969005,
+      "grad_norm": 0.12003884448897992,
+      "learning_rate": 8.923451991283279e-05,
+      "loss": 1.2532,
+      "step": 7607
+    },
+    {
+      "epoch": 0.5484033734592374,
+      "grad_norm": 0.12204270248611453,
+      "learning_rate": 8.921130874117921e-05,
+      "loss": 1.3023,
+      "step": 7608
+    },
+    {
+      "epoch": 0.5484754559215743,
+      "grad_norm": 0.11481208333581538,
+      "learning_rate": 8.918809815760585e-05,
+      "loss": 1.4149,
+      "step": 7609
+    },
+    {
+      "epoch": 0.5485475383839112,
+      "grad_norm": 0.18990179020324643,
+      "learning_rate": 8.916488816337782e-05,
+      "loss": 1.3124,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5486196208462482,
+      "grad_norm": 0.11912255201905465,
+      "learning_rate": 8.914167875976034e-05,
+      "loss": 1.4056,
+      "step": 7611
+    },
+    {
+      "epoch": 0.548691703308585,
+      "grad_norm": 0.11559922209684025,
+      "learning_rate": 8.911846994801851e-05,
+      "loss": 1.2717,
+      "step": 7612
+    },
+    {
+      "epoch": 0.5487637857709219,
+      "grad_norm": 0.12304212453954427,
+      "learning_rate": 8.909526172941745e-05,
+      "loss": 1.3272,
+      "step": 7613
+    },
+    {
+      "epoch": 0.5488358682332588,
+      "grad_norm": 0.1345112915932042,
+      "learning_rate": 8.907205410522212e-05,
+      "loss": 1.3357,
+      "step": 7614
+    },
+    {
+      "epoch": 0.5489079506955957,
+      "grad_norm": 0.11762530784317024,
+      "learning_rate": 8.904884707669764e-05,
+      "loss": 1.2579,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5489800331579326,
+      "grad_norm": 0.09959504672537711,
+      "learning_rate": 8.902564064510897e-05,
+      "loss": 1.1998,
+      "step": 7616
+    },
+    {
+      "epoch": 0.5490521156202696,
+      "grad_norm": 0.11111538321990912,
+      "learning_rate": 8.900243481172107e-05,
+      "loss": 1.2312,
+      "step": 7617
+    },
+    {
+      "epoch": 0.5491241980826065,
+      "grad_norm": 0.12680429469605725,
+      "learning_rate": 8.897922957779885e-05,
+      "loss": 1.3434,
+      "step": 7618
+    },
+    {
+      "epoch": 0.5491962805449434,
+      "grad_norm": 0.13376509346044899,
+      "learning_rate": 8.895602494460719e-05,
+      "loss": 1.2609,
+      "step": 7619
+    },
+    {
+      "epoch": 0.5492683630072803,
+      "grad_norm": 0.11580991837491224,
+      "learning_rate": 8.893282091341099e-05,
+      "loss": 1.2365,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5493404454696172,
+      "grad_norm": 0.11386767992441493,
+      "learning_rate": 8.890961748547508e-05,
+      "loss": 1.2901,
+      "step": 7621
+    },
+    {
+      "epoch": 0.5494125279319542,
+      "grad_norm": 0.11855513243656202,
+      "learning_rate": 8.888641466206421e-05,
+      "loss": 1.2012,
+      "step": 7622
+    },
+    {
+      "epoch": 0.5494846103942911,
+      "grad_norm": 0.11369063914699426,
+      "learning_rate": 8.886321244444316e-05,
+      "loss": 1.2665,
+      "step": 7623
+    },
+    {
+      "epoch": 0.549556692856628,
+      "grad_norm": 0.12081232505619938,
+      "learning_rate": 8.884001083387667e-05,
+      "loss": 1.2295,
+      "step": 7624
+    },
+    {
+      "epoch": 0.5496287753189649,
+      "grad_norm": 0.11141825250694881,
+      "learning_rate": 8.881680983162946e-05,
+      "loss": 1.3381,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5497008577813018,
+      "grad_norm": 0.11422845686317498,
+      "learning_rate": 8.879360943896613e-05,
+      "loss": 1.3502,
+      "step": 7626
+    },
+    {
+      "epoch": 0.5497729402436388,
+      "grad_norm": 0.11120195317547547,
+      "learning_rate": 8.877040965715137e-05,
+      "loss": 1.315,
+      "step": 7627
+    },
+    {
+      "epoch": 0.5498450227059757,
+      "grad_norm": 0.11847804517666125,
+      "learning_rate": 8.874721048744968e-05,
+      "loss": 1.3185,
+      "step": 7628
+    },
+    {
+      "epoch": 0.5499171051683126,
+      "grad_norm": 0.10624120621512702,
+      "learning_rate": 8.872401193112576e-05,
+      "loss": 1.2231,
+      "step": 7629
+    },
+    {
+      "epoch": 0.5499891876306494,
+      "grad_norm": 0.11700502096635043,
+      "learning_rate": 8.870081398944405e-05,
+      "loss": 1.3067,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5500612700929863,
+      "grad_norm": 0.1216822351549558,
+      "learning_rate": 8.867761666366907e-05,
+      "loss": 1.2862,
+      "step": 7631
+    },
+    {
+      "epoch": 0.5501333525553233,
+      "grad_norm": 0.1328836610244233,
+      "learning_rate": 8.86544199550652e-05,
+      "loss": 1.2941,
+      "step": 7632
+    },
+    {
+      "epoch": 0.5502054350176602,
+      "grad_norm": 0.1121557374611903,
+      "learning_rate": 8.863122386489704e-05,
+      "loss": 1.2676,
+      "step": 7633
+    },
+    {
+      "epoch": 0.5502775174799971,
+      "grad_norm": 0.11748224261790585,
+      "learning_rate": 8.860802839442885e-05,
+      "loss": 1.3626,
+      "step": 7634
+    },
+    {
+      "epoch": 0.550349599942334,
+      "grad_norm": 0.10147371986167245,
+      "learning_rate": 8.858483354492504e-05,
+      "loss": 1.2728,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5504216824046709,
+      "grad_norm": 0.11933463858922398,
+      "learning_rate": 8.856163931764987e-05,
+      "loss": 1.2984,
+      "step": 7636
+    },
+    {
+      "epoch": 0.5504937648670079,
+      "grad_norm": 0.13529956935397922,
+      "learning_rate": 8.853844571386776e-05,
+      "loss": 1.3779,
+      "step": 7637
+    },
+    {
+      "epoch": 0.5505658473293448,
+      "grad_norm": 0.12225298750117428,
+      "learning_rate": 8.851525273484286e-05,
+      "loss": 1.3283,
+      "step": 7638
+    },
+    {
+      "epoch": 0.5506379297916817,
+      "grad_norm": 0.11849958716295597,
+      "learning_rate": 8.849206038183945e-05,
+      "loss": 1.2841,
+      "step": 7639
+    },
+    {
+      "epoch": 0.5507100122540186,
+      "grad_norm": 0.11799884606819146,
+      "learning_rate": 8.846886865612169e-05,
+      "loss": 1.2427,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5507820947163555,
+      "grad_norm": 0.11807527171832483,
+      "learning_rate": 8.844567755895374e-05,
+      "loss": 1.1287,
+      "step": 7641
+    },
+    {
+      "epoch": 0.5508541771786925,
+      "grad_norm": 0.1058966737189468,
+      "learning_rate": 8.842248709159972e-05,
+      "loss": 1.2672,
+      "step": 7642
+    },
+    {
+      "epoch": 0.5509262596410294,
+      "grad_norm": 0.11383528082257195,
+      "learning_rate": 8.839929725532378e-05,
+      "loss": 1.2416,
+      "step": 7643
+    },
+    {
+      "epoch": 0.5509983421033663,
+      "grad_norm": 0.18912428272195106,
+      "learning_rate": 8.837610805138989e-05,
+      "loss": 1.2972,
+      "step": 7644
+    },
+    {
+      "epoch": 0.5510704245657032,
+      "grad_norm": 0.12307652224198211,
+      "learning_rate": 8.835291948106207e-05,
+      "loss": 1.3219,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5511425070280401,
+      "grad_norm": 0.10122365440300365,
+      "learning_rate": 8.832973154560438e-05,
+      "loss": 1.2597,
+      "step": 7646
+    },
+    {
+      "epoch": 0.5512145894903769,
+      "grad_norm": 0.17033285084770422,
+      "learning_rate": 8.830654424628074e-05,
+      "loss": 1.3639,
+      "step": 7647
+    },
+    {
+      "epoch": 0.5512866719527139,
+      "grad_norm": 0.11500973376723848,
+      "learning_rate": 8.828335758435502e-05,
+      "loss": 1.3384,
+      "step": 7648
+    },
+    {
+      "epoch": 0.5513587544150508,
+      "grad_norm": 0.1090567798995786,
+      "learning_rate": 8.826017156109116e-05,
+      "loss": 1.4199,
+      "step": 7649
+    },
+    {
+      "epoch": 0.5514308368773877,
+      "grad_norm": 0.1137856249338571,
+      "learning_rate": 8.823698617775292e-05,
+      "loss": 1.3542,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5515029193397246,
+      "grad_norm": 0.11060696315863733,
+      "learning_rate": 8.821380143560422e-05,
+      "loss": 1.2986,
+      "step": 7651
+    },
+    {
+      "epoch": 0.5515750018020615,
+      "grad_norm": 0.12040066479060038,
+      "learning_rate": 8.819061733590876e-05,
+      "loss": 1.2911,
+      "step": 7652
+    },
+    {
+      "epoch": 0.5516470842643985,
+      "grad_norm": 0.13041709055733092,
+      "learning_rate": 8.816743387993034e-05,
+      "loss": 1.2958,
+      "step": 7653
+    },
+    {
+      "epoch": 0.5517191667267354,
+      "grad_norm": 0.10884266702891393,
+      "learning_rate": 8.814425106893256e-05,
+      "loss": 1.3035,
+      "step": 7654
+    },
+    {
+      "epoch": 0.5517912491890723,
+      "grad_norm": 0.10586967849939449,
+      "learning_rate": 8.812106890417923e-05,
+      "loss": 1.1928,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5518633316514092,
+      "grad_norm": 0.11257472889764852,
+      "learning_rate": 8.809788738693388e-05,
+      "loss": 1.2871,
+      "step": 7656
+    },
+    {
+      "epoch": 0.5519354141137461,
+      "grad_norm": 0.12042393548153005,
+      "learning_rate": 8.807470651846018e-05,
+      "loss": 1.3293,
+      "step": 7657
+    },
+    {
+      "epoch": 0.5520074965760831,
+      "grad_norm": 0.1074475742735251,
+      "learning_rate": 8.80515263000216e-05,
+      "loss": 1.2979,
+      "step": 7658
+    },
+    {
+      "epoch": 0.55207957903842,
+      "grad_norm": 0.1200872924558861,
+      "learning_rate": 8.802834673288178e-05,
+      "loss": 1.2754,
+      "step": 7659
+    },
+    {
+      "epoch": 0.5521516615007569,
+      "grad_norm": 0.13493039523158698,
+      "learning_rate": 8.800516781830417e-05,
+      "loss": 1.2716,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5522237439630938,
+      "grad_norm": 0.11532178646009388,
+      "learning_rate": 8.798198955755223e-05,
+      "loss": 1.3395,
+      "step": 7661
+    },
+    {
+      "epoch": 0.5522958264254307,
+      "grad_norm": 0.11075441235377566,
+      "learning_rate": 8.795881195188934e-05,
+      "loss": 1.2271,
+      "step": 7662
+    },
+    {
+      "epoch": 0.5523679088877677,
+      "grad_norm": 0.13569720060081672,
+      "learning_rate": 8.793563500257893e-05,
+      "loss": 1.3173,
+      "step": 7663
+    },
+    {
+      "epoch": 0.5524399913501045,
+      "grad_norm": 0.12313132320874318,
+      "learning_rate": 8.791245871088433e-05,
+      "loss": 1.3216,
+      "step": 7664
+    },
+    {
+      "epoch": 0.5525120738124414,
+      "grad_norm": 0.22997993055815588,
+      "learning_rate": 8.78892830780689e-05,
+      "loss": 1.3066,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5525841562747783,
+      "grad_norm": 0.1193667386011618,
+      "learning_rate": 8.786610810539588e-05,
+      "loss": 1.3055,
+      "step": 7666
+    },
+    {
+      "epoch": 0.5526562387371152,
+      "grad_norm": 0.10896778008897559,
+      "learning_rate": 8.78429337941285e-05,
+      "loss": 1.2906,
+      "step": 7667
+    },
+    {
+      "epoch": 0.5527283211994521,
+      "grad_norm": 0.11348718438191567,
+      "learning_rate": 8.781976014553001e-05,
+      "loss": 1.2683,
+      "step": 7668
+    },
+    {
+      "epoch": 0.5528004036617891,
+      "grad_norm": 0.11979787074326975,
+      "learning_rate": 8.779658716086356e-05,
+      "loss": 1.2594,
+      "step": 7669
+    },
+    {
+      "epoch": 0.552872486124126,
+      "grad_norm": 0.11795887600902889,
+      "learning_rate": 8.777341484139229e-05,
+      "loss": 1.3056,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5529445685864629,
+      "grad_norm": 0.1318349982977022,
+      "learning_rate": 8.775024318837928e-05,
+      "loss": 1.2513,
+      "step": 7671
+    },
+    {
+      "epoch": 0.5530166510487998,
+      "grad_norm": 0.22429153704447113,
+      "learning_rate": 8.772707220308758e-05,
+      "loss": 1.2844,
+      "step": 7672
+    },
+    {
+      "epoch": 0.5530887335111367,
+      "grad_norm": 0.11603607286633379,
+      "learning_rate": 8.770390188678029e-05,
+      "loss": 1.2175,
+      "step": 7673
+    },
+    {
+      "epoch": 0.5531608159734737,
+      "grad_norm": 0.12783684001039308,
+      "learning_rate": 8.768073224072033e-05,
+      "loss": 1.3162,
+      "step": 7674
+    },
+    {
+      "epoch": 0.5532328984358106,
+      "grad_norm": 0.1337534963661725,
+      "learning_rate": 8.765756326617068e-05,
+      "loss": 1.2863,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5533049808981475,
+      "grad_norm": 0.18450437133384806,
+      "learning_rate": 8.763439496439422e-05,
+      "loss": 1.2634,
+      "step": 7676
+    },
+    {
+      "epoch": 0.5533770633604844,
+      "grad_norm": 0.11940841890646521,
+      "learning_rate": 8.76112273366539e-05,
+      "loss": 1.2865,
+      "step": 7677
+    },
+    {
+      "epoch": 0.5534491458228213,
+      "grad_norm": 0.11709972549984284,
+      "learning_rate": 8.758806038421251e-05,
+      "loss": 1.3916,
+      "step": 7678
+    },
+    {
+      "epoch": 0.5535212282851583,
+      "grad_norm": 0.27592734153860493,
+      "learning_rate": 8.756489410833287e-05,
+      "loss": 1.2737,
+      "step": 7679
+    },
+    {
+      "epoch": 0.5535933107474952,
+      "grad_norm": 0.11795389745297508,
+      "learning_rate": 8.754172851027772e-05,
+      "loss": 1.2672,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5536653932098321,
+      "grad_norm": 0.11174078600655972,
+      "learning_rate": 8.751856359130988e-05,
+      "loss": 1.3481,
+      "step": 7681
+    },
+    {
+      "epoch": 0.5537374756721689,
+      "grad_norm": 0.10451717414670897,
+      "learning_rate": 8.749539935269195e-05,
+      "loss": 1.2904,
+      "step": 7682
+    },
+    {
+      "epoch": 0.5538095581345058,
+      "grad_norm": 0.11074094760189211,
+      "learning_rate": 8.747223579568662e-05,
+      "loss": 1.1828,
+      "step": 7683
+    },
+    {
+      "epoch": 0.5538816405968428,
+      "grad_norm": 0.10993545978662522,
+      "learning_rate": 8.744907292155655e-05,
+      "loss": 1.2854,
+      "step": 7684
+    },
+    {
+      "epoch": 0.5539537230591797,
+      "grad_norm": 0.11166068193610539,
+      "learning_rate": 8.742591073156424e-05,
+      "loss": 1.2411,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5540258055215166,
+      "grad_norm": 0.13544123211672224,
+      "learning_rate": 8.740274922697231e-05,
+      "loss": 1.2635,
+      "step": 7686
+    },
+    {
+      "epoch": 0.5540978879838535,
+      "grad_norm": 0.13389952088130522,
+      "learning_rate": 8.737958840904324e-05,
+      "loss": 1.2895,
+      "step": 7687
+    },
+    {
+      "epoch": 0.5541699704461904,
+      "grad_norm": 0.12178600124617947,
+      "learning_rate": 8.735642827903953e-05,
+      "loss": 1.2727,
+      "step": 7688
+    },
+    {
+      "epoch": 0.5542420529085273,
+      "grad_norm": 0.1196473253821643,
+      "learning_rate": 8.733326883822356e-05,
+      "loss": 1.4324,
+      "step": 7689
+    },
+    {
+      "epoch": 0.5543141353708643,
+      "grad_norm": 0.1198564883515347,
+      "learning_rate": 8.731011008785776e-05,
+      "loss": 1.3002,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5543862178332012,
+      "grad_norm": 0.20035291857043339,
+      "learning_rate": 8.728695202920451e-05,
+      "loss": 1.267,
+      "step": 7691
+    },
+    {
+      "epoch": 0.5544583002955381,
+      "grad_norm": 0.11349410851109122,
+      "learning_rate": 8.726379466352613e-05,
+      "loss": 1.3126,
+      "step": 7692
+    },
+    {
+      "epoch": 0.554530382757875,
+      "grad_norm": 0.11444293093639261,
+      "learning_rate": 8.724063799208483e-05,
+      "loss": 1.2031,
+      "step": 7693
+    },
+    {
+      "epoch": 0.554602465220212,
+      "grad_norm": 0.12078710646109257,
+      "learning_rate": 8.721748201614292e-05,
+      "loss": 1.2681,
+      "step": 7694
+    },
+    {
+      "epoch": 0.5546745476825489,
+      "grad_norm": 0.11427890824070532,
+      "learning_rate": 8.719432673696259e-05,
+      "loss": 1.2681,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5547466301448858,
+      "grad_norm": 0.12766288782524632,
+      "learning_rate": 8.717117215580606e-05,
+      "loss": 1.3557,
+      "step": 7696
+    },
+    {
+      "epoch": 0.5548187126072227,
+      "grad_norm": 0.15878615961023765,
+      "learning_rate": 8.714801827393538e-05,
+      "loss": 1.3147,
+      "step": 7697
+    },
+    {
+      "epoch": 0.5548907950695596,
+      "grad_norm": 0.12390780597186872,
+      "learning_rate": 8.712486509261267e-05,
+      "loss": 1.3021,
+      "step": 7698
+    },
+    {
+      "epoch": 0.5549628775318964,
+      "grad_norm": 0.10997864880688658,
+      "learning_rate": 8.710171261310002e-05,
+      "loss": 1.2785,
+      "step": 7699
+    },
+    {
+      "epoch": 0.5550349599942334,
+      "grad_norm": 0.1306242744980595,
+      "learning_rate": 8.707856083665944e-05,
+      "loss": 1.3052,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5551070424565703,
+      "grad_norm": 0.12554903549520452,
+      "learning_rate": 8.705540976455287e-05,
+      "loss": 1.2039,
+      "step": 7701
+    },
+    {
+      "epoch": 0.5551791249189072,
+      "grad_norm": 0.12033208762065668,
+      "learning_rate": 8.703225939804225e-05,
+      "loss": 1.3062,
+      "step": 7702
+    },
+    {
+      "epoch": 0.5552512073812441,
+      "grad_norm": 0.1258103075068518,
+      "learning_rate": 8.700910973838952e-05,
+      "loss": 1.2988,
+      "step": 7703
+    },
+    {
+      "epoch": 0.555323289843581,
+      "grad_norm": 0.12116400527613358,
+      "learning_rate": 8.698596078685656e-05,
+      "loss": 1.386,
+      "step": 7704
+    },
+    {
+      "epoch": 0.555395372305918,
+      "grad_norm": 0.10359847004549141,
+      "learning_rate": 8.696281254470516e-05,
+      "loss": 1.2999,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5554674547682549,
+      "grad_norm": 0.10738264141544149,
+      "learning_rate": 8.693966501319711e-05,
+      "loss": 1.296,
+      "step": 7706
+    },
+    {
+      "epoch": 0.5555395372305918,
+      "grad_norm": 0.10837772287413054,
+      "learning_rate": 8.691651819359409e-05,
+      "loss": 1.3351,
+      "step": 7707
+    },
+    {
+      "epoch": 0.5556116196929287,
+      "grad_norm": 0.15434199289213793,
+      "learning_rate": 8.689337208715796e-05,
+      "loss": 1.3413,
+      "step": 7708
+    },
+    {
+      "epoch": 0.5556837021552656,
+      "grad_norm": 0.13225123164836242,
+      "learning_rate": 8.687022669515027e-05,
+      "loss": 1.2453,
+      "step": 7709
+    },
+    {
+      "epoch": 0.5557557846176026,
+      "grad_norm": 0.11528805976202593,
+      "learning_rate": 8.684708201883272e-05,
+      "loss": 1.3317,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5558278670799395,
+      "grad_norm": 0.10532075416590743,
+      "learning_rate": 8.682393805946679e-05,
+      "loss": 1.2484,
+      "step": 7711
+    },
+    {
+      "epoch": 0.5558999495422764,
+      "grad_norm": 0.11804808512260487,
+      "learning_rate": 8.680079481831419e-05,
+      "loss": 1.3107,
+      "step": 7712
+    },
+    {
+      "epoch": 0.5559720320046133,
+      "grad_norm": 0.12893567019984767,
+      "learning_rate": 8.677765229663633e-05,
+      "loss": 1.3273,
+      "step": 7713
+    },
+    {
+      "epoch": 0.5560441144669502,
+      "grad_norm": 0.12654264774200363,
+      "learning_rate": 8.675451049569472e-05,
+      "loss": 1.3251,
+      "step": 7714
+    },
+    {
+      "epoch": 0.5561161969292872,
+      "grad_norm": 0.12219401315196492,
+      "learning_rate": 8.673136941675079e-05,
+      "loss": 1.284,
+      "step": 7715
+    },
+    {
+      "epoch": 0.556188279391624,
+      "grad_norm": 0.13626511149800366,
+      "learning_rate": 8.670822906106589e-05,
+      "loss": 1.2623,
+      "step": 7716
+    },
+    {
+      "epoch": 0.5562603618539609,
+      "grad_norm": 0.11314408357467157,
+      "learning_rate": 8.668508942990145e-05,
+      "loss": 1.3149,
+      "step": 7717
+    },
+    {
+      "epoch": 0.5563324443162978,
+      "grad_norm": 0.1087170929800134,
+      "learning_rate": 8.666195052451877e-05,
+      "loss": 1.2767,
+      "step": 7718
+    },
+    {
+      "epoch": 0.5564045267786347,
+      "grad_norm": 0.11921008174287592,
+      "learning_rate": 8.663881234617909e-05,
+      "loss": 1.2942,
+      "step": 7719
+    },
+    {
+      "epoch": 0.5564766092409716,
+      "grad_norm": 0.10919752235088766,
+      "learning_rate": 8.661567489614363e-05,
+      "loss": 1.4093,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5565486917033086,
+      "grad_norm": 0.1382727227348337,
+      "learning_rate": 8.659253817567369e-05,
+      "loss": 1.3419,
+      "step": 7721
+    },
+    {
+      "epoch": 0.5566207741656455,
+      "grad_norm": 0.11271085791640126,
+      "learning_rate": 8.656940218603036e-05,
+      "loss": 1.1741,
+      "step": 7722
+    },
+    {
+      "epoch": 0.5566928566279824,
+      "grad_norm": 0.22012027303954015,
+      "learning_rate": 8.654626692847475e-05,
+      "loss": 1.3696,
+      "step": 7723
+    },
+    {
+      "epoch": 0.5567649390903193,
+      "grad_norm": 0.13295511548308295,
+      "learning_rate": 8.652313240426797e-05,
+      "loss": 1.2456,
+      "step": 7724
+    },
+    {
+      "epoch": 0.5568370215526562,
+      "grad_norm": 0.11763173253118625,
+      "learning_rate": 8.649999861467099e-05,
+      "loss": 1.2638,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5569091040149932,
+      "grad_norm": 0.10627576616946909,
+      "learning_rate": 8.64768655609449e-05,
+      "loss": 1.248,
+      "step": 7726
+    },
+    {
+      "epoch": 0.5569811864773301,
+      "grad_norm": 0.11123367478807357,
+      "learning_rate": 8.645373324435062e-05,
+      "loss": 1.1795,
+      "step": 7727
+    },
+    {
+      "epoch": 0.557053268939667,
+      "grad_norm": 0.10342518529078541,
+      "learning_rate": 8.643060166614908e-05,
+      "loss": 1.2591,
+      "step": 7728
+    },
+    {
+      "epoch": 0.5571253514020039,
+      "grad_norm": 0.11282609080630697,
+      "learning_rate": 8.640747082760108e-05,
+      "loss": 1.1797,
+      "step": 7729
+    },
+    {
+      "epoch": 0.5571974338643408,
+      "grad_norm": 0.1239297580484943,
+      "learning_rate": 8.638434072996759e-05,
+      "loss": 1.2861,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5572695163266778,
+      "grad_norm": 0.12013229094833193,
+      "learning_rate": 8.636121137450931e-05,
+      "loss": 1.2804,
+      "step": 7731
+    },
+    {
+      "epoch": 0.5573415987890147,
+      "grad_norm": 0.12020522195469244,
+      "learning_rate": 8.633808276248705e-05,
+      "loss": 1.3181,
+      "step": 7732
+    },
+    {
+      "epoch": 0.5574136812513516,
+      "grad_norm": 0.11320730083920798,
+      "learning_rate": 8.631495489516144e-05,
+      "loss": 1.2272,
+      "step": 7733
+    },
+    {
+      "epoch": 0.5574857637136884,
+      "grad_norm": 0.1114185416333173,
+      "learning_rate": 8.629182777379328e-05,
+      "loss": 1.3114,
+      "step": 7734
+    },
+    {
+      "epoch": 0.5575578461760253,
+      "grad_norm": 0.11971810172088257,
+      "learning_rate": 8.626870139964313e-05,
+      "loss": 1.3371,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5576299286383622,
+      "grad_norm": 0.11711862857251845,
+      "learning_rate": 8.62455757739716e-05,
+      "loss": 1.2856,
+      "step": 7736
+    },
+    {
+      "epoch": 0.5577020111006992,
+      "grad_norm": 0.1094339263138447,
+      "learning_rate": 8.622245089803925e-05,
+      "loss": 1.2757,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5577740935630361,
+      "grad_norm": 0.1222058942481018,
+      "learning_rate": 8.619932677310653e-05,
+      "loss": 1.1835,
+      "step": 7738
+    },
+    {
+      "epoch": 0.557846176025373,
+      "grad_norm": 0.11660540384602075,
+      "learning_rate": 8.617620340043401e-05,
+      "loss": 1.2862,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5579182584877099,
+      "grad_norm": 0.10904081519137737,
+      "learning_rate": 8.61530807812821e-05,
+      "loss": 1.2391,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5579903409500468,
+      "grad_norm": 0.10201765610553733,
+      "learning_rate": 8.612995891691113e-05,
+      "loss": 1.2516,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5580624234123838,
+      "grad_norm": 0.12449642374230943,
+      "learning_rate": 8.610683780858147e-05,
+      "loss": 1.2037,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5581345058747207,
+      "grad_norm": 0.11936181791406764,
+      "learning_rate": 8.608371745755345e-05,
+      "loss": 1.2348,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5582065883370576,
+      "grad_norm": 0.13242178538237923,
+      "learning_rate": 8.606059786508735e-05,
+      "loss": 1.3278,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5582786707993945,
+      "grad_norm": 0.10921448967559431,
+      "learning_rate": 8.603747903244335e-05,
+      "loss": 1.242,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5583507532617314,
+      "grad_norm": 0.11813840193312714,
+      "learning_rate": 8.601436096088166e-05,
+      "loss": 1.2695,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5584228357240684,
+      "grad_norm": 0.16663788814478714,
+      "learning_rate": 8.59912436516624e-05,
+      "loss": 1.3,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5584949181864053,
+      "grad_norm": 0.11846762122642134,
+      "learning_rate": 8.596812710604574e-05,
+      "loss": 1.2994,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5585670006487422,
+      "grad_norm": 0.1375195004177613,
+      "learning_rate": 8.594501132529165e-05,
+      "loss": 1.288,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5586390831110791,
+      "grad_norm": 0.12240952789531781,
+      "learning_rate": 8.592189631066018e-05,
+      "loss": 1.2846,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5587111655734159,
+      "grad_norm": 0.11004126003536359,
+      "learning_rate": 8.589878206341129e-05,
+      "loss": 1.2446,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5587832480357529,
+      "grad_norm": 0.10810857333690828,
+      "learning_rate": 8.587566858480496e-05,
+      "loss": 1.2754,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5588553304980898,
+      "grad_norm": 0.10412858505815877,
+      "learning_rate": 8.585255587610105e-05,
+      "loss": 1.2394,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5589274129604267,
+      "grad_norm": 0.10525335034203738,
+      "learning_rate": 8.582944393855941e-05,
+      "loss": 1.2272,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5589994954227636,
+      "grad_norm": 0.1149479499356176,
+      "learning_rate": 8.580633277343982e-05,
+      "loss": 1.1938,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5590715778851005,
+      "grad_norm": 0.1233217979138401,
+      "learning_rate": 8.578322238200211e-05,
+      "loss": 1.2002,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5591436603474375,
+      "grad_norm": 0.11435029637866198,
+      "learning_rate": 8.576011276550596e-05,
+      "loss": 1.3723,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5592157428097744,
+      "grad_norm": 0.1143485109815911,
+      "learning_rate": 8.573700392521106e-05,
+      "loss": 1.2806,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5592878252721113,
+      "grad_norm": 0.10602829349347373,
+      "learning_rate": 8.571389586237706e-05,
+      "loss": 1.2654,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5593599077344482,
+      "grad_norm": 0.1311386478260663,
+      "learning_rate": 8.56907885782635e-05,
+      "loss": 1.1861,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5594319901967851,
+      "grad_norm": 0.18129103219328235,
+      "learning_rate": 8.566768207413e-05,
+      "loss": 1.2778,
+      "step": 7761
+    },
+    {
+      "epoch": 0.559504072659122,
+      "grad_norm": 0.11724757023228327,
+      "learning_rate": 8.564457635123603e-05,
+      "loss": 1.3033,
+      "step": 7762
+    },
+    {
+      "epoch": 0.559576155121459,
+      "grad_norm": 0.19668222820623038,
+      "learning_rate": 8.562147141084112e-05,
+      "loss": 1.3097,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5596482375837959,
+      "grad_norm": 0.10869003501273686,
+      "learning_rate": 8.55983672542046e-05,
+      "loss": 1.3014,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5597203200461328,
+      "grad_norm": 0.12006519841123775,
+      "learning_rate": 8.557526388258593e-05,
+      "loss": 1.3508,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5597924025084697,
+      "grad_norm": 0.11677721008643703,
+      "learning_rate": 8.555216129724442e-05,
+      "loss": 1.2651,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5598644849708067,
+      "grad_norm": 0.11831024548662063,
+      "learning_rate": 8.55290594994394e-05,
+      "loss": 1.2843,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5599365674331435,
+      "grad_norm": 0.11813097320251068,
+      "learning_rate": 8.550595849043008e-05,
+      "loss": 1.1604,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5600086498954804,
+      "grad_norm": 0.11679104325970391,
+      "learning_rate": 8.548285827147568e-05,
+      "loss": 1.2681,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5600807323578173,
+      "grad_norm": 0.11279093794089368,
+      "learning_rate": 8.545975884383537e-05,
+      "loss": 1.2834,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5601528148201542,
+      "grad_norm": 0.1261442866446143,
+      "learning_rate": 8.543666020876832e-05,
+      "loss": 1.2498,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5602248972824911,
+      "grad_norm": 0.1169539645522041,
+      "learning_rate": 8.541356236753357e-05,
+      "loss": 1.2677,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5602969797448281,
+      "grad_norm": 0.1163715155215034,
+      "learning_rate": 8.539046532139011e-05,
+      "loss": 1.291,
+      "step": 7773
+    },
+    {
+      "epoch": 0.560369062207165,
+      "grad_norm": 0.1072221189979291,
+      "learning_rate": 8.536736907159704e-05,
+      "loss": 1.3033,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5604411446695019,
+      "grad_norm": 0.12022240893680698,
+      "learning_rate": 8.534427361941327e-05,
+      "loss": 1.4156,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5605132271318388,
+      "grad_norm": 0.11810691970124663,
+      "learning_rate": 8.532117896609768e-05,
+      "loss": 1.2953,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5605853095941757,
+      "grad_norm": 0.10704138479509029,
+      "learning_rate": 8.529808511290915e-05,
+      "loss": 1.2406,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5606573920565127,
+      "grad_norm": 0.11076520659234372,
+      "learning_rate": 8.527499206110652e-05,
+      "loss": 1.3293,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5607294745188496,
+      "grad_norm": 0.11284604664092175,
+      "learning_rate": 8.525189981194857e-05,
+      "loss": 1.321,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5608015569811865,
+      "grad_norm": 0.13086187955264583,
+      "learning_rate": 8.522880836669401e-05,
+      "loss": 1.3828,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5608736394435234,
+      "grad_norm": 0.11749344999234551,
+      "learning_rate": 8.520571772660155e-05,
+      "loss": 1.2703,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5609457219058603,
+      "grad_norm": 0.11560871726154373,
+      "learning_rate": 8.518262789292977e-05,
+      "loss": 1.3326,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5610178043681973,
+      "grad_norm": 0.11220426591548605,
+      "learning_rate": 8.515953886693739e-05,
+      "loss": 1.2168,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5610898868305342,
+      "grad_norm": 0.1127715296033893,
+      "learning_rate": 8.513645064988291e-05,
+      "loss": 1.3217,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5611619692928711,
+      "grad_norm": 0.1158105587720566,
+      "learning_rate": 8.511336324302483e-05,
+      "loss": 1.2969,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5612340517552079,
+      "grad_norm": 0.12228883446858693,
+      "learning_rate": 8.509027664762159e-05,
+      "loss": 1.2829,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5613061342175448,
+      "grad_norm": 0.11458784935364805,
+      "learning_rate": 8.506719086493172e-05,
+      "loss": 1.3513,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5613782166798817,
+      "grad_norm": 0.11369145206584301,
+      "learning_rate": 8.504410589621352e-05,
+      "loss": 1.3166,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5614502991422187,
+      "grad_norm": 0.11155731397210157,
+      "learning_rate": 8.502102174272538e-05,
+      "loss": 1.3446,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5615223816045556,
+      "grad_norm": 0.12598909677010264,
+      "learning_rate": 8.499793840572552e-05,
+      "loss": 1.2261,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5615944640668925,
+      "grad_norm": 0.1130353590485283,
+      "learning_rate": 8.49748558864722e-05,
+      "loss": 1.3825,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5616665465292294,
+      "grad_norm": 0.23088094596377184,
+      "learning_rate": 8.49517741862237e-05,
+      "loss": 1.1824,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5617386289915663,
+      "grad_norm": 0.11804053867962971,
+      "learning_rate": 8.492869330623814e-05,
+      "loss": 1.3277,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5618107114539033,
+      "grad_norm": 0.13826081497919615,
+      "learning_rate": 8.490561324777361e-05,
+      "loss": 1.3045,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5618827939162402,
+      "grad_norm": 0.11396360747618799,
+      "learning_rate": 8.488253401208816e-05,
+      "loss": 1.199,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5619548763785771,
+      "grad_norm": 0.11353490683235881,
+      "learning_rate": 8.485945560043988e-05,
+      "loss": 1.3031,
+      "step": 7796
+    },
+    {
+      "epoch": 0.562026958840914,
+      "grad_norm": 0.2239152064878169,
+      "learning_rate": 8.483637801408676e-05,
+      "loss": 1.3069,
+      "step": 7797
+    },
+    {
+      "epoch": 0.562099041303251,
+      "grad_norm": 0.11810879213368214,
+      "learning_rate": 8.481330125428666e-05,
+      "loss": 1.2461,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5621711237655879,
+      "grad_norm": 0.12156031588272076,
+      "learning_rate": 8.479022532229747e-05,
+      "loss": 1.3134,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5622432062279248,
+      "grad_norm": 0.1180680929587807,
+      "learning_rate": 8.476715021937712e-05,
+      "loss": 1.2576,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5623152886902617,
+      "grad_norm": 0.12250870156338681,
+      "learning_rate": 8.474407594678337e-05,
+      "loss": 1.3554,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5623873711525986,
+      "grad_norm": 0.11517093336296746,
+      "learning_rate": 8.472100250577395e-05,
+      "loss": 1.2877,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5624594536149354,
+      "grad_norm": 0.10853693790559875,
+      "learning_rate": 8.469792989760658e-05,
+      "loss": 1.3442,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5625315360772724,
+      "grad_norm": 0.11469397447672915,
+      "learning_rate": 8.46748581235389e-05,
+      "loss": 1.2073,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5626036185396093,
+      "grad_norm": 0.15841479723846913,
+      "learning_rate": 8.465178718482859e-05,
+      "loss": 1.2687,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5626757010019462,
+      "grad_norm": 0.12084521874089268,
+      "learning_rate": 8.462871708273316e-05,
+      "loss": 1.3437,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5627477834642831,
+      "grad_norm": 0.11161588275474459,
+      "learning_rate": 8.460564781851021e-05,
+      "loss": 1.275,
+      "step": 7807
+    },
+    {
+      "epoch": 0.56281986592662,
+      "grad_norm": 0.11275633932904044,
+      "learning_rate": 8.458257939341711e-05,
+      "loss": 1.2533,
+      "step": 7808
+    },
+    {
+      "epoch": 0.562891948388957,
+      "grad_norm": 0.11486615123080245,
+      "learning_rate": 8.455951180871141e-05,
+      "loss": 1.325,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5629640308512939,
+      "grad_norm": 0.3284632884413201,
+      "learning_rate": 8.453644506565044e-05,
+      "loss": 1.2625,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5630361133136308,
+      "grad_norm": 0.10517051024630614,
+      "learning_rate": 8.451337916549157e-05,
+      "loss": 1.2071,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5631081957759677,
+      "grad_norm": 0.11853935734232451,
+      "learning_rate": 8.449031410949206e-05,
+      "loss": 1.2656,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5631802782383046,
+      "grad_norm": 0.11234984974434696,
+      "learning_rate": 8.446724989890917e-05,
+      "loss": 1.3735,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5632523607006416,
+      "grad_norm": 0.10415736969575991,
+      "learning_rate": 8.444418653500014e-05,
+      "loss": 1.3423,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5633244431629785,
+      "grad_norm": 0.11794468038850565,
+      "learning_rate": 8.442112401902214e-05,
+      "loss": 1.2533,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5633965256253154,
+      "grad_norm": 0.1113234760152216,
+      "learning_rate": 8.439806235223221e-05,
+      "loss": 1.2869,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5634686080876523,
+      "grad_norm": 0.13515176289993222,
+      "learning_rate": 8.437500153588746e-05,
+      "loss": 1.2994,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5635406905499892,
+      "grad_norm": 0.1951587542443224,
+      "learning_rate": 8.435194157124492e-05,
+      "loss": 1.3225,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5636127730123262,
+      "grad_norm": 0.11170412224384436,
+      "learning_rate": 8.432888245956158e-05,
+      "loss": 1.1845,
+      "step": 7819
+    },
+    {
+      "epoch": 0.563684855474663,
+      "grad_norm": 0.12625512684316323,
+      "learning_rate": 8.430582420209432e-05,
+      "loss": 1.2587,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5637569379369999,
+      "grad_norm": 0.1250353051472128,
+      "learning_rate": 8.428276680010004e-05,
+      "loss": 1.2315,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5638290203993368,
+      "grad_norm": 0.1274438582141109,
+      "learning_rate": 8.425971025483558e-05,
+      "loss": 1.2931,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5639011028616737,
+      "grad_norm": 0.12616175510259842,
+      "learning_rate": 8.423665456755776e-05,
+      "loss": 1.2979,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5639731853240106,
+      "grad_norm": 0.1172529036139641,
+      "learning_rate": 8.421359973952328e-05,
+      "loss": 1.192,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5640452677863476,
+      "grad_norm": 0.13031489657111223,
+      "learning_rate": 8.419054577198882e-05,
+      "loss": 1.3478,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5641173502486845,
+      "grad_norm": 0.10916059103222998,
+      "learning_rate": 8.416749266621105e-05,
+      "loss": 1.3049,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5641894327110214,
+      "grad_norm": 0.12141293329756088,
+      "learning_rate": 8.414444042344659e-05,
+      "loss": 1.3487,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5642615151733583,
+      "grad_norm": 0.12349801458094939,
+      "learning_rate": 8.412138904495196e-05,
+      "loss": 1.2988,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5643335976356952,
+      "grad_norm": 0.1150323271822502,
+      "learning_rate": 8.40983385319837e-05,
+      "loss": 1.2573,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5644056800980322,
+      "grad_norm": 0.11255556462058287,
+      "learning_rate": 8.407528888579821e-05,
+      "loss": 1.2589,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5644777625603691,
+      "grad_norm": 0.11641084115102877,
+      "learning_rate": 8.4052240107652e-05,
+      "loss": 1.2769,
+      "step": 7831
+    },
+    {
+      "epoch": 0.564549845022706,
+      "grad_norm": 0.12049801930730053,
+      "learning_rate": 8.402919219880134e-05,
+      "loss": 1.1488,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5646219274850429,
+      "grad_norm": 0.11210175134294549,
+      "learning_rate": 8.400614516050259e-05,
+      "loss": 1.2585,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5646940099473798,
+      "grad_norm": 0.10508161200089526,
+      "learning_rate": 8.398309899401204e-05,
+      "loss": 1.2096,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5647660924097168,
+      "grad_norm": 0.11531374174168224,
+      "learning_rate": 8.396005370058584e-05,
+      "loss": 1.3003,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5648381748720537,
+      "grad_norm": 0.12259629643456038,
+      "learning_rate": 8.393700928148024e-05,
+      "loss": 1.2741,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5649102573343906,
+      "grad_norm": 0.12339702445707106,
+      "learning_rate": 8.391396573795133e-05,
+      "loss": 1.3055,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5649823397967274,
+      "grad_norm": 0.11590049757247578,
+      "learning_rate": 8.389092307125521e-05,
+      "loss": 1.2745,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5650544222590643,
+      "grad_norm": 0.1278151502931907,
+      "learning_rate": 8.386788128264787e-05,
+      "loss": 1.2653,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5651265047214012,
+      "grad_norm": 0.14233570386759523,
+      "learning_rate": 8.384484037338533e-05,
+      "loss": 1.3099,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5651985871837382,
+      "grad_norm": 0.11722694718920082,
+      "learning_rate": 8.382180034472353e-05,
+      "loss": 1.1773,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5652706696460751,
+      "grad_norm": 0.12222410839363998,
+      "learning_rate": 8.379876119791836e-05,
+      "loss": 1.2807,
+      "step": 7842
+    },
+    {
+      "epoch": 0.565342752108412,
+      "grad_norm": 0.33582185123067815,
+      "learning_rate": 8.37757229342256e-05,
+      "loss": 1.256,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5654148345707489,
+      "grad_norm": 0.11611843048978933,
+      "learning_rate": 8.375268555490112e-05,
+      "loss": 1.3047,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5654869170330858,
+      "grad_norm": 0.10921984143721533,
+      "learning_rate": 8.372964906120062e-05,
+      "loss": 1.3205,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5655589994954228,
+      "grad_norm": 0.11859648306584511,
+      "learning_rate": 8.370661345437984e-05,
+      "loss": 1.2225,
+      "step": 7846
+    },
+    {
+      "epoch": 0.5656310819577597,
+      "grad_norm": 0.11639759896126603,
+      "learning_rate": 8.368357873569436e-05,
+      "loss": 1.294,
+      "step": 7847
+    },
+    {
+      "epoch": 0.5657031644200966,
+      "grad_norm": 0.11239053712746254,
+      "learning_rate": 8.366054490639978e-05,
+      "loss": 1.4017,
+      "step": 7848
+    },
+    {
+      "epoch": 0.5657752468824335,
+      "grad_norm": 0.12699408678340826,
+      "learning_rate": 8.363751196775172e-05,
+      "loss": 1.2958,
+      "step": 7849
+    },
+    {
+      "epoch": 0.5658473293447704,
+      "grad_norm": 0.10779304835872398,
+      "learning_rate": 8.361447992100564e-05,
+      "loss": 1.212,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5659194118071074,
+      "grad_norm": 0.11547273442851623,
+      "learning_rate": 8.359144876741698e-05,
+      "loss": 1.2532,
+      "step": 7851
+    },
+    {
+      "epoch": 0.5659914942694443,
+      "grad_norm": 0.11937850227556164,
+      "learning_rate": 8.356841850824115e-05,
+      "loss": 1.3001,
+      "step": 7852
+    },
+    {
+      "epoch": 0.5660635767317812,
+      "grad_norm": 0.11164162662665067,
+      "learning_rate": 8.354538914473351e-05,
+      "loss": 1.2233,
+      "step": 7853
+    },
+    {
+      "epoch": 0.5661356591941181,
+      "grad_norm": 0.11217566415820131,
+      "learning_rate": 8.35223606781494e-05,
+      "loss": 1.307,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5662077416564549,
+      "grad_norm": 0.10626847363313392,
+      "learning_rate": 8.349933310974402e-05,
+      "loss": 1.2353,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5662798241187919,
+      "grad_norm": 0.10711795163045096,
+      "learning_rate": 8.347630644077264e-05,
+      "loss": 1.2313,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5663519065811288,
+      "grad_norm": 0.12078786736684174,
+      "learning_rate": 8.345328067249032e-05,
+      "loss": 1.2837,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5664239890434657,
+      "grad_norm": 0.12090067589027641,
+      "learning_rate": 8.34302558061523e-05,
+      "loss": 1.3535,
+      "step": 7858
+    },
+    {
+      "epoch": 0.5664960715058026,
+      "grad_norm": 0.13544686540463544,
+      "learning_rate": 8.340723184301353e-05,
+      "loss": 1.2282,
+      "step": 7859
+    },
+    {
+      "epoch": 0.5665681539681395,
+      "grad_norm": 0.12117651547123795,
+      "learning_rate": 8.338420878432911e-05,
+      "loss": 1.2624,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5666402364304765,
+      "grad_norm": 0.11161742280796629,
+      "learning_rate": 8.336118663135389e-05,
+      "loss": 1.35,
+      "step": 7861
+    },
+    {
+      "epoch": 0.5667123188928134,
+      "grad_norm": 0.11801650598123678,
+      "learning_rate": 8.333816538534292e-05,
+      "loss": 1.4141,
+      "step": 7862
+    },
+    {
+      "epoch": 0.5667844013551503,
+      "grad_norm": 0.12419837587168223,
+      "learning_rate": 8.331514504755098e-05,
+      "loss": 1.2788,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5668564838174872,
+      "grad_norm": 0.12840799015352808,
+      "learning_rate": 8.329212561923291e-05,
+      "loss": 1.3177,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5669285662798241,
+      "grad_norm": 0.13108993326863147,
+      "learning_rate": 8.326910710164342e-05,
+      "loss": 1.3347,
+      "step": 7865
+    },
+    {
+      "epoch": 0.567000648742161,
+      "grad_norm": 0.20098436484711057,
+      "learning_rate": 8.324608949603733e-05,
+      "loss": 1.2254,
+      "step": 7866
+    },
+    {
+      "epoch": 0.567072731204498,
+      "grad_norm": 0.11225041625696,
+      "learning_rate": 8.322307280366922e-05,
+      "loss": 1.2946,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5671448136668349,
+      "grad_norm": 0.11363691385659397,
+      "learning_rate": 8.320005702579377e-05,
+      "loss": 1.3115,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5672168961291718,
+      "grad_norm": 0.11705928605181233,
+      "learning_rate": 8.317704216366547e-05,
+      "loss": 1.2637,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5672889785915087,
+      "grad_norm": 0.12345977569451567,
+      "learning_rate": 8.315402821853886e-05,
+      "loss": 1.2671,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5673610610538457,
+      "grad_norm": 0.1281017967258741,
+      "learning_rate": 8.313101519166844e-05,
+      "loss": 1.2671,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5674331435161825,
+      "grad_norm": 0.12332187140444013,
+      "learning_rate": 8.310800308430861e-05,
+      "loss": 1.2757,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5675052259785194,
+      "grad_norm": 0.114435049881203,
+      "learning_rate": 8.308499189771374e-05,
+      "loss": 1.2532,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5675773084408563,
+      "grad_norm": 0.12120295138639321,
+      "learning_rate": 8.30619816331381e-05,
+      "loss": 1.2414,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5676493909031932,
+      "grad_norm": 0.10776915859303204,
+      "learning_rate": 8.303897229183602e-05,
+      "loss": 1.2819,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5677214733655301,
+      "grad_norm": 0.11916917039215041,
+      "learning_rate": 8.30159638750617e-05,
+      "loss": 1.2316,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5677935558278671,
+      "grad_norm": 0.10896809767116387,
+      "learning_rate": 8.299295638406929e-05,
+      "loss": 1.3829,
+      "step": 7877
+    },
+    {
+      "epoch": 0.567865638290204,
+      "grad_norm": 0.10476395334303874,
+      "learning_rate": 8.296994982011291e-05,
+      "loss": 1.2757,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5679377207525409,
+      "grad_norm": 0.12801755722674263,
+      "learning_rate": 8.294694418444657e-05,
+      "loss": 1.1962,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5680098032148778,
+      "grad_norm": 0.11464355234852974,
+      "learning_rate": 8.292393947832439e-05,
+      "loss": 1.2563,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5680818856772147,
+      "grad_norm": 0.10424571390120128,
+      "learning_rate": 8.290093570300025e-05,
+      "loss": 1.1406,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5681539681395517,
+      "grad_norm": 0.11273232261838878,
+      "learning_rate": 8.287793285972812e-05,
+      "loss": 1.2996,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5682260506018886,
+      "grad_norm": 0.12153431506380263,
+      "learning_rate": 8.285493094976174e-05,
+      "loss": 1.2772,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5682981330642255,
+      "grad_norm": 0.10984699611159772,
+      "learning_rate": 8.28319299743551e-05,
+      "loss": 1.2351,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5683702155265624,
+      "grad_norm": 0.128896903614762,
+      "learning_rate": 8.280892993476183e-05,
+      "loss": 1.2044,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5684422979888993,
+      "grad_norm": 0.1122499865052793,
+      "learning_rate": 8.278593083223569e-05,
+      "loss": 1.3758,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5685143804512363,
+      "grad_norm": 0.10896754548315367,
+      "learning_rate": 8.276293266803027e-05,
+      "loss": 1.2559,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5685864629135732,
+      "grad_norm": 0.11768249924607807,
+      "learning_rate": 8.273993544339929e-05,
+      "loss": 1.221,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5686585453759101,
+      "grad_norm": 0.12291716808670619,
+      "learning_rate": 8.271693915959621e-05,
+      "loss": 1.3469,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5687306278382469,
+      "grad_norm": 0.11191387580292408,
+      "learning_rate": 8.269394381787459e-05,
+      "loss": 1.3603,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5688027103005838,
+      "grad_norm": 0.10741947570954499,
+      "learning_rate": 8.267094941948783e-05,
+      "loss": 1.1186,
+      "step": 7891
+    },
+    {
+      "epoch": 0.5688747927629207,
+      "grad_norm": 0.12267992483087682,
+      "learning_rate": 8.264795596568935e-05,
+      "loss": 1.3274,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5689468752252577,
+      "grad_norm": 0.1277189814087623,
+      "learning_rate": 8.262496345773251e-05,
+      "loss": 1.2599,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5690189576875946,
+      "grad_norm": 0.12159175665126791,
+      "learning_rate": 8.260197189687063e-05,
+      "loss": 1.2632,
+      "step": 7894
+    },
+    {
+      "epoch": 0.5690910401499315,
+      "grad_norm": 0.11783120720242335,
+      "learning_rate": 8.257898128435692e-05,
+      "loss": 1.2363,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5691631226122684,
+      "grad_norm": 0.11659813814505292,
+      "learning_rate": 8.255599162144456e-05,
+      "loss": 1.2761,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5692352050746053,
+      "grad_norm": 0.1106385468143504,
+      "learning_rate": 8.253300290938671e-05,
+      "loss": 1.2467,
+      "step": 7897
+    },
+    {
+      "epoch": 0.5693072875369423,
+      "grad_norm": 0.11864147377139826,
+      "learning_rate": 8.251001514943651e-05,
+      "loss": 1.2324,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5693793699992792,
+      "grad_norm": 0.11176705660777546,
+      "learning_rate": 8.248702834284693e-05,
+      "loss": 1.3101,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5694514524616161,
+      "grad_norm": 0.12549151897533378,
+      "learning_rate": 8.246404249087097e-05,
+      "loss": 1.3215,
+      "step": 7900
+    },
+    {
+      "epoch": 0.569523534923953,
+      "grad_norm": 0.12353094228303878,
+      "learning_rate": 8.244105759476157e-05,
+      "loss": 1.3269,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5695956173862899,
+      "grad_norm": 0.11105543437909783,
+      "learning_rate": 8.241807365577165e-05,
+      "loss": 1.2739,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5696676998486269,
+      "grad_norm": 0.11430309512222303,
+      "learning_rate": 8.239509067515399e-05,
+      "loss": 1.2816,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5697397823109638,
+      "grad_norm": 0.1224597013379216,
+      "learning_rate": 8.237210865416138e-05,
+      "loss": 1.2878,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5698118647733007,
+      "grad_norm": 0.11355630333128709,
+      "learning_rate": 8.234912759404656e-05,
+      "loss": 1.2271,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5698839472356376,
+      "grad_norm": 0.10627510252164701,
+      "learning_rate": 8.23261474960622e-05,
+      "loss": 1.311,
+      "step": 7906
+    },
+    {
+      "epoch": 0.5699560296979744,
+      "grad_norm": 0.11431275341781291,
+      "learning_rate": 8.230316836146091e-05,
+      "loss": 1.3325,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5700281121603114,
+      "grad_norm": 0.11676319285245135,
+      "learning_rate": 8.228019019149526e-05,
+      "loss": 1.3102,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5701001946226483,
+      "grad_norm": 0.10850935736234099,
+      "learning_rate": 8.225721298741776e-05,
+      "loss": 1.2268,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5701722770849852,
+      "grad_norm": 0.11977554049419914,
+      "learning_rate": 8.223423675048091e-05,
+      "loss": 1.2986,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5702443595473221,
+      "grad_norm": 0.11361830711930457,
+      "learning_rate": 8.22112614819371e-05,
+      "loss": 1.2612,
+      "step": 7911
+    },
+    {
+      "epoch": 0.570316442009659,
+      "grad_norm": 0.11151860475865567,
+      "learning_rate": 8.218828718303865e-05,
+      "loss": 1.4819,
+      "step": 7912
+    },
+    {
+      "epoch": 0.570388524471996,
+      "grad_norm": 0.12434413320811243,
+      "learning_rate": 8.216531385503794e-05,
+      "loss": 1.2719,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5704606069343329,
+      "grad_norm": 0.119866726513284,
+      "learning_rate": 8.214234149918714e-05,
+      "loss": 1.3824,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5705326893966698,
+      "grad_norm": 0.12701663593691057,
+      "learning_rate": 8.211937011673849e-05,
+      "loss": 1.3146,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5706047718590067,
+      "grad_norm": 0.10774148726444437,
+      "learning_rate": 8.209639970894413e-05,
+      "loss": 1.307,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5706768543213436,
+      "grad_norm": 0.1180322274624751,
+      "learning_rate": 8.207343027705621e-05,
+      "loss": 1.3306,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5707489367836805,
+      "grad_norm": 0.11262257484165712,
+      "learning_rate": 8.205046182232664e-05,
+      "loss": 1.3184,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5708210192460175,
+      "grad_norm": 0.17354605788161462,
+      "learning_rate": 8.202749434600752e-05,
+      "loss": 1.2832,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5708931017083544,
+      "grad_norm": 0.12442486324088596,
+      "learning_rate": 8.200452784935073e-05,
+      "loss": 1.2604,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5709651841706913,
+      "grad_norm": 0.11684738514518077,
+      "learning_rate": 8.19815623336082e-05,
+      "loss": 1.3126,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5710372666330282,
+      "grad_norm": 0.11001632148617792,
+      "learning_rate": 8.19585978000317e-05,
+      "loss": 1.2756,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5711093490953651,
+      "grad_norm": 0.12271330381359685,
+      "learning_rate": 8.193563424987298e-05,
+      "loss": 1.2353,
+      "step": 7923
+    },
+    {
+      "epoch": 0.571181431557702,
+      "grad_norm": 0.11795336040181928,
+      "learning_rate": 8.191267168438385e-05,
+      "loss": 1.2367,
+      "step": 7924
+    },
+    {
+      "epoch": 0.5712535140200389,
+      "grad_norm": 0.1166628668275816,
+      "learning_rate": 8.188971010481593e-05,
+      "loss": 1.2348,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5713255964823758,
+      "grad_norm": 0.12033812484690078,
+      "learning_rate": 8.186674951242082e-05,
+      "loss": 1.2389,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5713976789447127,
+      "grad_norm": 0.11336522873855977,
+      "learning_rate": 8.184378990845005e-05,
+      "loss": 1.2591,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5714697614070496,
+      "grad_norm": 0.11522328545729216,
+      "learning_rate": 8.18208312941552e-05,
+      "loss": 1.3261,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5715418438693866,
+      "grad_norm": 0.1228429429728614,
+      "learning_rate": 8.17978736707877e-05,
+      "loss": 1.3312,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5716139263317235,
+      "grad_norm": 0.11014402682731464,
+      "learning_rate": 8.177491703959889e-05,
+      "loss": 1.2831,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5716860087940604,
+      "grad_norm": 0.11748789743329581,
+      "learning_rate": 8.175196140184012e-05,
+      "loss": 1.2554,
+      "step": 7931
+    },
+    {
+      "epoch": 0.5717580912563973,
+      "grad_norm": 0.10806889685507699,
+      "learning_rate": 8.172900675876276e-05,
+      "loss": 1.2686,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5718301737187342,
+      "grad_norm": 0.10437674906221994,
+      "learning_rate": 8.170605311161798e-05,
+      "loss": 1.3179,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5719022561810712,
+      "grad_norm": 0.112674194206267,
+      "learning_rate": 8.168310046165695e-05,
+      "loss": 1.1828,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5719743386434081,
+      "grad_norm": 0.10776507002912171,
+      "learning_rate": 8.166014881013084e-05,
+      "loss": 1.2922,
+      "step": 7935
+    },
+    {
+      "epoch": 0.572046421105745,
+      "grad_norm": 0.11925595041580803,
+      "learning_rate": 8.163719815829063e-05,
+      "loss": 1.2809,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5721185035680819,
+      "grad_norm": 0.11406885708561283,
+      "learning_rate": 8.161424850738746e-05,
+      "loss": 1.2148,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5721905860304188,
+      "grad_norm": 0.1342184234719843,
+      "learning_rate": 8.15912998586722e-05,
+      "loss": 1.234,
+      "step": 7938
+    },
+    {
+      "epoch": 0.5722626684927558,
+      "grad_norm": 0.11682451339543563,
+      "learning_rate": 8.156835221339582e-05,
+      "loss": 1.3619,
+      "step": 7939
+    },
+    {
+      "epoch": 0.5723347509550927,
+      "grad_norm": 0.10575639128054067,
+      "learning_rate": 8.154540557280908e-05,
+      "loss": 1.1491,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5724068334174296,
+      "grad_norm": 0.1254548811075419,
+      "learning_rate": 8.15224599381629e-05,
+      "loss": 1.2955,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5724789158797664,
+      "grad_norm": 0.1107395804291808,
+      "learning_rate": 8.149951531070792e-05,
+      "loss": 1.2488,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5725509983421033,
+      "grad_norm": 0.11204860325972094,
+      "learning_rate": 8.14765716916949e-05,
+      "loss": 1.2268,
+      "step": 7943
+    },
+    {
+      "epoch": 0.5726230808044402,
+      "grad_norm": 0.11956628494026894,
+      "learning_rate": 8.145362908237442e-05,
+      "loss": 1.2855,
+      "step": 7944
+    },
+    {
+      "epoch": 0.5726951632667772,
+      "grad_norm": 0.13016512543022915,
+      "learning_rate": 8.143068748399704e-05,
+      "loss": 1.2209,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5727672457291141,
+      "grad_norm": 0.10703940439128815,
+      "learning_rate": 8.140774689781335e-05,
+      "loss": 1.2767,
+      "step": 7946
+    },
+    {
+      "epoch": 0.572839328191451,
+      "grad_norm": 0.1324704787091278,
+      "learning_rate": 8.13848073250738e-05,
+      "loss": 1.2315,
+      "step": 7947
+    },
+    {
+      "epoch": 0.5729114106537879,
+      "grad_norm": 0.14852422897718023,
+      "learning_rate": 8.136186876702876e-05,
+      "loss": 1.2337,
+      "step": 7948
+    },
+    {
+      "epoch": 0.5729834931161248,
+      "grad_norm": 0.13613629880094122,
+      "learning_rate": 8.13389312249286e-05,
+      "loss": 1.3282,
+      "step": 7949
+    },
+    {
+      "epoch": 0.5730555755784618,
+      "grad_norm": 0.1357776220968841,
+      "learning_rate": 8.131599470002367e-05,
+      "loss": 1.1814,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5731276580407987,
+      "grad_norm": 0.13022530257687137,
+      "learning_rate": 8.129305919356418e-05,
+      "loss": 1.1788,
+      "step": 7951
+    },
+    {
+      "epoch": 0.5731997405031356,
+      "grad_norm": 0.11250488121572517,
+      "learning_rate": 8.12701247068003e-05,
+      "loss": 1.3877,
+      "step": 7952
+    },
+    {
+      "epoch": 0.5732718229654725,
+      "grad_norm": 0.11611732242818565,
+      "learning_rate": 8.124719124098217e-05,
+      "loss": 1.2268,
+      "step": 7953
+    },
+    {
+      "epoch": 0.5733439054278094,
+      "grad_norm": 0.14091951711632056,
+      "learning_rate": 8.122425879735992e-05,
+      "loss": 1.28,
+      "step": 7954
+    },
+    {
+      "epoch": 0.5734159878901464,
+      "grad_norm": 0.12373713628553008,
+      "learning_rate": 8.120132737718356e-05,
+      "loss": 1.463,
+      "step": 7955
+    },
+    {
+      "epoch": 0.5734880703524833,
+      "grad_norm": 0.12282683327389794,
+      "learning_rate": 8.117839698170299e-05,
+      "loss": 1.2978,
+      "step": 7956
+    },
+    {
+      "epoch": 0.5735601528148202,
+      "grad_norm": 0.11299670609556234,
+      "learning_rate": 8.115546761216822e-05,
+      "loss": 1.2908,
+      "step": 7957
+    },
+    {
+      "epoch": 0.5736322352771571,
+      "grad_norm": 0.109968670289948,
+      "learning_rate": 8.113253926982896e-05,
+      "loss": 1.3272,
+      "step": 7958
+    },
+    {
+      "epoch": 0.5737043177394939,
+      "grad_norm": 0.13205566151136153,
+      "learning_rate": 8.110961195593517e-05,
+      "loss": 1.2711,
+      "step": 7959
+    },
+    {
+      "epoch": 0.5737764002018308,
+      "grad_norm": 0.1068918277004998,
+      "learning_rate": 8.108668567173651e-05,
+      "loss": 1.33,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5738484826641678,
+      "grad_norm": 0.16050231839791881,
+      "learning_rate": 8.106376041848272e-05,
+      "loss": 1.1868,
+      "step": 7961
+    },
+    {
+      "epoch": 0.5739205651265047,
+      "grad_norm": 0.11504744773462171,
+      "learning_rate": 8.104083619742331e-05,
+      "loss": 1.3232,
+      "step": 7962
+    },
+    {
+      "epoch": 0.5739926475888416,
+      "grad_norm": 0.1222730614761369,
+      "learning_rate": 8.101791300980802e-05,
+      "loss": 1.245,
+      "step": 7963
+    },
+    {
+      "epoch": 0.5740647300511785,
+      "grad_norm": 0.13188759513195827,
+      "learning_rate": 8.099499085688626e-05,
+      "loss": 1.3607,
+      "step": 7964
+    },
+    {
+      "epoch": 0.5741368125135154,
+      "grad_norm": 0.12042158516103309,
+      "learning_rate": 8.097206973990756e-05,
+      "loss": 1.3852,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5742088949758524,
+      "grad_norm": 0.1371036198588017,
+      "learning_rate": 8.094914966012124e-05,
+      "loss": 1.1946,
+      "step": 7966
+    },
+    {
+      "epoch": 0.5742809774381893,
+      "grad_norm": 0.15412436339301186,
+      "learning_rate": 8.092623061877668e-05,
+      "loss": 1.3001,
+      "step": 7967
+    },
+    {
+      "epoch": 0.5743530599005262,
+      "grad_norm": 0.12375648785973632,
+      "learning_rate": 8.090331261712322e-05,
+      "loss": 1.3412,
+      "step": 7968
+    },
+    {
+      "epoch": 0.5744251423628631,
+      "grad_norm": 0.1269498065532288,
+      "learning_rate": 8.08803956564101e-05,
+      "loss": 1.2872,
+      "step": 7969
+    },
+    {
+      "epoch": 0.5744972248252,
+      "grad_norm": 0.11778723641838382,
+      "learning_rate": 8.085747973788644e-05,
+      "loss": 1.2337,
+      "step": 7970
+    },
+    {
+      "epoch": 0.574569307287537,
+      "grad_norm": 0.12807355537584306,
+      "learning_rate": 8.083456486280136e-05,
+      "loss": 1.3447,
+      "step": 7971
+    },
+    {
+      "epoch": 0.5746413897498739,
+      "grad_norm": 0.09542640892955062,
+      "learning_rate": 8.081165103240397e-05,
+      "loss": 1.2521,
+      "step": 7972
+    },
+    {
+      "epoch": 0.5747134722122108,
+      "grad_norm": 0.21407316515659613,
+      "learning_rate": 8.078873824794329e-05,
+      "loss": 1.3221,
+      "step": 7973
+    },
+    {
+      "epoch": 0.5747855546745477,
+      "grad_norm": 0.10823292998526539,
+      "learning_rate": 8.076582651066822e-05,
+      "loss": 1.3129,
+      "step": 7974
+    },
+    {
+      "epoch": 0.5748576371368846,
+      "grad_norm": 0.11661933838736359,
+      "learning_rate": 8.074291582182766e-05,
+      "loss": 1.331,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5749297195992215,
+      "grad_norm": 0.12099682690931601,
+      "learning_rate": 8.072000618267048e-05,
+      "loss": 1.2228,
+      "step": 7976
+    },
+    {
+      "epoch": 0.5750018020615584,
+      "grad_norm": 0.10690054557929025,
+      "learning_rate": 8.069709759444548e-05,
+      "loss": 1.3038,
+      "step": 7977
+    },
+    {
+      "epoch": 0.5750738845238953,
+      "grad_norm": 0.11782946375897231,
+      "learning_rate": 8.067419005840132e-05,
+      "loss": 1.2245,
+      "step": 7978
+    },
+    {
+      "epoch": 0.5751459669862322,
+      "grad_norm": 0.1097296259233567,
+      "learning_rate": 8.065128357578671e-05,
+      "loss": 1.2656,
+      "step": 7979
+    },
+    {
+      "epoch": 0.5752180494485691,
+      "grad_norm": 0.11300023962256066,
+      "learning_rate": 8.06283781478502e-05,
+      "loss": 1.2745,
+      "step": 7980
+    },
+    {
+      "epoch": 0.575290131910906,
+      "grad_norm": 0.138880257927294,
+      "learning_rate": 8.060547377584045e-05,
+      "loss": 1.3212,
+      "step": 7981
+    },
+    {
+      "epoch": 0.575362214373243,
+      "grad_norm": 0.11635648094183254,
+      "learning_rate": 8.058257046100585e-05,
+      "loss": 1.3309,
+      "step": 7982
+    },
+    {
+      "epoch": 0.5754342968355799,
+      "grad_norm": 0.11651574527497177,
+      "learning_rate": 8.055966820459487e-05,
+      "loss": 1.2676,
+      "step": 7983
+    },
+    {
+      "epoch": 0.5755063792979168,
+      "grad_norm": 0.12417816269634668,
+      "learning_rate": 8.053676700785589e-05,
+      "loss": 1.2053,
+      "step": 7984
+    },
+    {
+      "epoch": 0.5755784617602537,
+      "grad_norm": 0.11150741102838829,
+      "learning_rate": 8.051386687203725e-05,
+      "loss": 1.3073,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5756505442225907,
+      "grad_norm": 0.11653909059062265,
+      "learning_rate": 8.049096779838719e-05,
+      "loss": 1.1904,
+      "step": 7986
+    },
+    {
+      "epoch": 0.5757226266849276,
+      "grad_norm": 0.12948503452501295,
+      "learning_rate": 8.046806978815391e-05,
+      "loss": 1.2721,
+      "step": 7987
+    },
+    {
+      "epoch": 0.5757947091472645,
+      "grad_norm": 0.11739981953511579,
+      "learning_rate": 8.044517284258559e-05,
+      "loss": 1.4566,
+      "step": 7988
+    },
+    {
+      "epoch": 0.5758667916096014,
+      "grad_norm": 0.11941218234150601,
+      "learning_rate": 8.042227696293023e-05,
+      "loss": 1.3052,
+      "step": 7989
+    },
+    {
+      "epoch": 0.5759388740719383,
+      "grad_norm": 0.11474828035855456,
+      "learning_rate": 8.039938215043597e-05,
+      "loss": 1.2744,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5760109565342753,
+      "grad_norm": 0.11402127347331098,
+      "learning_rate": 8.037648840635073e-05,
+      "loss": 1.242,
+      "step": 7991
+    },
+    {
+      "epoch": 0.5760830389966122,
+      "grad_norm": 0.11592402004569248,
+      "learning_rate": 8.035359573192243e-05,
+      "loss": 1.2955,
+      "step": 7992
+    },
+    {
+      "epoch": 0.576155121458949,
+      "grad_norm": 0.12296716383651735,
+      "learning_rate": 8.033070412839891e-05,
+      "loss": 1.225,
+      "step": 7993
+    },
+    {
+      "epoch": 0.5762272039212859,
+      "grad_norm": 0.11392854356428579,
+      "learning_rate": 8.0307813597028e-05,
+      "loss": 1.2728,
+      "step": 7994
+    },
+    {
+      "epoch": 0.5762992863836228,
+      "grad_norm": 0.11565583657561937,
+      "learning_rate": 8.02849241390574e-05,
+      "loss": 1.1667,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5763713688459597,
+      "grad_norm": 0.12884580169659463,
+      "learning_rate": 8.026203575573484e-05,
+      "loss": 1.2231,
+      "step": 7996
+    },
+    {
+      "epoch": 0.5764434513082967,
+      "grad_norm": 0.11721656798081455,
+      "learning_rate": 8.023914844830789e-05,
+      "loss": 1.365,
+      "step": 7997
+    },
+    {
+      "epoch": 0.5765155337706336,
+      "grad_norm": 0.12496423128690326,
+      "learning_rate": 8.021626221802411e-05,
+      "loss": 1.1787,
+      "step": 7998
+    },
+    {
+      "epoch": 0.5765876162329705,
+      "grad_norm": 0.11537902167415184,
+      "learning_rate": 8.019337706613107e-05,
+      "loss": 1.3344,
+      "step": 7999
+    },
+    {
+      "epoch": 0.5766596986953074,
+      "grad_norm": 0.11896317250911322,
+      "learning_rate": 8.017049299387618e-05,
+      "loss": 1.3008,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5767317811576443,
+      "grad_norm": 0.12188812683942343,
+      "learning_rate": 8.014761000250682e-05,
+      "loss": 1.2139,
+      "step": 8001
+    },
+    {
+      "epoch": 0.5768038636199813,
+      "grad_norm": 0.12202470251979898,
+      "learning_rate": 8.012472809327028e-05,
+      "loss": 1.3641,
+      "step": 8002
+    },
+    {
+      "epoch": 0.5768759460823182,
+      "grad_norm": 0.1077095231029038,
+      "learning_rate": 8.010184726741387e-05,
+      "loss": 1.3053,
+      "step": 8003
+    },
+    {
+      "epoch": 0.5769480285446551,
+      "grad_norm": 0.11106095442592535,
+      "learning_rate": 8.007896752618486e-05,
+      "loss": 1.2727,
+      "step": 8004
+    },
+    {
+      "epoch": 0.577020111006992,
+      "grad_norm": 0.11370679787141239,
+      "learning_rate": 8.00560888708303e-05,
+      "loss": 1.2733,
+      "step": 8005
+    },
+    {
+      "epoch": 0.5770921934693289,
+      "grad_norm": 0.12412539059232855,
+      "learning_rate": 8.003321130259728e-05,
+      "loss": 1.2925,
+      "step": 8006
+    },
+    {
+      "epoch": 0.5771642759316659,
+      "grad_norm": 0.12911853239842846,
+      "learning_rate": 8.001033482273292e-05,
+      "loss": 1.2522,
+      "step": 8007
+    },
+    {
+      "epoch": 0.5772363583940028,
+      "grad_norm": 0.10634675444534222,
+      "learning_rate": 7.998745943248415e-05,
+      "loss": 1.1872,
+      "step": 8008
+    },
+    {
+      "epoch": 0.5773084408563397,
+      "grad_norm": 0.11236314544798152,
+      "learning_rate": 7.996458513309787e-05,
+      "loss": 1.2534,
+      "step": 8009
+    },
+    {
+      "epoch": 0.5773805233186766,
+      "grad_norm": 0.12146503956901264,
+      "learning_rate": 7.994171192582096e-05,
+      "loss": 1.35,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5774526057810134,
+      "grad_norm": 0.11508902976034305,
+      "learning_rate": 7.991883981190012e-05,
+      "loss": 1.275,
+      "step": 8011
+    },
+    {
+      "epoch": 0.5775246882433503,
+      "grad_norm": 0.11479582652076196,
+      "learning_rate": 7.989596879258224e-05,
+      "loss": 1.2307,
+      "step": 8012
+    },
+    {
+      "epoch": 0.5775967707056873,
+      "grad_norm": 0.134838497433226,
+      "learning_rate": 7.987309886911389e-05,
+      "loss": 1.2741,
+      "step": 8013
+    },
+    {
+      "epoch": 0.5776688531680242,
+      "grad_norm": 0.1082614087318794,
+      "learning_rate": 7.985023004274174e-05,
+      "loss": 1.3514,
+      "step": 8014
+    },
+    {
+      "epoch": 0.5777409356303611,
+      "grad_norm": 0.12302116897230331,
+      "learning_rate": 7.982736231471224e-05,
+      "loss": 1.257,
+      "step": 8015
+    },
+    {
+      "epoch": 0.577813018092698,
+      "grad_norm": 0.11230579323509392,
+      "learning_rate": 7.980449568627204e-05,
+      "loss": 1.3005,
+      "step": 8016
+    },
+    {
+      "epoch": 0.577885100555035,
+      "grad_norm": 0.12048862748998594,
+      "learning_rate": 7.978163015866748e-05,
+      "loss": 1.2147,
+      "step": 8017
+    },
+    {
+      "epoch": 0.5779571830173719,
+      "grad_norm": 0.11777535896194916,
+      "learning_rate": 7.975876573314496e-05,
+      "loss": 1.29,
+      "step": 8018
+    },
+    {
+      "epoch": 0.5780292654797088,
+      "grad_norm": 0.11705954049804078,
+      "learning_rate": 7.973590241095078e-05,
+      "loss": 1.2919,
+      "step": 8019
+    },
+    {
+      "epoch": 0.5781013479420457,
+      "grad_norm": 0.11728676313731322,
+      "learning_rate": 7.97130401933312e-05,
+      "loss": 1.3242,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5781734304043826,
+      "grad_norm": 0.10986126123906938,
+      "learning_rate": 7.969017908153242e-05,
+      "loss": 1.3288,
+      "step": 8021
+    },
+    {
+      "epoch": 0.5782455128667195,
+      "grad_norm": 0.11908808340459621,
+      "learning_rate": 7.96673190768006e-05,
+      "loss": 1.2984,
+      "step": 8022
+    },
+    {
+      "epoch": 0.5783175953290565,
+      "grad_norm": 0.11926241666890783,
+      "learning_rate": 7.964446018038177e-05,
+      "loss": 1.2876,
+      "step": 8023
+    },
+    {
+      "epoch": 0.5783896777913934,
+      "grad_norm": 0.11096181722939069,
+      "learning_rate": 7.962160239352195e-05,
+      "loss": 1.2494,
+      "step": 8024
+    },
+    {
+      "epoch": 0.5784617602537303,
+      "grad_norm": 0.11976380843315242,
+      "learning_rate": 7.959874571746713e-05,
+      "loss": 1.3853,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5785338427160672,
+      "grad_norm": 0.19675504401861418,
+      "learning_rate": 7.957589015346321e-05,
+      "loss": 1.3125,
+      "step": 8026
+    },
+    {
+      "epoch": 0.5786059251784041,
+      "grad_norm": 0.14193316214807422,
+      "learning_rate": 7.955303570275598e-05,
+      "loss": 1.3489,
+      "step": 8027
+    },
+    {
+      "epoch": 0.578678007640741,
+      "grad_norm": 0.13831111467490526,
+      "learning_rate": 7.95301823665912e-05,
+      "loss": 1.1804,
+      "step": 8028
+    },
+    {
+      "epoch": 0.5787500901030779,
+      "grad_norm": 0.10928416734915548,
+      "learning_rate": 7.95073301462146e-05,
+      "loss": 1.3044,
+      "step": 8029
+    },
+    {
+      "epoch": 0.5788221725654148,
+      "grad_norm": 0.1185770549394011,
+      "learning_rate": 7.948447904287191e-05,
+      "loss": 1.2617,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5788942550277517,
+      "grad_norm": 0.11788686375742115,
+      "learning_rate": 7.94616290578086e-05,
+      "loss": 1.3316,
+      "step": 8031
+    },
+    {
+      "epoch": 0.5789663374900886,
+      "grad_norm": 0.1244348927854137,
+      "learning_rate": 7.943878019227028e-05,
+      "loss": 1.2657,
+      "step": 8032
+    },
+    {
+      "epoch": 0.5790384199524256,
+      "grad_norm": 0.11348971599838323,
+      "learning_rate": 7.941593244750232e-05,
+      "loss": 1.3518,
+      "step": 8033
+    },
+    {
+      "epoch": 0.5791105024147625,
+      "grad_norm": 0.11956841293270119,
+      "learning_rate": 7.939308582475025e-05,
+      "loss": 1.3297,
+      "step": 8034
+    },
+    {
+      "epoch": 0.5791825848770994,
+      "grad_norm": 0.12094542847786643,
+      "learning_rate": 7.937024032525933e-05,
+      "loss": 1.2817,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5792546673394363,
+      "grad_norm": 0.11216752464143817,
+      "learning_rate": 7.934739595027491e-05,
+      "loss": 1.28,
+      "step": 8036
+    },
+    {
+      "epoch": 0.5793267498017732,
+      "grad_norm": 0.11652637583087373,
+      "learning_rate": 7.93245527010421e-05,
+      "loss": 1.1589,
+      "step": 8037
+    },
+    {
+      "epoch": 0.5793988322641102,
+      "grad_norm": 0.11172472506476767,
+      "learning_rate": 7.93017105788062e-05,
+      "loss": 1.256,
+      "step": 8038
+    },
+    {
+      "epoch": 0.5794709147264471,
+      "grad_norm": 0.11903926088259635,
+      "learning_rate": 7.927886958481223e-05,
+      "loss": 1.268,
+      "step": 8039
+    },
+    {
+      "epoch": 0.579542997188784,
+      "grad_norm": 0.11120168714924947,
+      "learning_rate": 7.925602972030525e-05,
+      "loss": 1.2144,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5796150796511209,
+      "grad_norm": 0.12940344725418157,
+      "learning_rate": 7.923319098653024e-05,
+      "loss": 1.3529,
+      "step": 8041
+    },
+    {
+      "epoch": 0.5796871621134578,
+      "grad_norm": 0.13425479818461536,
+      "learning_rate": 7.921035338473206e-05,
+      "loss": 1.1584,
+      "step": 8042
+    },
+    {
+      "epoch": 0.5797592445757948,
+      "grad_norm": 0.1128464389463256,
+      "learning_rate": 7.918751691615563e-05,
+      "loss": 1.185,
+      "step": 8043
+    },
+    {
+      "epoch": 0.5798313270381317,
+      "grad_norm": 0.10850223291723866,
+      "learning_rate": 7.916468158204576e-05,
+      "loss": 1.3771,
+      "step": 8044
+    },
+    {
+      "epoch": 0.5799034095004685,
+      "grad_norm": 0.11893009605326106,
+      "learning_rate": 7.91418473836471e-05,
+      "loss": 1.3559,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5799754919628054,
+      "grad_norm": 0.11994342668316472,
+      "learning_rate": 7.911901432220435e-05,
+      "loss": 1.2966,
+      "step": 8046
+    },
+    {
+      "epoch": 0.5800475744251423,
+      "grad_norm": 0.11304880279545029,
+      "learning_rate": 7.909618239896216e-05,
+      "loss": 1.2103,
+      "step": 8047
+    },
+    {
+      "epoch": 0.5801196568874792,
+      "grad_norm": 0.11294831087699761,
+      "learning_rate": 7.907335161516505e-05,
+      "loss": 1.3807,
+      "step": 8048
+    },
+    {
+      "epoch": 0.5801917393498162,
+      "grad_norm": 0.11816983142603477,
+      "learning_rate": 7.905052197205747e-05,
+      "loss": 1.2455,
+      "step": 8049
+    },
+    {
+      "epoch": 0.5802638218121531,
+      "grad_norm": 0.11989243029149758,
+      "learning_rate": 7.902769347088384e-05,
+      "loss": 1.2753,
+      "step": 8050
+    },
+    {
+      "epoch": 0.58033590427449,
+      "grad_norm": 0.11317765375910525,
+      "learning_rate": 7.900486611288855e-05,
+      "loss": 1.359,
+      "step": 8051
+    },
+    {
+      "epoch": 0.5804079867368269,
+      "grad_norm": 0.23177593464029486,
+      "learning_rate": 7.898203989931592e-05,
+      "loss": 1.3108,
+      "step": 8052
+    },
+    {
+      "epoch": 0.5804800691991638,
+      "grad_norm": 0.11520981426751221,
+      "learning_rate": 7.895921483141014e-05,
+      "loss": 1.2456,
+      "step": 8053
+    },
+    {
+      "epoch": 0.5805521516615008,
+      "grad_norm": 0.1604150850604023,
+      "learning_rate": 7.893639091041536e-05,
+      "loss": 1.2608,
+      "step": 8054
+    },
+    {
+      "epoch": 0.5806242341238377,
+      "grad_norm": 0.12806836064326782,
+      "learning_rate": 7.89135681375757e-05,
+      "loss": 1.2501,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5806963165861746,
+      "grad_norm": 0.12449243630641665,
+      "learning_rate": 7.889074651413527e-05,
+      "loss": 1.2256,
+      "step": 8056
+    },
+    {
+      "epoch": 0.5807683990485115,
+      "grad_norm": 0.11518653019632051,
+      "learning_rate": 7.886792604133799e-05,
+      "loss": 1.213,
+      "step": 8057
+    },
+    {
+      "epoch": 0.5808404815108484,
+      "grad_norm": 0.12320086681134788,
+      "learning_rate": 7.884510672042778e-05,
+      "loss": 1.3996,
+      "step": 8058
+    },
+    {
+      "epoch": 0.5809125639731854,
+      "grad_norm": 0.1286663921366504,
+      "learning_rate": 7.882228855264849e-05,
+      "loss": 1.3608,
+      "step": 8059
+    },
+    {
+      "epoch": 0.5809846464355223,
+      "grad_norm": 0.11138295301720504,
+      "learning_rate": 7.879947153924399e-05,
+      "loss": 1.2818,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5810567288978592,
+      "grad_norm": 0.10919306911789674,
+      "learning_rate": 7.877665568145793e-05,
+      "loss": 1.3299,
+      "step": 8061
+    },
+    {
+      "epoch": 0.5811288113601961,
+      "grad_norm": 0.11826258862157021,
+      "learning_rate": 7.8753840980534e-05,
+      "loss": 1.1723,
+      "step": 8062
+    },
+    {
+      "epoch": 0.5812008938225329,
+      "grad_norm": 0.16047579769188502,
+      "learning_rate": 7.873102743771584e-05,
+      "loss": 1.2733,
+      "step": 8063
+    },
+    {
+      "epoch": 0.5812729762848698,
+      "grad_norm": 0.10744334108182373,
+      "learning_rate": 7.87082150542469e-05,
+      "loss": 1.2889,
+      "step": 8064
+    },
+    {
+      "epoch": 0.5813450587472068,
+      "grad_norm": 0.10921454288194223,
+      "learning_rate": 7.868540383137076e-05,
+      "loss": 1.2774,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5814171412095437,
+      "grad_norm": 0.11658252253661909,
+      "learning_rate": 7.866259377033081e-05,
+      "loss": 1.3177,
+      "step": 8066
+    },
+    {
+      "epoch": 0.5814892236718806,
+      "grad_norm": 0.12095660059262062,
+      "learning_rate": 7.863978487237039e-05,
+      "loss": 1.315,
+      "step": 8067
+    },
+    {
+      "epoch": 0.5815613061342175,
+      "grad_norm": 0.12062849596444772,
+      "learning_rate": 7.861697713873274e-05,
+      "loss": 1.2631,
+      "step": 8068
+    },
+    {
+      "epoch": 0.5816333885965544,
+      "grad_norm": 0.11413496894585574,
+      "learning_rate": 7.859417057066117e-05,
+      "loss": 1.2731,
+      "step": 8069
+    },
+    {
+      "epoch": 0.5817054710588914,
+      "grad_norm": 0.11173712530967622,
+      "learning_rate": 7.85713651693988e-05,
+      "loss": 1.2803,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5817775535212283,
+      "grad_norm": 0.11364011562958443,
+      "learning_rate": 7.854856093618877e-05,
+      "loss": 1.2612,
+      "step": 8071
+    },
+    {
+      "epoch": 0.5818496359835652,
+      "grad_norm": 0.11869279669174214,
+      "learning_rate": 7.852575787227401e-05,
+      "loss": 1.2967,
+      "step": 8072
+    },
+    {
+      "epoch": 0.5819217184459021,
+      "grad_norm": 0.10776654282215545,
+      "learning_rate": 7.85029559788976e-05,
+      "loss": 1.3521,
+      "step": 8073
+    },
+    {
+      "epoch": 0.581993800908239,
+      "grad_norm": 0.11038645518506293,
+      "learning_rate": 7.84801552573024e-05,
+      "loss": 1.2905,
+      "step": 8074
+    },
+    {
+      "epoch": 0.582065883370576,
+      "grad_norm": 0.1136496846188598,
+      "learning_rate": 7.845735570873128e-05,
+      "loss": 1.2231,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5821379658329129,
+      "grad_norm": 0.25030867533219076,
+      "learning_rate": 7.843455733442698e-05,
+      "loss": 1.2868,
+      "step": 8076
+    },
+    {
+      "epoch": 0.5822100482952498,
+      "grad_norm": 0.11504316886958504,
+      "learning_rate": 7.841176013563222e-05,
+      "loss": 1.2795,
+      "step": 8077
+    },
+    {
+      "epoch": 0.5822821307575867,
+      "grad_norm": 0.1092520646192449,
+      "learning_rate": 7.838896411358969e-05,
+      "loss": 1.3513,
+      "step": 8078
+    },
+    {
+      "epoch": 0.5823542132199236,
+      "grad_norm": 0.11029650541203806,
+      "learning_rate": 7.836616926954196e-05,
+      "loss": 1.304,
+      "step": 8079
+    },
+    {
+      "epoch": 0.5824262956822605,
+      "grad_norm": 0.11829364524184069,
+      "learning_rate": 7.834337560473154e-05,
+      "loss": 1.27,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5824983781445974,
+      "grad_norm": 0.10580314563867883,
+      "learning_rate": 7.832058312040087e-05,
+      "loss": 1.1642,
+      "step": 8081
+    },
+    {
+      "epoch": 0.5825704606069343,
+      "grad_norm": 0.11912900514879843,
+      "learning_rate": 7.829779181779238e-05,
+      "loss": 1.2172,
+      "step": 8082
+    },
+    {
+      "epoch": 0.5826425430692712,
+      "grad_norm": 0.10508651041588478,
+      "learning_rate": 7.827500169814842e-05,
+      "loss": 1.2529,
+      "step": 8083
+    },
+    {
+      "epoch": 0.5827146255316081,
+      "grad_norm": 0.11338510564902674,
+      "learning_rate": 7.825221276271121e-05,
+      "loss": 1.2916,
+      "step": 8084
+    },
+    {
+      "epoch": 0.582786707993945,
+      "grad_norm": 0.10533934609602215,
+      "learning_rate": 7.822942501272297e-05,
+      "loss": 1.3316,
+      "step": 8085
+    },
+    {
+      "epoch": 0.582858790456282,
+      "grad_norm": 0.10759023371305743,
+      "learning_rate": 7.820663844942579e-05,
+      "loss": 1.2642,
+      "step": 8086
+    },
+    {
+      "epoch": 0.5829308729186189,
+      "grad_norm": 0.12643432069630048,
+      "learning_rate": 7.818385307406184e-05,
+      "loss": 1.1618,
+      "step": 8087
+    },
+    {
+      "epoch": 0.5830029553809558,
+      "grad_norm": 0.11290068721053466,
+      "learning_rate": 7.816106888787304e-05,
+      "loss": 1.2094,
+      "step": 8088
+    },
+    {
+      "epoch": 0.5830750378432927,
+      "grad_norm": 0.1268342289325438,
+      "learning_rate": 7.813828589210138e-05,
+      "loss": 1.2057,
+      "step": 8089
+    },
+    {
+      "epoch": 0.5831471203056297,
+      "grad_norm": 0.11310195798788673,
+      "learning_rate": 7.811550408798867e-05,
+      "loss": 1.2075,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5832192027679666,
+      "grad_norm": 0.11811650976769705,
+      "learning_rate": 7.809272347677683e-05,
+      "loss": 1.2753,
+      "step": 8091
+    },
+    {
+      "epoch": 0.5832912852303035,
+      "grad_norm": 0.1319265359563061,
+      "learning_rate": 7.806994405970753e-05,
+      "loss": 1.2626,
+      "step": 8092
+    },
+    {
+      "epoch": 0.5833633676926404,
+      "grad_norm": 0.12380986823473622,
+      "learning_rate": 7.804716583802248e-05,
+      "loss": 1.2824,
+      "step": 8093
+    },
+    {
+      "epoch": 0.5834354501549773,
+      "grad_norm": 0.11980059755297144,
+      "learning_rate": 7.802438881296323e-05,
+      "loss": 1.1802,
+      "step": 8094
+    },
+    {
+      "epoch": 0.5835075326173143,
+      "grad_norm": 0.11903402742922432,
+      "learning_rate": 7.800161298577145e-05,
+      "loss": 1.2476,
+      "step": 8095
+    },
+    {
+      "epoch": 0.5835796150796512,
+      "grad_norm": 0.11031561801748441,
+      "learning_rate": 7.797883835768856e-05,
+      "loss": 1.3086,
+      "step": 8096
+    },
+    {
+      "epoch": 0.583651697541988,
+      "grad_norm": 0.10949723467021616,
+      "learning_rate": 7.7956064929956e-05,
+      "loss": 1.2654,
+      "step": 8097
+    },
+    {
+      "epoch": 0.5837237800043249,
+      "grad_norm": 0.14269654025860198,
+      "learning_rate": 7.793329270381508e-05,
+      "loss": 1.2262,
+      "step": 8098
+    },
+    {
+      "epoch": 0.5837958624666618,
+      "grad_norm": 0.11503289427501488,
+      "learning_rate": 7.79105216805071e-05,
+      "loss": 1.3493,
+      "step": 8099
+    },
+    {
+      "epoch": 0.5838679449289987,
+      "grad_norm": 0.1145067458892188,
+      "learning_rate": 7.788775186127335e-05,
+      "loss": 1.2317,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5839400273913357,
+      "grad_norm": 0.1220357052347663,
+      "learning_rate": 7.786498324735495e-05,
+      "loss": 1.3441,
+      "step": 8101
+    },
+    {
+      "epoch": 0.5840121098536726,
+      "grad_norm": 0.11269042525815204,
+      "learning_rate": 7.784221583999298e-05,
+      "loss": 1.3139,
+      "step": 8102
+    },
+    {
+      "epoch": 0.5840841923160095,
+      "grad_norm": 0.12086972756007196,
+      "learning_rate": 7.781944964042843e-05,
+      "loss": 1.1955,
+      "step": 8103
+    },
+    {
+      "epoch": 0.5841562747783464,
+      "grad_norm": 0.11191480155242295,
+      "learning_rate": 7.779668464990235e-05,
+      "loss": 1.2917,
+      "step": 8104
+    },
+    {
+      "epoch": 0.5842283572406833,
+      "grad_norm": 0.10626270022269609,
+      "learning_rate": 7.77739208696556e-05,
+      "loss": 1.2477,
+      "step": 8105
+    },
+    {
+      "epoch": 0.5843004397030203,
+      "grad_norm": 0.11355706946395248,
+      "learning_rate": 7.7751158300929e-05,
+      "loss": 1.2073,
+      "step": 8106
+    },
+    {
+      "epoch": 0.5843725221653572,
+      "grad_norm": 0.11700457155522836,
+      "learning_rate": 7.772839694496332e-05,
+      "loss": 1.3326,
+      "step": 8107
+    },
+    {
+      "epoch": 0.5844446046276941,
+      "grad_norm": 0.11065638153036891,
+      "learning_rate": 7.770563680299921e-05,
+      "loss": 1.2713,
+      "step": 8108
+    },
+    {
+      "epoch": 0.584516687090031,
+      "grad_norm": 0.10874547809355974,
+      "learning_rate": 7.768287787627739e-05,
+      "loss": 1.3337,
+      "step": 8109
+    },
+    {
+      "epoch": 0.5845887695523679,
+      "grad_norm": 0.1160830002714094,
+      "learning_rate": 7.766012016603836e-05,
+      "loss": 1.2361,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5846608520147049,
+      "grad_norm": 0.12723855699020867,
+      "learning_rate": 7.763736367352267e-05,
+      "loss": 1.2994,
+      "step": 8111
+    },
+    {
+      "epoch": 0.5847329344770418,
+      "grad_norm": 0.12112429375885916,
+      "learning_rate": 7.761460839997065e-05,
+      "loss": 1.3879,
+      "step": 8112
+    },
+    {
+      "epoch": 0.5848050169393787,
+      "grad_norm": 0.12302432800517356,
+      "learning_rate": 7.759185434662281e-05,
+      "loss": 1.3245,
+      "step": 8113
+    },
+    {
+      "epoch": 0.5848770994017156,
+      "grad_norm": 0.12071332793550699,
+      "learning_rate": 7.756910151471935e-05,
+      "loss": 1.2069,
+      "step": 8114
+    },
+    {
+      "epoch": 0.5849491818640524,
+      "grad_norm": 0.12885038795485418,
+      "learning_rate": 7.754634990550055e-05,
+      "loss": 1.1357,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5850212643263893,
+      "grad_norm": 0.11643404196578368,
+      "learning_rate": 7.752359952020649e-05,
+      "loss": 1.336,
+      "step": 8116
+    },
+    {
+      "epoch": 0.5850933467887263,
+      "grad_norm": 0.13298010913086705,
+      "learning_rate": 7.750085036007741e-05,
+      "loss": 1.2733,
+      "step": 8117
+    },
+    {
+      "epoch": 0.5851654292510632,
+      "grad_norm": 0.12544702546283895,
+      "learning_rate": 7.747810242635325e-05,
+      "loss": 1.2006,
+      "step": 8118
+    },
+    {
+      "epoch": 0.5852375117134001,
+      "grad_norm": 0.1129070547715939,
+      "learning_rate": 7.7455355720274e-05,
+      "loss": 1.3014,
+      "step": 8119
+    },
+    {
+      "epoch": 0.585309594175737,
+      "grad_norm": 0.11104093049145687,
+      "learning_rate": 7.743261024307954e-05,
+      "loss": 1.2988,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5853816766380739,
+      "grad_norm": 0.11027166806999117,
+      "learning_rate": 7.740986599600969e-05,
+      "loss": 1.2855,
+      "step": 8121
+    },
+    {
+      "epoch": 0.5854537591004109,
+      "grad_norm": 0.12463604354649721,
+      "learning_rate": 7.738712298030427e-05,
+      "loss": 1.2401,
+      "step": 8122
+    },
+    {
+      "epoch": 0.5855258415627478,
+      "grad_norm": 0.10960501599793777,
+      "learning_rate": 7.736438119720297e-05,
+      "loss": 1.3185,
+      "step": 8123
+    },
+    {
+      "epoch": 0.5855979240250847,
+      "grad_norm": 0.11603786274518435,
+      "learning_rate": 7.734164064794537e-05,
+      "loss": 1.2136,
+      "step": 8124
+    },
+    {
+      "epoch": 0.5856700064874216,
+      "grad_norm": 0.12624825518892582,
+      "learning_rate": 7.731890133377104e-05,
+      "loss": 1.2919,
+      "step": 8125
+    },
+    {
+      "epoch": 0.5857420889497585,
+      "grad_norm": 0.11616801584675995,
+      "learning_rate": 7.729616325591951e-05,
+      "loss": 1.4449,
+      "step": 8126
+    },
+    {
+      "epoch": 0.5858141714120955,
+      "grad_norm": 0.11826780183023962,
+      "learning_rate": 7.727342641563024e-05,
+      "loss": 1.3548,
+      "step": 8127
+    },
+    {
+      "epoch": 0.5858862538744324,
+      "grad_norm": 0.11248433601055564,
+      "learning_rate": 7.725069081414252e-05,
+      "loss": 1.3172,
+      "step": 8128
+    },
+    {
+      "epoch": 0.5859583363367693,
+      "grad_norm": 0.11891255510900452,
+      "learning_rate": 7.722795645269567e-05,
+      "loss": 1.237,
+      "step": 8129
+    },
+    {
+      "epoch": 0.5860304187991062,
+      "grad_norm": 0.1170669611333012,
+      "learning_rate": 7.720522333252889e-05,
+      "loss": 1.2835,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5861025012614431,
+      "grad_norm": 0.1150223093466644,
+      "learning_rate": 7.718249145488142e-05,
+      "loss": 1.3742,
+      "step": 8131
+    },
+    {
+      "epoch": 0.58617458372378,
+      "grad_norm": 0.12499638088331308,
+      "learning_rate": 7.715976082099226e-05,
+      "loss": 1.2725,
+      "step": 8132
+    },
+    {
+      "epoch": 0.5862466661861169,
+      "grad_norm": 0.10880210808548127,
+      "learning_rate": 7.71370314321005e-05,
+      "loss": 1.311,
+      "step": 8133
+    },
+    {
+      "epoch": 0.5863187486484538,
+      "grad_norm": 0.12219720747355751,
+      "learning_rate": 7.711430328944501e-05,
+      "loss": 1.277,
+      "step": 8134
+    },
+    {
+      "epoch": 0.5863908311107907,
+      "grad_norm": 0.11935679695743197,
+      "learning_rate": 7.709157639426478e-05,
+      "loss": 1.2796,
+      "step": 8135
+    },
+    {
+      "epoch": 0.5864629135731276,
+      "grad_norm": 0.11237873151896624,
+      "learning_rate": 7.706885074779859e-05,
+      "loss": 1.2493,
+      "step": 8136
+    },
+    {
+      "epoch": 0.5865349960354646,
+      "grad_norm": 0.15559684383530287,
+      "learning_rate": 7.704612635128515e-05,
+      "loss": 1.3074,
+      "step": 8137
+    },
+    {
+      "epoch": 0.5866070784978015,
+      "grad_norm": 0.11477097441581705,
+      "learning_rate": 7.702340320596316e-05,
+      "loss": 1.2784,
+      "step": 8138
+    },
+    {
+      "epoch": 0.5866791609601384,
+      "grad_norm": 0.13061519984328046,
+      "learning_rate": 7.70006813130713e-05,
+      "loss": 1.3053,
+      "step": 8139
+    },
+    {
+      "epoch": 0.5867512434224753,
+      "grad_norm": 0.13137796171847482,
+      "learning_rate": 7.697796067384804e-05,
+      "loss": 1.3031,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5868233258848122,
+      "grad_norm": 0.10226736191825829,
+      "learning_rate": 7.695524128953189e-05,
+      "loss": 1.1993,
+      "step": 8141
+    },
+    {
+      "epoch": 0.5868954083471492,
+      "grad_norm": 0.11096830306488265,
+      "learning_rate": 7.693252316136127e-05,
+      "loss": 1.2281,
+      "step": 8142
+    },
+    {
+      "epoch": 0.5869674908094861,
+      "grad_norm": 0.11767543683330241,
+      "learning_rate": 7.690980629057447e-05,
+      "loss": 1.4301,
+      "step": 8143
+    },
+    {
+      "epoch": 0.587039573271823,
+      "grad_norm": 0.11868704384517745,
+      "learning_rate": 7.68870906784098e-05,
+      "loss": 1.2171,
+      "step": 8144
+    },
+    {
+      "epoch": 0.5871116557341599,
+      "grad_norm": 0.11599922077169088,
+      "learning_rate": 7.686437632610548e-05,
+      "loss": 1.3062,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5871837381964968,
+      "grad_norm": 0.1292840410467575,
+      "learning_rate": 7.684166323489965e-05,
+      "loss": 1.3269,
+      "step": 8146
+    },
+    {
+      "epoch": 0.5872558206588337,
+      "grad_norm": 0.11020436572914179,
+      "learning_rate": 7.68189514060303e-05,
+      "loss": 1.3312,
+      "step": 8147
+    },
+    {
+      "epoch": 0.5873279031211707,
+      "grad_norm": 0.11115044691307178,
+      "learning_rate": 7.679624084073551e-05,
+      "loss": 1.2653,
+      "step": 8148
+    },
+    {
+      "epoch": 0.5873999855835075,
+      "grad_norm": 0.10694328634281369,
+      "learning_rate": 7.677353154025317e-05,
+      "loss": 1.2578,
+      "step": 8149
+    },
+    {
+      "epoch": 0.5874720680458444,
+      "grad_norm": 0.10942402863488,
+      "learning_rate": 7.675082350582119e-05,
+      "loss": 1.2576,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5875441505081813,
+      "grad_norm": 0.11726675760600005,
+      "learning_rate": 7.672811673867729e-05,
+      "loss": 1.2786,
+      "step": 8151
+    },
+    {
+      "epoch": 0.5876162329705182,
+      "grad_norm": 0.11128729481736725,
+      "learning_rate": 7.670541124005921e-05,
+      "loss": 1.3868,
+      "step": 8152
+    },
+    {
+      "epoch": 0.5876883154328552,
+      "grad_norm": 0.10593009568113203,
+      "learning_rate": 7.668270701120464e-05,
+      "loss": 1.2799,
+      "step": 8153
+    },
+    {
+      "epoch": 0.5877603978951921,
+      "grad_norm": 0.10338164877830167,
+      "learning_rate": 7.666000405335115e-05,
+      "loss": 1.3397,
+      "step": 8154
+    },
+    {
+      "epoch": 0.587832480357529,
+      "grad_norm": 0.12788878214571067,
+      "learning_rate": 7.663730236773623e-05,
+      "loss": 1.3097,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5879045628198659,
+      "grad_norm": 0.11281807915899704,
+      "learning_rate": 7.661460195559732e-05,
+      "loss": 1.2101,
+      "step": 8156
+    },
+    {
+      "epoch": 0.5879766452822028,
+      "grad_norm": 0.20010141758530592,
+      "learning_rate": 7.659190281817182e-05,
+      "loss": 1.2744,
+      "step": 8157
+    },
+    {
+      "epoch": 0.5880487277445398,
+      "grad_norm": 0.12262172339856882,
+      "learning_rate": 7.656920495669707e-05,
+      "loss": 1.2197,
+      "step": 8158
+    },
+    {
+      "epoch": 0.5881208102068767,
+      "grad_norm": 0.11134523188833557,
+      "learning_rate": 7.654650837241024e-05,
+      "loss": 1.2847,
+      "step": 8159
+    },
+    {
+      "epoch": 0.5881928926692136,
+      "grad_norm": 0.11818831130614797,
+      "learning_rate": 7.652381306654851e-05,
+      "loss": 1.2819,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5882649751315505,
+      "grad_norm": 0.1167879652298812,
+      "learning_rate": 7.6501119040349e-05,
+      "loss": 1.2964,
+      "step": 8161
+    },
+    {
+      "epoch": 0.5883370575938874,
+      "grad_norm": 0.1488009793461574,
+      "learning_rate": 7.647842629504876e-05,
+      "loss": 1.3451,
+      "step": 8162
+    },
+    {
+      "epoch": 0.5884091400562244,
+      "grad_norm": 0.11063242755886966,
+      "learning_rate": 7.645573483188471e-05,
+      "loss": 1.2823,
+      "step": 8163
+    },
+    {
+      "epoch": 0.5884812225185613,
+      "grad_norm": 0.18448819821643572,
+      "learning_rate": 7.643304465209374e-05,
+      "loss": 1.2786,
+      "step": 8164
+    },
+    {
+      "epoch": 0.5885533049808982,
+      "grad_norm": 0.1145483784383,
+      "learning_rate": 7.641035575691263e-05,
+      "loss": 1.1602,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5886253874432351,
+      "grad_norm": 0.12654611053228296,
+      "learning_rate": 7.638766814757821e-05,
+      "loss": 1.2988,
+      "step": 8166
+    },
+    {
+      "epoch": 0.5886974699055719,
+      "grad_norm": 0.11603093785214794,
+      "learning_rate": 7.636498182532712e-05,
+      "loss": 1.2001,
+      "step": 8167
+    },
+    {
+      "epoch": 0.5887695523679088,
+      "grad_norm": 0.11901110343515364,
+      "learning_rate": 7.634229679139598e-05,
+      "loss": 1.2049,
+      "step": 8168
+    },
+    {
+      "epoch": 0.5888416348302458,
+      "grad_norm": 0.11427571328132971,
+      "learning_rate": 7.631961304702124e-05,
+      "loss": 1.3137,
+      "step": 8169
+    },
+    {
+      "epoch": 0.5889137172925827,
+      "grad_norm": 0.12450370846106867,
+      "learning_rate": 7.629693059343951e-05,
+      "loss": 1.2979,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5889857997549196,
+      "grad_norm": 0.11338879177367756,
+      "learning_rate": 7.627424943188709e-05,
+      "loss": 1.357,
+      "step": 8171
+    },
+    {
+      "epoch": 0.5890578822172565,
+      "grad_norm": 0.11405678080453319,
+      "learning_rate": 7.625156956360036e-05,
+      "loss": 1.1866,
+      "step": 8172
+    },
+    {
+      "epoch": 0.5891299646795934,
+      "grad_norm": 0.1164638854488389,
+      "learning_rate": 7.622889098981553e-05,
+      "loss": 1.287,
+      "step": 8173
+    },
+    {
+      "epoch": 0.5892020471419304,
+      "grad_norm": 0.1109489515162917,
+      "learning_rate": 7.620621371176878e-05,
+      "loss": 1.2229,
+      "step": 8174
+    },
+    {
+      "epoch": 0.5892741296042673,
+      "grad_norm": 0.11275984291113257,
+      "learning_rate": 7.618353773069626e-05,
+      "loss": 1.2905,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5893462120666042,
+      "grad_norm": 0.12160408590934801,
+      "learning_rate": 7.616086304783403e-05,
+      "loss": 1.2772,
+      "step": 8176
+    },
+    {
+      "epoch": 0.5894182945289411,
+      "grad_norm": 0.11963518745303867,
+      "learning_rate": 7.613818966441801e-05,
+      "loss": 1.2848,
+      "step": 8177
+    },
+    {
+      "epoch": 0.589490376991278,
+      "grad_norm": 0.1129989235239824,
+      "learning_rate": 7.611551758168412e-05,
+      "loss": 1.3197,
+      "step": 8178
+    },
+    {
+      "epoch": 0.589562459453615,
+      "grad_norm": 0.1221359898000458,
+      "learning_rate": 7.60928468008682e-05,
+      "loss": 1.2589,
+      "step": 8179
+    },
+    {
+      "epoch": 0.5896345419159519,
+      "grad_norm": 0.10742479180934944,
+      "learning_rate": 7.607017732320604e-05,
+      "loss": 1.2465,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5897066243782888,
+      "grad_norm": 0.2045198650389439,
+      "learning_rate": 7.604750914993329e-05,
+      "loss": 1.2269,
+      "step": 8181
+    },
+    {
+      "epoch": 0.5897787068406257,
+      "grad_norm": 0.11208455000253996,
+      "learning_rate": 7.602484228228552e-05,
+      "loss": 1.2303,
+      "step": 8182
+    },
+    {
+      "epoch": 0.5898507893029626,
+      "grad_norm": 0.12291597135828256,
+      "learning_rate": 7.600217672149839e-05,
+      "loss": 1.3498,
+      "step": 8183
+    },
+    {
+      "epoch": 0.5899228717652995,
+      "grad_norm": 0.1258209202043684,
+      "learning_rate": 7.597951246880732e-05,
+      "loss": 1.1699,
+      "step": 8184
+    },
+    {
+      "epoch": 0.5899949542276364,
+      "grad_norm": 0.11357200815245898,
+      "learning_rate": 7.59568495254477e-05,
+      "loss": 1.2829,
+      "step": 8185
+    },
+    {
+      "epoch": 0.5900670366899733,
+      "grad_norm": 0.10476843731008696,
+      "learning_rate": 7.59341878926549e-05,
+      "loss": 1.3037,
+      "step": 8186
+    },
+    {
+      "epoch": 0.5901391191523102,
+      "grad_norm": 0.1224640616102367,
+      "learning_rate": 7.591152757166409e-05,
+      "loss": 1.265,
+      "step": 8187
+    },
+    {
+      "epoch": 0.5902112016146471,
+      "grad_norm": 0.11322122867569122,
+      "learning_rate": 7.58888685637106e-05,
+      "loss": 1.2351,
+      "step": 8188
+    },
+    {
+      "epoch": 0.590283284076984,
+      "grad_norm": 0.12001445745800011,
+      "learning_rate": 7.586621087002945e-05,
+      "loss": 1.281,
+      "step": 8189
+    },
+    {
+      "epoch": 0.590355366539321,
+      "grad_norm": 0.11275848864489345,
+      "learning_rate": 7.584355449185574e-05,
+      "loss": 1.2431,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5904274490016579,
+      "grad_norm": 0.10629920719055362,
+      "learning_rate": 7.582089943042435e-05,
+      "loss": 1.2005,
+      "step": 8191
+    },
+    {
+      "epoch": 0.5904995314639948,
+      "grad_norm": 0.11586150074006982,
+      "learning_rate": 7.579824568697032e-05,
+      "loss": 1.2813,
+      "step": 8192
+    },
+    {
+      "epoch": 0.5905716139263317,
+      "grad_norm": 0.11428277449676619,
+      "learning_rate": 7.57755932627284e-05,
+      "loss": 1.3635,
+      "step": 8193
+    },
+    {
+      "epoch": 0.5906436963886686,
+      "grad_norm": 0.12529426822215545,
+      "learning_rate": 7.575294215893337e-05,
+      "loss": 1.2063,
+      "step": 8194
+    },
+    {
+      "epoch": 0.5907157788510056,
+      "grad_norm": 0.13332071259641898,
+      "learning_rate": 7.573029237681991e-05,
+      "loss": 1.2873,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5907878613133425,
+      "grad_norm": 0.10687966546971979,
+      "learning_rate": 7.57076439176226e-05,
+      "loss": 1.2325,
+      "step": 8196
+    },
+    {
+      "epoch": 0.5908599437756794,
+      "grad_norm": 0.119660824613366,
+      "learning_rate": 7.568499678257605e-05,
+      "loss": 1.3003,
+      "step": 8197
+    },
+    {
+      "epoch": 0.5909320262380163,
+      "grad_norm": 0.10668976777071575,
+      "learning_rate": 7.566235097291475e-05,
+      "loss": 1.2798,
+      "step": 8198
+    },
+    {
+      "epoch": 0.5910041087003532,
+      "grad_norm": 0.1127947739846384,
+      "learning_rate": 7.563970648987299e-05,
+      "loss": 1.2322,
+      "step": 8199
+    },
+    {
+      "epoch": 0.5910761911626902,
+      "grad_norm": 0.11169198912436956,
+      "learning_rate": 7.561706333468517e-05,
+      "loss": 1.2546,
+      "step": 8200
+    },
+    {
+      "epoch": 0.591148273625027,
+      "grad_norm": 0.11001770439452256,
+      "learning_rate": 7.559442150858553e-05,
+      "loss": 1.2693,
+      "step": 8201
+    },
+    {
+      "epoch": 0.5912203560873639,
+      "grad_norm": 0.11006327327464402,
+      "learning_rate": 7.55717810128083e-05,
+      "loss": 1.3386,
+      "step": 8202
+    },
+    {
+      "epoch": 0.5912924385497008,
+      "grad_norm": 0.12415404167320639,
+      "learning_rate": 7.554914184858751e-05,
+      "loss": 1.2685,
+      "step": 8203
+    },
+    {
+      "epoch": 0.5913645210120377,
+      "grad_norm": 0.12019275192029612,
+      "learning_rate": 7.552650401715722e-05,
+      "loss": 1.2515,
+      "step": 8204
+    },
+    {
+      "epoch": 0.5914366034743747,
+      "grad_norm": 0.12358470673180111,
+      "learning_rate": 7.550386751975143e-05,
+      "loss": 1.3231,
+      "step": 8205
+    },
+    {
+      "epoch": 0.5915086859367116,
+      "grad_norm": 0.1255687624538222,
+      "learning_rate": 7.548123235760403e-05,
+      "loss": 1.3081,
+      "step": 8206
+    },
+    {
+      "epoch": 0.5915807683990485,
+      "grad_norm": 0.11694643262578293,
+      "learning_rate": 7.545859853194878e-05,
+      "loss": 1.3182,
+      "step": 8207
+    },
+    {
+      "epoch": 0.5916528508613854,
+      "grad_norm": 0.11034207511016182,
+      "learning_rate": 7.543596604401948e-05,
+      "loss": 1.2421,
+      "step": 8208
+    },
+    {
+      "epoch": 0.5917249333237223,
+      "grad_norm": 0.1208701425038323,
+      "learning_rate": 7.541333489504975e-05,
+      "loss": 1.2344,
+      "step": 8209
+    },
+    {
+      "epoch": 0.5917970157860593,
+      "grad_norm": 0.1175620532721735,
+      "learning_rate": 7.53907050862733e-05,
+      "loss": 1.2224,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5918690982483962,
+      "grad_norm": 0.11575032972723547,
+      "learning_rate": 7.536807661892353e-05,
+      "loss": 1.2339,
+      "step": 8211
+    },
+    {
+      "epoch": 0.5919411807107331,
+      "grad_norm": 0.17713463260523687,
+      "learning_rate": 7.534544949423397e-05,
+      "loss": 1.3035,
+      "step": 8212
+    },
+    {
+      "epoch": 0.59201326317307,
+      "grad_norm": 0.13507207244885872,
+      "learning_rate": 7.532282371343795e-05,
+      "loss": 1.3357,
+      "step": 8213
+    },
+    {
+      "epoch": 0.5920853456354069,
+      "grad_norm": 0.1091087580452526,
+      "learning_rate": 7.530019927776885e-05,
+      "loss": 1.2228,
+      "step": 8214
+    },
+    {
+      "epoch": 0.5921574280977439,
+      "grad_norm": 0.11254668128284728,
+      "learning_rate": 7.527757618845984e-05,
+      "loss": 1.2226,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5922295105600808,
+      "grad_norm": 0.10965855197138999,
+      "learning_rate": 7.525495444674409e-05,
+      "loss": 1.2582,
+      "step": 8216
+    },
+    {
+      "epoch": 0.5923015930224177,
+      "grad_norm": 0.12461972560198677,
+      "learning_rate": 7.523233405385473e-05,
+      "loss": 1.2393,
+      "step": 8217
+    },
+    {
+      "epoch": 0.5923736754847546,
+      "grad_norm": 0.12106984940262858,
+      "learning_rate": 7.520971501102469e-05,
+      "loss": 1.2545,
+      "step": 8218
+    },
+    {
+      "epoch": 0.5924457579470914,
+      "grad_norm": 0.1137350696365268,
+      "learning_rate": 7.518709731948699e-05,
+      "loss": 1.2549,
+      "step": 8219
+    },
+    {
+      "epoch": 0.5925178404094283,
+      "grad_norm": 0.12493252819136139,
+      "learning_rate": 7.516448098047446e-05,
+      "loss": 1.2854,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5925899228717653,
+      "grad_norm": 0.11832188193508163,
+      "learning_rate": 7.514186599521993e-05,
+      "loss": 1.2215,
+      "step": 8221
+    },
+    {
+      "epoch": 0.5926620053341022,
+      "grad_norm": 0.11439654615717303,
+      "learning_rate": 7.511925236495605e-05,
+      "loss": 1.264,
+      "step": 8222
+    },
+    {
+      "epoch": 0.5927340877964391,
+      "grad_norm": 0.13596305836910227,
+      "learning_rate": 7.509664009091552e-05,
+      "loss": 1.3031,
+      "step": 8223
+    },
+    {
+      "epoch": 0.592806170258776,
+      "grad_norm": 0.13529407063682664,
+      "learning_rate": 7.507402917433091e-05,
+      "loss": 1.3267,
+      "step": 8224
+    },
+    {
+      "epoch": 0.5928782527211129,
+      "grad_norm": 0.12669319847919416,
+      "learning_rate": 7.505141961643471e-05,
+      "loss": 1.2879,
+      "step": 8225
+    },
+    {
+      "epoch": 0.5929503351834499,
+      "grad_norm": 0.11384162626968254,
+      "learning_rate": 7.502881141845929e-05,
+      "loss": 1.3512,
+      "step": 8226
+    },
+    {
+      "epoch": 0.5930224176457868,
+      "grad_norm": 0.12175544880267489,
+      "learning_rate": 7.500620458163709e-05,
+      "loss": 1.338,
+      "step": 8227
+    },
+    {
+      "epoch": 0.5930945001081237,
+      "grad_norm": 0.10682985151064607,
+      "learning_rate": 7.498359910720034e-05,
+      "loss": 1.3052,
+      "step": 8228
+    },
+    {
+      "epoch": 0.5931665825704606,
+      "grad_norm": 0.12942489491042244,
+      "learning_rate": 7.496099499638125e-05,
+      "loss": 1.2553,
+      "step": 8229
+    },
+    {
+      "epoch": 0.5932386650327975,
+      "grad_norm": 0.1116067923709528,
+      "learning_rate": 7.493839225041193e-05,
+      "loss": 1.2177,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5933107474951345,
+      "grad_norm": 0.11838659079533344,
+      "learning_rate": 7.491579087052441e-05,
+      "loss": 1.2396,
+      "step": 8231
+    },
+    {
+      "epoch": 0.5933828299574714,
+      "grad_norm": 0.12850051830668743,
+      "learning_rate": 7.48931908579507e-05,
+      "loss": 1.2587,
+      "step": 8232
+    },
+    {
+      "epoch": 0.5934549124198083,
+      "grad_norm": 0.11875686056871522,
+      "learning_rate": 7.487059221392274e-05,
+      "loss": 1.2062,
+      "step": 8233
+    },
+    {
+      "epoch": 0.5935269948821452,
+      "grad_norm": 0.1134429581952051,
+      "learning_rate": 7.48479949396723e-05,
+      "loss": 1.3086,
+      "step": 8234
+    },
+    {
+      "epoch": 0.5935990773444821,
+      "grad_norm": 0.11990211119907673,
+      "learning_rate": 7.482539903643114e-05,
+      "loss": 1.3316,
+      "step": 8235
+    },
+    {
+      "epoch": 0.593671159806819,
+      "grad_norm": 0.15219540272014093,
+      "learning_rate": 7.480280450543095e-05,
+      "loss": 1.3068,
+      "step": 8236
+    },
+    {
+      "epoch": 0.5937432422691559,
+      "grad_norm": 0.46704323999764924,
+      "learning_rate": 7.478021134790336e-05,
+      "loss": 1.3031,
+      "step": 8237
+    },
+    {
+      "epoch": 0.5938153247314928,
+      "grad_norm": 0.10642614402623937,
+      "learning_rate": 7.475761956507987e-05,
+      "loss": 1.2405,
+      "step": 8238
+    },
+    {
+      "epoch": 0.5938874071938297,
+      "grad_norm": 0.10888579240891412,
+      "learning_rate": 7.473502915819194e-05,
+      "loss": 1.2552,
+      "step": 8239
+    },
+    {
+      "epoch": 0.5939594896561666,
+      "grad_norm": 0.10521157640694359,
+      "learning_rate": 7.47124401284709e-05,
+      "loss": 1.2158,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5940315721185035,
+      "grad_norm": 0.12879979670635963,
+      "learning_rate": 7.468985247714815e-05,
+      "loss": 1.2743,
+      "step": 8241
+    },
+    {
+      "epoch": 0.5941036545808405,
+      "grad_norm": 0.1266515573383367,
+      "learning_rate": 7.466726620545485e-05,
+      "loss": 1.304,
+      "step": 8242
+    },
+    {
+      "epoch": 0.5941757370431774,
+      "grad_norm": 0.11106840876671892,
+      "learning_rate": 7.464468131462222e-05,
+      "loss": 1.209,
+      "step": 8243
+    },
+    {
+      "epoch": 0.5942478195055143,
+      "grad_norm": 0.12004873590166379,
+      "learning_rate": 7.462209780588121e-05,
+      "loss": 1.2573,
+      "step": 8244
+    },
+    {
+      "epoch": 0.5943199019678512,
+      "grad_norm": 0.11319045113886989,
+      "learning_rate": 7.459951568046297e-05,
+      "loss": 1.1963,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5943919844301881,
+      "grad_norm": 0.1265090531640231,
+      "learning_rate": 7.457693493959834e-05,
+      "loss": 1.2253,
+      "step": 8246
+    },
+    {
+      "epoch": 0.5944640668925251,
+      "grad_norm": 0.12311624147996204,
+      "learning_rate": 7.455435558451823e-05,
+      "loss": 1.2313,
+      "step": 8247
+    },
+    {
+      "epoch": 0.594536149354862,
+      "grad_norm": 0.11451627647060986,
+      "learning_rate": 7.453177761645331e-05,
+      "loss": 1.2619,
+      "step": 8248
+    },
+    {
+      "epoch": 0.5946082318171989,
+      "grad_norm": 0.10873818682163768,
+      "learning_rate": 7.450920103663443e-05,
+      "loss": 1.239,
+      "step": 8249
+    },
+    {
+      "epoch": 0.5946803142795358,
+      "grad_norm": 0.10527710616336936,
+      "learning_rate": 7.448662584629211e-05,
+      "loss": 1.3055,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5947523967418727,
+      "grad_norm": 0.1770091893874333,
+      "learning_rate": 7.446405204665697e-05,
+      "loss": 1.2407,
+      "step": 8251
+    },
+    {
+      "epoch": 0.5948244792042097,
+      "grad_norm": 0.12093574420135553,
+      "learning_rate": 7.44414796389594e-05,
+      "loss": 1.3274,
+      "step": 8252
+    },
+    {
+      "epoch": 0.5948965616665465,
+      "grad_norm": 0.11273365849715646,
+      "learning_rate": 7.441890862442984e-05,
+      "loss": 1.2956,
+      "step": 8253
+    },
+    {
+      "epoch": 0.5949686441288834,
+      "grad_norm": 0.11610677953155646,
+      "learning_rate": 7.439633900429862e-05,
+      "loss": 1.284,
+      "step": 8254
+    },
+    {
+      "epoch": 0.5950407265912203,
+      "grad_norm": 0.1092568349099545,
+      "learning_rate": 7.437377077979601e-05,
+      "loss": 1.2304,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5951128090535572,
+      "grad_norm": 0.11482157163251253,
+      "learning_rate": 7.435120395215214e-05,
+      "loss": 1.2438,
+      "step": 8256
+    },
+    {
+      "epoch": 0.5951848915158942,
+      "grad_norm": 0.11361711926259119,
+      "learning_rate": 7.432863852259708e-05,
+      "loss": 1.3581,
+      "step": 8257
+    },
+    {
+      "epoch": 0.5952569739782311,
+      "grad_norm": 0.11151146114965053,
+      "learning_rate": 7.43060744923609e-05,
+      "loss": 1.2597,
+      "step": 8258
+    },
+    {
+      "epoch": 0.595329056440568,
+      "grad_norm": 0.12509280316487353,
+      "learning_rate": 7.428351186267356e-05,
+      "loss": 1.3076,
+      "step": 8259
+    },
+    {
+      "epoch": 0.5954011389029049,
+      "grad_norm": 0.11294926147659803,
+      "learning_rate": 7.426095063476486e-05,
+      "loss": 1.2755,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5954732213652418,
+      "grad_norm": 0.12128032511594772,
+      "learning_rate": 7.423839080986464e-05,
+      "loss": 1.2669,
+      "step": 8261
+    },
+    {
+      "epoch": 0.5955453038275788,
+      "grad_norm": 0.1098872330504074,
+      "learning_rate": 7.421583238920253e-05,
+      "loss": 1.2789,
+      "step": 8262
+    },
+    {
+      "epoch": 0.5956173862899157,
+      "grad_norm": 0.12346942001222687,
+      "learning_rate": 7.419327537400829e-05,
+      "loss": 1.2534,
+      "step": 8263
+    },
+    {
+      "epoch": 0.5956894687522526,
+      "grad_norm": 0.12288333347343942,
+      "learning_rate": 7.41707197655114e-05,
+      "loss": 1.2011,
+      "step": 8264
+    },
+    {
+      "epoch": 0.5957615512145895,
+      "grad_norm": 0.12482607723623806,
+      "learning_rate": 7.414816556494138e-05,
+      "loss": 1.2108,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5958336336769264,
+      "grad_norm": 0.12900079321192415,
+      "learning_rate": 7.412561277352753e-05,
+      "loss": 1.1383,
+      "step": 8266
+    },
+    {
+      "epoch": 0.5959057161392634,
+      "grad_norm": 0.12267609097001418,
+      "learning_rate": 7.410306139249934e-05,
+      "loss": 1.2134,
+      "step": 8267
+    },
+    {
+      "epoch": 0.5959777986016003,
+      "grad_norm": 0.11945275201151548,
+      "learning_rate": 7.408051142308596e-05,
+      "loss": 1.2379,
+      "step": 8268
+    },
+    {
+      "epoch": 0.5960498810639372,
+      "grad_norm": 0.12026776765406676,
+      "learning_rate": 7.405796286651662e-05,
+      "loss": 1.3234,
+      "step": 8269
+    },
+    {
+      "epoch": 0.5961219635262741,
+      "grad_norm": 0.12573302637954978,
+      "learning_rate": 7.403541572402036e-05,
+      "loss": 1.2456,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5961940459886109,
+      "grad_norm": 0.12379957560031878,
+      "learning_rate": 7.40128699968262e-05,
+      "loss": 1.3097,
+      "step": 8271
+    },
+    {
+      "epoch": 0.5962661284509478,
+      "grad_norm": 0.10980342248579679,
+      "learning_rate": 7.399032568616315e-05,
+      "loss": 1.338,
+      "step": 8272
+    },
+    {
+      "epoch": 0.5963382109132848,
+      "grad_norm": 0.1207876754960535,
+      "learning_rate": 7.396778279326005e-05,
+      "loss": 1.2673,
+      "step": 8273
+    },
+    {
+      "epoch": 0.5964102933756217,
+      "grad_norm": 0.12614920111818062,
+      "learning_rate": 7.394524131934566e-05,
+      "loss": 1.2403,
+      "step": 8274
+    },
+    {
+      "epoch": 0.5964823758379586,
+      "grad_norm": 0.12234545073432039,
+      "learning_rate": 7.392270126564869e-05,
+      "loss": 1.2974,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5965544583002955,
+      "grad_norm": 0.12190879404232881,
+      "learning_rate": 7.390016263339782e-05,
+      "loss": 1.3232,
+      "step": 8276
+    },
+    {
+      "epoch": 0.5966265407626324,
+      "grad_norm": 0.1301379309486666,
+      "learning_rate": 7.387762542382159e-05,
+      "loss": 1.3238,
+      "step": 8277
+    },
+    {
+      "epoch": 0.5966986232249694,
+      "grad_norm": 0.11798935996138744,
+      "learning_rate": 7.385508963814847e-05,
+      "loss": 1.3049,
+      "step": 8278
+    },
+    {
+      "epoch": 0.5967707056873063,
+      "grad_norm": 0.1132464894894101,
+      "learning_rate": 7.383255527760683e-05,
+      "loss": 1.2918,
+      "step": 8279
+    },
+    {
+      "epoch": 0.5968427881496432,
+      "grad_norm": 0.10448559082629454,
+      "learning_rate": 7.381002234342506e-05,
+      "loss": 1.2862,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5969148706119801,
+      "grad_norm": 0.10951781285751098,
+      "learning_rate": 7.378749083683139e-05,
+      "loss": 1.3135,
+      "step": 8281
+    },
+    {
+      "epoch": 0.596986953074317,
+      "grad_norm": 0.11245601003653304,
+      "learning_rate": 7.376496075905395e-05,
+      "loss": 1.2909,
+      "step": 8282
+    },
+    {
+      "epoch": 0.597059035536654,
+      "grad_norm": 0.11162937392612428,
+      "learning_rate": 7.374243211132085e-05,
+      "loss": 1.1646,
+      "step": 8283
+    },
+    {
+      "epoch": 0.5971311179989909,
+      "grad_norm": 0.12258123146152886,
+      "learning_rate": 7.37199048948601e-05,
+      "loss": 1.3144,
+      "step": 8284
+    },
+    {
+      "epoch": 0.5972032004613278,
+      "grad_norm": 0.11352942887174232,
+      "learning_rate": 7.369737911089967e-05,
+      "loss": 1.2859,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5972752829236647,
+      "grad_norm": 0.11687018757032917,
+      "learning_rate": 7.367485476066736e-05,
+      "loss": 1.2302,
+      "step": 8286
+    },
+    {
+      "epoch": 0.5973473653860016,
+      "grad_norm": 0.11264324296005872,
+      "learning_rate": 7.3652331845391e-05,
+      "loss": 1.2138,
+      "step": 8287
+    },
+    {
+      "epoch": 0.5974194478483384,
+      "grad_norm": 0.11311379989736689,
+      "learning_rate": 7.362981036629823e-05,
+      "loss": 1.2603,
+      "step": 8288
+    },
+    {
+      "epoch": 0.5974915303106754,
+      "grad_norm": 0.1211929199737724,
+      "learning_rate": 7.360729032461675e-05,
+      "loss": 1.3177,
+      "step": 8289
+    },
+    {
+      "epoch": 0.5975636127730123,
+      "grad_norm": 0.12239482429432447,
+      "learning_rate": 7.358477172157405e-05,
+      "loss": 1.2787,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5976356952353492,
+      "grad_norm": 0.12183271825314324,
+      "learning_rate": 7.35622545583976e-05,
+      "loss": 1.3671,
+      "step": 8291
+    },
+    {
+      "epoch": 0.5977077776976861,
+      "grad_norm": 0.10782817662274262,
+      "learning_rate": 7.353973883631483e-05,
+      "loss": 1.2549,
+      "step": 8292
+    },
+    {
+      "epoch": 0.597779860160023,
+      "grad_norm": 0.1244974598598908,
+      "learning_rate": 7.351722455655297e-05,
+      "loss": 1.2404,
+      "step": 8293
+    },
+    {
+      "epoch": 0.59785194262236,
+      "grad_norm": 0.13112341832697844,
+      "learning_rate": 7.34947117203393e-05,
+      "loss": 1.2483,
+      "step": 8294
+    },
+    {
+      "epoch": 0.5979240250846969,
+      "grad_norm": 0.11926198525088116,
+      "learning_rate": 7.347220032890097e-05,
+      "loss": 1.258,
+      "step": 8295
+    },
+    {
+      "epoch": 0.5979961075470338,
+      "grad_norm": 0.11081211850004534,
+      "learning_rate": 7.344969038346507e-05,
+      "loss": 1.2709,
+      "step": 8296
+    },
+    {
+      "epoch": 0.5980681900093707,
+      "grad_norm": 0.11216932714679805,
+      "learning_rate": 7.342718188525854e-05,
+      "loss": 1.3165,
+      "step": 8297
+    },
+    {
+      "epoch": 0.5981402724717076,
+      "grad_norm": 0.1146658390177065,
+      "learning_rate": 7.340467483550832e-05,
+      "loss": 1.2926,
+      "step": 8298
+    },
+    {
+      "epoch": 0.5982123549340446,
+      "grad_norm": 0.11831149380083683,
+      "learning_rate": 7.338216923544128e-05,
+      "loss": 1.316,
+      "step": 8299
+    },
+    {
+      "epoch": 0.5982844373963815,
+      "grad_norm": 0.11904514207306555,
+      "learning_rate": 7.335966508628414e-05,
+      "loss": 1.2843,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5983565198587184,
+      "grad_norm": 0.14166412683802132,
+      "learning_rate": 7.333716238926355e-05,
+      "loss": 1.2446,
+      "step": 8301
+    },
+    {
+      "epoch": 0.5984286023210553,
+      "grad_norm": 0.12255718671625128,
+      "learning_rate": 7.331466114560618e-05,
+      "loss": 1.2995,
+      "step": 8302
+    },
+    {
+      "epoch": 0.5985006847833922,
+      "grad_norm": 0.13590060671447107,
+      "learning_rate": 7.329216135653848e-05,
+      "loss": 1.4058,
+      "step": 8303
+    },
+    {
+      "epoch": 0.5985727672457292,
+      "grad_norm": 0.11968253160338023,
+      "learning_rate": 7.326966302328697e-05,
+      "loss": 1.2275,
+      "step": 8304
+    },
+    {
+      "epoch": 0.598644849708066,
+      "grad_norm": 0.12781657210532069,
+      "learning_rate": 7.324716614707793e-05,
+      "loss": 1.2165,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5987169321704029,
+      "grad_norm": 0.11262047554732259,
+      "learning_rate": 7.322467072913765e-05,
+      "loss": 1.2683,
+      "step": 8306
+    },
+    {
+      "epoch": 0.5987890146327398,
+      "grad_norm": 0.11862448207346903,
+      "learning_rate": 7.320217677069238e-05,
+      "loss": 1.3635,
+      "step": 8307
+    },
+    {
+      "epoch": 0.5988610970950767,
+      "grad_norm": 0.24998301123874672,
+      "learning_rate": 7.317968427296824e-05,
+      "loss": 1.253,
+      "step": 8308
+    },
+    {
+      "epoch": 0.5989331795574137,
+      "grad_norm": 0.11874915252784657,
+      "learning_rate": 7.315719323719124e-05,
+      "loss": 1.2724,
+      "step": 8309
+    },
+    {
+      "epoch": 0.5990052620197506,
+      "grad_norm": 0.1060031912532573,
+      "learning_rate": 7.31347036645873e-05,
+      "loss": 1.2834,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5990773444820875,
+      "grad_norm": 0.1188201059720975,
+      "learning_rate": 7.31122155563824e-05,
+      "loss": 1.3319,
+      "step": 8311
+    },
+    {
+      "epoch": 0.5991494269444244,
+      "grad_norm": 0.12049723794408548,
+      "learning_rate": 7.30897289138023e-05,
+      "loss": 1.2739,
+      "step": 8312
+    },
+    {
+      "epoch": 0.5992215094067613,
+      "grad_norm": 0.11680142456544508,
+      "learning_rate": 7.30672437380727e-05,
+      "loss": 1.236,
+      "step": 8313
+    },
+    {
+      "epoch": 0.5992935918690983,
+      "grad_norm": 0.11547222373980265,
+      "learning_rate": 7.304476003041928e-05,
+      "loss": 1.3475,
+      "step": 8314
+    },
+    {
+      "epoch": 0.5993656743314352,
+      "grad_norm": 0.11655348592301985,
+      "learning_rate": 7.302227779206755e-05,
+      "loss": 1.2507,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5994377567937721,
+      "grad_norm": 0.131815156321868,
+      "learning_rate": 7.299979702424308e-05,
+      "loss": 1.2795,
+      "step": 8316
+    },
+    {
+      "epoch": 0.599509839256109,
+      "grad_norm": 0.12110282364356488,
+      "learning_rate": 7.29773177281712e-05,
+      "loss": 1.3303,
+      "step": 8317
+    },
+    {
+      "epoch": 0.5995819217184459,
+      "grad_norm": 0.11602542440736151,
+      "learning_rate": 7.295483990507727e-05,
+      "loss": 1.2178,
+      "step": 8318
+    },
+    {
+      "epoch": 0.5996540041807829,
+      "grad_norm": 0.11478607336640614,
+      "learning_rate": 7.293236355618647e-05,
+      "loss": 1.3162,
+      "step": 8319
+    },
+    {
+      "epoch": 0.5997260866431198,
+      "grad_norm": 0.11483332134057221,
+      "learning_rate": 7.290988868272406e-05,
+      "loss": 1.2494,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5997981691054567,
+      "grad_norm": 0.12534710400550308,
+      "learning_rate": 7.288741528591505e-05,
+      "loss": 1.3147,
+      "step": 8321
+    },
+    {
+      "epoch": 0.5998702515677936,
+      "grad_norm": 0.11321843343699158,
+      "learning_rate": 7.286494336698449e-05,
+      "loss": 1.2612,
+      "step": 8322
+    },
+    {
+      "epoch": 0.5999423340301304,
+      "grad_norm": 0.11682704355734196,
+      "learning_rate": 7.284247292715721e-05,
+      "loss": 1.2787,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6000144164924673,
+      "grad_norm": 0.11566526769564442,
+      "learning_rate": 7.282000396765818e-05,
+      "loss": 1.2163,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6000864989548043,
+      "grad_norm": 0.12242295687647924,
+      "learning_rate": 7.279753648971206e-05,
+      "loss": 1.2712,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6001585814171412,
+      "grad_norm": 0.11687219491922808,
+      "learning_rate": 7.27750704945436e-05,
+      "loss": 1.33,
+      "step": 8326
+    },
+    {
+      "epoch": 0.6002306638794781,
+      "grad_norm": 0.11603242662437023,
+      "learning_rate": 7.275260598337733e-05,
+      "loss": 1.2191,
+      "step": 8327
+    },
+    {
+      "epoch": 0.600302746341815,
+      "grad_norm": 0.1142898432613424,
+      "learning_rate": 7.273014295743776e-05,
+      "loss": 1.3451,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6003748288041519,
+      "grad_norm": 0.1160008946980655,
+      "learning_rate": 7.270768141794942e-05,
+      "loss": 1.2719,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6004469112664889,
+      "grad_norm": 0.11961620709803746,
+      "learning_rate": 7.268522136613659e-05,
+      "loss": 1.3365,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6005189937288258,
+      "grad_norm": 0.12570429552969767,
+      "learning_rate": 7.266276280322354e-05,
+      "loss": 1.1681,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6005910761911627,
+      "grad_norm": 0.12893587831168565,
+      "learning_rate": 7.26403057304345e-05,
+      "loss": 1.3366,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6006631586534996,
+      "grad_norm": 0.13036838741648563,
+      "learning_rate": 7.261785014899356e-05,
+      "loss": 1.2949,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6007352411158365,
+      "grad_norm": 0.12324851890077909,
+      "learning_rate": 7.259539606012478e-05,
+      "loss": 1.3521,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6008073235781735,
+      "grad_norm": 0.12441938881012038,
+      "learning_rate": 7.257294346505205e-05,
+      "loss": 1.2656,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6008794060405104,
+      "grad_norm": 0.1313671950679522,
+      "learning_rate": 7.255049236499932e-05,
+      "loss": 1.2806,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6009514885028473,
+      "grad_norm": 0.18936176804709767,
+      "learning_rate": 7.252804276119023e-05,
+      "loss": 1.3137,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6010235709651842,
+      "grad_norm": 0.11892090912998436,
+      "learning_rate": 7.250559465484865e-05,
+      "loss": 1.2784,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6010956534275211,
+      "grad_norm": 0.10949815736110313,
+      "learning_rate": 7.248314804719813e-05,
+      "loss": 1.2468,
+      "step": 8339
+    },
+    {
+      "epoch": 0.601167735889858,
+      "grad_norm": 0.10820652761048465,
+      "learning_rate": 7.246070293946223e-05,
+      "loss": 1.2352,
+      "step": 8340
+    },
+    {
+      "epoch": 0.6012398183521949,
+      "grad_norm": 0.1340675696755367,
+      "learning_rate": 7.243825933286432e-05,
+      "loss": 1.2653,
+      "step": 8341
+    },
+    {
+      "epoch": 0.6013119008145318,
+      "grad_norm": 0.16737059487803782,
+      "learning_rate": 7.241581722862791e-05,
+      "loss": 1.2756,
+      "step": 8342
+    },
+    {
+      "epoch": 0.6013839832768687,
+      "grad_norm": 0.10873750744249405,
+      "learning_rate": 7.239337662797624e-05,
+      "loss": 1.3291,
+      "step": 8343
+    },
+    {
+      "epoch": 0.6014560657392056,
+      "grad_norm": 0.11332722325170756,
+      "learning_rate": 7.237093753213252e-05,
+      "loss": 1.2605,
+      "step": 8344
+    },
+    {
+      "epoch": 0.6015281482015425,
+      "grad_norm": 0.11375746918418769,
+      "learning_rate": 7.234849994231982e-05,
+      "loss": 1.1638,
+      "step": 8345
+    },
+    {
+      "epoch": 0.6016002306638795,
+      "grad_norm": 0.13875180950459617,
+      "learning_rate": 7.232606385976132e-05,
+      "loss": 1.3357,
+      "step": 8346
+    },
+    {
+      "epoch": 0.6016723131262164,
+      "grad_norm": 0.11479613830992282,
+      "learning_rate": 7.230362928567991e-05,
+      "loss": 1.2448,
+      "step": 8347
+    },
+    {
+      "epoch": 0.6017443955885533,
+      "grad_norm": 0.10878583685301424,
+      "learning_rate": 7.228119622129848e-05,
+      "loss": 1.2773,
+      "step": 8348
+    },
+    {
+      "epoch": 0.6018164780508902,
+      "grad_norm": 0.14493174435053205,
+      "learning_rate": 7.225876466783983e-05,
+      "loss": 1.2846,
+      "step": 8349
+    },
+    {
+      "epoch": 0.6018885605132271,
+      "grad_norm": 0.11171882225763823,
+      "learning_rate": 7.223633462652665e-05,
+      "loss": 1.2934,
+      "step": 8350
+    },
+    {
+      "epoch": 0.6019606429755641,
+      "grad_norm": 0.10921632060576913,
+      "learning_rate": 7.221390609858164e-05,
+      "loss": 1.2563,
+      "step": 8351
+    },
+    {
+      "epoch": 0.602032725437901,
+      "grad_norm": 0.12184967562013435,
+      "learning_rate": 7.219147908522735e-05,
+      "loss": 1.3358,
+      "step": 8352
+    },
+    {
+      "epoch": 0.6021048079002379,
+      "grad_norm": 0.11466947617172737,
+      "learning_rate": 7.216905358768622e-05,
+      "loss": 1.2375,
+      "step": 8353
+    },
+    {
+      "epoch": 0.6021768903625748,
+      "grad_norm": 0.11926366329874953,
+      "learning_rate": 7.214662960718061e-05,
+      "loss": 1.2994,
+      "step": 8354
+    },
+    {
+      "epoch": 0.6022489728249117,
+      "grad_norm": 0.11648600345210768,
+      "learning_rate": 7.212420714493291e-05,
+      "loss": 1.2939,
+      "step": 8355
+    },
+    {
+      "epoch": 0.6023210552872487,
+      "grad_norm": 0.10718878295259382,
+      "learning_rate": 7.210178620216532e-05,
+      "loss": 1.2172,
+      "step": 8356
+    },
+    {
+      "epoch": 0.6023931377495855,
+      "grad_norm": 0.1265128822322177,
+      "learning_rate": 7.207936678009996e-05,
+      "loss": 1.2695,
+      "step": 8357
+    },
+    {
+      "epoch": 0.6024652202119224,
+      "grad_norm": 0.1300388312747018,
+      "learning_rate": 7.205694887995888e-05,
+      "loss": 1.3046,
+      "step": 8358
+    },
+    {
+      "epoch": 0.6025373026742593,
+      "grad_norm": 0.11653550719150872,
+      "learning_rate": 7.203453250296404e-05,
+      "loss": 1.4068,
+      "step": 8359
+    },
+    {
+      "epoch": 0.6026093851365962,
+      "grad_norm": 0.10779655065624266,
+      "learning_rate": 7.201211765033743e-05,
+      "loss": 1.2312,
+      "step": 8360
+    },
+    {
+      "epoch": 0.6026814675989332,
+      "grad_norm": 0.10980230198713366,
+      "learning_rate": 7.198970432330076e-05,
+      "loss": 1.3575,
+      "step": 8361
+    },
+    {
+      "epoch": 0.6027535500612701,
+      "grad_norm": 0.12380632957685647,
+      "learning_rate": 7.196729252307579e-05,
+      "loss": 1.2624,
+      "step": 8362
+    },
+    {
+      "epoch": 0.602825632523607,
+      "grad_norm": 0.11534159683573011,
+      "learning_rate": 7.194488225088417e-05,
+      "loss": 1.341,
+      "step": 8363
+    },
+    {
+      "epoch": 0.6028977149859439,
+      "grad_norm": 0.11439755000291928,
+      "learning_rate": 7.192247350794746e-05,
+      "loss": 1.2518,
+      "step": 8364
+    },
+    {
+      "epoch": 0.6029697974482808,
+      "grad_norm": 0.11432436138993149,
+      "learning_rate": 7.190006629548714e-05,
+      "loss": 1.2003,
+      "step": 8365
+    },
+    {
+      "epoch": 0.6030418799106178,
+      "grad_norm": 0.11939885279916455,
+      "learning_rate": 7.187766061472459e-05,
+      "loss": 1.3251,
+      "step": 8366
+    },
+    {
+      "epoch": 0.6031139623729547,
+      "grad_norm": 0.12104536205045345,
+      "learning_rate": 7.18552564668811e-05,
+      "loss": 1.1274,
+      "step": 8367
+    },
+    {
+      "epoch": 0.6031860448352916,
+      "grad_norm": 0.2960870733564666,
+      "learning_rate": 7.183285385317795e-05,
+      "loss": 1.371,
+      "step": 8368
+    },
+    {
+      "epoch": 0.6032581272976285,
+      "grad_norm": 0.12243168998142004,
+      "learning_rate": 7.181045277483625e-05,
+      "loss": 1.2405,
+      "step": 8369
+    },
+    {
+      "epoch": 0.6033302097599654,
+      "grad_norm": 0.11267608186420519,
+      "learning_rate": 7.178805323307707e-05,
+      "loss": 1.2622,
+      "step": 8370
+    },
+    {
+      "epoch": 0.6034022922223023,
+      "grad_norm": 0.11831370723579726,
+      "learning_rate": 7.176565522912138e-05,
+      "loss": 1.2068,
+      "step": 8371
+    },
+    {
+      "epoch": 0.6034743746846393,
+      "grad_norm": 0.1106552686685881,
+      "learning_rate": 7.174325876419006e-05,
+      "loss": 1.1603,
+      "step": 8372
+    },
+    {
+      "epoch": 0.6035464571469762,
+      "grad_norm": 0.11874361530323307,
+      "learning_rate": 7.172086383950393e-05,
+      "loss": 1.2721,
+      "step": 8373
+    },
+    {
+      "epoch": 0.6036185396093131,
+      "grad_norm": 0.12443104458384834,
+      "learning_rate": 7.16984704562837e-05,
+      "loss": 1.2384,
+      "step": 8374
+    },
+    {
+      "epoch": 0.6036906220716499,
+      "grad_norm": 0.11765455194873345,
+      "learning_rate": 7.167607861575007e-05,
+      "loss": 1.2501,
+      "step": 8375
+    },
+    {
+      "epoch": 0.6037627045339868,
+      "grad_norm": 0.12386021508468817,
+      "learning_rate": 7.16536883191235e-05,
+      "loss": 1.2633,
+      "step": 8376
+    },
+    {
+      "epoch": 0.6038347869963238,
+      "grad_norm": 0.10816112300033946,
+      "learning_rate": 7.163129956762454e-05,
+      "loss": 1.2345,
+      "step": 8377
+    },
+    {
+      "epoch": 0.6039068694586607,
+      "grad_norm": 0.13762195857053086,
+      "learning_rate": 7.160891236247353e-05,
+      "loss": 1.2318,
+      "step": 8378
+    },
+    {
+      "epoch": 0.6039789519209976,
+      "grad_norm": 0.11133013011889951,
+      "learning_rate": 7.15865267048908e-05,
+      "loss": 1.1996,
+      "step": 8379
+    },
+    {
+      "epoch": 0.6040510343833345,
+      "grad_norm": 0.11720119121581521,
+      "learning_rate": 7.156414259609657e-05,
+      "loss": 1.3027,
+      "step": 8380
+    },
+    {
+      "epoch": 0.6041231168456714,
+      "grad_norm": 0.10684713160288309,
+      "learning_rate": 7.154176003731092e-05,
+      "loss": 1.2666,
+      "step": 8381
+    },
+    {
+      "epoch": 0.6041951993080084,
+      "grad_norm": 0.11539414103280073,
+      "learning_rate": 7.151937902975398e-05,
+      "loss": 1.2808,
+      "step": 8382
+    },
+    {
+      "epoch": 0.6042672817703453,
+      "grad_norm": 0.11666723897341005,
+      "learning_rate": 7.149699957464568e-05,
+      "loss": 1.2837,
+      "step": 8383
+    },
+    {
+      "epoch": 0.6043393642326822,
+      "grad_norm": 0.11507368760801497,
+      "learning_rate": 7.147462167320589e-05,
+      "loss": 1.348,
+      "step": 8384
+    },
+    {
+      "epoch": 0.6044114466950191,
+      "grad_norm": 0.13195164660140515,
+      "learning_rate": 7.14522453266544e-05,
+      "loss": 1.3013,
+      "step": 8385
+    },
+    {
+      "epoch": 0.604483529157356,
+      "grad_norm": 0.12288067422701271,
+      "learning_rate": 7.142987053621094e-05,
+      "loss": 1.3562,
+      "step": 8386
+    },
+    {
+      "epoch": 0.604555611619693,
+      "grad_norm": 0.11544143742447173,
+      "learning_rate": 7.140749730309516e-05,
+      "loss": 1.324,
+      "step": 8387
+    },
+    {
+      "epoch": 0.6046276940820299,
+      "grad_norm": 0.10579573109493869,
+      "learning_rate": 7.138512562852655e-05,
+      "loss": 1.3142,
+      "step": 8388
+    },
+    {
+      "epoch": 0.6046997765443668,
+      "grad_norm": 0.1079373323199125,
+      "learning_rate": 7.136275551372458e-05,
+      "loss": 1.2136,
+      "step": 8389
+    },
+    {
+      "epoch": 0.6047718590067037,
+      "grad_norm": 0.12588963461922564,
+      "learning_rate": 7.134038695990863e-05,
+      "loss": 1.2584,
+      "step": 8390
+    },
+    {
+      "epoch": 0.6048439414690406,
+      "grad_norm": 0.11859023789450039,
+      "learning_rate": 7.131801996829802e-05,
+      "loss": 1.2578,
+      "step": 8391
+    },
+    {
+      "epoch": 0.6049160239313774,
+      "grad_norm": 0.10721958854503741,
+      "learning_rate": 7.129565454011189e-05,
+      "loss": 1.2764,
+      "step": 8392
+    },
+    {
+      "epoch": 0.6049881063937144,
+      "grad_norm": 0.11288671686452573,
+      "learning_rate": 7.127329067656941e-05,
+      "loss": 1.308,
+      "step": 8393
+    },
+    {
+      "epoch": 0.6050601888560513,
+      "grad_norm": 0.13321347460556873,
+      "learning_rate": 7.125092837888952e-05,
+      "loss": 1.3427,
+      "step": 8394
+    },
+    {
+      "epoch": 0.6051322713183882,
+      "grad_norm": 0.11634541812481335,
+      "learning_rate": 7.12285676482913e-05,
+      "loss": 1.3193,
+      "step": 8395
+    },
+    {
+      "epoch": 0.6052043537807251,
+      "grad_norm": 0.11088221420735955,
+      "learning_rate": 7.12062084859935e-05,
+      "loss": 1.357,
+      "step": 8396
+    },
+    {
+      "epoch": 0.605276436243062,
+      "grad_norm": 0.11290036808997854,
+      "learning_rate": 7.118385089321498e-05,
+      "loss": 1.2509,
+      "step": 8397
+    },
+    {
+      "epoch": 0.605348518705399,
+      "grad_norm": 0.1213968977686018,
+      "learning_rate": 7.116149487117432e-05,
+      "loss": 1.3631,
+      "step": 8398
+    },
+    {
+      "epoch": 0.6054206011677359,
+      "grad_norm": 0.16176539736799952,
+      "learning_rate": 7.113914042109023e-05,
+      "loss": 1.2677,
+      "step": 8399
+    },
+    {
+      "epoch": 0.6054926836300728,
+      "grad_norm": 0.4189745857284686,
+      "learning_rate": 7.111678754418117e-05,
+      "loss": 1.288,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6055647660924097,
+      "grad_norm": 0.12462788236300582,
+      "learning_rate": 7.10944362416656e-05,
+      "loss": 1.1934,
+      "step": 8401
+    },
+    {
+      "epoch": 0.6056368485547466,
+      "grad_norm": 0.10305536112706577,
+      "learning_rate": 7.107208651476185e-05,
+      "loss": 1.3281,
+      "step": 8402
+    },
+    {
+      "epoch": 0.6057089310170836,
+      "grad_norm": 0.11970152361294897,
+      "learning_rate": 7.104973836468815e-05,
+      "loss": 1.2812,
+      "step": 8403
+    },
+    {
+      "epoch": 0.6057810134794205,
+      "grad_norm": 0.12466171024434125,
+      "learning_rate": 7.102739179266271e-05,
+      "loss": 1.2456,
+      "step": 8404
+    },
+    {
+      "epoch": 0.6058530959417574,
+      "grad_norm": 0.12432909983075109,
+      "learning_rate": 7.100504679990365e-05,
+      "loss": 1.3613,
+      "step": 8405
+    },
+    {
+      "epoch": 0.6059251784040943,
+      "grad_norm": 0.10764018652049213,
+      "learning_rate": 7.098270338762891e-05,
+      "loss": 1.2576,
+      "step": 8406
+    },
+    {
+      "epoch": 0.6059972608664312,
+      "grad_norm": 0.11459220071145527,
+      "learning_rate": 7.096036155705643e-05,
+      "loss": 1.2677,
+      "step": 8407
+    },
+    {
+      "epoch": 0.6060693433287682,
+      "grad_norm": 0.09958209227067757,
+      "learning_rate": 7.093802130940403e-05,
+      "loss": 1.2927,
+      "step": 8408
+    },
+    {
+      "epoch": 0.606141425791105,
+      "grad_norm": 0.11893542805426792,
+      "learning_rate": 7.091568264588952e-05,
+      "loss": 1.2622,
+      "step": 8409
+    },
+    {
+      "epoch": 0.6062135082534419,
+      "grad_norm": 0.11195718175784794,
+      "learning_rate": 7.089334556773046e-05,
+      "loss": 1.334,
+      "step": 8410
+    },
+    {
+      "epoch": 0.6062855907157788,
+      "grad_norm": 0.11784017683922152,
+      "learning_rate": 7.087101007614443e-05,
+      "loss": 1.3242,
+      "step": 8411
+    },
+    {
+      "epoch": 0.6063576731781157,
+      "grad_norm": 0.11950427131281464,
+      "learning_rate": 7.084867617234898e-05,
+      "loss": 1.31,
+      "step": 8412
+    },
+    {
+      "epoch": 0.6064297556404527,
+      "grad_norm": 0.12314380567407489,
+      "learning_rate": 7.08263438575615e-05,
+      "loss": 1.2483,
+      "step": 8413
+    },
+    {
+      "epoch": 0.6065018381027896,
+      "grad_norm": 0.20842722924976304,
+      "learning_rate": 7.080401313299925e-05,
+      "loss": 1.2352,
+      "step": 8414
+    },
+    {
+      "epoch": 0.6065739205651265,
+      "grad_norm": 0.12514292515641076,
+      "learning_rate": 7.07816839998795e-05,
+      "loss": 1.2635,
+      "step": 8415
+    },
+    {
+      "epoch": 0.6066460030274634,
+      "grad_norm": 0.11787351727482988,
+      "learning_rate": 7.075935645941932e-05,
+      "loss": 1.2283,
+      "step": 8416
+    },
+    {
+      "epoch": 0.6067180854898003,
+      "grad_norm": 0.19342942727804663,
+      "learning_rate": 7.073703051283584e-05,
+      "loss": 1.3212,
+      "step": 8417
+    },
+    {
+      "epoch": 0.6067901679521372,
+      "grad_norm": 0.1168673522437923,
+      "learning_rate": 7.071470616134599e-05,
+      "loss": 1.3034,
+      "step": 8418
+    },
+    {
+      "epoch": 0.6068622504144742,
+      "grad_norm": 0.1316971028909596,
+      "learning_rate": 7.069238340616666e-05,
+      "loss": 1.3323,
+      "step": 8419
+    },
+    {
+      "epoch": 0.6069343328768111,
+      "grad_norm": 0.1101595075028963,
+      "learning_rate": 7.067006224851456e-05,
+      "loss": 1.2372,
+      "step": 8420
+    },
+    {
+      "epoch": 0.607006415339148,
+      "grad_norm": 0.24724058791385164,
+      "learning_rate": 7.064774268960653e-05,
+      "loss": 1.2254,
+      "step": 8421
+    },
+    {
+      "epoch": 0.6070784978014849,
+      "grad_norm": 0.11284843704059064,
+      "learning_rate": 7.06254247306591e-05,
+      "loss": 1.2848,
+      "step": 8422
+    },
+    {
+      "epoch": 0.6071505802638218,
+      "grad_norm": 0.15176953094149445,
+      "learning_rate": 7.060310837288884e-05,
+      "loss": 1.3014,
+      "step": 8423
+    },
+    {
+      "epoch": 0.6072226627261588,
+      "grad_norm": 0.139162177856279,
+      "learning_rate": 7.058079361751213e-05,
+      "loss": 1.3785,
+      "step": 8424
+    },
+    {
+      "epoch": 0.6072947451884957,
+      "grad_norm": 0.10960506444337395,
+      "learning_rate": 7.055848046574534e-05,
+      "loss": 1.3764,
+      "step": 8425
+    },
+    {
+      "epoch": 0.6073668276508325,
+      "grad_norm": 0.11549363841906608,
+      "learning_rate": 7.05361689188048e-05,
+      "loss": 1.3266,
+      "step": 8426
+    },
+    {
+      "epoch": 0.6074389101131694,
+      "grad_norm": 0.11936066370958637,
+      "learning_rate": 7.051385897790663e-05,
+      "loss": 1.2584,
+      "step": 8427
+    },
+    {
+      "epoch": 0.6075109925755063,
+      "grad_norm": 0.1264774914966905,
+      "learning_rate": 7.049155064426693e-05,
+      "loss": 1.1786,
+      "step": 8428
+    },
+    {
+      "epoch": 0.6075830750378433,
+      "grad_norm": 0.11518991973086787,
+      "learning_rate": 7.04692439191017e-05,
+      "loss": 1.2551,
+      "step": 8429
+    },
+    {
+      "epoch": 0.6076551575001802,
+      "grad_norm": 0.11490319098443469,
+      "learning_rate": 7.044693880362687e-05,
+      "loss": 1.3477,
+      "step": 8430
+    },
+    {
+      "epoch": 0.6077272399625171,
+      "grad_norm": 0.10405137548972827,
+      "learning_rate": 7.042463529905829e-05,
+      "loss": 1.2694,
+      "step": 8431
+    },
+    {
+      "epoch": 0.607799322424854,
+      "grad_norm": 0.13630083352609065,
+      "learning_rate": 7.040233340661167e-05,
+      "loss": 1.3209,
+      "step": 8432
+    },
+    {
+      "epoch": 0.6078714048871909,
+      "grad_norm": 0.11822042045990742,
+      "learning_rate": 7.038003312750263e-05,
+      "loss": 1.2285,
+      "step": 8433
+    },
+    {
+      "epoch": 0.6079434873495279,
+      "grad_norm": 0.11307119915897951,
+      "learning_rate": 7.03577344629468e-05,
+      "loss": 1.2767,
+      "step": 8434
+    },
+    {
+      "epoch": 0.6080155698118648,
+      "grad_norm": 0.11769942459505027,
+      "learning_rate": 7.033543741415964e-05,
+      "loss": 1.2341,
+      "step": 8435
+    },
+    {
+      "epoch": 0.6080876522742017,
+      "grad_norm": 0.11814872373860442,
+      "learning_rate": 7.031314198235652e-05,
+      "loss": 1.3136,
+      "step": 8436
+    },
+    {
+      "epoch": 0.6081597347365386,
+      "grad_norm": 0.10166108810110572,
+      "learning_rate": 7.029084816875275e-05,
+      "loss": 1.2467,
+      "step": 8437
+    },
+    {
+      "epoch": 0.6082318171988755,
+      "grad_norm": 0.11379942000249821,
+      "learning_rate": 7.02685559745635e-05,
+      "loss": 1.1824,
+      "step": 8438
+    },
+    {
+      "epoch": 0.6083038996612125,
+      "grad_norm": 0.11759659696246466,
+      "learning_rate": 7.024626540100401e-05,
+      "loss": 1.3225,
+      "step": 8439
+    },
+    {
+      "epoch": 0.6083759821235494,
+      "grad_norm": 0.1275538464148093,
+      "learning_rate": 7.022397644928919e-05,
+      "loss": 1.2251,
+      "step": 8440
+    },
+    {
+      "epoch": 0.6084480645858863,
+      "grad_norm": 0.11849657445681962,
+      "learning_rate": 7.020168912063405e-05,
+      "loss": 1.2951,
+      "step": 8441
+    },
+    {
+      "epoch": 0.6085201470482232,
+      "grad_norm": 0.12137132409622599,
+      "learning_rate": 7.01794034162534e-05,
+      "loss": 1.2878,
+      "step": 8442
+    },
+    {
+      "epoch": 0.6085922295105601,
+      "grad_norm": 0.12335759194698125,
+      "learning_rate": 7.015711933736211e-05,
+      "loss": 1.2996,
+      "step": 8443
+    },
+    {
+      "epoch": 0.6086643119728969,
+      "grad_norm": 0.11949005769803143,
+      "learning_rate": 7.013483688517476e-05,
+      "loss": 1.2788,
+      "step": 8444
+    },
+    {
+      "epoch": 0.6087363944352339,
+      "grad_norm": 0.12583165138919733,
+      "learning_rate": 7.011255606090599e-05,
+      "loss": 1.3152,
+      "step": 8445
+    },
+    {
+      "epoch": 0.6088084768975708,
+      "grad_norm": 0.11428231860242095,
+      "learning_rate": 7.009027686577032e-05,
+      "loss": 1.1576,
+      "step": 8446
+    },
+    {
+      "epoch": 0.6088805593599077,
+      "grad_norm": 0.11055284666342945,
+      "learning_rate": 7.00679993009821e-05,
+      "loss": 1.2114,
+      "step": 8447
+    },
+    {
+      "epoch": 0.6089526418222446,
+      "grad_norm": 0.18237008839216223,
+      "learning_rate": 7.00457233677557e-05,
+      "loss": 1.391,
+      "step": 8448
+    },
+    {
+      "epoch": 0.6090247242845815,
+      "grad_norm": 0.10840961150409714,
+      "learning_rate": 7.002344906730536e-05,
+      "loss": 1.1563,
+      "step": 8449
+    },
+    {
+      "epoch": 0.6090968067469185,
+      "grad_norm": 0.10926157054743751,
+      "learning_rate": 7.000117640084526e-05,
+      "loss": 1.2382,
+      "step": 8450
+    },
+    {
+      "epoch": 0.6091688892092554,
+      "grad_norm": 0.12376179656390705,
+      "learning_rate": 6.997890536958937e-05,
+      "loss": 1.3061,
+      "step": 8451
+    },
+    {
+      "epoch": 0.6092409716715923,
+      "grad_norm": 0.11278858581681299,
+      "learning_rate": 6.995663597475174e-05,
+      "loss": 1.2476,
+      "step": 8452
+    },
+    {
+      "epoch": 0.6093130541339292,
+      "grad_norm": 0.11756629291798135,
+      "learning_rate": 6.993436821754621e-05,
+      "loss": 1.2997,
+      "step": 8453
+    },
+    {
+      "epoch": 0.6093851365962661,
+      "grad_norm": 0.11071767769376296,
+      "learning_rate": 6.991210209918663e-05,
+      "loss": 1.1871,
+      "step": 8454
+    },
+    {
+      "epoch": 0.6094572190586031,
+      "grad_norm": 0.12437237753884096,
+      "learning_rate": 6.988983762088661e-05,
+      "loss": 1.2481,
+      "step": 8455
+    },
+    {
+      "epoch": 0.60952930152094,
+      "grad_norm": 0.11182329857000806,
+      "learning_rate": 6.986757478385983e-05,
+      "loss": 1.3333,
+      "step": 8456
+    },
+    {
+      "epoch": 0.6096013839832769,
+      "grad_norm": 0.14068816852950247,
+      "learning_rate": 6.98453135893198e-05,
+      "loss": 1.3091,
+      "step": 8457
+    },
+    {
+      "epoch": 0.6096734664456138,
+      "grad_norm": 0.12689995043284027,
+      "learning_rate": 6.982305403847997e-05,
+      "loss": 1.304,
+      "step": 8458
+    },
+    {
+      "epoch": 0.6097455489079507,
+      "grad_norm": 0.1555624142239978,
+      "learning_rate": 6.980079613255364e-05,
+      "loss": 1.3205,
+      "step": 8459
+    },
+    {
+      "epoch": 0.6098176313702877,
+      "grad_norm": 0.1332240109737626,
+      "learning_rate": 6.977853987275408e-05,
+      "loss": 1.2696,
+      "step": 8460
+    },
+    {
+      "epoch": 0.6098897138326245,
+      "grad_norm": 0.11148694757437999,
+      "learning_rate": 6.975628526029447e-05,
+      "loss": 1.1986,
+      "step": 8461
+    },
+    {
+      "epoch": 0.6099617962949614,
+      "grad_norm": 0.11717042493614588,
+      "learning_rate": 6.973403229638793e-05,
+      "loss": 1.2353,
+      "step": 8462
+    },
+    {
+      "epoch": 0.6100338787572983,
+      "grad_norm": 0.11868615948376936,
+      "learning_rate": 6.971178098224735e-05,
+      "loss": 1.3005,
+      "step": 8463
+    },
+    {
+      "epoch": 0.6101059612196352,
+      "grad_norm": 0.11845416196317912,
+      "learning_rate": 6.968953131908567e-05,
+      "loss": 1.3273,
+      "step": 8464
+    },
+    {
+      "epoch": 0.6101780436819721,
+      "grad_norm": 0.10991321125683584,
+      "learning_rate": 6.96672833081157e-05,
+      "loss": 1.2139,
+      "step": 8465
+    },
+    {
+      "epoch": 0.6102501261443091,
+      "grad_norm": 0.11624056915873937,
+      "learning_rate": 6.964503695055019e-05,
+      "loss": 1.2264,
+      "step": 8466
+    },
+    {
+      "epoch": 0.610322208606646,
+      "grad_norm": 0.11621687069210461,
+      "learning_rate": 6.962279224760172e-05,
+      "loss": 1.3247,
+      "step": 8467
+    },
+    {
+      "epoch": 0.6103942910689829,
+      "grad_norm": 0.1106016476710265,
+      "learning_rate": 6.960054920048282e-05,
+      "loss": 1.2319,
+      "step": 8468
+    },
+    {
+      "epoch": 0.6104663735313198,
+      "grad_norm": 0.1182424872602721,
+      "learning_rate": 6.957830781040591e-05,
+      "loss": 1.2675,
+      "step": 8469
+    },
+    {
+      "epoch": 0.6105384559936567,
+      "grad_norm": 0.3336402975115834,
+      "learning_rate": 6.955606807858343e-05,
+      "loss": 1.2735,
+      "step": 8470
+    },
+    {
+      "epoch": 0.6106105384559937,
+      "grad_norm": 0.11647282114306672,
+      "learning_rate": 6.953383000622761e-05,
+      "loss": 1.2315,
+      "step": 8471
+    },
+    {
+      "epoch": 0.6106826209183306,
+      "grad_norm": 0.11110821675150498,
+      "learning_rate": 6.951159359455062e-05,
+      "loss": 1.3485,
+      "step": 8472
+    },
+    {
+      "epoch": 0.6107547033806675,
+      "grad_norm": 0.12623966474963905,
+      "learning_rate": 6.948935884476446e-05,
+      "loss": 1.3224,
+      "step": 8473
+    },
+    {
+      "epoch": 0.6108267858430044,
+      "grad_norm": 0.12045389887968343,
+      "learning_rate": 6.946712575808128e-05,
+      "loss": 1.2942,
+      "step": 8474
+    },
+    {
+      "epoch": 0.6108988683053413,
+      "grad_norm": 0.11145270387490491,
+      "learning_rate": 6.944489433571286e-05,
+      "loss": 1.3327,
+      "step": 8475
+    },
+    {
+      "epoch": 0.6109709507676783,
+      "grad_norm": 0.11894343761305737,
+      "learning_rate": 6.942266457887109e-05,
+      "loss": 1.2911,
+      "step": 8476
+    },
+    {
+      "epoch": 0.6110430332300152,
+      "grad_norm": 0.10643417331093523,
+      "learning_rate": 6.940043648876758e-05,
+      "loss": 1.2356,
+      "step": 8477
+    },
+    {
+      "epoch": 0.611115115692352,
+      "grad_norm": 0.11984029852781164,
+      "learning_rate": 6.93782100666141e-05,
+      "loss": 1.3001,
+      "step": 8478
+    },
+    {
+      "epoch": 0.6111871981546889,
+      "grad_norm": 0.11825124441330519,
+      "learning_rate": 6.93559853136221e-05,
+      "loss": 1.2902,
+      "step": 8479
+    },
+    {
+      "epoch": 0.6112592806170258,
+      "grad_norm": 0.26178045905420944,
+      "learning_rate": 6.933376223100307e-05,
+      "loss": 1.3538,
+      "step": 8480
+    },
+    {
+      "epoch": 0.6113313630793628,
+      "grad_norm": 0.11994094807733932,
+      "learning_rate": 6.931154081996833e-05,
+      "loss": 1.3107,
+      "step": 8481
+    },
+    {
+      "epoch": 0.6114034455416997,
+      "grad_norm": 0.12618141653053935,
+      "learning_rate": 6.928932108172914e-05,
+      "loss": 1.2585,
+      "step": 8482
+    },
+    {
+      "epoch": 0.6114755280040366,
+      "grad_norm": 0.1221385978227431,
+      "learning_rate": 6.926710301749673e-05,
+      "loss": 1.3415,
+      "step": 8483
+    },
+    {
+      "epoch": 0.6115476104663735,
+      "grad_norm": 0.11354136727565749,
+      "learning_rate": 6.924488662848214e-05,
+      "loss": 1.2809,
+      "step": 8484
+    },
+    {
+      "epoch": 0.6116196929287104,
+      "grad_norm": 0.11241439008657464,
+      "learning_rate": 6.922267191589637e-05,
+      "loss": 1.2359,
+      "step": 8485
+    },
+    {
+      "epoch": 0.6116917753910474,
+      "grad_norm": 0.12855830171793817,
+      "learning_rate": 6.92004588809503e-05,
+      "loss": 1.2354,
+      "step": 8486
+    },
+    {
+      "epoch": 0.6117638578533843,
+      "grad_norm": 0.1168831871851611,
+      "learning_rate": 6.917824752485477e-05,
+      "loss": 1.2849,
+      "step": 8487
+    },
+    {
+      "epoch": 0.6118359403157212,
+      "grad_norm": 0.11884718391630618,
+      "learning_rate": 6.915603784882054e-05,
+      "loss": 1.3306,
+      "step": 8488
+    },
+    {
+      "epoch": 0.6119080227780581,
+      "grad_norm": 0.12207942198014675,
+      "learning_rate": 6.913382985405814e-05,
+      "loss": 1.3002,
+      "step": 8489
+    },
+    {
+      "epoch": 0.611980105240395,
+      "grad_norm": 0.11327304209725543,
+      "learning_rate": 6.911162354177818e-05,
+      "loss": 1.2672,
+      "step": 8490
+    },
+    {
+      "epoch": 0.612052187702732,
+      "grad_norm": 0.12864107768084002,
+      "learning_rate": 6.9089418913191e-05,
+      "loss": 1.2628,
+      "step": 8491
+    },
+    {
+      "epoch": 0.6121242701650689,
+      "grad_norm": 0.12147035125345798,
+      "learning_rate": 6.90672159695071e-05,
+      "loss": 1.3931,
+      "step": 8492
+    },
+    {
+      "epoch": 0.6121963526274058,
+      "grad_norm": 0.16588859479267215,
+      "learning_rate": 6.904501471193663e-05,
+      "loss": 1.3104,
+      "step": 8493
+    },
+    {
+      "epoch": 0.6122684350897427,
+      "grad_norm": 0.11734674542836841,
+      "learning_rate": 6.902281514168983e-05,
+      "loss": 1.282,
+      "step": 8494
+    },
+    {
+      "epoch": 0.6123405175520796,
+      "grad_norm": 0.11050334846149973,
+      "learning_rate": 6.900061725997664e-05,
+      "loss": 1.2168,
+      "step": 8495
+    },
+    {
+      "epoch": 0.6124126000144164,
+      "grad_norm": 0.11061155371015077,
+      "learning_rate": 6.897842106800722e-05,
+      "loss": 1.2406,
+      "step": 8496
+    },
+    {
+      "epoch": 0.6124846824767534,
+      "grad_norm": 0.13244053224321278,
+      "learning_rate": 6.895622656699135e-05,
+      "loss": 1.2327,
+      "step": 8497
+    },
+    {
+      "epoch": 0.6125567649390903,
+      "grad_norm": 0.1779934491538715,
+      "learning_rate": 6.893403375813888e-05,
+      "loss": 1.3163,
+      "step": 8498
+    },
+    {
+      "epoch": 0.6126288474014272,
+      "grad_norm": 0.12110741170425642,
+      "learning_rate": 6.891184264265944e-05,
+      "loss": 1.1592,
+      "step": 8499
+    },
+    {
+      "epoch": 0.6127009298637641,
+      "grad_norm": 0.11336958758543554,
+      "learning_rate": 6.888965322176276e-05,
+      "loss": 1.3737,
+      "step": 8500
+    },
+    {
+      "epoch": 0.612773012326101,
+      "grad_norm": 0.1196251556238169,
+      "learning_rate": 6.886746549665827e-05,
+      "loss": 1.2323,
+      "step": 8501
+    },
+    {
+      "epoch": 0.612845094788438,
+      "grad_norm": 0.11925316218730321,
+      "learning_rate": 6.884527946855546e-05,
+      "loss": 1.2234,
+      "step": 8502
+    },
+    {
+      "epoch": 0.6129171772507749,
+      "grad_norm": 0.11934552747300554,
+      "learning_rate": 6.88230951386636e-05,
+      "loss": 1.2315,
+      "step": 8503
+    },
+    {
+      "epoch": 0.6129892597131118,
+      "grad_norm": 0.12118798525198435,
+      "learning_rate": 6.880091250819197e-05,
+      "loss": 1.2022,
+      "step": 8504
+    },
+    {
+      "epoch": 0.6130613421754487,
+      "grad_norm": 0.12362248350631047,
+      "learning_rate": 6.877873157834974e-05,
+      "loss": 1.2313,
+      "step": 8505
+    },
+    {
+      "epoch": 0.6131334246377856,
+      "grad_norm": 0.12040213949612334,
+      "learning_rate": 6.875655235034597e-05,
+      "loss": 1.3,
+      "step": 8506
+    },
+    {
+      "epoch": 0.6132055071001226,
+      "grad_norm": 0.1250598802587639,
+      "learning_rate": 6.873437482538958e-05,
+      "loss": 1.224,
+      "step": 8507
+    },
+    {
+      "epoch": 0.6132775895624595,
+      "grad_norm": 0.10859021646612198,
+      "learning_rate": 6.871219900468947e-05,
+      "loss": 1.3268,
+      "step": 8508
+    },
+    {
+      "epoch": 0.6133496720247964,
+      "grad_norm": 0.12685088784866078,
+      "learning_rate": 6.869002488945442e-05,
+      "loss": 1.2179,
+      "step": 8509
+    },
+    {
+      "epoch": 0.6134217544871333,
+      "grad_norm": 0.11574468482416811,
+      "learning_rate": 6.866785248089315e-05,
+      "loss": 1.3743,
+      "step": 8510
+    },
+    {
+      "epoch": 0.6134938369494702,
+      "grad_norm": 0.1430974018190425,
+      "learning_rate": 6.864568178021423e-05,
+      "loss": 1.2851,
+      "step": 8511
+    },
+    {
+      "epoch": 0.6135659194118072,
+      "grad_norm": 0.11422317168544152,
+      "learning_rate": 6.862351278862613e-05,
+      "loss": 1.261,
+      "step": 8512
+    },
+    {
+      "epoch": 0.613638001874144,
+      "grad_norm": 0.12282103299132868,
+      "learning_rate": 6.860134550733726e-05,
+      "loss": 1.2801,
+      "step": 8513
+    },
+    {
+      "epoch": 0.6137100843364809,
+      "grad_norm": 0.11421632296467625,
+      "learning_rate": 6.857917993755602e-05,
+      "loss": 1.3302,
+      "step": 8514
+    },
+    {
+      "epoch": 0.6137821667988178,
+      "grad_norm": 0.13990581659366272,
+      "learning_rate": 6.855701608049054e-05,
+      "loss": 1.2523,
+      "step": 8515
+    },
+    {
+      "epoch": 0.6138542492611547,
+      "grad_norm": 0.11671373932342631,
+      "learning_rate": 6.853485393734899e-05,
+      "loss": 1.3431,
+      "step": 8516
+    },
+    {
+      "epoch": 0.6139263317234916,
+      "grad_norm": 0.12346906043970932,
+      "learning_rate": 6.851269350933936e-05,
+      "loss": 1.3494,
+      "step": 8517
+    },
+    {
+      "epoch": 0.6139984141858286,
+      "grad_norm": 0.11657014444327221,
+      "learning_rate": 6.849053479766968e-05,
+      "loss": 1.3855,
+      "step": 8518
+    },
+    {
+      "epoch": 0.6140704966481655,
+      "grad_norm": 0.16129472070579318,
+      "learning_rate": 6.846837780354772e-05,
+      "loss": 1.2497,
+      "step": 8519
+    },
+    {
+      "epoch": 0.6141425791105024,
+      "grad_norm": 0.11313462144161457,
+      "learning_rate": 6.844622252818127e-05,
+      "loss": 1.2888,
+      "step": 8520
+    },
+    {
+      "epoch": 0.6142146615728393,
+      "grad_norm": 0.11369788856223276,
+      "learning_rate": 6.842406897277795e-05,
+      "loss": 1.2856,
+      "step": 8521
+    },
+    {
+      "epoch": 0.6142867440351762,
+      "grad_norm": 0.12330738710462778,
+      "learning_rate": 6.84019171385454e-05,
+      "loss": 1.3051,
+      "step": 8522
+    },
+    {
+      "epoch": 0.6143588264975132,
+      "grad_norm": 0.12330262128129493,
+      "learning_rate": 6.837976702669103e-05,
+      "loss": 1.3398,
+      "step": 8523
+    },
+    {
+      "epoch": 0.6144309089598501,
+      "grad_norm": 0.11719203266167362,
+      "learning_rate": 6.835761863842224e-05,
+      "loss": 1.3358,
+      "step": 8524
+    },
+    {
+      "epoch": 0.614502991422187,
+      "grad_norm": 0.16439915257392668,
+      "learning_rate": 6.833547197494632e-05,
+      "loss": 1.3781,
+      "step": 8525
+    },
+    {
+      "epoch": 0.6145750738845239,
+      "grad_norm": 0.1123610637514884,
+      "learning_rate": 6.831332703747044e-05,
+      "loss": 1.284,
+      "step": 8526
+    },
+    {
+      "epoch": 0.6146471563468608,
+      "grad_norm": 0.11343427246802408,
+      "learning_rate": 6.829118382720173e-05,
+      "loss": 1.1766,
+      "step": 8527
+    },
+    {
+      "epoch": 0.6147192388091978,
+      "grad_norm": 0.11320992976180723,
+      "learning_rate": 6.826904234534717e-05,
+      "loss": 1.2317,
+      "step": 8528
+    },
+    {
+      "epoch": 0.6147913212715347,
+      "grad_norm": 0.1231897011332532,
+      "learning_rate": 6.824690259311369e-05,
+      "loss": 1.3244,
+      "step": 8529
+    },
+    {
+      "epoch": 0.6148634037338715,
+      "grad_norm": 0.12517979630787496,
+      "learning_rate": 6.822476457170806e-05,
+      "loss": 1.2601,
+      "step": 8530
+    },
+    {
+      "epoch": 0.6149354861962084,
+      "grad_norm": 0.12198528151985676,
+      "learning_rate": 6.820262828233704e-05,
+      "loss": 1.2442,
+      "step": 8531
+    },
+    {
+      "epoch": 0.6150075686585453,
+      "grad_norm": 0.11721781267815878,
+      "learning_rate": 6.818049372620726e-05,
+      "loss": 1.2817,
+      "step": 8532
+    },
+    {
+      "epoch": 0.6150796511208823,
+      "grad_norm": 0.1262414327482851,
+      "learning_rate": 6.815836090452525e-05,
+      "loss": 1.2676,
+      "step": 8533
+    },
+    {
+      "epoch": 0.6151517335832192,
+      "grad_norm": 0.1305678502377985,
+      "learning_rate": 6.813622981849742e-05,
+      "loss": 1.3332,
+      "step": 8534
+    },
+    {
+      "epoch": 0.6152238160455561,
+      "grad_norm": 0.12013946694684281,
+      "learning_rate": 6.811410046933009e-05,
+      "loss": 1.2975,
+      "step": 8535
+    },
+    {
+      "epoch": 0.615295898507893,
+      "grad_norm": 0.12822470609560682,
+      "learning_rate": 6.809197285822959e-05,
+      "loss": 1.2345,
+      "step": 8536
+    },
+    {
+      "epoch": 0.6153679809702299,
+      "grad_norm": 0.12627424161055764,
+      "learning_rate": 6.806984698640202e-05,
+      "loss": 1.2901,
+      "step": 8537
+    },
+    {
+      "epoch": 0.6154400634325669,
+      "grad_norm": 0.12687934491018926,
+      "learning_rate": 6.804772285505344e-05,
+      "loss": 1.2528,
+      "step": 8538
+    },
+    {
+      "epoch": 0.6155121458949038,
+      "grad_norm": 0.15000990799458344,
+      "learning_rate": 6.80256004653898e-05,
+      "loss": 1.2944,
+      "step": 8539
+    },
+    {
+      "epoch": 0.6155842283572407,
+      "grad_norm": 0.11551094625381755,
+      "learning_rate": 6.800347981861699e-05,
+      "loss": 1.2771,
+      "step": 8540
+    },
+    {
+      "epoch": 0.6156563108195776,
+      "grad_norm": 0.11619964426802025,
+      "learning_rate": 6.798136091594081e-05,
+      "loss": 1.2856,
+      "step": 8541
+    },
+    {
+      "epoch": 0.6157283932819145,
+      "grad_norm": 0.11866578974929223,
+      "learning_rate": 6.795924375856688e-05,
+      "loss": 1.2497,
+      "step": 8542
+    },
+    {
+      "epoch": 0.6158004757442515,
+      "grad_norm": 0.11348375033992501,
+      "learning_rate": 6.793712834770079e-05,
+      "loss": 1.2904,
+      "step": 8543
+    },
+    {
+      "epoch": 0.6158725582065884,
+      "grad_norm": 0.11613515065336942,
+      "learning_rate": 6.791501468454808e-05,
+      "loss": 1.3374,
+      "step": 8544
+    },
+    {
+      "epoch": 0.6159446406689253,
+      "grad_norm": 0.11526014647362491,
+      "learning_rate": 6.789290277031413e-05,
+      "loss": 1.2862,
+      "step": 8545
+    },
+    {
+      "epoch": 0.6160167231312622,
+      "grad_norm": 0.12844985200907055,
+      "learning_rate": 6.787079260620418e-05,
+      "loss": 1.2728,
+      "step": 8546
+    },
+    {
+      "epoch": 0.6160888055935991,
+      "grad_norm": 0.11871915224535234,
+      "learning_rate": 6.784868419342349e-05,
+      "loss": 1.2788,
+      "step": 8547
+    },
+    {
+      "epoch": 0.6161608880559359,
+      "grad_norm": 0.11313239361460636,
+      "learning_rate": 6.782657753317711e-05,
+      "loss": 1.3515,
+      "step": 8548
+    },
+    {
+      "epoch": 0.6162329705182729,
+      "grad_norm": 0.11792318286385144,
+      "learning_rate": 6.780447262667012e-05,
+      "loss": 1.1878,
+      "step": 8549
+    },
+    {
+      "epoch": 0.6163050529806098,
+      "grad_norm": 0.12070392503603086,
+      "learning_rate": 6.778236947510739e-05,
+      "loss": 1.3326,
+      "step": 8550
+    },
+    {
+      "epoch": 0.6163771354429467,
+      "grad_norm": 0.11634497637071402,
+      "learning_rate": 6.776026807969378e-05,
+      "loss": 1.306,
+      "step": 8551
+    },
+    {
+      "epoch": 0.6164492179052836,
+      "grad_norm": 0.12036787045563055,
+      "learning_rate": 6.773816844163393e-05,
+      "loss": 1.3794,
+      "step": 8552
+    },
+    {
+      "epoch": 0.6165213003676205,
+      "grad_norm": 0.14265140848567537,
+      "learning_rate": 6.771607056213257e-05,
+      "loss": 1.2714,
+      "step": 8553
+    },
+    {
+      "epoch": 0.6165933828299575,
+      "grad_norm": 0.11027758780154857,
+      "learning_rate": 6.769397444239416e-05,
+      "loss": 1.3098,
+      "step": 8554
+    },
+    {
+      "epoch": 0.6166654652922944,
+      "grad_norm": 0.11967536089862835,
+      "learning_rate": 6.767188008362318e-05,
+      "loss": 1.1807,
+      "step": 8555
+    },
+    {
+      "epoch": 0.6167375477546313,
+      "grad_norm": 0.12493201574640804,
+      "learning_rate": 6.764978748702393e-05,
+      "loss": 1.1729,
+      "step": 8556
+    },
+    {
+      "epoch": 0.6168096302169682,
+      "grad_norm": 0.12015890204292067,
+      "learning_rate": 6.762769665380067e-05,
+      "loss": 1.3141,
+      "step": 8557
+    },
+    {
+      "epoch": 0.6168817126793051,
+      "grad_norm": 0.11587203089594132,
+      "learning_rate": 6.760560758515756e-05,
+      "loss": 1.2583,
+      "step": 8558
+    },
+    {
+      "epoch": 0.6169537951416421,
+      "grad_norm": 0.11082252244563609,
+      "learning_rate": 6.758352028229867e-05,
+      "loss": 1.2428,
+      "step": 8559
+    },
+    {
+      "epoch": 0.617025877603979,
+      "grad_norm": 0.11487838398204735,
+      "learning_rate": 6.75614347464279e-05,
+      "loss": 1.2361,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6170979600663159,
+      "grad_norm": 0.11855679736059713,
+      "learning_rate": 6.753935097874913e-05,
+      "loss": 1.2786,
+      "step": 8561
+    },
+    {
+      "epoch": 0.6171700425286528,
+      "grad_norm": 0.12680215007627607,
+      "learning_rate": 6.751726898046617e-05,
+      "loss": 1.1717,
+      "step": 8562
+    },
+    {
+      "epoch": 0.6172421249909897,
+      "grad_norm": 0.12116209685852491,
+      "learning_rate": 6.749518875278264e-05,
+      "loss": 1.1702,
+      "step": 8563
+    },
+    {
+      "epoch": 0.6173142074533267,
+      "grad_norm": 0.11263324474083161,
+      "learning_rate": 6.747311029690211e-05,
+      "loss": 1.2413,
+      "step": 8564
+    },
+    {
+      "epoch": 0.6173862899156635,
+      "grad_norm": 0.1216454083453875,
+      "learning_rate": 6.74510336140281e-05,
+      "loss": 1.2083,
+      "step": 8565
+    },
+    {
+      "epoch": 0.6174583723780004,
+      "grad_norm": 0.12506451702182858,
+      "learning_rate": 6.742895870536388e-05,
+      "loss": 1.3656,
+      "step": 8566
+    },
+    {
+      "epoch": 0.6175304548403373,
+      "grad_norm": 0.11648178689245332,
+      "learning_rate": 6.740688557211284e-05,
+      "loss": 1.288,
+      "step": 8567
+    },
+    {
+      "epoch": 0.6176025373026742,
+      "grad_norm": 0.19880212688407786,
+      "learning_rate": 6.738481421547812e-05,
+      "loss": 1.2983,
+      "step": 8568
+    },
+    {
+      "epoch": 0.6176746197650111,
+      "grad_norm": 0.11642462888878652,
+      "learning_rate": 6.736274463666283e-05,
+      "loss": 1.323,
+      "step": 8569
+    },
+    {
+      "epoch": 0.6177467022273481,
+      "grad_norm": 0.11808296694795067,
+      "learning_rate": 6.734067683686986e-05,
+      "loss": 1.3172,
+      "step": 8570
+    },
+    {
+      "epoch": 0.617818784689685,
+      "grad_norm": 0.11872117155645927,
+      "learning_rate": 6.731861081730226e-05,
+      "loss": 1.2943,
+      "step": 8571
+    },
+    {
+      "epoch": 0.6178908671520219,
+      "grad_norm": 0.12241783188479098,
+      "learning_rate": 6.729654657916272e-05,
+      "loss": 1.3936,
+      "step": 8572
+    },
+    {
+      "epoch": 0.6179629496143588,
+      "grad_norm": 0.1149837269005374,
+      "learning_rate": 6.727448412365399e-05,
+      "loss": 1.26,
+      "step": 8573
+    },
+    {
+      "epoch": 0.6180350320766957,
+      "grad_norm": 0.12296385026500377,
+      "learning_rate": 6.725242345197859e-05,
+      "loss": 1.2774,
+      "step": 8574
+    },
+    {
+      "epoch": 0.6181071145390327,
+      "grad_norm": 0.11875728258245795,
+      "learning_rate": 6.723036456533913e-05,
+      "loss": 1.2892,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6181791970013696,
+      "grad_norm": 0.11202325898360153,
+      "learning_rate": 6.720830746493795e-05,
+      "loss": 1.231,
+      "step": 8576
+    },
+    {
+      "epoch": 0.6182512794637065,
+      "grad_norm": 0.11326272941112284,
+      "learning_rate": 6.718625215197739e-05,
+      "loss": 1.1974,
+      "step": 8577
+    },
+    {
+      "epoch": 0.6183233619260434,
+      "grad_norm": 0.1096932706056103,
+      "learning_rate": 6.716419862765963e-05,
+      "loss": 1.3566,
+      "step": 8578
+    },
+    {
+      "epoch": 0.6183954443883803,
+      "grad_norm": 0.11588920309931475,
+      "learning_rate": 6.714214689318678e-05,
+      "loss": 1.2926,
+      "step": 8579
+    },
+    {
+      "epoch": 0.6184675268507173,
+      "grad_norm": 0.10711824540705533,
+      "learning_rate": 6.712009694976092e-05,
+      "loss": 1.2921,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6185396093130542,
+      "grad_norm": 0.11647301873640538,
+      "learning_rate": 6.709804879858392e-05,
+      "loss": 1.2317,
+      "step": 8581
+    },
+    {
+      "epoch": 0.618611691775391,
+      "grad_norm": 0.11598267177141988,
+      "learning_rate": 6.707600244085761e-05,
+      "loss": 1.3253,
+      "step": 8582
+    },
+    {
+      "epoch": 0.6186837742377279,
+      "grad_norm": 0.12803584220387834,
+      "learning_rate": 6.705395787778368e-05,
+      "loss": 1.34,
+      "step": 8583
+    },
+    {
+      "epoch": 0.6187558567000648,
+      "grad_norm": 0.1230130741984956,
+      "learning_rate": 6.703191511056381e-05,
+      "loss": 1.2873,
+      "step": 8584
+    },
+    {
+      "epoch": 0.6188279391624018,
+      "grad_norm": 0.11922246088966724,
+      "learning_rate": 6.700987414039955e-05,
+      "loss": 1.3248,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6189000216247387,
+      "grad_norm": 0.11639335589298588,
+      "learning_rate": 6.698783496849224e-05,
+      "loss": 1.2875,
+      "step": 8586
+    },
+    {
+      "epoch": 0.6189721040870756,
+      "grad_norm": 0.1126023477305488,
+      "learning_rate": 6.696579759604327e-05,
+      "loss": 1.2575,
+      "step": 8587
+    },
+    {
+      "epoch": 0.6190441865494125,
+      "grad_norm": 0.11969821769000669,
+      "learning_rate": 6.694376202425385e-05,
+      "loss": 1.2746,
+      "step": 8588
+    },
+    {
+      "epoch": 0.6191162690117494,
+      "grad_norm": 0.12352334525029236,
+      "learning_rate": 6.692172825432516e-05,
+      "loss": 1.3317,
+      "step": 8589
+    },
+    {
+      "epoch": 0.6191883514740864,
+      "grad_norm": 0.11125360933626308,
+      "learning_rate": 6.689969628745819e-05,
+      "loss": 1.3042,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6192604339364233,
+      "grad_norm": 0.10029446947456795,
+      "learning_rate": 6.687766612485389e-05,
+      "loss": 1.2373,
+      "step": 8591
+    },
+    {
+      "epoch": 0.6193325163987602,
+      "grad_norm": 0.11646180845584315,
+      "learning_rate": 6.68556377677131e-05,
+      "loss": 1.2704,
+      "step": 8592
+    },
+    {
+      "epoch": 0.6194045988610971,
+      "grad_norm": 0.11780330430338996,
+      "learning_rate": 6.68336112172366e-05,
+      "loss": 1.4209,
+      "step": 8593
+    },
+    {
+      "epoch": 0.619476681323434,
+      "grad_norm": 0.1352015285243312,
+      "learning_rate": 6.681158647462499e-05,
+      "loss": 1.1459,
+      "step": 8594
+    },
+    {
+      "epoch": 0.619548763785771,
+      "grad_norm": 0.11013008252193883,
+      "learning_rate": 6.678956354107882e-05,
+      "loss": 1.2756,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6196208462481079,
+      "grad_norm": 0.11283414873867682,
+      "learning_rate": 6.676754241779852e-05,
+      "loss": 1.3014,
+      "step": 8596
+    },
+    {
+      "epoch": 0.6196929287104448,
+      "grad_norm": 0.1239474315706601,
+      "learning_rate": 6.674552310598453e-05,
+      "loss": 1.3042,
+      "step": 8597
+    },
+    {
+      "epoch": 0.6197650111727817,
+      "grad_norm": 0.11553962505773153,
+      "learning_rate": 6.6723505606837e-05,
+      "loss": 1.3283,
+      "step": 8598
+    },
+    {
+      "epoch": 0.6198370936351186,
+      "grad_norm": 0.11517310204934449,
+      "learning_rate": 6.670148992155612e-05,
+      "loss": 1.2742,
+      "step": 8599
+    },
+    {
+      "epoch": 0.6199091760974554,
+      "grad_norm": 0.15275363624210728,
+      "learning_rate": 6.667947605134193e-05,
+      "loss": 1.2182,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6199812585597924,
+      "grad_norm": 0.10780145110037814,
+      "learning_rate": 6.665746399739438e-05,
+      "loss": 1.1907,
+      "step": 8601
+    },
+    {
+      "epoch": 0.6200533410221293,
+      "grad_norm": 0.12316103422310025,
+      "learning_rate": 6.663545376091334e-05,
+      "loss": 1.2845,
+      "step": 8602
+    },
+    {
+      "epoch": 0.6201254234844662,
+      "grad_norm": 0.11861283421142682,
+      "learning_rate": 6.661344534309856e-05,
+      "loss": 1.3815,
+      "step": 8603
+    },
+    {
+      "epoch": 0.6201975059468031,
+      "grad_norm": 0.10644485394377742,
+      "learning_rate": 6.659143874514971e-05,
+      "loss": 1.3355,
+      "step": 8604
+    },
+    {
+      "epoch": 0.62026958840914,
+      "grad_norm": 0.10396901101284545,
+      "learning_rate": 6.656943396826629e-05,
+      "loss": 1.2629,
+      "step": 8605
+    },
+    {
+      "epoch": 0.620341670871477,
+      "grad_norm": 0.13241148009067988,
+      "learning_rate": 6.654743101364783e-05,
+      "loss": 1.2913,
+      "step": 8606
+    },
+    {
+      "epoch": 0.6204137533338139,
+      "grad_norm": 0.12216403287282553,
+      "learning_rate": 6.652542988249363e-05,
+      "loss": 1.2896,
+      "step": 8607
+    },
+    {
+      "epoch": 0.6204858357961508,
+      "grad_norm": 0.11460976749718474,
+      "learning_rate": 6.650343057600302e-05,
+      "loss": 1.2714,
+      "step": 8608
+    },
+    {
+      "epoch": 0.6205579182584877,
+      "grad_norm": 0.12616993081310654,
+      "learning_rate": 6.648143309537508e-05,
+      "loss": 1.3507,
+      "step": 8609
+    },
+    {
+      "epoch": 0.6206300007208246,
+      "grad_norm": 0.13121110634041316,
+      "learning_rate": 6.645943744180889e-05,
+      "loss": 1.2712,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6207020831831616,
+      "grad_norm": 0.10748697149340032,
+      "learning_rate": 6.643744361650342e-05,
+      "loss": 1.3157,
+      "step": 8611
+    },
+    {
+      "epoch": 0.6207741656454985,
+      "grad_norm": 0.11284841064075649,
+      "learning_rate": 6.641545162065757e-05,
+      "loss": 1.3235,
+      "step": 8612
+    },
+    {
+      "epoch": 0.6208462481078354,
+      "grad_norm": 0.11948339532724486,
+      "learning_rate": 6.639346145547005e-05,
+      "loss": 1.1835,
+      "step": 8613
+    },
+    {
+      "epoch": 0.6209183305701723,
+      "grad_norm": 0.1182765282385424,
+      "learning_rate": 6.63714731221395e-05,
+      "loss": 1.2725,
+      "step": 8614
+    },
+    {
+      "epoch": 0.6209904130325092,
+      "grad_norm": 0.128541608832498,
+      "learning_rate": 6.634948662186453e-05,
+      "loss": 1.2072,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6210624954948462,
+      "grad_norm": 0.12342868818036977,
+      "learning_rate": 6.632750195584362e-05,
+      "loss": 1.3605,
+      "step": 8616
+    },
+    {
+      "epoch": 0.621134577957183,
+      "grad_norm": 0.12189205993376385,
+      "learning_rate": 6.630551912527507e-05,
+      "loss": 1.3253,
+      "step": 8617
+    },
+    {
+      "epoch": 0.6212066604195199,
+      "grad_norm": 0.11302571305735724,
+      "learning_rate": 6.628353813135714e-05,
+      "loss": 1.3014,
+      "step": 8618
+    },
+    {
+      "epoch": 0.6212787428818568,
+      "grad_norm": 0.14231860416971742,
+      "learning_rate": 6.626155897528805e-05,
+      "loss": 1.2437,
+      "step": 8619
+    },
+    {
+      "epoch": 0.6213508253441937,
+      "grad_norm": 0.11550952583747812,
+      "learning_rate": 6.623958165826584e-05,
+      "loss": 1.2409,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6214229078065306,
+      "grad_norm": 0.11273043498137361,
+      "learning_rate": 6.621760618148845e-05,
+      "loss": 1.2779,
+      "step": 8621
+    },
+    {
+      "epoch": 0.6214949902688676,
+      "grad_norm": 0.1218516748706754,
+      "learning_rate": 6.619563254615375e-05,
+      "loss": 1.2664,
+      "step": 8622
+    },
+    {
+      "epoch": 0.6215670727312045,
+      "grad_norm": 0.1146135935657861,
+      "learning_rate": 6.617366075345944e-05,
+      "loss": 1.3126,
+      "step": 8623
+    },
+    {
+      "epoch": 0.6216391551935414,
+      "grad_norm": 0.11746541482814801,
+      "learning_rate": 6.615169080460331e-05,
+      "loss": 1.267,
+      "step": 8624
+    },
+    {
+      "epoch": 0.6217112376558783,
+      "grad_norm": 0.12552156624309943,
+      "learning_rate": 6.612972270078282e-05,
+      "loss": 1.3071,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6217833201182152,
+      "grad_norm": 0.125290460064071,
+      "learning_rate": 6.610775644319549e-05,
+      "loss": 1.3164,
+      "step": 8626
+    },
+    {
+      "epoch": 0.6218554025805522,
+      "grad_norm": 0.11366484446408003,
+      "learning_rate": 6.608579203303858e-05,
+      "loss": 1.2308,
+      "step": 8627
+    },
+    {
+      "epoch": 0.6219274850428891,
+      "grad_norm": 0.11577223701597181,
+      "learning_rate": 6.606382947150944e-05,
+      "loss": 1.2762,
+      "step": 8628
+    },
+    {
+      "epoch": 0.621999567505226,
+      "grad_norm": 0.11989306383710079,
+      "learning_rate": 6.60418687598052e-05,
+      "loss": 1.2736,
+      "step": 8629
+    },
+    {
+      "epoch": 0.6220716499675629,
+      "grad_norm": 0.11965751689948133,
+      "learning_rate": 6.601990989912293e-05,
+      "loss": 1.3556,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6221437324298998,
+      "grad_norm": 0.12247993833255506,
+      "learning_rate": 6.599795289065954e-05,
+      "loss": 1.3915,
+      "step": 8631
+    },
+    {
+      "epoch": 0.6222158148922368,
+      "grad_norm": 0.12048152221395568,
+      "learning_rate": 6.59759977356119e-05,
+      "loss": 1.2346,
+      "step": 8632
+    },
+    {
+      "epoch": 0.6222878973545737,
+      "grad_norm": 0.1693105819303625,
+      "learning_rate": 6.595404443517678e-05,
+      "loss": 1.1508,
+      "step": 8633
+    },
+    {
+      "epoch": 0.6223599798169105,
+      "grad_norm": 0.11148647984500047,
+      "learning_rate": 6.593209299055083e-05,
+      "loss": 1.2621,
+      "step": 8634
+    },
+    {
+      "epoch": 0.6224320622792474,
+      "grad_norm": 0.13026849125249762,
+      "learning_rate": 6.591014340293059e-05,
+      "loss": 1.3249,
+      "step": 8635
+    },
+    {
+      "epoch": 0.6225041447415843,
+      "grad_norm": 0.1139855672730904,
+      "learning_rate": 6.588819567351248e-05,
+      "loss": 1.257,
+      "step": 8636
+    },
+    {
+      "epoch": 0.6225762272039213,
+      "grad_norm": 0.1347784082531949,
+      "learning_rate": 6.586624980349292e-05,
+      "loss": 1.3519,
+      "step": 8637
+    },
+    {
+      "epoch": 0.6226483096662582,
+      "grad_norm": 0.12589293383970554,
+      "learning_rate": 6.584430579406812e-05,
+      "loss": 1.3265,
+      "step": 8638
+    },
+    {
+      "epoch": 0.6227203921285951,
+      "grad_norm": 0.15811267833008275,
+      "learning_rate": 6.58223636464342e-05,
+      "loss": 1.1873,
+      "step": 8639
+    },
+    {
+      "epoch": 0.622792474590932,
+      "grad_norm": 0.13444894380417965,
+      "learning_rate": 6.58004233617872e-05,
+      "loss": 1.2192,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6228645570532689,
+      "grad_norm": 0.15623102714622944,
+      "learning_rate": 6.57784849413231e-05,
+      "loss": 1.3363,
+      "step": 8641
+    },
+    {
+      "epoch": 0.6229366395156059,
+      "grad_norm": 0.10975151228645748,
+      "learning_rate": 6.575654838623776e-05,
+      "loss": 1.2197,
+      "step": 8642
+    },
+    {
+      "epoch": 0.6230087219779428,
+      "grad_norm": 0.1389142486391399,
+      "learning_rate": 6.573461369772685e-05,
+      "loss": 1.3049,
+      "step": 8643
+    },
+    {
+      "epoch": 0.6230808044402797,
+      "grad_norm": 0.11450635522563735,
+      "learning_rate": 6.571268087698608e-05,
+      "loss": 1.3732,
+      "step": 8644
+    },
+    {
+      "epoch": 0.6231528869026166,
+      "grad_norm": 0.12186872671523777,
+      "learning_rate": 6.569074992521087e-05,
+      "loss": 1.3171,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6232249693649535,
+      "grad_norm": 0.11102562850783011,
+      "learning_rate": 6.566882084359678e-05,
+      "loss": 1.2603,
+      "step": 8646
+    },
+    {
+      "epoch": 0.6232970518272904,
+      "grad_norm": 0.12008879063497513,
+      "learning_rate": 6.56468936333391e-05,
+      "loss": 1.3415,
+      "step": 8647
+    },
+    {
+      "epoch": 0.6233691342896274,
+      "grad_norm": 0.12258582741664138,
+      "learning_rate": 6.562496829563304e-05,
+      "loss": 1.3666,
+      "step": 8648
+    },
+    {
+      "epoch": 0.6234412167519643,
+      "grad_norm": 0.11130808844569974,
+      "learning_rate": 6.560304483167368e-05,
+      "loss": 1.3145,
+      "step": 8649
+    },
+    {
+      "epoch": 0.6235132992143012,
+      "grad_norm": 0.11527742039806126,
+      "learning_rate": 6.558112324265617e-05,
+      "loss": 1.2847,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6235853816766381,
+      "grad_norm": 0.11220463201979591,
+      "learning_rate": 6.555920352977534e-05,
+      "loss": 1.3508,
+      "step": 8651
+    },
+    {
+      "epoch": 0.6236574641389749,
+      "grad_norm": 0.1211289636416692,
+      "learning_rate": 6.553728569422604e-05,
+      "loss": 1.2375,
+      "step": 8652
+    },
+    {
+      "epoch": 0.6237295466013119,
+      "grad_norm": 0.12453487587583398,
+      "learning_rate": 6.551536973720298e-05,
+      "loss": 1.2155,
+      "step": 8653
+    },
+    {
+      "epoch": 0.6238016290636488,
+      "grad_norm": 0.1095252779595874,
+      "learning_rate": 6.549345565990075e-05,
+      "loss": 1.2739,
+      "step": 8654
+    },
+    {
+      "epoch": 0.6238737115259857,
+      "grad_norm": 0.12911605115112845,
+      "learning_rate": 6.547154346351392e-05,
+      "loss": 1.267,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6239457939883226,
+      "grad_norm": 0.11695014973839951,
+      "learning_rate": 6.544963314923689e-05,
+      "loss": 1.3019,
+      "step": 8656
+    },
+    {
+      "epoch": 0.6240178764506595,
+      "grad_norm": 0.11649561481444706,
+      "learning_rate": 6.542772471826391e-05,
+      "loss": 1.2167,
+      "step": 8657
+    },
+    {
+      "epoch": 0.6240899589129965,
+      "grad_norm": 0.11153169708927559,
+      "learning_rate": 6.540581817178923e-05,
+      "loss": 1.2756,
+      "step": 8658
+    },
+    {
+      "epoch": 0.6241620413753334,
+      "grad_norm": 0.11606319127375148,
+      "learning_rate": 6.538391351100696e-05,
+      "loss": 1.2465,
+      "step": 8659
+    },
+    {
+      "epoch": 0.6242341238376703,
+      "grad_norm": 0.11397799619976416,
+      "learning_rate": 6.53620107371111e-05,
+      "loss": 1.3047,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6243062063000072,
+      "grad_norm": 0.12799348910863825,
+      "learning_rate": 6.534010985129552e-05,
+      "loss": 1.1767,
+      "step": 8661
+    },
+    {
+      "epoch": 0.6243782887623441,
+      "grad_norm": 0.11515983666675995,
+      "learning_rate": 6.531821085475401e-05,
+      "loss": 1.2916,
+      "step": 8662
+    },
+    {
+      "epoch": 0.6244503712246811,
+      "grad_norm": 0.13029976071742283,
+      "learning_rate": 6.529631374868031e-05,
+      "loss": 1.3342,
+      "step": 8663
+    },
+    {
+      "epoch": 0.624522453687018,
+      "grad_norm": 0.11201958144977411,
+      "learning_rate": 6.527441853426799e-05,
+      "loss": 1.3472,
+      "step": 8664
+    },
+    {
+      "epoch": 0.6245945361493549,
+      "grad_norm": 0.12263837223431841,
+      "learning_rate": 6.52525252127105e-05,
+      "loss": 1.3006,
+      "step": 8665
+    },
+    {
+      "epoch": 0.6246666186116918,
+      "grad_norm": 0.11426283665040773,
+      "learning_rate": 6.523063378520124e-05,
+      "loss": 1.3325,
+      "step": 8666
+    },
+    {
+      "epoch": 0.6247387010740287,
+      "grad_norm": 0.10344190376124275,
+      "learning_rate": 6.520874425293348e-05,
+      "loss": 1.2545,
+      "step": 8667
+    },
+    {
+      "epoch": 0.6248107835363657,
+      "grad_norm": 0.1119400074321556,
+      "learning_rate": 6.518685661710046e-05,
+      "loss": 1.2831,
+      "step": 8668
+    },
+    {
+      "epoch": 0.6248828659987025,
+      "grad_norm": 0.11187183445629839,
+      "learning_rate": 6.516497087889517e-05,
+      "loss": 1.3353,
+      "step": 8669
+    },
+    {
+      "epoch": 0.6249549484610394,
+      "grad_norm": 0.11721354224286909,
+      "learning_rate": 6.514308703951062e-05,
+      "loss": 1.333,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6250270309233763,
+      "grad_norm": 0.12132269135870197,
+      "learning_rate": 6.512120510013963e-05,
+      "loss": 1.35,
+      "step": 8671
+    },
+    {
+      "epoch": 0.6250991133857132,
+      "grad_norm": 0.1193445333403056,
+      "learning_rate": 6.509932506197505e-05,
+      "loss": 1.3047,
+      "step": 8672
+    },
+    {
+      "epoch": 0.6251711958480501,
+      "grad_norm": 0.11337297488288194,
+      "learning_rate": 6.507744692620947e-05,
+      "loss": 1.1321,
+      "step": 8673
+    },
+    {
+      "epoch": 0.6252432783103871,
+      "grad_norm": 0.11499068402779548,
+      "learning_rate": 6.505557069403545e-05,
+      "loss": 1.3731,
+      "step": 8674
+    },
+    {
+      "epoch": 0.625315360772724,
+      "grad_norm": 0.12002214185917992,
+      "learning_rate": 6.503369636664547e-05,
+      "loss": 1.3179,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6253874432350609,
+      "grad_norm": 0.11352212854592884,
+      "learning_rate": 6.501182394523182e-05,
+      "loss": 1.2988,
+      "step": 8676
+    },
+    {
+      "epoch": 0.6254595256973978,
+      "grad_norm": 0.12054156166739372,
+      "learning_rate": 6.498995343098679e-05,
+      "loss": 1.3511,
+      "step": 8677
+    },
+    {
+      "epoch": 0.6255316081597347,
+      "grad_norm": 0.1157834015853403,
+      "learning_rate": 6.49680848251025e-05,
+      "loss": 1.2645,
+      "step": 8678
+    },
+    {
+      "epoch": 0.6256036906220717,
+      "grad_norm": 0.12390839938238354,
+      "learning_rate": 6.494621812877102e-05,
+      "loss": 1.3758,
+      "step": 8679
+    },
+    {
+      "epoch": 0.6256757730844086,
+      "grad_norm": 0.17700754136607583,
+      "learning_rate": 6.492435334318422e-05,
+      "loss": 1.2702,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6257478555467455,
+      "grad_norm": 0.11360330181779328,
+      "learning_rate": 6.490249046953397e-05,
+      "loss": 1.3199,
+      "step": 8681
+    },
+    {
+      "epoch": 0.6258199380090824,
+      "grad_norm": 0.12007355739154574,
+      "learning_rate": 6.488062950901198e-05,
+      "loss": 1.2597,
+      "step": 8682
+    },
+    {
+      "epoch": 0.6258920204714193,
+      "grad_norm": 0.11629049891156072,
+      "learning_rate": 6.485877046280989e-05,
+      "loss": 1.2399,
+      "step": 8683
+    },
+    {
+      "epoch": 0.6259641029337563,
+      "grad_norm": 0.11849964702173107,
+      "learning_rate": 6.483691333211915e-05,
+      "loss": 1.2481,
+      "step": 8684
+    },
+    {
+      "epoch": 0.6260361853960932,
+      "grad_norm": 0.1246872961231172,
+      "learning_rate": 6.481505811813125e-05,
+      "loss": 1.3677,
+      "step": 8685
+    },
+    {
+      "epoch": 0.62610826785843,
+      "grad_norm": 0.12080140933041694,
+      "learning_rate": 6.479320482203744e-05,
+      "loss": 1.287,
+      "step": 8686
+    },
+    {
+      "epoch": 0.6261803503207669,
+      "grad_norm": 0.1247709234345406,
+      "learning_rate": 6.477135344502895e-05,
+      "loss": 1.2449,
+      "step": 8687
+    },
+    {
+      "epoch": 0.6262524327831038,
+      "grad_norm": 0.12469000341549993,
+      "learning_rate": 6.474950398829684e-05,
+      "loss": 1.2996,
+      "step": 8688
+    },
+    {
+      "epoch": 0.6263245152454407,
+      "grad_norm": 0.13384078493933146,
+      "learning_rate": 6.472765645303212e-05,
+      "loss": 1.2512,
+      "step": 8689
+    },
+    {
+      "epoch": 0.6263965977077777,
+      "grad_norm": 0.11236341693681769,
+      "learning_rate": 6.470581084042568e-05,
+      "loss": 1.2957,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6264686801701146,
+      "grad_norm": 0.11504381080006218,
+      "learning_rate": 6.468396715166834e-05,
+      "loss": 1.2872,
+      "step": 8691
+    },
+    {
+      "epoch": 0.6265407626324515,
+      "grad_norm": 0.11542829968526544,
+      "learning_rate": 6.46621253879507e-05,
+      "loss": 1.3007,
+      "step": 8692
+    },
+    {
+      "epoch": 0.6266128450947884,
+      "grad_norm": 0.10823591233943071,
+      "learning_rate": 6.464028555046335e-05,
+      "loss": 1.2613,
+      "step": 8693
+    },
+    {
+      "epoch": 0.6266849275571253,
+      "grad_norm": 0.11535010131632445,
+      "learning_rate": 6.46184476403968e-05,
+      "loss": 1.2093,
+      "step": 8694
+    },
+    {
+      "epoch": 0.6267570100194623,
+      "grad_norm": 0.10872707773582556,
+      "learning_rate": 6.459661165894141e-05,
+      "loss": 1.2535,
+      "step": 8695
+    },
+    {
+      "epoch": 0.6268290924817992,
+      "grad_norm": 0.11684326611428476,
+      "learning_rate": 6.457477760728738e-05,
+      "loss": 1.3384,
+      "step": 8696
+    },
+    {
+      "epoch": 0.6269011749441361,
+      "grad_norm": 0.12047969057626971,
+      "learning_rate": 6.455294548662492e-05,
+      "loss": 1.1888,
+      "step": 8697
+    },
+    {
+      "epoch": 0.626973257406473,
+      "grad_norm": 0.11625807240241905,
+      "learning_rate": 6.4531115298144e-05,
+      "loss": 1.294,
+      "step": 8698
+    },
+    {
+      "epoch": 0.62704533986881,
+      "grad_norm": 0.11944225199210978,
+      "learning_rate": 6.450928704303465e-05,
+      "loss": 1.2534,
+      "step": 8699
+    },
+    {
+      "epoch": 0.6271174223311469,
+      "grad_norm": 0.14495183454989724,
+      "learning_rate": 6.448746072248666e-05,
+      "loss": 1.307,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6271895047934838,
+      "grad_norm": 0.11701128313016435,
+      "learning_rate": 6.446563633768979e-05,
+      "loss": 1.2879,
+      "step": 8701
+    },
+    {
+      "epoch": 0.6272615872558207,
+      "grad_norm": 0.10908868849762297,
+      "learning_rate": 6.444381388983356e-05,
+      "loss": 1.2198,
+      "step": 8702
+    },
+    {
+      "epoch": 0.6273336697181576,
+      "grad_norm": 0.12040631387482055,
+      "learning_rate": 6.442199338010764e-05,
+      "loss": 1.2827,
+      "step": 8703
+    },
+    {
+      "epoch": 0.6274057521804944,
+      "grad_norm": 0.10879205427208667,
+      "learning_rate": 6.440017480970135e-05,
+      "loss": 1.2597,
+      "step": 8704
+    },
+    {
+      "epoch": 0.6274778346428314,
+      "grad_norm": 0.11239404455602171,
+      "learning_rate": 6.437835817980403e-05,
+      "loss": 1.2405,
+      "step": 8705
+    },
+    {
+      "epoch": 0.6275499171051683,
+      "grad_norm": 0.10791942952174971,
+      "learning_rate": 6.435654349160482e-05,
+      "loss": 1.3212,
+      "step": 8706
+    },
+    {
+      "epoch": 0.6276219995675052,
+      "grad_norm": 0.13327777260453164,
+      "learning_rate": 6.433473074629291e-05,
+      "loss": 1.2794,
+      "step": 8707
+    },
+    {
+      "epoch": 0.6276940820298421,
+      "grad_norm": 0.11763116416405496,
+      "learning_rate": 6.431291994505724e-05,
+      "loss": 1.3656,
+      "step": 8708
+    },
+    {
+      "epoch": 0.627766164492179,
+      "grad_norm": 0.12658581674476083,
+      "learning_rate": 6.429111108908673e-05,
+      "loss": 1.3732,
+      "step": 8709
+    },
+    {
+      "epoch": 0.627838246954516,
+      "grad_norm": 0.11937617144749772,
+      "learning_rate": 6.426930417957007e-05,
+      "loss": 1.3512,
+      "step": 8710
+    },
+    {
+      "epoch": 0.6279103294168529,
+      "grad_norm": 0.12091963727735024,
+      "learning_rate": 6.4247499217696e-05,
+      "loss": 1.2944,
+      "step": 8711
+    },
+    {
+      "epoch": 0.6279824118791898,
+      "grad_norm": 0.10949162718604498,
+      "learning_rate": 6.422569620465308e-05,
+      "loss": 1.2948,
+      "step": 8712
+    },
+    {
+      "epoch": 0.6280544943415267,
+      "grad_norm": 0.12329063399825248,
+      "learning_rate": 6.420389514162979e-05,
+      "loss": 1.3883,
+      "step": 8713
+    },
+    {
+      "epoch": 0.6281265768038636,
+      "grad_norm": 0.1310346966922714,
+      "learning_rate": 6.418209602981442e-05,
+      "loss": 1.4055,
+      "step": 8714
+    },
+    {
+      "epoch": 0.6281986592662006,
+      "grad_norm": 0.12887835072680345,
+      "learning_rate": 6.416029887039525e-05,
+      "loss": 1.2558,
+      "step": 8715
+    },
+    {
+      "epoch": 0.6282707417285375,
+      "grad_norm": 0.1147230833535405,
+      "learning_rate": 6.413850366456043e-05,
+      "loss": 1.2839,
+      "step": 8716
+    },
+    {
+      "epoch": 0.6283428241908744,
+      "grad_norm": 0.13714036834605806,
+      "learning_rate": 6.411671041349801e-05,
+      "loss": 1.2856,
+      "step": 8717
+    },
+    {
+      "epoch": 0.6284149066532113,
+      "grad_norm": 0.1117590236549213,
+      "learning_rate": 6.409491911839588e-05,
+      "loss": 1.2869,
+      "step": 8718
+    },
+    {
+      "epoch": 0.6284869891155482,
+      "grad_norm": 0.1290132773165584,
+      "learning_rate": 6.407312978044188e-05,
+      "loss": 1.2806,
+      "step": 8719
+    },
+    {
+      "epoch": 0.6285590715778852,
+      "grad_norm": 0.11843144427873475,
+      "learning_rate": 6.405134240082368e-05,
+      "loss": 1.3191,
+      "step": 8720
+    },
+    {
+      "epoch": 0.628631154040222,
+      "grad_norm": 0.10524617272601372,
+      "learning_rate": 6.402955698072896e-05,
+      "loss": 1.3116,
+      "step": 8721
+    },
+    {
+      "epoch": 0.6287032365025589,
+      "grad_norm": 0.11176920392506653,
+      "learning_rate": 6.400777352134519e-05,
+      "loss": 1.1548,
+      "step": 8722
+    },
+    {
+      "epoch": 0.6287753189648958,
+      "grad_norm": 0.1223112061802004,
+      "learning_rate": 6.398599202385978e-05,
+      "loss": 1.252,
+      "step": 8723
+    },
+    {
+      "epoch": 0.6288474014272327,
+      "grad_norm": 0.10692433900267677,
+      "learning_rate": 6.39642124894599e-05,
+      "loss": 1.2262,
+      "step": 8724
+    },
+    {
+      "epoch": 0.6289194838895696,
+      "grad_norm": 0.13224857006731564,
+      "learning_rate": 6.394243491933292e-05,
+      "loss": 1.2757,
+      "step": 8725
+    },
+    {
+      "epoch": 0.6289915663519066,
+      "grad_norm": 0.11503623957952964,
+      "learning_rate": 6.392065931466578e-05,
+      "loss": 1.2664,
+      "step": 8726
+    },
+    {
+      "epoch": 0.6290636488142435,
+      "grad_norm": 0.12508410514054893,
+      "learning_rate": 6.38988856766455e-05,
+      "loss": 1.2823,
+      "step": 8727
+    },
+    {
+      "epoch": 0.6291357312765804,
+      "grad_norm": 0.12668270915769794,
+      "learning_rate": 6.387711400645888e-05,
+      "loss": 1.3112,
+      "step": 8728
+    },
+    {
+      "epoch": 0.6292078137389173,
+      "grad_norm": 0.12683251179531402,
+      "learning_rate": 6.385534430529275e-05,
+      "loss": 1.265,
+      "step": 8729
+    },
+    {
+      "epoch": 0.6292798962012542,
+      "grad_norm": 0.11352237571665864,
+      "learning_rate": 6.383357657433371e-05,
+      "loss": 1.3344,
+      "step": 8730
+    },
+    {
+      "epoch": 0.6293519786635912,
+      "grad_norm": 0.12077400019031519,
+      "learning_rate": 6.38118108147683e-05,
+      "loss": 1.2362,
+      "step": 8731
+    },
+    {
+      "epoch": 0.6294240611259281,
+      "grad_norm": 0.10870567961262245,
+      "learning_rate": 6.379004702778295e-05,
+      "loss": 1.2533,
+      "step": 8732
+    },
+    {
+      "epoch": 0.629496143588265,
+      "grad_norm": 0.1181213435463219,
+      "learning_rate": 6.376828521456395e-05,
+      "loss": 1.3067,
+      "step": 8733
+    },
+    {
+      "epoch": 0.6295682260506019,
+      "grad_norm": 0.11057641793620276,
+      "learning_rate": 6.374652537629757e-05,
+      "loss": 1.2558,
+      "step": 8734
+    },
+    {
+      "epoch": 0.6296403085129388,
+      "grad_norm": 0.12350425387207097,
+      "learning_rate": 6.372476751416991e-05,
+      "loss": 1.2709,
+      "step": 8735
+    },
+    {
+      "epoch": 0.6297123909752758,
+      "grad_norm": 0.12259496666513368,
+      "learning_rate": 6.370301162936693e-05,
+      "loss": 1.2324,
+      "step": 8736
+    },
+    {
+      "epoch": 0.6297844734376127,
+      "grad_norm": 0.12153843211137001,
+      "learning_rate": 6.368125772307452e-05,
+      "loss": 1.3189,
+      "step": 8737
+    },
+    {
+      "epoch": 0.6298565558999495,
+      "grad_norm": 0.11830036087788226,
+      "learning_rate": 6.365950579647848e-05,
+      "loss": 1.3074,
+      "step": 8738
+    },
+    {
+      "epoch": 0.6299286383622864,
+      "grad_norm": 0.1102903162959682,
+      "learning_rate": 6.363775585076452e-05,
+      "loss": 1.2491,
+      "step": 8739
+    },
+    {
+      "epoch": 0.6300007208246233,
+      "grad_norm": 0.11873988755123077,
+      "learning_rate": 6.361600788711816e-05,
+      "loss": 1.4253,
+      "step": 8740
+    },
+    {
+      "epoch": 0.6300728032869602,
+      "grad_norm": 0.11898556207913279,
+      "learning_rate": 6.359426190672487e-05,
+      "loss": 1.2825,
+      "step": 8741
+    },
+    {
+      "epoch": 0.6301448857492972,
+      "grad_norm": 0.12103739702539397,
+      "learning_rate": 6.357251791076997e-05,
+      "loss": 1.2631,
+      "step": 8742
+    },
+    {
+      "epoch": 0.6302169682116341,
+      "grad_norm": 0.10467628605274533,
+      "learning_rate": 6.355077590043876e-05,
+      "loss": 1.215,
+      "step": 8743
+    },
+    {
+      "epoch": 0.630289050673971,
+      "grad_norm": 0.11905789368336346,
+      "learning_rate": 6.352903587691635e-05,
+      "loss": 1.2135,
+      "step": 8744
+    },
+    {
+      "epoch": 0.6303611331363079,
+      "grad_norm": 0.21329141995906892,
+      "learning_rate": 6.350729784138776e-05,
+      "loss": 1.2852,
+      "step": 8745
+    },
+    {
+      "epoch": 0.6304332155986448,
+      "grad_norm": 0.12671573861845498,
+      "learning_rate": 6.348556179503788e-05,
+      "loss": 1.3219,
+      "step": 8746
+    },
+    {
+      "epoch": 0.6305052980609818,
+      "grad_norm": 0.10477186167092967,
+      "learning_rate": 6.346382773905162e-05,
+      "loss": 1.2442,
+      "step": 8747
+    },
+    {
+      "epoch": 0.6305773805233187,
+      "grad_norm": 0.14809894584421288,
+      "learning_rate": 6.344209567461357e-05,
+      "loss": 1.3156,
+      "step": 8748
+    },
+    {
+      "epoch": 0.6306494629856556,
+      "grad_norm": 0.1276269703094078,
+      "learning_rate": 6.342036560290835e-05,
+      "loss": 1.279,
+      "step": 8749
+    },
+    {
+      "epoch": 0.6307215454479925,
+      "grad_norm": 0.11352640778638416,
+      "learning_rate": 6.339863752512046e-05,
+      "loss": 1.3638,
+      "step": 8750
+    },
+    {
+      "epoch": 0.6307936279103294,
+      "grad_norm": 0.13215569707262748,
+      "learning_rate": 6.337691144243429e-05,
+      "loss": 1.2291,
+      "step": 8751
+    },
+    {
+      "epoch": 0.6308657103726664,
+      "grad_norm": 0.111067969266845,
+      "learning_rate": 6.335518735603407e-05,
+      "loss": 1.255,
+      "step": 8752
+    },
+    {
+      "epoch": 0.6309377928350033,
+      "grad_norm": 0.1128707642899132,
+      "learning_rate": 6.333346526710398e-05,
+      "loss": 1.3087,
+      "step": 8753
+    },
+    {
+      "epoch": 0.6310098752973402,
+      "grad_norm": 0.1286044851969513,
+      "learning_rate": 6.331174517682806e-05,
+      "loss": 1.2902,
+      "step": 8754
+    },
+    {
+      "epoch": 0.6310819577596771,
+      "grad_norm": 0.11594966607319086,
+      "learning_rate": 6.329002708639022e-05,
+      "loss": 1.2723,
+      "step": 8755
+    },
+    {
+      "epoch": 0.6311540402220139,
+      "grad_norm": 0.12282184590068752,
+      "learning_rate": 6.326831099697435e-05,
+      "loss": 1.1714,
+      "step": 8756
+    },
+    {
+      "epoch": 0.6312261226843509,
+      "grad_norm": 0.14001072444419105,
+      "learning_rate": 6.324659690976414e-05,
+      "loss": 1.2852,
+      "step": 8757
+    },
+    {
+      "epoch": 0.6312982051466878,
+      "grad_norm": 0.12233960347084771,
+      "learning_rate": 6.322488482594323e-05,
+      "loss": 1.3193,
+      "step": 8758
+    },
+    {
+      "epoch": 0.6313702876090247,
+      "grad_norm": 0.12577346924973687,
+      "learning_rate": 6.320317474669504e-05,
+      "loss": 1.3261,
+      "step": 8759
+    },
+    {
+      "epoch": 0.6314423700713616,
+      "grad_norm": 0.11632266051314247,
+      "learning_rate": 6.318146667320306e-05,
+      "loss": 1.2426,
+      "step": 8760
+    },
+    {
+      "epoch": 0.6315144525336985,
+      "grad_norm": 0.11531698318379836,
+      "learning_rate": 6.315976060665052e-05,
+      "loss": 1.3163,
+      "step": 8761
+    },
+    {
+      "epoch": 0.6315865349960355,
+      "grad_norm": 0.13134145966706556,
+      "learning_rate": 6.313805654822063e-05,
+      "loss": 1.2454,
+      "step": 8762
+    },
+    {
+      "epoch": 0.6316586174583724,
+      "grad_norm": 0.14343699622712314,
+      "learning_rate": 6.311635449909642e-05,
+      "loss": 1.275,
+      "step": 8763
+    },
+    {
+      "epoch": 0.6317306999207093,
+      "grad_norm": 0.142373873182071,
+      "learning_rate": 6.309465446046083e-05,
+      "loss": 1.2825,
+      "step": 8764
+    },
+    {
+      "epoch": 0.6318027823830462,
+      "grad_norm": 0.12012456597538142,
+      "learning_rate": 6.307295643349678e-05,
+      "loss": 1.2364,
+      "step": 8765
+    },
+    {
+      "epoch": 0.6318748648453831,
+      "grad_norm": 0.12972512906334052,
+      "learning_rate": 6.305126041938696e-05,
+      "loss": 1.3054,
+      "step": 8766
+    },
+    {
+      "epoch": 0.63194694730772,
+      "grad_norm": 0.12310470380663623,
+      "learning_rate": 6.3029566419314e-05,
+      "loss": 1.2564,
+      "step": 8767
+    },
+    {
+      "epoch": 0.632019029770057,
+      "grad_norm": 0.1196658538709967,
+      "learning_rate": 6.300787443446036e-05,
+      "loss": 1.3923,
+      "step": 8768
+    },
+    {
+      "epoch": 0.6320911122323939,
+      "grad_norm": 0.11835174985459422,
+      "learning_rate": 6.298618446600856e-05,
+      "loss": 1.2757,
+      "step": 8769
+    },
+    {
+      "epoch": 0.6321631946947308,
+      "grad_norm": 0.15025854535220423,
+      "learning_rate": 6.296449651514085e-05,
+      "loss": 1.2188,
+      "step": 8770
+    },
+    {
+      "epoch": 0.6322352771570677,
+      "grad_norm": 0.11592354130372975,
+      "learning_rate": 6.294281058303939e-05,
+      "loss": 1.224,
+      "step": 8771
+    },
+    {
+      "epoch": 0.6323073596194047,
+      "grad_norm": 0.11711187513347249,
+      "learning_rate": 6.292112667088625e-05,
+      "loss": 1.279,
+      "step": 8772
+    },
+    {
+      "epoch": 0.6323794420817415,
+      "grad_norm": 0.11476843081971856,
+      "learning_rate": 6.289944477986345e-05,
+      "loss": 1.3141,
+      "step": 8773
+    },
+    {
+      "epoch": 0.6324515245440784,
+      "grad_norm": 0.11744719964973838,
+      "learning_rate": 6.287776491115285e-05,
+      "loss": 1.3503,
+      "step": 8774
+    },
+    {
+      "epoch": 0.6325236070064153,
+      "grad_norm": 0.12664056813801858,
+      "learning_rate": 6.285608706593614e-05,
+      "loss": 1.3304,
+      "step": 8775
+    },
+    {
+      "epoch": 0.6325956894687522,
+      "grad_norm": 0.11527844232430251,
+      "learning_rate": 6.2834411245395e-05,
+      "loss": 1.3428,
+      "step": 8776
+    },
+    {
+      "epoch": 0.6326677719310891,
+      "grad_norm": 0.11875389770755272,
+      "learning_rate": 6.281273745071088e-05,
+      "loss": 1.29,
+      "step": 8777
+    },
+    {
+      "epoch": 0.6327398543934261,
+      "grad_norm": 0.11667577919543419,
+      "learning_rate": 6.279106568306531e-05,
+      "loss": 1.2145,
+      "step": 8778
+    },
+    {
+      "epoch": 0.632811936855763,
+      "grad_norm": 0.11923199113655597,
+      "learning_rate": 6.276939594363954e-05,
+      "loss": 1.2941,
+      "step": 8779
+    },
+    {
+      "epoch": 0.6328840193180999,
+      "grad_norm": 0.12943080764614606,
+      "learning_rate": 6.274772823361477e-05,
+      "loss": 1.3036,
+      "step": 8780
+    },
+    {
+      "epoch": 0.6329561017804368,
+      "grad_norm": 0.10901559525044532,
+      "learning_rate": 6.272606255417202e-05,
+      "loss": 1.3029,
+      "step": 8781
+    },
+    {
+      "epoch": 0.6330281842427737,
+      "grad_norm": 0.11094544093276482,
+      "learning_rate": 6.270439890649238e-05,
+      "loss": 1.3251,
+      "step": 8782
+    },
+    {
+      "epoch": 0.6331002667051107,
+      "grad_norm": 0.11474537260268267,
+      "learning_rate": 6.268273729175664e-05,
+      "loss": 1.3072,
+      "step": 8783
+    },
+    {
+      "epoch": 0.6331723491674476,
+      "grad_norm": 0.12487140004319502,
+      "learning_rate": 6.266107771114559e-05,
+      "loss": 1.3148,
+      "step": 8784
+    },
+    {
+      "epoch": 0.6332444316297845,
+      "grad_norm": 0.10416542272100145,
+      "learning_rate": 6.263942016583982e-05,
+      "loss": 1.2801,
+      "step": 8785
+    },
+    {
+      "epoch": 0.6333165140921214,
+      "grad_norm": 0.12058492677286138,
+      "learning_rate": 6.261776465701987e-05,
+      "loss": 1.1703,
+      "step": 8786
+    },
+    {
+      "epoch": 0.6333885965544583,
+      "grad_norm": 0.1153098999737247,
+      "learning_rate": 6.259611118586621e-05,
+      "loss": 1.269,
+      "step": 8787
+    },
+    {
+      "epoch": 0.6334606790167953,
+      "grad_norm": 0.1212230712206225,
+      "learning_rate": 6.257445975355913e-05,
+      "loss": 1.3283,
+      "step": 8788
+    },
+    {
+      "epoch": 0.6335327614791322,
+      "grad_norm": 0.10900012208249032,
+      "learning_rate": 6.255281036127879e-05,
+      "loss": 1.2648,
+      "step": 8789
+    },
+    {
+      "epoch": 0.633604843941469,
+      "grad_norm": 0.13556611908957708,
+      "learning_rate": 6.253116301020527e-05,
+      "loss": 1.2855,
+      "step": 8790
+    },
+    {
+      "epoch": 0.6336769264038059,
+      "grad_norm": 0.11382766689580517,
+      "learning_rate": 6.250951770151861e-05,
+      "loss": 1.3206,
+      "step": 8791
+    },
+    {
+      "epoch": 0.6337490088661428,
+      "grad_norm": 0.12502956320061048,
+      "learning_rate": 6.248787443639863e-05,
+      "loss": 1.3724,
+      "step": 8792
+    },
+    {
+      "epoch": 0.6338210913284797,
+      "grad_norm": 0.12951172836570632,
+      "learning_rate": 6.24662332160251e-05,
+      "loss": 1.178,
+      "step": 8793
+    },
+    {
+      "epoch": 0.6338931737908167,
+      "grad_norm": 0.11483138463128713,
+      "learning_rate": 6.244459404157758e-05,
+      "loss": 1.2601,
+      "step": 8794
+    },
+    {
+      "epoch": 0.6339652562531536,
+      "grad_norm": 0.12199844416467279,
+      "learning_rate": 6.242295691423572e-05,
+      "loss": 1.266,
+      "step": 8795
+    },
+    {
+      "epoch": 0.6340373387154905,
+      "grad_norm": 0.11625199692540474,
+      "learning_rate": 6.24013218351789e-05,
+      "loss": 1.2925,
+      "step": 8796
+    },
+    {
+      "epoch": 0.6341094211778274,
+      "grad_norm": 0.11721852306053239,
+      "learning_rate": 6.237968880558637e-05,
+      "loss": 1.1884,
+      "step": 8797
+    },
+    {
+      "epoch": 0.6341815036401643,
+      "grad_norm": 0.11462336059345167,
+      "learning_rate": 6.23580578266374e-05,
+      "loss": 1.2742,
+      "step": 8798
+    },
+    {
+      "epoch": 0.6342535861025013,
+      "grad_norm": 0.11366550788780475,
+      "learning_rate": 6.233642889951097e-05,
+      "loss": 1.3237,
+      "step": 8799
+    },
+    {
+      "epoch": 0.6343256685648382,
+      "grad_norm": 0.10898323888183373,
+      "learning_rate": 6.231480202538615e-05,
+      "loss": 1.2103,
+      "step": 8800
+    },
+    {
+      "epoch": 0.6343977510271751,
+      "grad_norm": 0.11334140560095846,
+      "learning_rate": 6.229317720544176e-05,
+      "loss": 1.3842,
+      "step": 8801
+    },
+    {
+      "epoch": 0.634469833489512,
+      "grad_norm": 0.12889601315501042,
+      "learning_rate": 6.227155444085658e-05,
+      "loss": 1.3219,
+      "step": 8802
+    },
+    {
+      "epoch": 0.634541915951849,
+      "grad_norm": 0.10464598104291384,
+      "learning_rate": 6.224993373280912e-05,
+      "loss": 1.1773,
+      "step": 8803
+    },
+    {
+      "epoch": 0.6346139984141859,
+      "grad_norm": 0.12230089319951218,
+      "learning_rate": 6.222831508247808e-05,
+      "loss": 1.1996,
+      "step": 8804
+    },
+    {
+      "epoch": 0.6346860808765228,
+      "grad_norm": 0.11520916831387051,
+      "learning_rate": 6.220669849104175e-05,
+      "loss": 1.2322,
+      "step": 8805
+    },
+    {
+      "epoch": 0.6347581633388597,
+      "grad_norm": 0.11159751213698434,
+      "learning_rate": 6.218508395967849e-05,
+      "loss": 1.312,
+      "step": 8806
+    },
+    {
+      "epoch": 0.6348302458011966,
+      "grad_norm": 0.11468216549043653,
+      "learning_rate": 6.216347148956643e-05,
+      "loss": 1.3447,
+      "step": 8807
+    },
+    {
+      "epoch": 0.6349023282635334,
+      "grad_norm": 0.11960237233864728,
+      "learning_rate": 6.214186108188364e-05,
+      "loss": 1.2715,
+      "step": 8808
+    },
+    {
+      "epoch": 0.6349744107258704,
+      "grad_norm": 0.12042607047520859,
+      "learning_rate": 6.212025273780815e-05,
+      "loss": 1.2546,
+      "step": 8809
+    },
+    {
+      "epoch": 0.6350464931882073,
+      "grad_norm": 0.11397592395071715,
+      "learning_rate": 6.209864645851777e-05,
+      "loss": 1.3315,
+      "step": 8810
+    },
+    {
+      "epoch": 0.6351185756505442,
+      "grad_norm": 0.12009466292026079,
+      "learning_rate": 6.207704224519022e-05,
+      "loss": 1.2432,
+      "step": 8811
+    },
+    {
+      "epoch": 0.6351906581128811,
+      "grad_norm": 0.11273750936461459,
+      "learning_rate": 6.205544009900311e-05,
+      "loss": 1.2537,
+      "step": 8812
+    },
+    {
+      "epoch": 0.635262740575218,
+      "grad_norm": 0.12089897948648388,
+      "learning_rate": 6.2033840021134e-05,
+      "loss": 1.2412,
+      "step": 8813
+    },
+    {
+      "epoch": 0.635334823037555,
+      "grad_norm": 0.11853580993358762,
+      "learning_rate": 6.201224201276028e-05,
+      "loss": 1.3957,
+      "step": 8814
+    },
+    {
+      "epoch": 0.6354069054998919,
+      "grad_norm": 0.11302186724204127,
+      "learning_rate": 6.199064607505922e-05,
+      "loss": 1.3387,
+      "step": 8815
+    },
+    {
+      "epoch": 0.6354789879622288,
+      "grad_norm": 0.1103060585405058,
+      "learning_rate": 6.196905220920795e-05,
+      "loss": 1.2685,
+      "step": 8816
+    },
+    {
+      "epoch": 0.6355510704245657,
+      "grad_norm": 0.11692489699518813,
+      "learning_rate": 6.19474604163836e-05,
+      "loss": 1.2175,
+      "step": 8817
+    },
+    {
+      "epoch": 0.6356231528869026,
+      "grad_norm": 0.11040137823055893,
+      "learning_rate": 6.19258706977631e-05,
+      "loss": 1.3608,
+      "step": 8818
+    },
+    {
+      "epoch": 0.6356952353492396,
+      "grad_norm": 0.1190788519754099,
+      "learning_rate": 6.190428305452324e-05,
+      "loss": 1.2942,
+      "step": 8819
+    },
+    {
+      "epoch": 0.6357673178115765,
+      "grad_norm": 0.11183297395745978,
+      "learning_rate": 6.188269748784077e-05,
+      "loss": 1.277,
+      "step": 8820
+    },
+    {
+      "epoch": 0.6358394002739134,
+      "grad_norm": 0.1298945100370436,
+      "learning_rate": 6.186111399889229e-05,
+      "loss": 1.3294,
+      "step": 8821
+    },
+    {
+      "epoch": 0.6359114827362503,
+      "grad_norm": 0.11726748209849003,
+      "learning_rate": 6.183953258885433e-05,
+      "loss": 1.2578,
+      "step": 8822
+    },
+    {
+      "epoch": 0.6359835651985872,
+      "grad_norm": 0.13158139053798634,
+      "learning_rate": 6.18179532589032e-05,
+      "loss": 1.3249,
+      "step": 8823
+    },
+    {
+      "epoch": 0.6360556476609242,
+      "grad_norm": 0.12020777608178439,
+      "learning_rate": 6.179637601021522e-05,
+      "loss": 1.2614,
+      "step": 8824
+    },
+    {
+      "epoch": 0.636127730123261,
+      "grad_norm": 0.11593942212826998,
+      "learning_rate": 6.17748008439665e-05,
+      "loss": 1.2787,
+      "step": 8825
+    },
+    {
+      "epoch": 0.6361998125855979,
+      "grad_norm": 0.1147745112395971,
+      "learning_rate": 6.175322776133314e-05,
+      "loss": 1.2467,
+      "step": 8826
+    },
+    {
+      "epoch": 0.6362718950479348,
+      "grad_norm": 0.12808955471521272,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 1.2904,
+      "step": 8827
+    },
+    {
+      "epoch": 0.6363439775102717,
+      "grad_norm": 0.12760549266212562,
+      "learning_rate": 6.171008785161599e-05,
+      "loss": 1.3423,
+      "step": 8828
+    },
+    {
+      "epoch": 0.6364160599726086,
+      "grad_norm": 0.12095063965597612,
+      "learning_rate": 6.168852102688372e-05,
+      "loss": 1.3465,
+      "step": 8829
+    },
+    {
+      "epoch": 0.6364881424349456,
+      "grad_norm": 0.11593921548608534,
+      "learning_rate": 6.166695629046975e-05,
+      "loss": 1.2085,
+      "step": 8830
+    },
+    {
+      "epoch": 0.6365602248972825,
+      "grad_norm": 0.10558978287428353,
+      "learning_rate": 6.164539364354962e-05,
+      "loss": 1.3626,
+      "step": 8831
+    },
+    {
+      "epoch": 0.6366323073596194,
+      "grad_norm": 0.12605994978554982,
+      "learning_rate": 6.162383308729869e-05,
+      "loss": 1.3282,
+      "step": 8832
+    },
+    {
+      "epoch": 0.6367043898219563,
+      "grad_norm": 0.134889120684799,
+      "learning_rate": 6.160227462289217e-05,
+      "loss": 1.2614,
+      "step": 8833
+    },
+    {
+      "epoch": 0.6367764722842932,
+      "grad_norm": 0.10964374744103196,
+      "learning_rate": 6.158071825150518e-05,
+      "loss": 1.2498,
+      "step": 8834
+    },
+    {
+      "epoch": 0.6368485547466302,
+      "grad_norm": 0.11757802052428501,
+      "learning_rate": 6.155916397431276e-05,
+      "loss": 1.2959,
+      "step": 8835
+    },
+    {
+      "epoch": 0.6369206372089671,
+      "grad_norm": 0.11618777284613559,
+      "learning_rate": 6.153761179248983e-05,
+      "loss": 1.3395,
+      "step": 8836
+    },
+    {
+      "epoch": 0.636992719671304,
+      "grad_norm": 0.12030254432724777,
+      "learning_rate": 6.151606170721113e-05,
+      "loss": 1.2201,
+      "step": 8837
+    },
+    {
+      "epoch": 0.6370648021336409,
+      "grad_norm": 0.12343948055975344,
+      "learning_rate": 6.149451371965136e-05,
+      "loss": 1.2213,
+      "step": 8838
+    },
+    {
+      "epoch": 0.6371368845959778,
+      "grad_norm": 0.11447416008914849,
+      "learning_rate": 6.147296783098506e-05,
+      "loss": 1.2541,
+      "step": 8839
+    },
+    {
+      "epoch": 0.6372089670583148,
+      "grad_norm": 0.1504422553866375,
+      "learning_rate": 6.14514240423867e-05,
+      "loss": 1.24,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6372810495206517,
+      "grad_norm": 0.19334361374347633,
+      "learning_rate": 6.14298823550306e-05,
+      "loss": 1.2489,
+      "step": 8841
+    },
+    {
+      "epoch": 0.6373531319829885,
+      "grad_norm": 0.11658721404031751,
+      "learning_rate": 6.140834277009098e-05,
+      "loss": 1.2389,
+      "step": 8842
+    },
+    {
+      "epoch": 0.6374252144453254,
+      "grad_norm": 0.1220538216567391,
+      "learning_rate": 6.138680528874188e-05,
+      "loss": 1.2724,
+      "step": 8843
+    },
+    {
+      "epoch": 0.6374972969076623,
+      "grad_norm": 0.12480314505022311,
+      "learning_rate": 6.136526991215737e-05,
+      "loss": 1.2434,
+      "step": 8844
+    },
+    {
+      "epoch": 0.6375693793699992,
+      "grad_norm": 0.12115099925728509,
+      "learning_rate": 6.13437366415113e-05,
+      "loss": 1.3573,
+      "step": 8845
+    },
+    {
+      "epoch": 0.6376414618323362,
+      "grad_norm": 0.11695949361950735,
+      "learning_rate": 6.132220547797739e-05,
+      "loss": 1.2518,
+      "step": 8846
+    },
+    {
+      "epoch": 0.6377135442946731,
+      "grad_norm": 0.12761158167111422,
+      "learning_rate": 6.130067642272928e-05,
+      "loss": 1.3892,
+      "step": 8847
+    },
+    {
+      "epoch": 0.63778562675701,
+      "grad_norm": 0.11820662960108501,
+      "learning_rate": 6.127914947694056e-05,
+      "loss": 1.2088,
+      "step": 8848
+    },
+    {
+      "epoch": 0.6378577092193469,
+      "grad_norm": 0.11680213771548582,
+      "learning_rate": 6.12576246417846e-05,
+      "loss": 1.2386,
+      "step": 8849
+    },
+    {
+      "epoch": 0.6379297916816838,
+      "grad_norm": 0.3248207336228661,
+      "learning_rate": 6.123610191843468e-05,
+      "loss": 1.3219,
+      "step": 8850
+    },
+    {
+      "epoch": 0.6380018741440208,
+      "grad_norm": 0.11423067336907508,
+      "learning_rate": 6.121458130806402e-05,
+      "loss": 1.2615,
+      "step": 8851
+    },
+    {
+      "epoch": 0.6380739566063577,
+      "grad_norm": 0.1089364727982282,
+      "learning_rate": 6.11930628118456e-05,
+      "loss": 1.1986,
+      "step": 8852
+    },
+    {
+      "epoch": 0.6381460390686946,
+      "grad_norm": 0.12245444762004812,
+      "learning_rate": 6.11715464309525e-05,
+      "loss": 1.3968,
+      "step": 8853
+    },
+    {
+      "epoch": 0.6382181215310315,
+      "grad_norm": 0.12102633535259891,
+      "learning_rate": 6.115003216655746e-05,
+      "loss": 1.1734,
+      "step": 8854
+    },
+    {
+      "epoch": 0.6382902039933684,
+      "grad_norm": 0.126224016407463,
+      "learning_rate": 6.112852001983325e-05,
+      "loss": 1.2684,
+      "step": 8855
+    },
+    {
+      "epoch": 0.6383622864557054,
+      "grad_norm": 0.1343593864150149,
+      "learning_rate": 6.11070099919524e-05,
+      "loss": 1.2674,
+      "step": 8856
+    },
+    {
+      "epoch": 0.6384343689180423,
+      "grad_norm": 0.1299637012129175,
+      "learning_rate": 6.108550208408752e-05,
+      "loss": 1.363,
+      "step": 8857
+    },
+    {
+      "epoch": 0.6385064513803792,
+      "grad_norm": 0.12040392719675608,
+      "learning_rate": 6.106399629741087e-05,
+      "loss": 1.2241,
+      "step": 8858
+    },
+    {
+      "epoch": 0.638578533842716,
+      "grad_norm": 0.117604317740091,
+      "learning_rate": 6.104249263309479e-05,
+      "loss": 1.2948,
+      "step": 8859
+    },
+    {
+      "epoch": 0.6386506163050529,
+      "grad_norm": 0.12303360458019762,
+      "learning_rate": 6.1020991092311364e-05,
+      "loss": 1.2737,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6387226987673899,
+      "grad_norm": 0.1269860970948824,
+      "learning_rate": 6.099949167623262e-05,
+      "loss": 1.1906,
+      "step": 8861
+    },
+    {
+      "epoch": 0.6387947812297268,
+      "grad_norm": 0.14783175071059912,
+      "learning_rate": 6.097799438603051e-05,
+      "loss": 1.333,
+      "step": 8862
+    },
+    {
+      "epoch": 0.6388668636920637,
+      "grad_norm": 0.1281489804546493,
+      "learning_rate": 6.095649922287683e-05,
+      "loss": 1.2244,
+      "step": 8863
+    },
+    {
+      "epoch": 0.6389389461544006,
+      "grad_norm": 0.1146721912962272,
+      "learning_rate": 6.09350061879432e-05,
+      "loss": 1.2703,
+      "step": 8864
+    },
+    {
+      "epoch": 0.6390110286167375,
+      "grad_norm": 0.12149623185094605,
+      "learning_rate": 6.091351528240121e-05,
+      "loss": 1.2907,
+      "step": 8865
+    },
+    {
+      "epoch": 0.6390831110790745,
+      "grad_norm": 0.11563850735919527,
+      "learning_rate": 6.089202650742233e-05,
+      "loss": 1.3543,
+      "step": 8866
+    },
+    {
+      "epoch": 0.6391551935414114,
+      "grad_norm": 0.12736730166815763,
+      "learning_rate": 6.08705398641779e-05,
+      "loss": 1.2301,
+      "step": 8867
+    },
+    {
+      "epoch": 0.6392272760037483,
+      "grad_norm": 0.1283821093806247,
+      "learning_rate": 6.084905535383908e-05,
+      "loss": 1.2131,
+      "step": 8868
+    },
+    {
+      "epoch": 0.6392993584660852,
+      "grad_norm": 0.11930493967815305,
+      "learning_rate": 6.0827572977576984e-05,
+      "loss": 1.3,
+      "step": 8869
+    },
+    {
+      "epoch": 0.6393714409284221,
+      "grad_norm": 0.11609506922545478,
+      "learning_rate": 6.0806092736562624e-05,
+      "loss": 1.209,
+      "step": 8870
+    },
+    {
+      "epoch": 0.639443523390759,
+      "grad_norm": 0.11673935747156375,
+      "learning_rate": 6.0784614631966876e-05,
+      "loss": 1.319,
+      "step": 8871
+    },
+    {
+      "epoch": 0.639515605853096,
+      "grad_norm": 0.12407812569308968,
+      "learning_rate": 6.076313866496044e-05,
+      "loss": 1.2659,
+      "step": 8872
+    },
+    {
+      "epoch": 0.6395876883154329,
+      "grad_norm": 0.12810629178071367,
+      "learning_rate": 6.074166483671399e-05,
+      "loss": 1.2648,
+      "step": 8873
+    },
+    {
+      "epoch": 0.6396597707777698,
+      "grad_norm": 0.16025464224999153,
+      "learning_rate": 6.072019314839795e-05,
+      "loss": 1.1907,
+      "step": 8874
+    },
+    {
+      "epoch": 0.6397318532401067,
+      "grad_norm": 0.11240970091472254,
+      "learning_rate": 6.069872360118286e-05,
+      "loss": 1.22,
+      "step": 8875
+    },
+    {
+      "epoch": 0.6398039357024436,
+      "grad_norm": 0.12018328803372674,
+      "learning_rate": 6.067725619623892e-05,
+      "loss": 1.3159,
+      "step": 8876
+    },
+    {
+      "epoch": 0.6398760181647805,
+      "grad_norm": 0.14282756343501693,
+      "learning_rate": 6.065579093473631e-05,
+      "loss": 1.2315,
+      "step": 8877
+    },
+    {
+      "epoch": 0.6399481006271174,
+      "grad_norm": 0.14735158402480406,
+      "learning_rate": 6.0634327817845036e-05,
+      "loss": 1.2288,
+      "step": 8878
+    },
+    {
+      "epoch": 0.6400201830894543,
+      "grad_norm": 0.11215274810139943,
+      "learning_rate": 6.0612866846735114e-05,
+      "loss": 1.2725,
+      "step": 8879
+    },
+    {
+      "epoch": 0.6400922655517912,
+      "grad_norm": 0.15240714907846142,
+      "learning_rate": 6.0591408022576314e-05,
+      "loss": 1.2939,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6401643480141281,
+      "grad_norm": 0.12762219644221806,
+      "learning_rate": 6.056995134653835e-05,
+      "loss": 1.185,
+      "step": 8881
+    },
+    {
+      "epoch": 0.6402364304764651,
+      "grad_norm": 0.20981508865051682,
+      "learning_rate": 6.0548496819790755e-05,
+      "loss": 1.2892,
+      "step": 8882
+    },
+    {
+      "epoch": 0.640308512938802,
+      "grad_norm": 0.10637537240470106,
+      "learning_rate": 6.052704444350302e-05,
+      "loss": 1.2555,
+      "step": 8883
+    },
+    {
+      "epoch": 0.6403805954011389,
+      "grad_norm": 0.12394100184986555,
+      "learning_rate": 6.050559421884452e-05,
+      "loss": 1.199,
+      "step": 8884
+    },
+    {
+      "epoch": 0.6404526778634758,
+      "grad_norm": 0.13262058513257183,
+      "learning_rate": 6.048414614698448e-05,
+      "loss": 1.3004,
+      "step": 8885
+    },
+    {
+      "epoch": 0.6405247603258127,
+      "grad_norm": 0.10527677536325819,
+      "learning_rate": 6.046270022909197e-05,
+      "loss": 1.253,
+      "step": 8886
+    },
+    {
+      "epoch": 0.6405968427881497,
+      "grad_norm": 0.12059064182363176,
+      "learning_rate": 6.0441256466335995e-05,
+      "loss": 1.2372,
+      "step": 8887
+    },
+    {
+      "epoch": 0.6406689252504866,
+      "grad_norm": 0.1453848571414907,
+      "learning_rate": 6.041981485988546e-05,
+      "loss": 1.3395,
+      "step": 8888
+    },
+    {
+      "epoch": 0.6407410077128235,
+      "grad_norm": 0.12605833217842566,
+      "learning_rate": 6.039837541090913e-05,
+      "loss": 1.2518,
+      "step": 8889
+    },
+    {
+      "epoch": 0.6408130901751604,
+      "grad_norm": 0.11168512345968824,
+      "learning_rate": 6.037693812057561e-05,
+      "loss": 1.346,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6408851726374973,
+      "grad_norm": 0.11521014596430827,
+      "learning_rate": 6.035550299005342e-05,
+      "loss": 1.2884,
+      "step": 8891
+    },
+    {
+      "epoch": 0.6409572550998343,
+      "grad_norm": 0.11261982908148387,
+      "learning_rate": 6.033407002051099e-05,
+      "loss": 1.2848,
+      "step": 8892
+    },
+    {
+      "epoch": 0.6410293375621712,
+      "grad_norm": 0.11181578265082713,
+      "learning_rate": 6.031263921311665e-05,
+      "loss": 1.2428,
+      "step": 8893
+    },
+    {
+      "epoch": 0.641101420024508,
+      "grad_norm": 0.11984559806286564,
+      "learning_rate": 6.029121056903849e-05,
+      "loss": 1.2377,
+      "step": 8894
+    },
+    {
+      "epoch": 0.6411735024868449,
+      "grad_norm": 0.10789789784849248,
+      "learning_rate": 6.0269784089444605e-05,
+      "loss": 1.2013,
+      "step": 8895
+    },
+    {
+      "epoch": 0.6412455849491818,
+      "grad_norm": 0.12679094385333362,
+      "learning_rate": 6.0248359775502904e-05,
+      "loss": 1.2402,
+      "step": 8896
+    },
+    {
+      "epoch": 0.6413176674115187,
+      "grad_norm": 0.1148280301134645,
+      "learning_rate": 6.0226937628381264e-05,
+      "loss": 1.2129,
+      "step": 8897
+    },
+    {
+      "epoch": 0.6413897498738557,
+      "grad_norm": 0.1301859264126733,
+      "learning_rate": 6.020551764924732e-05,
+      "loss": 1.2586,
+      "step": 8898
+    },
+    {
+      "epoch": 0.6414618323361926,
+      "grad_norm": 0.11972066020930777,
+      "learning_rate": 6.018409983926867e-05,
+      "loss": 1.2942,
+      "step": 8899
+    },
+    {
+      "epoch": 0.6415339147985295,
+      "grad_norm": 0.12596439151657374,
+      "learning_rate": 6.0162684199612776e-05,
+      "loss": 1.2614,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6416059972608664,
+      "grad_norm": 0.11382938858134573,
+      "learning_rate": 6.014127073144702e-05,
+      "loss": 1.318,
+      "step": 8901
+    },
+    {
+      "epoch": 0.6416780797232033,
+      "grad_norm": 0.11525511959793629,
+      "learning_rate": 6.0119859435938586e-05,
+      "loss": 1.278,
+      "step": 8902
+    },
+    {
+      "epoch": 0.6417501621855403,
+      "grad_norm": 0.12642175352434354,
+      "learning_rate": 6.009845031425459e-05,
+      "loss": 1.2836,
+      "step": 8903
+    },
+    {
+      "epoch": 0.6418222446478772,
+      "grad_norm": 0.13718443547791204,
+      "learning_rate": 6.007704336756201e-05,
+      "loss": 1.2438,
+      "step": 8904
+    },
+    {
+      "epoch": 0.6418943271102141,
+      "grad_norm": 0.12325087400271556,
+      "learning_rate": 6.005563859702772e-05,
+      "loss": 1.2599,
+      "step": 8905
+    },
+    {
+      "epoch": 0.641966409572551,
+      "grad_norm": 0.11264236984475072,
+      "learning_rate": 6.0034236003818486e-05,
+      "loss": 1.3719,
+      "step": 8906
+    },
+    {
+      "epoch": 0.6420384920348879,
+      "grad_norm": 0.12850700793490463,
+      "learning_rate": 6.001283558910094e-05,
+      "loss": 1.2347,
+      "step": 8907
+    },
+    {
+      "epoch": 0.6421105744972249,
+      "grad_norm": 0.1272319946216312,
+      "learning_rate": 5.9991437354041616e-05,
+      "loss": 1.2875,
+      "step": 8908
+    },
+    {
+      "epoch": 0.6421826569595618,
+      "grad_norm": 0.12407926202105252,
+      "learning_rate": 5.997004129980683e-05,
+      "loss": 1.3794,
+      "step": 8909
+    },
+    {
+      "epoch": 0.6422547394218987,
+      "grad_norm": 0.1166369197667031,
+      "learning_rate": 5.994864742756293e-05,
+      "loss": 1.3126,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6423268218842355,
+      "grad_norm": 0.11545280103021326,
+      "learning_rate": 5.992725573847606e-05,
+      "loss": 1.1805,
+      "step": 8911
+    },
+    {
+      "epoch": 0.6423989043465724,
+      "grad_norm": 0.11481533612027842,
+      "learning_rate": 5.990586623371228e-05,
+      "loss": 1.2789,
+      "step": 8912
+    },
+    {
+      "epoch": 0.6424709868089094,
+      "grad_norm": 0.1532839290510674,
+      "learning_rate": 5.9884478914437436e-05,
+      "loss": 1.3412,
+      "step": 8913
+    },
+    {
+      "epoch": 0.6425430692712463,
+      "grad_norm": 0.13108026475497772,
+      "learning_rate": 5.9863093781817394e-05,
+      "loss": 1.1627,
+      "step": 8914
+    },
+    {
+      "epoch": 0.6426151517335832,
+      "grad_norm": 0.12205838500820253,
+      "learning_rate": 5.984171083701782e-05,
+      "loss": 1.3085,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6426872341959201,
+      "grad_norm": 0.11846817938828132,
+      "learning_rate": 5.982033008120429e-05,
+      "loss": 1.2681,
+      "step": 8916
+    },
+    {
+      "epoch": 0.642759316658257,
+      "grad_norm": 0.11644644746368399,
+      "learning_rate": 5.97989515155422e-05,
+      "loss": 1.2631,
+      "step": 8917
+    },
+    {
+      "epoch": 0.642831399120594,
+      "grad_norm": 0.1167384790444034,
+      "learning_rate": 5.977757514119688e-05,
+      "loss": 1.2832,
+      "step": 8918
+    },
+    {
+      "epoch": 0.6429034815829309,
+      "grad_norm": 0.1225069944947297,
+      "learning_rate": 5.9756200959333594e-05,
+      "loss": 1.3522,
+      "step": 8919
+    },
+    {
+      "epoch": 0.6429755640452678,
+      "grad_norm": 0.12839265130915475,
+      "learning_rate": 5.97348289711174e-05,
+      "loss": 1.3183,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6430476465076047,
+      "grad_norm": 0.1069111858386003,
+      "learning_rate": 5.971345917771323e-05,
+      "loss": 1.2134,
+      "step": 8921
+    },
+    {
+      "epoch": 0.6431197289699416,
+      "grad_norm": 0.13198028307372425,
+      "learning_rate": 5.969209158028593e-05,
+      "loss": 1.3053,
+      "step": 8922
+    },
+    {
+      "epoch": 0.6431918114322785,
+      "grad_norm": 0.12188346809580207,
+      "learning_rate": 5.967072618000025e-05,
+      "loss": 1.2766,
+      "step": 8923
+    },
+    {
+      "epoch": 0.6432638938946155,
+      "grad_norm": 0.11443621508033439,
+      "learning_rate": 5.964936297802082e-05,
+      "loss": 1.2668,
+      "step": 8924
+    },
+    {
+      "epoch": 0.6433359763569524,
+      "grad_norm": 0.11834587773208459,
+      "learning_rate": 5.962800197551206e-05,
+      "loss": 1.3632,
+      "step": 8925
+    },
+    {
+      "epoch": 0.6434080588192893,
+      "grad_norm": 0.12201921882675033,
+      "learning_rate": 5.960664317363841e-05,
+      "loss": 1.3482,
+      "step": 8926
+    },
+    {
+      "epoch": 0.6434801412816262,
+      "grad_norm": 0.10692890647983129,
+      "learning_rate": 5.958528657356401e-05,
+      "loss": 1.1911,
+      "step": 8927
+    },
+    {
+      "epoch": 0.6435522237439631,
+      "grad_norm": 0.12101333052347858,
+      "learning_rate": 5.9563932176453106e-05,
+      "loss": 1.269,
+      "step": 8928
+    },
+    {
+      "epoch": 0.6436243062063,
+      "grad_norm": 0.11205453043222309,
+      "learning_rate": 5.954257998346964e-05,
+      "loss": 1.2238,
+      "step": 8929
+    },
+    {
+      "epoch": 0.6436963886686369,
+      "grad_norm": 0.13164352246711467,
+      "learning_rate": 5.952122999577752e-05,
+      "loss": 1.3188,
+      "step": 8930
+    },
+    {
+      "epoch": 0.6437684711309738,
+      "grad_norm": 0.11395364148750468,
+      "learning_rate": 5.9499882214540435e-05,
+      "loss": 1.2333,
+      "step": 8931
+    },
+    {
+      "epoch": 0.6438405535933107,
+      "grad_norm": 0.1171427681620401,
+      "learning_rate": 5.9478536640922165e-05,
+      "loss": 1.1421,
+      "step": 8932
+    },
+    {
+      "epoch": 0.6439126360556476,
+      "grad_norm": 0.11832626115863362,
+      "learning_rate": 5.9457193276086145e-05,
+      "loss": 1.2225,
+      "step": 8933
+    },
+    {
+      "epoch": 0.6439847185179846,
+      "grad_norm": 0.10437253477707706,
+      "learning_rate": 5.94358521211958e-05,
+      "loss": 1.2899,
+      "step": 8934
+    },
+    {
+      "epoch": 0.6440568009803215,
+      "grad_norm": 0.11320907912826844,
+      "learning_rate": 5.9414513177414365e-05,
+      "loss": 1.2594,
+      "step": 8935
+    },
+    {
+      "epoch": 0.6441288834426584,
+      "grad_norm": 0.12662939084576286,
+      "learning_rate": 5.9393176445905106e-05,
+      "loss": 1.3112,
+      "step": 8936
+    },
+    {
+      "epoch": 0.6442009659049953,
+      "grad_norm": 0.10993289926572943,
+      "learning_rate": 5.9371841927830985e-05,
+      "loss": 1.3024,
+      "step": 8937
+    },
+    {
+      "epoch": 0.6442730483673322,
+      "grad_norm": 0.14082547867918582,
+      "learning_rate": 5.935050962435498e-05,
+      "loss": 1.2827,
+      "step": 8938
+    },
+    {
+      "epoch": 0.6443451308296692,
+      "grad_norm": 0.11791741320441305,
+      "learning_rate": 5.9329179536639835e-05,
+      "loss": 1.2462,
+      "step": 8939
+    },
+    {
+      "epoch": 0.6444172132920061,
+      "grad_norm": 0.11528204934398582,
+      "learning_rate": 5.9307851665848244e-05,
+      "loss": 1.2281,
+      "step": 8940
+    },
+    {
+      "epoch": 0.644489295754343,
+      "grad_norm": 0.11631663655857147,
+      "learning_rate": 5.9286526013142794e-05,
+      "loss": 1.2311,
+      "step": 8941
+    },
+    {
+      "epoch": 0.6445613782166799,
+      "grad_norm": 0.1091544044120495,
+      "learning_rate": 5.926520257968593e-05,
+      "loss": 1.2875,
+      "step": 8942
+    },
+    {
+      "epoch": 0.6446334606790168,
+      "grad_norm": 0.12930650604504368,
+      "learning_rate": 5.924388136663992e-05,
+      "loss": 1.22,
+      "step": 8943
+    },
+    {
+      "epoch": 0.6447055431413538,
+      "grad_norm": 0.14233358361916035,
+      "learning_rate": 5.922256237516699e-05,
+      "loss": 1.2837,
+      "step": 8944
+    },
+    {
+      "epoch": 0.6447776256036907,
+      "grad_norm": 0.1142946613077718,
+      "learning_rate": 5.9201245606429215e-05,
+      "loss": 1.247,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6448497080660275,
+      "grad_norm": 0.12989489693954348,
+      "learning_rate": 5.917993106158859e-05,
+      "loss": 1.2779,
+      "step": 8946
+    },
+    {
+      "epoch": 0.6449217905283644,
+      "grad_norm": 0.11898229081385292,
+      "learning_rate": 5.915861874180687e-05,
+      "loss": 1.274,
+      "step": 8947
+    },
+    {
+      "epoch": 0.6449938729907013,
+      "grad_norm": 0.11437511000067468,
+      "learning_rate": 5.9137308648245837e-05,
+      "loss": 1.3635,
+      "step": 8948
+    },
+    {
+      "epoch": 0.6450659554530382,
+      "grad_norm": 0.13526065885818594,
+      "learning_rate": 5.9116000782066985e-05,
+      "loss": 1.2752,
+      "step": 8949
+    },
+    {
+      "epoch": 0.6451380379153752,
+      "grad_norm": 0.12261370508322245,
+      "learning_rate": 5.909469514443191e-05,
+      "loss": 1.2758,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6452101203777121,
+      "grad_norm": 0.1167690541259254,
+      "learning_rate": 5.907339173650188e-05,
+      "loss": 1.268,
+      "step": 8951
+    },
+    {
+      "epoch": 0.645282202840049,
+      "grad_norm": 0.12151030103710066,
+      "learning_rate": 5.9052090559438166e-05,
+      "loss": 1.2017,
+      "step": 8952
+    },
+    {
+      "epoch": 0.6453542853023859,
+      "grad_norm": 0.1353706223646288,
+      "learning_rate": 5.903079161440179e-05,
+      "loss": 1.2615,
+      "step": 8953
+    },
+    {
+      "epoch": 0.6454263677647228,
+      "grad_norm": 0.11225318104604505,
+      "learning_rate": 5.9009494902553844e-05,
+      "loss": 1.2709,
+      "step": 8954
+    },
+    {
+      "epoch": 0.6454984502270598,
+      "grad_norm": 0.12196687781498086,
+      "learning_rate": 5.8988200425055104e-05,
+      "loss": 1.302,
+      "step": 8955
+    },
+    {
+      "epoch": 0.6455705326893967,
+      "grad_norm": 0.12561287885205047,
+      "learning_rate": 5.896690818306638e-05,
+      "loss": 1.2457,
+      "step": 8956
+    },
+    {
+      "epoch": 0.6456426151517336,
+      "grad_norm": 0.10831137681565411,
+      "learning_rate": 5.894561817774819e-05,
+      "loss": 1.2827,
+      "step": 8957
+    },
+    {
+      "epoch": 0.6457146976140705,
+      "grad_norm": 0.11316121897885503,
+      "learning_rate": 5.892433041026115e-05,
+      "loss": 1.3205,
+      "step": 8958
+    },
+    {
+      "epoch": 0.6457867800764074,
+      "grad_norm": 0.12143873872740206,
+      "learning_rate": 5.890304488176557e-05,
+      "loss": 1.2909,
+      "step": 8959
+    },
+    {
+      "epoch": 0.6458588625387444,
+      "grad_norm": 0.1156702552955781,
+      "learning_rate": 5.888176159342171e-05,
+      "loss": 1.2086,
+      "step": 8960
+    },
+    {
+      "epoch": 0.6459309450010813,
+      "grad_norm": 0.11229901472849818,
+      "learning_rate": 5.8860480546389684e-05,
+      "loss": 1.2612,
+      "step": 8961
+    },
+    {
+      "epoch": 0.6460030274634182,
+      "grad_norm": 0.12018626873369356,
+      "learning_rate": 5.8839201741829483e-05,
+      "loss": 1.2526,
+      "step": 8962
+    },
+    {
+      "epoch": 0.646075109925755,
+      "grad_norm": 0.10875472405853043,
+      "learning_rate": 5.881792518090106e-05,
+      "loss": 1.2015,
+      "step": 8963
+    },
+    {
+      "epoch": 0.6461471923880919,
+      "grad_norm": 0.11234606287071355,
+      "learning_rate": 5.8796650864764154e-05,
+      "loss": 1.2871,
+      "step": 8964
+    },
+    {
+      "epoch": 0.6462192748504288,
+      "grad_norm": 0.11384481946022808,
+      "learning_rate": 5.877537879457837e-05,
+      "loss": 1.2936,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6462913573127658,
+      "grad_norm": 0.11730830781131707,
+      "learning_rate": 5.875410897150322e-05,
+      "loss": 1.2737,
+      "step": 8966
+    },
+    {
+      "epoch": 0.6463634397751027,
+      "grad_norm": 0.11935213589745108,
+      "learning_rate": 5.873284139669816e-05,
+      "loss": 1.2738,
+      "step": 8967
+    },
+    {
+      "epoch": 0.6464355222374396,
+      "grad_norm": 0.11408838570845697,
+      "learning_rate": 5.871157607132245e-05,
+      "loss": 1.2735,
+      "step": 8968
+    },
+    {
+      "epoch": 0.6465076046997765,
+      "grad_norm": 0.12992183941073276,
+      "learning_rate": 5.869031299653518e-05,
+      "loss": 1.2808,
+      "step": 8969
+    },
+    {
+      "epoch": 0.6465796871621134,
+      "grad_norm": 0.1264113935083685,
+      "learning_rate": 5.866905217349544e-05,
+      "loss": 1.2745,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6466517696244504,
+      "grad_norm": 0.12255418265370663,
+      "learning_rate": 5.864779360336209e-05,
+      "loss": 1.2522,
+      "step": 8971
+    },
+    {
+      "epoch": 0.6467238520867873,
+      "grad_norm": 0.12082558810493407,
+      "learning_rate": 5.862653728729397e-05,
+      "loss": 1.2692,
+      "step": 8972
+    },
+    {
+      "epoch": 0.6467959345491242,
+      "grad_norm": 0.11493952351246585,
+      "learning_rate": 5.860528322644969e-05,
+      "loss": 1.2357,
+      "step": 8973
+    },
+    {
+      "epoch": 0.6468680170114611,
+      "grad_norm": 0.10849308960238174,
+      "learning_rate": 5.858403142198781e-05,
+      "loss": 1.2315,
+      "step": 8974
+    },
+    {
+      "epoch": 0.646940099473798,
+      "grad_norm": 0.11041339154181862,
+      "learning_rate": 5.85627818750667e-05,
+      "loss": 1.2745,
+      "step": 8975
+    },
+    {
+      "epoch": 0.647012181936135,
+      "grad_norm": 0.11810699719175229,
+      "learning_rate": 5.854153458684473e-05,
+      "loss": 1.342,
+      "step": 8976
+    },
+    {
+      "epoch": 0.6470842643984719,
+      "grad_norm": 0.13197393184777959,
+      "learning_rate": 5.852028955848001e-05,
+      "loss": 1.2953,
+      "step": 8977
+    },
+    {
+      "epoch": 0.6471563468608088,
+      "grad_norm": 0.10826693386305478,
+      "learning_rate": 5.8499046791130585e-05,
+      "loss": 1.1644,
+      "step": 8978
+    },
+    {
+      "epoch": 0.6472284293231457,
+      "grad_norm": 0.13113333534850707,
+      "learning_rate": 5.847780628595439e-05,
+      "loss": 1.2647,
+      "step": 8979
+    },
+    {
+      "epoch": 0.6473005117854826,
+      "grad_norm": 0.17071584998972641,
+      "learning_rate": 5.845656804410924e-05,
+      "loss": 1.2964,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6473725942478195,
+      "grad_norm": 0.11692471781413731,
+      "learning_rate": 5.843533206675277e-05,
+      "loss": 1.2545,
+      "step": 8981
+    },
+    {
+      "epoch": 0.6474446767101564,
+      "grad_norm": 0.11614677141100792,
+      "learning_rate": 5.8414098355042555e-05,
+      "loss": 1.283,
+      "step": 8982
+    },
+    {
+      "epoch": 0.6475167591724933,
+      "grad_norm": 0.12581222246318235,
+      "learning_rate": 5.8392866910136036e-05,
+      "loss": 1.2597,
+      "step": 8983
+    },
+    {
+      "epoch": 0.6475888416348302,
+      "grad_norm": 0.11772431607868973,
+      "learning_rate": 5.837163773319046e-05,
+      "loss": 1.2101,
+      "step": 8984
+    },
+    {
+      "epoch": 0.6476609240971671,
+      "grad_norm": 0.1201218292757571,
+      "learning_rate": 5.835041082536308e-05,
+      "loss": 1.2904,
+      "step": 8985
+    },
+    {
+      "epoch": 0.647733006559504,
+      "grad_norm": 0.11907892109918741,
+      "learning_rate": 5.832918618781089e-05,
+      "loss": 1.2621,
+      "step": 8986
+    },
+    {
+      "epoch": 0.647805089021841,
+      "grad_norm": 0.11309450946981421,
+      "learning_rate": 5.830796382169087e-05,
+      "loss": 1.3109,
+      "step": 8987
+    },
+    {
+      "epoch": 0.6478771714841779,
+      "grad_norm": 0.12211695837851677,
+      "learning_rate": 5.828674372815981e-05,
+      "loss": 1.191,
+      "step": 8988
+    },
+    {
+      "epoch": 0.6479492539465148,
+      "grad_norm": 0.12234387700687384,
+      "learning_rate": 5.8265525908374384e-05,
+      "loss": 1.2701,
+      "step": 8989
+    },
+    {
+      "epoch": 0.6480213364088517,
+      "grad_norm": 0.1130157238809864,
+      "learning_rate": 5.8244310363491174e-05,
+      "loss": 1.367,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6480934188711887,
+      "grad_norm": 0.12099003181000414,
+      "learning_rate": 5.82230970946666e-05,
+      "loss": 1.4043,
+      "step": 8991
+    },
+    {
+      "epoch": 0.6481655013335256,
+      "grad_norm": 0.12342922284182109,
+      "learning_rate": 5.820188610305701e-05,
+      "loss": 1.248,
+      "step": 8992
+    },
+    {
+      "epoch": 0.6482375837958625,
+      "grad_norm": 0.13076683490410324,
+      "learning_rate": 5.818067738981851e-05,
+      "loss": 1.1905,
+      "step": 8993
+    },
+    {
+      "epoch": 0.6483096662581994,
+      "grad_norm": 0.11340843300999462,
+      "learning_rate": 5.815947095610723e-05,
+      "loss": 1.2706,
+      "step": 8994
+    },
+    {
+      "epoch": 0.6483817487205363,
+      "grad_norm": 0.12710992641553445,
+      "learning_rate": 5.813826680307912e-05,
+      "loss": 1.211,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6484538311828733,
+      "grad_norm": 0.15308251062417821,
+      "learning_rate": 5.811706493189001e-05,
+      "loss": 1.3085,
+      "step": 8996
+    },
+    {
+      "epoch": 0.6485259136452102,
+      "grad_norm": 0.11860135718212034,
+      "learning_rate": 5.809586534369547e-05,
+      "loss": 1.2241,
+      "step": 8997
+    },
+    {
+      "epoch": 0.648597996107547,
+      "grad_norm": 0.1143118808764199,
+      "learning_rate": 5.8074668039651205e-05,
+      "loss": 1.2537,
+      "step": 8998
+    },
+    {
+      "epoch": 0.6486700785698839,
+      "grad_norm": 0.19324956040858143,
+      "learning_rate": 5.8053473020912595e-05,
+      "loss": 1.2936,
+      "step": 8999
+    },
+    {
+      "epoch": 0.6487421610322208,
+      "grad_norm": 0.11351991510569809,
+      "learning_rate": 5.803228028863501e-05,
+      "loss": 1.2099,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6488142434945577,
+      "grad_norm": 0.1303379475073411,
+      "learning_rate": 5.801108984397354e-05,
+      "loss": 1.2746,
+      "step": 9001
+    },
+    {
+      "epoch": 0.6488863259568947,
+      "grad_norm": 0.13743404140015797,
+      "learning_rate": 5.798990168808334e-05,
+      "loss": 1.1951,
+      "step": 9002
+    },
+    {
+      "epoch": 0.6489584084192316,
+      "grad_norm": 0.1286084054928108,
+      "learning_rate": 5.796871582211934e-05,
+      "loss": 1.2952,
+      "step": 9003
+    },
+    {
+      "epoch": 0.6490304908815685,
+      "grad_norm": 0.1270058928199597,
+      "learning_rate": 5.794753224723639e-05,
+      "loss": 1.2857,
+      "step": 9004
+    },
+    {
+      "epoch": 0.6491025733439054,
+      "grad_norm": 0.11609531778002431,
+      "learning_rate": 5.792635096458909e-05,
+      "loss": 1.3222,
+      "step": 9005
+    },
+    {
+      "epoch": 0.6491746558062423,
+      "grad_norm": 0.11533732143217783,
+      "learning_rate": 5.790517197533204e-05,
+      "loss": 1.2714,
+      "step": 9006
+    },
+    {
+      "epoch": 0.6492467382685793,
+      "grad_norm": 0.1709115301538874,
+      "learning_rate": 5.788399528061974e-05,
+      "loss": 1.2791,
+      "step": 9007
+    },
+    {
+      "epoch": 0.6493188207309162,
+      "grad_norm": 0.12907720943456205,
+      "learning_rate": 5.786282088160653e-05,
+      "loss": 1.2344,
+      "step": 9008
+    },
+    {
+      "epoch": 0.6493909031932531,
+      "grad_norm": 0.13768431378800833,
+      "learning_rate": 5.78416487794465e-05,
+      "loss": 1.3182,
+      "step": 9009
+    },
+    {
+      "epoch": 0.64946298565559,
+      "grad_norm": 0.11539713422967626,
+      "learning_rate": 5.782047897529373e-05,
+      "loss": 1.2294,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6495350681179269,
+      "grad_norm": 0.116227432956215,
+      "learning_rate": 5.7799311470302255e-05,
+      "loss": 1.195,
+      "step": 9011
+    },
+    {
+      "epoch": 0.6496071505802639,
+      "grad_norm": 0.12982799672643666,
+      "learning_rate": 5.7778146265625874e-05,
+      "loss": 1.2946,
+      "step": 9012
+    },
+    {
+      "epoch": 0.6496792330426008,
+      "grad_norm": 0.12313245580891904,
+      "learning_rate": 5.7756983362418215e-05,
+      "loss": 1.3287,
+      "step": 9013
+    },
+    {
+      "epoch": 0.6497513155049377,
+      "grad_norm": 0.11158223932487835,
+      "learning_rate": 5.773582276183287e-05,
+      "loss": 1.2711,
+      "step": 9014
+    },
+    {
+      "epoch": 0.6498233979672745,
+      "grad_norm": 0.11473266834075088,
+      "learning_rate": 5.771466446502326e-05,
+      "loss": 1.269,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6498954804296114,
+      "grad_norm": 0.13721215794890665,
+      "learning_rate": 5.769350847314281e-05,
+      "loss": 1.3014,
+      "step": 9016
+    },
+    {
+      "epoch": 0.6499675628919483,
+      "grad_norm": 0.1243476609179662,
+      "learning_rate": 5.767235478734458e-05,
+      "loss": 1.3376,
+      "step": 9017
+    },
+    {
+      "epoch": 0.6500396453542853,
+      "grad_norm": 0.11983546787343023,
+      "learning_rate": 5.7651203408781715e-05,
+      "loss": 1.262,
+      "step": 9018
+    },
+    {
+      "epoch": 0.6501117278166222,
+      "grad_norm": 0.12122547999777727,
+      "learning_rate": 5.763005433860706e-05,
+      "loss": 1.3249,
+      "step": 9019
+    },
+    {
+      "epoch": 0.6501838102789591,
+      "grad_norm": 0.12145894646869221,
+      "learning_rate": 5.760890757797359e-05,
+      "loss": 1.2356,
+      "step": 9020
+    },
+    {
+      "epoch": 0.650255892741296,
+      "grad_norm": 0.11961193263921188,
+      "learning_rate": 5.7587763128033865e-05,
+      "loss": 1.2704,
+      "step": 9021
+    },
+    {
+      "epoch": 0.650327975203633,
+      "grad_norm": 0.11840185704851097,
+      "learning_rate": 5.756662098994048e-05,
+      "loss": 1.1762,
+      "step": 9022
+    },
+    {
+      "epoch": 0.6504000576659699,
+      "grad_norm": 0.12291872874157521,
+      "learning_rate": 5.754548116484581e-05,
+      "loss": 1.2392,
+      "step": 9023
+    },
+    {
+      "epoch": 0.6504721401283068,
+      "grad_norm": 0.11591098312691366,
+      "learning_rate": 5.7524343653902325e-05,
+      "loss": 1.4206,
+      "step": 9024
+    },
+    {
+      "epoch": 0.6505442225906437,
+      "grad_norm": 0.10935667450789598,
+      "learning_rate": 5.750320845826206e-05,
+      "loss": 1.2961,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6506163050529806,
+      "grad_norm": 0.11277159479192951,
+      "learning_rate": 5.748207557907712e-05,
+      "loss": 1.2504,
+      "step": 9026
+    },
+    {
+      "epoch": 0.6506883875153175,
+      "grad_norm": 0.1278117911050367,
+      "learning_rate": 5.7460945017499444e-05,
+      "loss": 1.4139,
+      "step": 9027
+    },
+    {
+      "epoch": 0.6507604699776545,
+      "grad_norm": 0.11157134327382842,
+      "learning_rate": 5.743981677468082e-05,
+      "loss": 1.2944,
+      "step": 9028
+    },
+    {
+      "epoch": 0.6508325524399914,
+      "grad_norm": 0.14925323643740995,
+      "learning_rate": 5.741869085177295e-05,
+      "loss": 1.3428,
+      "step": 9029
+    },
+    {
+      "epoch": 0.6509046349023283,
+      "grad_norm": 0.12082585427813262,
+      "learning_rate": 5.739756724992736e-05,
+      "loss": 1.2469,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6509767173646652,
+      "grad_norm": 0.11985080233394012,
+      "learning_rate": 5.73764459702955e-05,
+      "loss": 1.3654,
+      "step": 9031
+    },
+    {
+      "epoch": 0.6510487998270021,
+      "grad_norm": 0.11899156141385527,
+      "learning_rate": 5.735532701402865e-05,
+      "loss": 1.2678,
+      "step": 9032
+    },
+    {
+      "epoch": 0.651120882289339,
+      "grad_norm": 0.11872440732576628,
+      "learning_rate": 5.7334210382277975e-05,
+      "loss": 1.3349,
+      "step": 9033
+    },
+    {
+      "epoch": 0.6511929647516759,
+      "grad_norm": 0.12188243078445288,
+      "learning_rate": 5.731309607619455e-05,
+      "loss": 1.3203,
+      "step": 9034
+    },
+    {
+      "epoch": 0.6512650472140128,
+      "grad_norm": 0.14133457193262933,
+      "learning_rate": 5.729198409692926e-05,
+      "loss": 1.2834,
+      "step": 9035
+    },
+    {
+      "epoch": 0.6513371296763497,
+      "grad_norm": 0.11090613348542439,
+      "learning_rate": 5.727087444563293e-05,
+      "loss": 1.3133,
+      "step": 9036
+    },
+    {
+      "epoch": 0.6514092121386866,
+      "grad_norm": 0.1184619624938345,
+      "learning_rate": 5.72497671234562e-05,
+      "loss": 1.264,
+      "step": 9037
+    },
+    {
+      "epoch": 0.6514812946010236,
+      "grad_norm": 0.1241273172744117,
+      "learning_rate": 5.722866213154962e-05,
+      "loss": 1.2785,
+      "step": 9038
+    },
+    {
+      "epoch": 0.6515533770633605,
+      "grad_norm": 0.1307539622472298,
+      "learning_rate": 5.7207559471063596e-05,
+      "loss": 1.3005,
+      "step": 9039
+    },
+    {
+      "epoch": 0.6516254595256974,
+      "grad_norm": 0.11385941886339547,
+      "learning_rate": 5.718645914314842e-05,
+      "loss": 1.2164,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6516975419880343,
+      "grad_norm": 0.12553993668179295,
+      "learning_rate": 5.7165361148954233e-05,
+      "loss": 1.2564,
+      "step": 9041
+    },
+    {
+      "epoch": 0.6517696244503712,
+      "grad_norm": 0.11501878644410168,
+      "learning_rate": 5.714426548963108e-05,
+      "loss": 1.3002,
+      "step": 9042
+    },
+    {
+      "epoch": 0.6518417069127082,
+      "grad_norm": 0.13156020094216697,
+      "learning_rate": 5.712317216632884e-05,
+      "loss": 1.3136,
+      "step": 9043
+    },
+    {
+      "epoch": 0.6519137893750451,
+      "grad_norm": 0.10830861578047557,
+      "learning_rate": 5.710208118019732e-05,
+      "loss": 1.2253,
+      "step": 9044
+    },
+    {
+      "epoch": 0.651985871837382,
+      "grad_norm": 0.11516171385996402,
+      "learning_rate": 5.708099253238614e-05,
+      "loss": 1.2654,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6520579542997189,
+      "grad_norm": 0.11764034057687198,
+      "learning_rate": 5.7059906224044845e-05,
+      "loss": 1.3169,
+      "step": 9046
+    },
+    {
+      "epoch": 0.6521300367620558,
+      "grad_norm": 0.10596256520083783,
+      "learning_rate": 5.70388222563228e-05,
+      "loss": 1.2716,
+      "step": 9047
+    },
+    {
+      "epoch": 0.6522021192243928,
+      "grad_norm": 0.11853894473874405,
+      "learning_rate": 5.701774063036929e-05,
+      "loss": 1.2595,
+      "step": 9048
+    },
+    {
+      "epoch": 0.6522742016867297,
+      "grad_norm": 0.12494711536598105,
+      "learning_rate": 5.699666134733349e-05,
+      "loss": 1.3144,
+      "step": 9049
+    },
+    {
+      "epoch": 0.6523462841490665,
+      "grad_norm": 0.1408906803545654,
+      "learning_rate": 5.697558440836428e-05,
+      "loss": 1.3205,
+      "step": 9050
+    },
+    {
+      "epoch": 0.6524183666114034,
+      "grad_norm": 0.12545388180752054,
+      "learning_rate": 5.695450981461067e-05,
+      "loss": 1.2295,
+      "step": 9051
+    },
+    {
+      "epoch": 0.6524904490737403,
+      "grad_norm": 0.11855816794880641,
+      "learning_rate": 5.6933437567221374e-05,
+      "loss": 1.2493,
+      "step": 9052
+    },
+    {
+      "epoch": 0.6525625315360772,
+      "grad_norm": 0.12062457139254251,
+      "learning_rate": 5.691236766734505e-05,
+      "loss": 1.2443,
+      "step": 9053
+    },
+    {
+      "epoch": 0.6526346139984142,
+      "grad_norm": 0.11395337664401095,
+      "learning_rate": 5.6891300116130076e-05,
+      "loss": 1.3576,
+      "step": 9054
+    },
+    {
+      "epoch": 0.6527066964607511,
+      "grad_norm": 0.11429283948414233,
+      "learning_rate": 5.6870234914724964e-05,
+      "loss": 1.275,
+      "step": 9055
+    },
+    {
+      "epoch": 0.652778778923088,
+      "grad_norm": 0.12331546337533884,
+      "learning_rate": 5.6849172064277876e-05,
+      "loss": 1.29,
+      "step": 9056
+    },
+    {
+      "epoch": 0.6528508613854249,
+      "grad_norm": 0.12401559278136427,
+      "learning_rate": 5.6828111565937e-05,
+      "loss": 1.2307,
+      "step": 9057
+    },
+    {
+      "epoch": 0.6529229438477618,
+      "grad_norm": 0.11821563053317277,
+      "learning_rate": 5.680705342085024e-05,
+      "loss": 1.293,
+      "step": 9058
+    },
+    {
+      "epoch": 0.6529950263100988,
+      "grad_norm": 0.12110448761233568,
+      "learning_rate": 5.6785997630165435e-05,
+      "loss": 1.3185,
+      "step": 9059
+    },
+    {
+      "epoch": 0.6530671087724357,
+      "grad_norm": 0.12189631752928093,
+      "learning_rate": 5.676494419503044e-05,
+      "loss": 1.1633,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6531391912347726,
+      "grad_norm": 0.12011073484345303,
+      "learning_rate": 5.674389311659274e-05,
+      "loss": 1.1445,
+      "step": 9061
+    },
+    {
+      "epoch": 0.6532112736971095,
+      "grad_norm": 0.11685815140242038,
+      "learning_rate": 5.672284439599984e-05,
+      "loss": 1.2713,
+      "step": 9062
+    },
+    {
+      "epoch": 0.6532833561594464,
+      "grad_norm": 0.10611313929971988,
+      "learning_rate": 5.670179803439906e-05,
+      "loss": 1.2262,
+      "step": 9063
+    },
+    {
+      "epoch": 0.6533554386217834,
+      "grad_norm": 0.11944983123479232,
+      "learning_rate": 5.668075403293771e-05,
+      "loss": 1.2996,
+      "step": 9064
+    },
+    {
+      "epoch": 0.6534275210841203,
+      "grad_norm": 0.12342799799728744,
+      "learning_rate": 5.665971239276279e-05,
+      "loss": 1.3215,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6534996035464572,
+      "grad_norm": 0.10505827388741015,
+      "learning_rate": 5.663867311502127e-05,
+      "loss": 1.2321,
+      "step": 9066
+    },
+    {
+      "epoch": 0.653571686008794,
+      "grad_norm": 0.11697423815402225,
+      "learning_rate": 5.6617636200859935e-05,
+      "loss": 1.2796,
+      "step": 9067
+    },
+    {
+      "epoch": 0.6536437684711309,
+      "grad_norm": 0.1269315861293628,
+      "learning_rate": 5.659660165142563e-05,
+      "loss": 1.3219,
+      "step": 9068
+    },
+    {
+      "epoch": 0.6537158509334678,
+      "grad_norm": 0.13235445386477543,
+      "learning_rate": 5.6575569467864784e-05,
+      "loss": 1.3341,
+      "step": 9069
+    },
+    {
+      "epoch": 0.6537879333958048,
+      "grad_norm": 0.11843926022381784,
+      "learning_rate": 5.655453965132389e-05,
+      "loss": 1.2778,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6538600158581417,
+      "grad_norm": 0.1242732291618286,
+      "learning_rate": 5.6533512202949267e-05,
+      "loss": 1.3106,
+      "step": 9071
+    },
+    {
+      "epoch": 0.6539320983204786,
+      "grad_norm": 0.12954937107741804,
+      "learning_rate": 5.651248712388708e-05,
+      "loss": 1.2271,
+      "step": 9072
+    },
+    {
+      "epoch": 0.6540041807828155,
+      "grad_norm": 0.1151727218415202,
+      "learning_rate": 5.6491464415283416e-05,
+      "loss": 1.2207,
+      "step": 9073
+    },
+    {
+      "epoch": 0.6540762632451524,
+      "grad_norm": 0.12497760026852206,
+      "learning_rate": 5.6470444078284155e-05,
+      "loss": 1.2411,
+      "step": 9074
+    },
+    {
+      "epoch": 0.6541483457074894,
+      "grad_norm": 0.11142420263062766,
+      "learning_rate": 5.644942611403513e-05,
+      "loss": 1.2173,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6542204281698263,
+      "grad_norm": 0.116376336337671,
+      "learning_rate": 5.642841052368199e-05,
+      "loss": 1.2243,
+      "step": 9076
+    },
+    {
+      "epoch": 0.6542925106321632,
+      "grad_norm": 0.10982893174280149,
+      "learning_rate": 5.640739730837028e-05,
+      "loss": 1.3752,
+      "step": 9077
+    },
+    {
+      "epoch": 0.6543645930945001,
+      "grad_norm": 0.11700783355392479,
+      "learning_rate": 5.638638646924541e-05,
+      "loss": 1.2732,
+      "step": 9078
+    },
+    {
+      "epoch": 0.654436675556837,
+      "grad_norm": 0.11737711031602946,
+      "learning_rate": 5.6365378007452664e-05,
+      "loss": 1.2501,
+      "step": 9079
+    },
+    {
+      "epoch": 0.654508758019174,
+      "grad_norm": 0.12856668956800793,
+      "learning_rate": 5.6344371924137173e-05,
+      "loss": 1.1607,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6545808404815109,
+      "grad_norm": 0.12463630146084588,
+      "learning_rate": 5.632336822044397e-05,
+      "loss": 1.2159,
+      "step": 9081
+    },
+    {
+      "epoch": 0.6546529229438478,
+      "grad_norm": 0.1158013218889403,
+      "learning_rate": 5.6302366897517946e-05,
+      "loss": 1.3295,
+      "step": 9082
+    },
+    {
+      "epoch": 0.6547250054061847,
+      "grad_norm": 0.11122619758847582,
+      "learning_rate": 5.628136795650385e-05,
+      "loss": 1.3452,
+      "step": 9083
+    },
+    {
+      "epoch": 0.6547970878685216,
+      "grad_norm": 0.12099490087351356,
+      "learning_rate": 5.626037139854632e-05,
+      "loss": 1.2126,
+      "step": 9084
+    },
+    {
+      "epoch": 0.6548691703308585,
+      "grad_norm": 0.12062844772040335,
+      "learning_rate": 5.6239377224789845e-05,
+      "loss": 1.2563,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6549412527931954,
+      "grad_norm": 0.11962104755685689,
+      "learning_rate": 5.621838543637883e-05,
+      "loss": 1.2896,
+      "step": 9086
+    },
+    {
+      "epoch": 0.6550133352555323,
+      "grad_norm": 0.1213230908409732,
+      "learning_rate": 5.6197396034457464e-05,
+      "loss": 1.254,
+      "step": 9087
+    },
+    {
+      "epoch": 0.6550854177178692,
+      "grad_norm": 0.11545531406160586,
+      "learning_rate": 5.61764090201699e-05,
+      "loss": 1.3129,
+      "step": 9088
+    },
+    {
+      "epoch": 0.6551575001802061,
+      "grad_norm": 0.11506024341380833,
+      "learning_rate": 5.615542439466008e-05,
+      "loss": 1.275,
+      "step": 9089
+    },
+    {
+      "epoch": 0.655229582642543,
+      "grad_norm": 0.13582163212473564,
+      "learning_rate": 5.613444215907191e-05,
+      "loss": 1.2609,
+      "step": 9090
+    },
+    {
+      "epoch": 0.65530166510488,
+      "grad_norm": 0.1249168670607649,
+      "learning_rate": 5.6113462314549046e-05,
+      "loss": 1.1212,
+      "step": 9091
+    },
+    {
+      "epoch": 0.6553737475672169,
+      "grad_norm": 0.12379302539103172,
+      "learning_rate": 5.609248486223512e-05,
+      "loss": 1.2403,
+      "step": 9092
+    },
+    {
+      "epoch": 0.6554458300295538,
+      "grad_norm": 0.10938980711179434,
+      "learning_rate": 5.6071509803273604e-05,
+      "loss": 1.2608,
+      "step": 9093
+    },
+    {
+      "epoch": 0.6555179124918907,
+      "grad_norm": 0.12668514439800765,
+      "learning_rate": 5.6050537138807723e-05,
+      "loss": 1.3053,
+      "step": 9094
+    },
+    {
+      "epoch": 0.6555899949542277,
+      "grad_norm": 0.11997753425879934,
+      "learning_rate": 5.60295668699808e-05,
+      "loss": 1.2966,
+      "step": 9095
+    },
+    {
+      "epoch": 0.6556620774165646,
+      "grad_norm": 0.11115518986499526,
+      "learning_rate": 5.600859899793584e-05,
+      "loss": 1.3037,
+      "step": 9096
+    },
+    {
+      "epoch": 0.6557341598789015,
+      "grad_norm": 0.12498458876097232,
+      "learning_rate": 5.598763352381585e-05,
+      "loss": 1.3267,
+      "step": 9097
+    },
+    {
+      "epoch": 0.6558062423412384,
+      "grad_norm": 0.14176170918399422,
+      "learning_rate": 5.596667044876349e-05,
+      "loss": 1.053,
+      "step": 9098
+    },
+    {
+      "epoch": 0.6558783248035753,
+      "grad_norm": 0.13674601471714418,
+      "learning_rate": 5.5945709773921576e-05,
+      "loss": 1.2494,
+      "step": 9099
+    },
+    {
+      "epoch": 0.6559504072659122,
+      "grad_norm": 0.10972411306040981,
+      "learning_rate": 5.5924751500432584e-05,
+      "loss": 1.2653,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6560224897282492,
+      "grad_norm": 0.10501801015063503,
+      "learning_rate": 5.590379562943899e-05,
+      "loss": 1.2822,
+      "step": 9101
+    },
+    {
+      "epoch": 0.656094572190586,
+      "grad_norm": 0.12224296949396225,
+      "learning_rate": 5.5882842162082995e-05,
+      "loss": 1.2701,
+      "step": 9102
+    },
+    {
+      "epoch": 0.6561666546529229,
+      "grad_norm": 0.15420665184948693,
+      "learning_rate": 5.5861891099506723e-05,
+      "loss": 1.2515,
+      "step": 9103
+    },
+    {
+      "epoch": 0.6562387371152598,
+      "grad_norm": 0.11852731347401403,
+      "learning_rate": 5.5840942442852316e-05,
+      "loss": 1.3448,
+      "step": 9104
+    },
+    {
+      "epoch": 0.6563108195775967,
+      "grad_norm": 0.10853198095955353,
+      "learning_rate": 5.581999619326164e-05,
+      "loss": 1.2796,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6563829020399337,
+      "grad_norm": 0.10956987097967358,
+      "learning_rate": 5.579905235187639e-05,
+      "loss": 1.3802,
+      "step": 9106
+    },
+    {
+      "epoch": 0.6564549845022706,
+      "grad_norm": 0.12365067218383582,
+      "learning_rate": 5.5778110919838154e-05,
+      "loss": 1.3208,
+      "step": 9107
+    },
+    {
+      "epoch": 0.6565270669646075,
+      "grad_norm": 0.11609553273747693,
+      "learning_rate": 5.5757171898288554e-05,
+      "loss": 1.3114,
+      "step": 9108
+    },
+    {
+      "epoch": 0.6565991494269444,
+      "grad_norm": 0.11559271644617125,
+      "learning_rate": 5.573623528836893e-05,
+      "loss": 1.2869,
+      "step": 9109
+    },
+    {
+      "epoch": 0.6566712318892813,
+      "grad_norm": 0.12174538485633281,
+      "learning_rate": 5.571530109122043e-05,
+      "loss": 1.2631,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6567433143516183,
+      "grad_norm": 0.11181702249886447,
+      "learning_rate": 5.5694369307984204e-05,
+      "loss": 1.2384,
+      "step": 9111
+    },
+    {
+      "epoch": 0.6568153968139552,
+      "grad_norm": 0.123790512458963,
+      "learning_rate": 5.567343993980119e-05,
+      "loss": 1.3301,
+      "step": 9112
+    },
+    {
+      "epoch": 0.6568874792762921,
+      "grad_norm": 0.11805976068663075,
+      "learning_rate": 5.565251298781233e-05,
+      "loss": 1.2657,
+      "step": 9113
+    },
+    {
+      "epoch": 0.656959561738629,
+      "grad_norm": 0.10810303111522185,
+      "learning_rate": 5.5631588453158214e-05,
+      "loss": 1.2484,
+      "step": 9114
+    },
+    {
+      "epoch": 0.6570316442009659,
+      "grad_norm": 0.1498958682104326,
+      "learning_rate": 5.5610666336979454e-05,
+      "loss": 1.3042,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6571037266633029,
+      "grad_norm": 0.12588695162993405,
+      "learning_rate": 5.5589746640416465e-05,
+      "loss": 1.3524,
+      "step": 9116
+    },
+    {
+      "epoch": 0.6571758091256398,
+      "grad_norm": 0.10774352320436494,
+      "learning_rate": 5.5568829364609664e-05,
+      "loss": 1.2652,
+      "step": 9117
+    },
+    {
+      "epoch": 0.6572478915879767,
+      "grad_norm": 0.11230366041722592,
+      "learning_rate": 5.554791451069912e-05,
+      "loss": 1.2841,
+      "step": 9118
+    },
+    {
+      "epoch": 0.6573199740503135,
+      "grad_norm": 0.12050332844328392,
+      "learning_rate": 5.552700207982491e-05,
+      "loss": 1.2405,
+      "step": 9119
+    },
+    {
+      "epoch": 0.6573920565126504,
+      "grad_norm": 0.10894129900528943,
+      "learning_rate": 5.5506092073126914e-05,
+      "loss": 1.2663,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6574641389749873,
+      "grad_norm": 0.11048614854041557,
+      "learning_rate": 5.5485184491745043e-05,
+      "loss": 1.2691,
+      "step": 9121
+    },
+    {
+      "epoch": 0.6575362214373243,
+      "grad_norm": 0.15449955756569794,
+      "learning_rate": 5.5464279336818816e-05,
+      "loss": 1.3153,
+      "step": 9122
+    },
+    {
+      "epoch": 0.6576083038996612,
+      "grad_norm": 0.12104162733402964,
+      "learning_rate": 5.544337660948779e-05,
+      "loss": 1.2914,
+      "step": 9123
+    },
+    {
+      "epoch": 0.6576803863619981,
+      "grad_norm": 0.12175244282480586,
+      "learning_rate": 5.542247631089135e-05,
+      "loss": 1.2951,
+      "step": 9124
+    },
+    {
+      "epoch": 0.657752468824335,
+      "grad_norm": 0.12014344331941629,
+      "learning_rate": 5.540157844216876e-05,
+      "loss": 1.295,
+      "step": 9125
+    },
+    {
+      "epoch": 0.6578245512866719,
+      "grad_norm": 0.12129627602369783,
+      "learning_rate": 5.5380683004459124e-05,
+      "loss": 1.1974,
+      "step": 9126
+    },
+    {
+      "epoch": 0.6578966337490089,
+      "grad_norm": 0.1225729952234999,
+      "learning_rate": 5.535978999890144e-05,
+      "loss": 1.2567,
+      "step": 9127
+    },
+    {
+      "epoch": 0.6579687162113458,
+      "grad_norm": 0.12769936833313278,
+      "learning_rate": 5.5338899426634575e-05,
+      "loss": 1.2693,
+      "step": 9128
+    },
+    {
+      "epoch": 0.6580407986736827,
+      "grad_norm": 0.11438283595795196,
+      "learning_rate": 5.5318011288797237e-05,
+      "loss": 1.2437,
+      "step": 9129
+    },
+    {
+      "epoch": 0.6581128811360196,
+      "grad_norm": 0.10887883720554786,
+      "learning_rate": 5.529712558652802e-05,
+      "loss": 1.275,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6581849635983565,
+      "grad_norm": 0.1411645627244447,
+      "learning_rate": 5.5276242320965374e-05,
+      "loss": 1.3163,
+      "step": 9131
+    },
+    {
+      "epoch": 0.6582570460606935,
+      "grad_norm": 0.12536984164327952,
+      "learning_rate": 5.5255361493247634e-05,
+      "loss": 1.3848,
+      "step": 9132
+    },
+    {
+      "epoch": 0.6583291285230304,
+      "grad_norm": 0.12469382761871867,
+      "learning_rate": 5.5234483104513005e-05,
+      "loss": 1.2268,
+      "step": 9133
+    },
+    {
+      "epoch": 0.6584012109853673,
+      "grad_norm": 0.11229372393601132,
+      "learning_rate": 5.521360715589952e-05,
+      "loss": 1.249,
+      "step": 9134
+    },
+    {
+      "epoch": 0.6584732934477042,
+      "grad_norm": 0.1057365632976729,
+      "learning_rate": 5.5192733648545116e-05,
+      "loss": 1.2833,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6585453759100411,
+      "grad_norm": 0.11138863532158805,
+      "learning_rate": 5.5171862583587576e-05,
+      "loss": 1.32,
+      "step": 9136
+    },
+    {
+      "epoch": 0.658617458372378,
+      "grad_norm": 0.12491505507663292,
+      "learning_rate": 5.5150993962164574e-05,
+      "loss": 1.2403,
+      "step": 9137
+    },
+    {
+      "epoch": 0.6586895408347149,
+      "grad_norm": 0.12118792216786509,
+      "learning_rate": 5.513012778541364e-05,
+      "loss": 1.3091,
+      "step": 9138
+    },
+    {
+      "epoch": 0.6587616232970518,
+      "grad_norm": 0.13898133743983732,
+      "learning_rate": 5.510926405447215e-05,
+      "loss": 1.295,
+      "step": 9139
+    },
+    {
+      "epoch": 0.6588337057593887,
+      "grad_norm": 0.11947811964670015,
+      "learning_rate": 5.5088402770477376e-05,
+      "loss": 1.241,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6589057882217256,
+      "grad_norm": 0.12582283632979913,
+      "learning_rate": 5.506754393456645e-05,
+      "loss": 1.296,
+      "step": 9141
+    },
+    {
+      "epoch": 0.6589778706840625,
+      "grad_norm": 0.11512325422394663,
+      "learning_rate": 5.504668754787634e-05,
+      "loss": 1.1856,
+      "step": 9142
+    },
+    {
+      "epoch": 0.6590499531463995,
+      "grad_norm": 0.11111432153648819,
+      "learning_rate": 5.5025833611543945e-05,
+      "loss": 1.258,
+      "step": 9143
+    },
+    {
+      "epoch": 0.6591220356087364,
+      "grad_norm": 0.12256006871110484,
+      "learning_rate": 5.500498212670595e-05,
+      "loss": 1.3299,
+      "step": 9144
+    },
+    {
+      "epoch": 0.6591941180710733,
+      "grad_norm": 0.10922868074104884,
+      "learning_rate": 5.498413309449897e-05,
+      "loss": 1.1689,
+      "step": 9145
+    },
+    {
+      "epoch": 0.6592662005334102,
+      "grad_norm": 0.11256410606421088,
+      "learning_rate": 5.4963286516059496e-05,
+      "loss": 1.197,
+      "step": 9146
+    },
+    {
+      "epoch": 0.6593382829957471,
+      "grad_norm": 0.11975814157167945,
+      "learning_rate": 5.4942442392523754e-05,
+      "loss": 1.246,
+      "step": 9147
+    },
+    {
+      "epoch": 0.6594103654580841,
+      "grad_norm": 0.10638610286586997,
+      "learning_rate": 5.492160072502801e-05,
+      "loss": 1.3477,
+      "step": 9148
+    },
+    {
+      "epoch": 0.659482447920421,
+      "grad_norm": 0.1192709791012276,
+      "learning_rate": 5.490076151470833e-05,
+      "loss": 1.1593,
+      "step": 9149
+    },
+    {
+      "epoch": 0.6595545303827579,
+      "grad_norm": 0.11000380227987605,
+      "learning_rate": 5.4879924762700655e-05,
+      "loss": 1.2416,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6596266128450948,
+      "grad_norm": 0.12079835591707655,
+      "learning_rate": 5.485909047014066e-05,
+      "loss": 1.2879,
+      "step": 9151
+    },
+    {
+      "epoch": 0.6596986953074317,
+      "grad_norm": 0.1179489776612575,
+      "learning_rate": 5.48382586381641e-05,
+      "loss": 1.2399,
+      "step": 9152
+    },
+    {
+      "epoch": 0.6597707777697687,
+      "grad_norm": 0.1106732033398615,
+      "learning_rate": 5.481742926790649e-05,
+      "loss": 1.2197,
+      "step": 9153
+    },
+    {
+      "epoch": 0.6598428602321055,
+      "grad_norm": 0.10896176738272428,
+      "learning_rate": 5.4796602360503235e-05,
+      "loss": 1.2359,
+      "step": 9154
+    },
+    {
+      "epoch": 0.6599149426944424,
+      "grad_norm": 0.1152453449434305,
+      "learning_rate": 5.47757779170895e-05,
+      "loss": 1.2548,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6599870251567793,
+      "grad_norm": 0.11496330409320055,
+      "learning_rate": 5.4754955938800426e-05,
+      "loss": 1.3398,
+      "step": 9156
+    },
+    {
+      "epoch": 0.6600591076191162,
+      "grad_norm": 0.11355657404857937,
+      "learning_rate": 5.473413642677108e-05,
+      "loss": 1.2578,
+      "step": 9157
+    },
+    {
+      "epoch": 0.6601311900814532,
+      "grad_norm": 0.11643083839034687,
+      "learning_rate": 5.4713319382136274e-05,
+      "loss": 1.3084,
+      "step": 9158
+    },
+    {
+      "epoch": 0.6602032725437901,
+      "grad_norm": 0.13583118157356155,
+      "learning_rate": 5.4692504806030674e-05,
+      "loss": 1.336,
+      "step": 9159
+    },
+    {
+      "epoch": 0.660275355006127,
+      "grad_norm": 0.12092856552758491,
+      "learning_rate": 5.467169269958887e-05,
+      "loss": 1.2527,
+      "step": 9160
+    },
+    {
+      "epoch": 0.6603474374684639,
+      "grad_norm": 0.1253191935773658,
+      "learning_rate": 5.4650883063945366e-05,
+      "loss": 1.2926,
+      "step": 9161
+    },
+    {
+      "epoch": 0.6604195199308008,
+      "grad_norm": 0.11673717317930182,
+      "learning_rate": 5.463007590023448e-05,
+      "loss": 1.1431,
+      "step": 9162
+    },
+    {
+      "epoch": 0.6604916023931378,
+      "grad_norm": 0.11783444601618898,
+      "learning_rate": 5.460927120959029e-05,
+      "loss": 1.3702,
+      "step": 9163
+    },
+    {
+      "epoch": 0.6605636848554747,
+      "grad_norm": 0.12523493422537138,
+      "learning_rate": 5.458846899314687e-05,
+      "loss": 1.2683,
+      "step": 9164
+    },
+    {
+      "epoch": 0.6606357673178116,
+      "grad_norm": 0.12848775310944474,
+      "learning_rate": 5.45676692520382e-05,
+      "loss": 1.2367,
+      "step": 9165
+    },
+    {
+      "epoch": 0.6607078497801485,
+      "grad_norm": 0.1172759375084241,
+      "learning_rate": 5.454687198739804e-05,
+      "loss": 1.3848,
+      "step": 9166
+    },
+    {
+      "epoch": 0.6607799322424854,
+      "grad_norm": 0.1204077247895604,
+      "learning_rate": 5.452607720035995e-05,
+      "loss": 1.3185,
+      "step": 9167
+    },
+    {
+      "epoch": 0.6608520147048224,
+      "grad_norm": 0.12025193199442387,
+      "learning_rate": 5.4505284892057476e-05,
+      "loss": 1.2443,
+      "step": 9168
+    },
+    {
+      "epoch": 0.6609240971671593,
+      "grad_norm": 0.10870733042854794,
+      "learning_rate": 5.448449506362394e-05,
+      "loss": 1.2763,
+      "step": 9169
+    },
+    {
+      "epoch": 0.6609961796294962,
+      "grad_norm": 0.11502077037467345,
+      "learning_rate": 5.4463707716192694e-05,
+      "loss": 1.3727,
+      "step": 9170
+    },
+    {
+      "epoch": 0.661068262091833,
+      "grad_norm": 0.11563393922434216,
+      "learning_rate": 5.4442922850896726e-05,
+      "loss": 1.3025,
+      "step": 9171
+    },
+    {
+      "epoch": 0.6611403445541699,
+      "grad_norm": 0.11717382732508497,
+      "learning_rate": 5.442214046886902e-05,
+      "loss": 1.3236,
+      "step": 9172
+    },
+    {
+      "epoch": 0.6612124270165068,
+      "grad_norm": 0.11039369030987754,
+      "learning_rate": 5.4401360571242375e-05,
+      "loss": 1.2318,
+      "step": 9173
+    },
+    {
+      "epoch": 0.6612845094788438,
+      "grad_norm": 0.11716687846311019,
+      "learning_rate": 5.438058315914959e-05,
+      "loss": 1.294,
+      "step": 9174
+    },
+    {
+      "epoch": 0.6613565919411807,
+      "grad_norm": 0.1282537416563885,
+      "learning_rate": 5.435980823372311e-05,
+      "loss": 1.162,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6614286744035176,
+      "grad_norm": 0.12825442875993887,
+      "learning_rate": 5.4339035796095386e-05,
+      "loss": 1.186,
+      "step": 9176
+    },
+    {
+      "epoch": 0.6615007568658545,
+      "grad_norm": 0.11614426893084372,
+      "learning_rate": 5.431826584739872e-05,
+      "loss": 1.2096,
+      "step": 9177
+    },
+    {
+      "epoch": 0.6615728393281914,
+      "grad_norm": 0.11291234269728412,
+      "learning_rate": 5.4297498388765234e-05,
+      "loss": 1.3263,
+      "step": 9178
+    },
+    {
+      "epoch": 0.6616449217905284,
+      "grad_norm": 0.11750441098965368,
+      "learning_rate": 5.427673342132695e-05,
+      "loss": 1.2738,
+      "step": 9179
+    },
+    {
+      "epoch": 0.6617170042528653,
+      "grad_norm": 0.11860140044723143,
+      "learning_rate": 5.425597094621575e-05,
+      "loss": 1.2462,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6617890867152022,
+      "grad_norm": 0.11379183366310955,
+      "learning_rate": 5.423521096456336e-05,
+      "loss": 1.2619,
+      "step": 9181
+    },
+    {
+      "epoch": 0.6618611691775391,
+      "grad_norm": 0.12176622685773876,
+      "learning_rate": 5.421445347750141e-05,
+      "loss": 1.2946,
+      "step": 9182
+    },
+    {
+      "epoch": 0.661933251639876,
+      "grad_norm": 0.11410434413992028,
+      "learning_rate": 5.419369848616136e-05,
+      "loss": 1.3232,
+      "step": 9183
+    },
+    {
+      "epoch": 0.662005334102213,
+      "grad_norm": 0.11972705321688933,
+      "learning_rate": 5.417294599167453e-05,
+      "loss": 1.2796,
+      "step": 9184
+    },
+    {
+      "epoch": 0.6620774165645499,
+      "grad_norm": 0.12720232903237402,
+      "learning_rate": 5.415219599517212e-05,
+      "loss": 1.2667,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6621494990268868,
+      "grad_norm": 0.13669495880255386,
+      "learning_rate": 5.4131448497785196e-05,
+      "loss": 1.2461,
+      "step": 9186
+    },
+    {
+      "epoch": 0.6622215814892237,
+      "grad_norm": 0.12608257044210652,
+      "learning_rate": 5.4110703500644676e-05,
+      "loss": 1.2259,
+      "step": 9187
+    },
+    {
+      "epoch": 0.6622936639515606,
+      "grad_norm": 0.11809020469450686,
+      "learning_rate": 5.408996100488135e-05,
+      "loss": 1.1976,
+      "step": 9188
+    },
+    {
+      "epoch": 0.6623657464138974,
+      "grad_norm": 0.10573216413575759,
+      "learning_rate": 5.4069221011625884e-05,
+      "loss": 1.3032,
+      "step": 9189
+    },
+    {
+      "epoch": 0.6624378288762344,
+      "grad_norm": 0.13331733587024186,
+      "learning_rate": 5.4048483522008776e-05,
+      "loss": 1.3925,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6625099113385713,
+      "grad_norm": 0.12129180986081778,
+      "learning_rate": 5.402774853716041e-05,
+      "loss": 1.2696,
+      "step": 9191
+    },
+    {
+      "epoch": 0.6625819938009082,
+      "grad_norm": 0.12567160699902294,
+      "learning_rate": 5.4007016058211035e-05,
+      "loss": 1.2645,
+      "step": 9192
+    },
+    {
+      "epoch": 0.6626540762632451,
+      "grad_norm": 0.12837396374029336,
+      "learning_rate": 5.3986286086290746e-05,
+      "loss": 1.2493,
+      "step": 9193
+    },
+    {
+      "epoch": 0.662726158725582,
+      "grad_norm": 0.11645623793023423,
+      "learning_rate": 5.396555862252952e-05,
+      "loss": 1.2744,
+      "step": 9194
+    },
+    {
+      "epoch": 0.662798241187919,
+      "grad_norm": 0.13398710339989295,
+      "learning_rate": 5.3944833668057185e-05,
+      "loss": 1.3194,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6628703236502559,
+      "grad_norm": 0.1245861406837509,
+      "learning_rate": 5.3924111224003424e-05,
+      "loss": 1.2104,
+      "step": 9196
+    },
+    {
+      "epoch": 0.6629424061125928,
+      "grad_norm": 0.11555857077463354,
+      "learning_rate": 5.3903391291497815e-05,
+      "loss": 1.3182,
+      "step": 9197
+    },
+    {
+      "epoch": 0.6630144885749297,
+      "grad_norm": 0.12295552833072529,
+      "learning_rate": 5.3882673871669766e-05,
+      "loss": 1.2403,
+      "step": 9198
+    },
+    {
+      "epoch": 0.6630865710372666,
+      "grad_norm": 0.13388012302658575,
+      "learning_rate": 5.386195896564862e-05,
+      "loss": 1.2682,
+      "step": 9199
+    },
+    {
+      "epoch": 0.6631586534996036,
+      "grad_norm": 0.11490840042527659,
+      "learning_rate": 5.384124657456339e-05,
+      "loss": 1.2104,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6632307359619405,
+      "grad_norm": 0.11768588784647782,
+      "learning_rate": 5.382053669954321e-05,
+      "loss": 1.1978,
+      "step": 9201
+    },
+    {
+      "epoch": 0.6633028184242774,
+      "grad_norm": 0.11900300865714711,
+      "learning_rate": 5.37998293417169e-05,
+      "loss": 1.2419,
+      "step": 9202
+    },
+    {
+      "epoch": 0.6633749008866143,
+      "grad_norm": 0.13404807273018143,
+      "learning_rate": 5.377912450221326e-05,
+      "loss": 1.222,
+      "step": 9203
+    },
+    {
+      "epoch": 0.6634469833489512,
+      "grad_norm": 0.11842179033521377,
+      "learning_rate": 5.375842218216076e-05,
+      "loss": 1.2164,
+      "step": 9204
+    },
+    {
+      "epoch": 0.6635190658112882,
+      "grad_norm": 0.12190214728979999,
+      "learning_rate": 5.373772238268798e-05,
+      "loss": 1.2993,
+      "step": 9205
+    },
+    {
+      "epoch": 0.663591148273625,
+      "grad_norm": 0.11707420317668218,
+      "learning_rate": 5.37170251049232e-05,
+      "loss": 1.2797,
+      "step": 9206
+    },
+    {
+      "epoch": 0.6636632307359619,
+      "grad_norm": 0.1503914918094998,
+      "learning_rate": 5.369633034999466e-05,
+      "loss": 1.1302,
+      "step": 9207
+    },
+    {
+      "epoch": 0.6637353131982988,
+      "grad_norm": 0.10912610247503868,
+      "learning_rate": 5.367563811903027e-05,
+      "loss": 1.21,
+      "step": 9208
+    },
+    {
+      "epoch": 0.6638073956606357,
+      "grad_norm": 0.11878716224917063,
+      "learning_rate": 5.365494841315808e-05,
+      "loss": 1.2859,
+      "step": 9209
+    },
+    {
+      "epoch": 0.6638794781229727,
+      "grad_norm": 0.12250688677226819,
+      "learning_rate": 5.363426123350586e-05,
+      "loss": 1.3414,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6639515605853096,
+      "grad_norm": 0.12975308337837407,
+      "learning_rate": 5.361357658120114e-05,
+      "loss": 1.3044,
+      "step": 9211
+    },
+    {
+      "epoch": 0.6640236430476465,
+      "grad_norm": 0.12131787306994764,
+      "learning_rate": 5.35928944573715e-05,
+      "loss": 1.3425,
+      "step": 9212
+    },
+    {
+      "epoch": 0.6640957255099834,
+      "grad_norm": 0.12618055344627183,
+      "learning_rate": 5.3572214863144246e-05,
+      "loss": 1.2588,
+      "step": 9213
+    },
+    {
+      "epoch": 0.6641678079723203,
+      "grad_norm": 0.1364245851295228,
+      "learning_rate": 5.3551537799646704e-05,
+      "loss": 1.2916,
+      "step": 9214
+    },
+    {
+      "epoch": 0.6642398904346573,
+      "grad_norm": 0.13566407569093358,
+      "learning_rate": 5.353086326800586e-05,
+      "loss": 1.3121,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6643119728969942,
+      "grad_norm": 0.12308972762135961,
+      "learning_rate": 5.351019126934869e-05,
+      "loss": 1.2386,
+      "step": 9216
+    },
+    {
+      "epoch": 0.6643840553593311,
+      "grad_norm": 0.2183819860921875,
+      "learning_rate": 5.348952180480196e-05,
+      "loss": 1.2089,
+      "step": 9217
+    },
+    {
+      "epoch": 0.664456137821668,
+      "grad_norm": 0.11415864481212966,
+      "learning_rate": 5.346885487549247e-05,
+      "loss": 1.2658,
+      "step": 9218
+    },
+    {
+      "epoch": 0.6645282202840049,
+      "grad_norm": 0.12416210786885169,
+      "learning_rate": 5.3448190482546634e-05,
+      "loss": 1.366,
+      "step": 9219
+    },
+    {
+      "epoch": 0.6646003027463419,
+      "grad_norm": 0.12197248992711152,
+      "learning_rate": 5.342752862709088e-05,
+      "loss": 1.2379,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6646723852086788,
+      "grad_norm": 0.12364809172789885,
+      "learning_rate": 5.340686931025146e-05,
+      "loss": 1.3,
+      "step": 9221
+    },
+    {
+      "epoch": 0.6647444676710157,
+      "grad_norm": 0.1205029955029666,
+      "learning_rate": 5.33862125331545e-05,
+      "loss": 1.2262,
+      "step": 9222
+    },
+    {
+      "epoch": 0.6648165501333525,
+      "grad_norm": 0.11495665168268059,
+      "learning_rate": 5.336555829692598e-05,
+      "loss": 1.252,
+      "step": 9223
+    },
+    {
+      "epoch": 0.6648886325956894,
+      "grad_norm": 0.12575335979522534,
+      "learning_rate": 5.3344906602691726e-05,
+      "loss": 1.1922,
+      "step": 9224
+    },
+    {
+      "epoch": 0.6649607150580263,
+      "grad_norm": 0.10844570388877753,
+      "learning_rate": 5.332425745157745e-05,
+      "loss": 1.2814,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6650327975203633,
+      "grad_norm": 0.15660591488381131,
+      "learning_rate": 5.3303610844708726e-05,
+      "loss": 1.2705,
+      "step": 9226
+    },
+    {
+      "epoch": 0.6651048799827002,
+      "grad_norm": 0.1310420961364187,
+      "learning_rate": 5.3282966783210964e-05,
+      "loss": 1.2907,
+      "step": 9227
+    },
+    {
+      "epoch": 0.6651769624450371,
+      "grad_norm": 0.12320658129698275,
+      "learning_rate": 5.3262325268209454e-05,
+      "loss": 1.2863,
+      "step": 9228
+    },
+    {
+      "epoch": 0.665249044907374,
+      "grad_norm": 0.13720588411200202,
+      "learning_rate": 5.324168630082935e-05,
+      "loss": 1.276,
+      "step": 9229
+    },
+    {
+      "epoch": 0.6653211273697109,
+      "grad_norm": 0.11529824700848076,
+      "learning_rate": 5.322104988219565e-05,
+      "loss": 1.2436,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6653932098320479,
+      "grad_norm": 0.1230116153468517,
+      "learning_rate": 5.320041601343322e-05,
+      "loss": 1.2699,
+      "step": 9231
+    },
+    {
+      "epoch": 0.6654652922943848,
+      "grad_norm": 0.11921317284745088,
+      "learning_rate": 5.317978469566681e-05,
+      "loss": 1.277,
+      "step": 9232
+    },
+    {
+      "epoch": 0.6655373747567217,
+      "grad_norm": 0.11829252663170688,
+      "learning_rate": 5.3159155930021e-05,
+      "loss": 1.2342,
+      "step": 9233
+    },
+    {
+      "epoch": 0.6656094572190586,
+      "grad_norm": 0.13011030522230121,
+      "learning_rate": 5.313852971762023e-05,
+      "loss": 1.2378,
+      "step": 9234
+    },
+    {
+      "epoch": 0.6656815396813955,
+      "grad_norm": 0.23440177739994159,
+      "learning_rate": 5.311790605958884e-05,
+      "loss": 1.3547,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6657536221437325,
+      "grad_norm": 0.11767774156434303,
+      "learning_rate": 5.3097284957050974e-05,
+      "loss": 1.3517,
+      "step": 9236
+    },
+    {
+      "epoch": 0.6658257046060694,
+      "grad_norm": 0.1256721726962244,
+      "learning_rate": 5.307666641113069e-05,
+      "loss": 1.28,
+      "step": 9237
+    },
+    {
+      "epoch": 0.6658977870684063,
+      "grad_norm": 0.11698086989129153,
+      "learning_rate": 5.3056050422951874e-05,
+      "loss": 1.4162,
+      "step": 9238
+    },
+    {
+      "epoch": 0.6659698695307432,
+      "grad_norm": 0.11263269848822884,
+      "learning_rate": 5.3035436993638286e-05,
+      "loss": 1.2603,
+      "step": 9239
+    },
+    {
+      "epoch": 0.6660419519930801,
+      "grad_norm": 0.11705718922426706,
+      "learning_rate": 5.301482612431354e-05,
+      "loss": 1.2239,
+      "step": 9240
+    },
+    {
+      "epoch": 0.666114034455417,
+      "grad_norm": 0.14570518888948947,
+      "learning_rate": 5.29942178161011e-05,
+      "loss": 1.2595,
+      "step": 9241
+    },
+    {
+      "epoch": 0.6661861169177539,
+      "grad_norm": 0.1233216204820908,
+      "learning_rate": 5.297361207012432e-05,
+      "loss": 1.2249,
+      "step": 9242
+    },
+    {
+      "epoch": 0.6662581993800908,
+      "grad_norm": 0.13364052657666053,
+      "learning_rate": 5.295300888750644e-05,
+      "loss": 1.3146,
+      "step": 9243
+    },
+    {
+      "epoch": 0.6663302818424277,
+      "grad_norm": 0.11489228608549622,
+      "learning_rate": 5.293240826937038e-05,
+      "loss": 1.2836,
+      "step": 9244
+    },
+    {
+      "epoch": 0.6664023643047646,
+      "grad_norm": 0.11581345211820503,
+      "learning_rate": 5.29118102168392e-05,
+      "loss": 1.2263,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6664744467671015,
+      "grad_norm": 0.10629585748681047,
+      "learning_rate": 5.289121473103562e-05,
+      "loss": 1.224,
+      "step": 9246
+    },
+    {
+      "epoch": 0.6665465292294385,
+      "grad_norm": 0.126296456352531,
+      "learning_rate": 5.287062181308231e-05,
+      "loss": 1.2559,
+      "step": 9247
+    },
+    {
+      "epoch": 0.6666186116917754,
+      "grad_norm": 0.1192960523349481,
+      "learning_rate": 5.285003146410168e-05,
+      "loss": 1.2652,
+      "step": 9248
+    },
+    {
+      "epoch": 0.6666906941541123,
+      "grad_norm": 0.11992978822530259,
+      "learning_rate": 5.282944368521618e-05,
+      "loss": 1.2039,
+      "step": 9249
+    },
+    {
+      "epoch": 0.6667627766164492,
+      "grad_norm": 0.12873131833776225,
+      "learning_rate": 5.2808858477548014e-05,
+      "loss": 1.2342,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6668348590787861,
+      "grad_norm": 0.12135841873659375,
+      "learning_rate": 5.278827584221928e-05,
+      "loss": 1.3019,
+      "step": 9251
+    },
+    {
+      "epoch": 0.6669069415411231,
+      "grad_norm": 0.11810900707888829,
+      "learning_rate": 5.276769578035179e-05,
+      "loss": 1.2608,
+      "step": 9252
+    },
+    {
+      "epoch": 0.66697902400346,
+      "grad_norm": 0.11622576990824973,
+      "learning_rate": 5.274711829306749e-05,
+      "loss": 1.2955,
+      "step": 9253
+    },
+    {
+      "epoch": 0.6670511064657969,
+      "grad_norm": 0.10720026817046567,
+      "learning_rate": 5.272654338148796e-05,
+      "loss": 1.2151,
+      "step": 9254
+    },
+    {
+      "epoch": 0.6671231889281338,
+      "grad_norm": 0.10808632911843483,
+      "learning_rate": 5.270597104673478e-05,
+      "loss": 1.2029,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6671952713904707,
+      "grad_norm": 0.1253176527220381,
+      "learning_rate": 5.2685401289929245e-05,
+      "loss": 1.2746,
+      "step": 9256
+    },
+    {
+      "epoch": 0.6672673538528077,
+      "grad_norm": 0.11283822983502419,
+      "learning_rate": 5.2664834112192594e-05,
+      "loss": 1.3037,
+      "step": 9257
+    },
+    {
+      "epoch": 0.6673394363151445,
+      "grad_norm": 0.13002665930206558,
+      "learning_rate": 5.264426951464599e-05,
+      "loss": 1.2344,
+      "step": 9258
+    },
+    {
+      "epoch": 0.6674115187774814,
+      "grad_norm": 0.13801175887899514,
+      "learning_rate": 5.26237074984104e-05,
+      "loss": 1.2959,
+      "step": 9259
+    },
+    {
+      "epoch": 0.6674836012398183,
+      "grad_norm": 0.12301884100103008,
+      "learning_rate": 5.260314806460655e-05,
+      "loss": 1.2028,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6675556837021552,
+      "grad_norm": 0.14096065537996302,
+      "learning_rate": 5.2582591214355113e-05,
+      "loss": 1.3045,
+      "step": 9261
+    },
+    {
+      "epoch": 0.6676277661644922,
+      "grad_norm": 0.1285133853890683,
+      "learning_rate": 5.25620369487767e-05,
+      "loss": 1.2645,
+      "step": 9262
+    },
+    {
+      "epoch": 0.6676998486268291,
+      "grad_norm": 0.12300504390363942,
+      "learning_rate": 5.254148526899171e-05,
+      "loss": 1.28,
+      "step": 9263
+    },
+    {
+      "epoch": 0.667771931089166,
+      "grad_norm": 0.11241376080491278,
+      "learning_rate": 5.252093617612031e-05,
+      "loss": 1.2859,
+      "step": 9264
+    },
+    {
+      "epoch": 0.6678440135515029,
+      "grad_norm": 0.11159250138322263,
+      "learning_rate": 5.250038967128264e-05,
+      "loss": 1.2245,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6679160960138398,
+      "grad_norm": 0.12084542549272055,
+      "learning_rate": 5.2479845755598656e-05,
+      "loss": 1.2092,
+      "step": 9266
+    },
+    {
+      "epoch": 0.6679881784761768,
+      "grad_norm": 0.11768444775101985,
+      "learning_rate": 5.245930443018826e-05,
+      "loss": 1.2673,
+      "step": 9267
+    },
+    {
+      "epoch": 0.6680602609385137,
+      "grad_norm": 0.13149488845860915,
+      "learning_rate": 5.2438765696171054e-05,
+      "loss": 1.3002,
+      "step": 9268
+    },
+    {
+      "epoch": 0.6681323434008506,
+      "grad_norm": 0.11932077965765431,
+      "learning_rate": 5.241822955466662e-05,
+      "loss": 1.3236,
+      "step": 9269
+    },
+    {
+      "epoch": 0.6682044258631875,
+      "grad_norm": 0.12693923757541858,
+      "learning_rate": 5.2397696006794306e-05,
+      "loss": 1.2545,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6682765083255244,
+      "grad_norm": 0.11482510469894615,
+      "learning_rate": 5.2377165053673505e-05,
+      "loss": 1.2744,
+      "step": 9271
+    },
+    {
+      "epoch": 0.6683485907878614,
+      "grad_norm": 0.12757995732746993,
+      "learning_rate": 5.235663669642321e-05,
+      "loss": 1.2188,
+      "step": 9272
+    },
+    {
+      "epoch": 0.6684206732501983,
+      "grad_norm": 0.12048084587417322,
+      "learning_rate": 5.233611093616245e-05,
+      "loss": 1.276,
+      "step": 9273
+    },
+    {
+      "epoch": 0.6684927557125352,
+      "grad_norm": 0.1140141637358262,
+      "learning_rate": 5.231558777401002e-05,
+      "loss": 1.2524,
+      "step": 9274
+    },
+    {
+      "epoch": 0.668564838174872,
+      "grad_norm": 0.1737627449314489,
+      "learning_rate": 5.229506721108473e-05,
+      "loss": 1.258,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6686369206372089,
+      "grad_norm": 0.11998742922707613,
+      "learning_rate": 5.227454924850501e-05,
+      "loss": 1.2461,
+      "step": 9276
+    },
+    {
+      "epoch": 0.6687090030995458,
+      "grad_norm": 0.11758426542384338,
+      "learning_rate": 5.225403388738933e-05,
+      "loss": 1.3139,
+      "step": 9277
+    },
+    {
+      "epoch": 0.6687810855618828,
+      "grad_norm": 0.11813963286630867,
+      "learning_rate": 5.2233521128855956e-05,
+      "loss": 1.2487,
+      "step": 9278
+    },
+    {
+      "epoch": 0.6688531680242197,
+      "grad_norm": 0.3005526645953934,
+      "learning_rate": 5.2213010974023004e-05,
+      "loss": 1.2471,
+      "step": 9279
+    },
+    {
+      "epoch": 0.6689252504865566,
+      "grad_norm": 0.11064897144307777,
+      "learning_rate": 5.219250342400845e-05,
+      "loss": 1.3062,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6689973329488935,
+      "grad_norm": 0.11702760159905393,
+      "learning_rate": 5.2171998479930185e-05,
+      "loss": 1.3159,
+      "step": 9281
+    },
+    {
+      "epoch": 0.6690694154112304,
+      "grad_norm": 0.11686093307555512,
+      "learning_rate": 5.2151496142905876e-05,
+      "loss": 1.3133,
+      "step": 9282
+    },
+    {
+      "epoch": 0.6691414978735674,
+      "grad_norm": 0.11706199568289158,
+      "learning_rate": 5.213099641405309e-05,
+      "loss": 1.3635,
+      "step": 9283
+    },
+    {
+      "epoch": 0.6692135803359043,
+      "grad_norm": 0.11087648338682339,
+      "learning_rate": 5.211049929448926e-05,
+      "loss": 1.2316,
+      "step": 9284
+    },
+    {
+      "epoch": 0.6692856627982412,
+      "grad_norm": 0.11477443134616326,
+      "learning_rate": 5.209000478533163e-05,
+      "loss": 1.2361,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6693577452605781,
+      "grad_norm": 0.11863599180179772,
+      "learning_rate": 5.206951288769736e-05,
+      "loss": 1.2,
+      "step": 9286
+    },
+    {
+      "epoch": 0.669429827722915,
+      "grad_norm": 0.11250942222492162,
+      "learning_rate": 5.204902360270344e-05,
+      "loss": 1.2125,
+      "step": 9287
+    },
+    {
+      "epoch": 0.669501910185252,
+      "grad_norm": 0.10820033255307383,
+      "learning_rate": 5.202853693146672e-05,
+      "loss": 1.2617,
+      "step": 9288
+    },
+    {
+      "epoch": 0.6695739926475889,
+      "grad_norm": 0.11294703823032062,
+      "learning_rate": 5.2008052875103886e-05,
+      "loss": 1.2767,
+      "step": 9289
+    },
+    {
+      "epoch": 0.6696460751099258,
+      "grad_norm": 0.12433568529754116,
+      "learning_rate": 5.198757143473154e-05,
+      "loss": 1.3366,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6697181575722627,
+      "grad_norm": 0.12254893161558912,
+      "learning_rate": 5.196709261146606e-05,
+      "loss": 1.2549,
+      "step": 9291
+    },
+    {
+      "epoch": 0.6697902400345995,
+      "grad_norm": 0.12176254925901793,
+      "learning_rate": 5.194661640642377e-05,
+      "loss": 1.2609,
+      "step": 9292
+    },
+    {
+      "epoch": 0.6698623224969364,
+      "grad_norm": 0.11503491464427103,
+      "learning_rate": 5.192614282072077e-05,
+      "loss": 1.2807,
+      "step": 9293
+    },
+    {
+      "epoch": 0.6699344049592734,
+      "grad_norm": 0.13243060925561773,
+      "learning_rate": 5.190567185547306e-05,
+      "loss": 1.4084,
+      "step": 9294
+    },
+    {
+      "epoch": 0.6700064874216103,
+      "grad_norm": 0.11215547951587578,
+      "learning_rate": 5.188520351179651e-05,
+      "loss": 1.3036,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6700785698839472,
+      "grad_norm": 0.11639726333987195,
+      "learning_rate": 5.186473779080682e-05,
+      "loss": 1.2786,
+      "step": 9296
+    },
+    {
+      "epoch": 0.6701506523462841,
+      "grad_norm": 0.11168445786234751,
+      "learning_rate": 5.184427469361954e-05,
+      "loss": 1.207,
+      "step": 9297
+    },
+    {
+      "epoch": 0.670222734808621,
+      "grad_norm": 0.12027002399017193,
+      "learning_rate": 5.182381422135011e-05,
+      "loss": 1.206,
+      "step": 9298
+    },
+    {
+      "epoch": 0.670294817270958,
+      "grad_norm": 0.12191069203990333,
+      "learning_rate": 5.18033563751138e-05,
+      "loss": 1.3204,
+      "step": 9299
+    },
+    {
+      "epoch": 0.6703668997332949,
+      "grad_norm": 0.11416564517046617,
+      "learning_rate": 5.178290115602578e-05,
+      "loss": 1.3455,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6704389821956318,
+      "grad_norm": 0.11705336355685389,
+      "learning_rate": 5.176244856520094e-05,
+      "loss": 1.1983,
+      "step": 9301
+    },
+    {
+      "epoch": 0.6705110646579687,
+      "grad_norm": 0.11575701477434197,
+      "learning_rate": 5.174199860375424e-05,
+      "loss": 1.2873,
+      "step": 9302
+    },
+    {
+      "epoch": 0.6705831471203056,
+      "grad_norm": 0.13681031940442595,
+      "learning_rate": 5.1721551272800326e-05,
+      "loss": 1.3168,
+      "step": 9303
+    },
+    {
+      "epoch": 0.6706552295826426,
+      "grad_norm": 0.1250681952380854,
+      "learning_rate": 5.1701106573453836e-05,
+      "loss": 1.3507,
+      "step": 9304
+    },
+    {
+      "epoch": 0.6707273120449795,
+      "grad_norm": 0.11222036070393411,
+      "learning_rate": 5.1680664506829045e-05,
+      "loss": 1.2826,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6707993945073164,
+      "grad_norm": 0.11765890721314905,
+      "learning_rate": 5.1660225074040355e-05,
+      "loss": 1.2103,
+      "step": 9306
+    },
+    {
+      "epoch": 0.6708714769696533,
+      "grad_norm": 0.11357940160374373,
+      "learning_rate": 5.163978827620186e-05,
+      "loss": 1.4045,
+      "step": 9307
+    },
+    {
+      "epoch": 0.6709435594319902,
+      "grad_norm": 0.1180174067985736,
+      "learning_rate": 5.1619354114427595e-05,
+      "loss": 1.2741,
+      "step": 9308
+    },
+    {
+      "epoch": 0.6710156418943272,
+      "grad_norm": 0.1141306679488559,
+      "learning_rate": 5.15989225898313e-05,
+      "loss": 1.2709,
+      "step": 9309
+    },
+    {
+      "epoch": 0.671087724356664,
+      "grad_norm": 0.11838767219916371,
+      "learning_rate": 5.157849370352671e-05,
+      "loss": 1.2813,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6711598068190009,
+      "grad_norm": 0.12067230747950083,
+      "learning_rate": 5.155806745662743e-05,
+      "loss": 1.3005,
+      "step": 9311
+    },
+    {
+      "epoch": 0.6712318892813378,
+      "grad_norm": 0.10703303523452824,
+      "learning_rate": 5.15376438502469e-05,
+      "loss": 1.268,
+      "step": 9312
+    },
+    {
+      "epoch": 0.6713039717436747,
+      "grad_norm": 0.12431947973774006,
+      "learning_rate": 5.151722288549828e-05,
+      "loss": 1.1928,
+      "step": 9313
+    },
+    {
+      "epoch": 0.6713760542060117,
+      "grad_norm": 0.11940040721283644,
+      "learning_rate": 5.149680456349473e-05,
+      "loss": 1.2618,
+      "step": 9314
+    },
+    {
+      "epoch": 0.6714481366683486,
+      "grad_norm": 0.1280819624644067,
+      "learning_rate": 5.147638888534929e-05,
+      "loss": 1.329,
+      "step": 9315
+    },
+    {
+      "epoch": 0.6715202191306855,
+      "grad_norm": 0.11105301848158702,
+      "learning_rate": 5.1455975852174784e-05,
+      "loss": 1.23,
+      "step": 9316
+    },
+    {
+      "epoch": 0.6715923015930224,
+      "grad_norm": 0.12347195083948422,
+      "learning_rate": 5.143556546508386e-05,
+      "loss": 1.236,
+      "step": 9317
+    },
+    {
+      "epoch": 0.6716643840553593,
+      "grad_norm": 0.11650482198110693,
+      "learning_rate": 5.141515772518903e-05,
+      "loss": 1.2371,
+      "step": 9318
+    },
+    {
+      "epoch": 0.6717364665176963,
+      "grad_norm": 0.12881060081721765,
+      "learning_rate": 5.1394752633602816e-05,
+      "loss": 1.2978,
+      "step": 9319
+    },
+    {
+      "epoch": 0.6718085489800332,
+      "grad_norm": 0.11471092574869585,
+      "learning_rate": 5.1374350191437446e-05,
+      "loss": 1.2782,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6718806314423701,
+      "grad_norm": 0.1251110193477613,
+      "learning_rate": 5.135395039980497e-05,
+      "loss": 1.2136,
+      "step": 9321
+    },
+    {
+      "epoch": 0.671952713904707,
+      "grad_norm": 0.1107010753640715,
+      "learning_rate": 5.133355325981741e-05,
+      "loss": 1.2753,
+      "step": 9322
+    },
+    {
+      "epoch": 0.6720247963670439,
+      "grad_norm": 0.11996249118182681,
+      "learning_rate": 5.131315877258653e-05,
+      "loss": 1.2495,
+      "step": 9323
+    },
+    {
+      "epoch": 0.6720968788293809,
+      "grad_norm": 0.12141680032351561,
+      "learning_rate": 5.1292766939224133e-05,
+      "loss": 1.3001,
+      "step": 9324
+    },
+    {
+      "epoch": 0.6721689612917178,
+      "grad_norm": 0.13119628522263752,
+      "learning_rate": 5.1272377760841664e-05,
+      "loss": 1.255,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6722410437540547,
+      "grad_norm": 0.11491815086249092,
+      "learning_rate": 5.125199123855051e-05,
+      "loss": 1.3961,
+      "step": 9326
+    },
+    {
+      "epoch": 0.6723131262163915,
+      "grad_norm": 0.12124999967275492,
+      "learning_rate": 5.1231607373461934e-05,
+      "loss": 1.2593,
+      "step": 9327
+    },
+    {
+      "epoch": 0.6723852086787284,
+      "grad_norm": 0.11733789722556986,
+      "learning_rate": 5.121122616668712e-05,
+      "loss": 1.3017,
+      "step": 9328
+    },
+    {
+      "epoch": 0.6724572911410653,
+      "grad_norm": 0.12710187236049653,
+      "learning_rate": 5.1190847619336904e-05,
+      "loss": 1.3021,
+      "step": 9329
+    },
+    {
+      "epoch": 0.6725293736034023,
+      "grad_norm": 0.12972155884787936,
+      "learning_rate": 5.1170471732522165e-05,
+      "loss": 1.2642,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6726014560657392,
+      "grad_norm": 0.12247983054170929,
+      "learning_rate": 5.115009850735355e-05,
+      "loss": 1.2542,
+      "step": 9331
+    },
+    {
+      "epoch": 0.6726735385280761,
+      "grad_norm": 0.13626258880418568,
+      "learning_rate": 5.1129727944941594e-05,
+      "loss": 1.2952,
+      "step": 9332
+    },
+    {
+      "epoch": 0.672745620990413,
+      "grad_norm": 0.12183350395329474,
+      "learning_rate": 5.110936004639668e-05,
+      "loss": 1.2308,
+      "step": 9333
+    },
+    {
+      "epoch": 0.6728177034527499,
+      "grad_norm": 0.11829221500971919,
+      "learning_rate": 5.108899481282903e-05,
+      "loss": 1.2766,
+      "step": 9334
+    },
+    {
+      "epoch": 0.6728897859150869,
+      "grad_norm": 0.12117716875871491,
+      "learning_rate": 5.1068632245348726e-05,
+      "loss": 1.2849,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6729618683774238,
+      "grad_norm": 0.11197175937557523,
+      "learning_rate": 5.104827234506572e-05,
+      "loss": 1.1697,
+      "step": 9336
+    },
+    {
+      "epoch": 0.6730339508397607,
+      "grad_norm": 0.11365908959220639,
+      "learning_rate": 5.1027915113089796e-05,
+      "loss": 1.2318,
+      "step": 9337
+    },
+    {
+      "epoch": 0.6731060333020976,
+      "grad_norm": 0.11753752939896625,
+      "learning_rate": 5.100756055053063e-05,
+      "loss": 1.2374,
+      "step": 9338
+    },
+    {
+      "epoch": 0.6731781157644345,
+      "grad_norm": 0.13922393883521522,
+      "learning_rate": 5.098720865849771e-05,
+      "loss": 1.3834,
+      "step": 9339
+    },
+    {
+      "epoch": 0.6732501982267715,
+      "grad_norm": 0.11582205646093291,
+      "learning_rate": 5.096685943810039e-05,
+      "loss": 1.3398,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6733222806891084,
+      "grad_norm": 0.12258992861412629,
+      "learning_rate": 5.0946512890447896e-05,
+      "loss": 1.2801,
+      "step": 9341
+    },
+    {
+      "epoch": 0.6733943631514453,
+      "grad_norm": 0.12048971342830267,
+      "learning_rate": 5.09261690166493e-05,
+      "loss": 1.353,
+      "step": 9342
+    },
+    {
+      "epoch": 0.6734664456137822,
+      "grad_norm": 0.11628053743189253,
+      "learning_rate": 5.0905827817813526e-05,
+      "loss": 1.2454,
+      "step": 9343
+    },
+    {
+      "epoch": 0.673538528076119,
+      "grad_norm": 0.12004313325464115,
+      "learning_rate": 5.088548929504933e-05,
+      "loss": 1.2598,
+      "step": 9344
+    },
+    {
+      "epoch": 0.673610610538456,
+      "grad_norm": 0.11441756251623557,
+      "learning_rate": 5.086515344946536e-05,
+      "loss": 1.316,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6736826930007929,
+      "grad_norm": 0.11262042225924553,
+      "learning_rate": 5.08448202821701e-05,
+      "loss": 1.2892,
+      "step": 9346
+    },
+    {
+      "epoch": 0.6737547754631298,
+      "grad_norm": 0.19772388513865635,
+      "learning_rate": 5.08244897942719e-05,
+      "loss": 1.3046,
+      "step": 9347
+    },
+    {
+      "epoch": 0.6738268579254667,
+      "grad_norm": 0.26811515529238583,
+      "learning_rate": 5.080416198687894e-05,
+      "loss": 1.2725,
+      "step": 9348
+    },
+    {
+      "epoch": 0.6738989403878036,
+      "grad_norm": 0.12871477950067955,
+      "learning_rate": 5.078383686109926e-05,
+      "loss": 1.3441,
+      "step": 9349
+    },
+    {
+      "epoch": 0.6739710228501405,
+      "grad_norm": 0.11530430622186563,
+      "learning_rate": 5.076351441804079e-05,
+      "loss": 1.2495,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6740431053124775,
+      "grad_norm": 0.12853419117182682,
+      "learning_rate": 5.074319465881126e-05,
+      "loss": 1.1977,
+      "step": 9351
+    },
+    {
+      "epoch": 0.6741151877748144,
+      "grad_norm": 0.1114034975702743,
+      "learning_rate": 5.0722877584518266e-05,
+      "loss": 1.2011,
+      "step": 9352
+    },
+    {
+      "epoch": 0.6741872702371513,
+      "grad_norm": 0.10878073541759342,
+      "learning_rate": 5.070256319626935e-05,
+      "loss": 1.3037,
+      "step": 9353
+    },
+    {
+      "epoch": 0.6742593526994882,
+      "grad_norm": 0.12237789297596206,
+      "learning_rate": 5.068225149517167e-05,
+      "loss": 1.2424,
+      "step": 9354
+    },
+    {
+      "epoch": 0.6743314351618251,
+      "grad_norm": 0.10872346333115503,
+      "learning_rate": 5.0661942482332556e-05,
+      "loss": 1.2597,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6744035176241621,
+      "grad_norm": 0.12241004050519372,
+      "learning_rate": 5.0641636158858954e-05,
+      "loss": 1.3547,
+      "step": 9356
+    },
+    {
+      "epoch": 0.674475600086499,
+      "grad_norm": 0.11605798344817689,
+      "learning_rate": 5.062133252585779e-05,
+      "loss": 1.2684,
+      "step": 9357
+    },
+    {
+      "epoch": 0.6745476825488359,
+      "grad_norm": 0.11115666035906699,
+      "learning_rate": 5.0601031584435675e-05,
+      "loss": 1.2124,
+      "step": 9358
+    },
+    {
+      "epoch": 0.6746197650111728,
+      "grad_norm": 0.11899629927718247,
+      "learning_rate": 5.058073333569933e-05,
+      "loss": 1.3045,
+      "step": 9359
+    },
+    {
+      "epoch": 0.6746918474735097,
+      "grad_norm": 0.1222136132086362,
+      "learning_rate": 5.056043778075512e-05,
+      "loss": 1.2917,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6747639299358467,
+      "grad_norm": 0.11723730519041377,
+      "learning_rate": 5.05401449207094e-05,
+      "loss": 1.1952,
+      "step": 9361
+    },
+    {
+      "epoch": 0.6748360123981835,
+      "grad_norm": 0.11627127750437566,
+      "learning_rate": 5.0519854756668184e-05,
+      "loss": 1.3128,
+      "step": 9362
+    },
+    {
+      "epoch": 0.6749080948605204,
+      "grad_norm": 0.12053214380255384,
+      "learning_rate": 5.0499567289737584e-05,
+      "loss": 1.2091,
+      "step": 9363
+    },
+    {
+      "epoch": 0.6749801773228573,
+      "grad_norm": 0.11617686488664511,
+      "learning_rate": 5.047928252102345e-05,
+      "loss": 1.3557,
+      "step": 9364
+    },
+    {
+      "epoch": 0.6750522597851942,
+      "grad_norm": 0.16263594085361338,
+      "learning_rate": 5.04590004516314e-05,
+      "loss": 1.3292,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6751243422475312,
+      "grad_norm": 0.11397565851079998,
+      "learning_rate": 5.043872108266703e-05,
+      "loss": 1.2193,
+      "step": 9366
+    },
+    {
+      "epoch": 0.6751964247098681,
+      "grad_norm": 0.21552573831346916,
+      "learning_rate": 5.041844441523571e-05,
+      "loss": 1.3538,
+      "step": 9367
+    },
+    {
+      "epoch": 0.675268507172205,
+      "grad_norm": 0.11653946076888005,
+      "learning_rate": 5.039817045044282e-05,
+      "loss": 1.1043,
+      "step": 9368
+    },
+    {
+      "epoch": 0.6753405896345419,
+      "grad_norm": 0.121625271868639,
+      "learning_rate": 5.037789918939335e-05,
+      "loss": 1.1839,
+      "step": 9369
+    },
+    {
+      "epoch": 0.6754126720968788,
+      "grad_norm": 0.11928897141415268,
+      "learning_rate": 5.035763063319232e-05,
+      "loss": 1.3405,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6754847545592157,
+      "grad_norm": 0.11217849215522517,
+      "learning_rate": 5.0337364782944475e-05,
+      "loss": 1.3051,
+      "step": 9371
+    },
+    {
+      "epoch": 0.6755568370215527,
+      "grad_norm": 0.10665678964165493,
+      "learning_rate": 5.031710163975463e-05,
+      "loss": 1.2106,
+      "step": 9372
+    },
+    {
+      "epoch": 0.6756289194838896,
+      "grad_norm": 0.1278852451063035,
+      "learning_rate": 5.029684120472719e-05,
+      "loss": 1.2969,
+      "step": 9373
+    },
+    {
+      "epoch": 0.6757010019462265,
+      "grad_norm": 0.12094163112584859,
+      "learning_rate": 5.0276583478966556e-05,
+      "loss": 1.3311,
+      "step": 9374
+    },
+    {
+      "epoch": 0.6757730844085634,
+      "grad_norm": 0.11894702573830607,
+      "learning_rate": 5.025632846357695e-05,
+      "loss": 1.2472,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6758451668709003,
+      "grad_norm": 0.1345767056756199,
+      "learning_rate": 5.023607615966247e-05,
+      "loss": 1.2983,
+      "step": 9376
+    },
+    {
+      "epoch": 0.6759172493332373,
+      "grad_norm": 0.12600061112978717,
+      "learning_rate": 5.021582656832704e-05,
+      "loss": 1.1759,
+      "step": 9377
+    },
+    {
+      "epoch": 0.6759893317955742,
+      "grad_norm": 0.13013017990808665,
+      "learning_rate": 5.0195579690674447e-05,
+      "loss": 1.3551,
+      "step": 9378
+    },
+    {
+      "epoch": 0.676061414257911,
+      "grad_norm": 0.1169366553713186,
+      "learning_rate": 5.017533552780832e-05,
+      "loss": 1.2423,
+      "step": 9379
+    },
+    {
+      "epoch": 0.6761334967202479,
+      "grad_norm": 0.13565371292743617,
+      "learning_rate": 5.015509408083215e-05,
+      "loss": 1.2944,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6762055791825848,
+      "grad_norm": 0.11689926027252065,
+      "learning_rate": 5.0134855350849284e-05,
+      "loss": 1.277,
+      "step": 9381
+    },
+    {
+      "epoch": 0.6762776616449218,
+      "grad_norm": 0.13012329374156748,
+      "learning_rate": 5.011461933896291e-05,
+      "loss": 1.3391,
+      "step": 9382
+    },
+    {
+      "epoch": 0.6763497441072587,
+      "grad_norm": 0.12368069645195057,
+      "learning_rate": 5.0094386046276054e-05,
+      "loss": 1.3103,
+      "step": 9383
+    },
+    {
+      "epoch": 0.6764218265695956,
+      "grad_norm": 0.11860811776717,
+      "learning_rate": 5.007415547389165e-05,
+      "loss": 1.2948,
+      "step": 9384
+    },
+    {
+      "epoch": 0.6764939090319325,
+      "grad_norm": 0.11512622166281845,
+      "learning_rate": 5.00539276229124e-05,
+      "loss": 1.2231,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6765659914942694,
+      "grad_norm": 0.12106434026552437,
+      "learning_rate": 5.003370249444093e-05,
+      "loss": 1.2152,
+      "step": 9386
+    },
+    {
+      "epoch": 0.6766380739566064,
+      "grad_norm": 0.12904171009464396,
+      "learning_rate": 5.001348008957969e-05,
+      "loss": 1.2375,
+      "step": 9387
+    },
+    {
+      "epoch": 0.6767101564189433,
+      "grad_norm": 0.11380628383907161,
+      "learning_rate": 4.999326040943098e-05,
+      "loss": 1.3868,
+      "step": 9388
+    },
+    {
+      "epoch": 0.6767822388812802,
+      "grad_norm": 0.11777227360063908,
+      "learning_rate": 4.997304345509695e-05,
+      "loss": 1.233,
+      "step": 9389
+    },
+    {
+      "epoch": 0.6768543213436171,
+      "grad_norm": 0.12069813758955876,
+      "learning_rate": 4.9952829227679597e-05,
+      "loss": 1.2761,
+      "step": 9390
+    },
+    {
+      "epoch": 0.676926403805954,
+      "grad_norm": 0.12103641233903785,
+      "learning_rate": 4.9932617728280795e-05,
+      "loss": 1.1921,
+      "step": 9391
+    },
+    {
+      "epoch": 0.676998486268291,
+      "grad_norm": 0.12034439269858696,
+      "learning_rate": 4.9912408958002234e-05,
+      "loss": 1.3157,
+      "step": 9392
+    },
+    {
+      "epoch": 0.6770705687306279,
+      "grad_norm": 0.11110977285962542,
+      "learning_rate": 4.9892202917945484e-05,
+      "loss": 1.3384,
+      "step": 9393
+    },
+    {
+      "epoch": 0.6771426511929648,
+      "grad_norm": 0.11424730166297346,
+      "learning_rate": 4.9871999609211964e-05,
+      "loss": 1.2053,
+      "step": 9394
+    },
+    {
+      "epoch": 0.6772147336553017,
+      "grad_norm": 0.12470600695335429,
+      "learning_rate": 4.9851799032902913e-05,
+      "loss": 1.26,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6772868161176385,
+      "grad_norm": 0.12658057935590927,
+      "learning_rate": 4.983160119011945e-05,
+      "loss": 1.2959,
+      "step": 9396
+    },
+    {
+      "epoch": 0.6773588985799754,
+      "grad_norm": 0.12111768219690847,
+      "learning_rate": 4.9811406081962586e-05,
+      "loss": 1.212,
+      "step": 9397
+    },
+    {
+      "epoch": 0.6774309810423124,
+      "grad_norm": 0.11347838644115002,
+      "learning_rate": 4.9791213709533006e-05,
+      "loss": 1.2425,
+      "step": 9398
+    },
+    {
+      "epoch": 0.6775030635046493,
+      "grad_norm": 0.11805185662521168,
+      "learning_rate": 4.97710240739315e-05,
+      "loss": 1.2578,
+      "step": 9399
+    },
+    {
+      "epoch": 0.6775751459669862,
+      "grad_norm": 0.10523328140341509,
+      "learning_rate": 4.975083717625855e-05,
+      "loss": 1.2497,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6776472284293231,
+      "grad_norm": 0.1273573830948896,
+      "learning_rate": 4.973065301761455e-05,
+      "loss": 1.3203,
+      "step": 9401
+    },
+    {
+      "epoch": 0.67771931089166,
+      "grad_norm": 0.13028015245723734,
+      "learning_rate": 4.9710471599099605e-05,
+      "loss": 1.2199,
+      "step": 9402
+    },
+    {
+      "epoch": 0.677791393353997,
+      "grad_norm": 0.12241940971705191,
+      "learning_rate": 4.969029292181391e-05,
+      "loss": 1.2992,
+      "step": 9403
+    },
+    {
+      "epoch": 0.6778634758163339,
+      "grad_norm": 0.11112202133213066,
+      "learning_rate": 4.967011698685732e-05,
+      "loss": 1.2121,
+      "step": 9404
+    },
+    {
+      "epoch": 0.6779355582786708,
+      "grad_norm": 0.12473270836995634,
+      "learning_rate": 4.9649943795329654e-05,
+      "loss": 1.3377,
+      "step": 9405
+    },
+    {
+      "epoch": 0.6780076407410077,
+      "grad_norm": 0.13074000639455705,
+      "learning_rate": 4.962977334833048e-05,
+      "loss": 1.2756,
+      "step": 9406
+    },
+    {
+      "epoch": 0.6780797232033446,
+      "grad_norm": 0.12143223377539974,
+      "learning_rate": 4.9609605646959226e-05,
+      "loss": 1.3094,
+      "step": 9407
+    },
+    {
+      "epoch": 0.6781518056656816,
+      "grad_norm": 0.14349484736638168,
+      "learning_rate": 4.958944069231531e-05,
+      "loss": 1.2351,
+      "step": 9408
+    },
+    {
+      "epoch": 0.6782238881280185,
+      "grad_norm": 0.1531174819208378,
+      "learning_rate": 4.9569278485497906e-05,
+      "loss": 1.3101,
+      "step": 9409
+    },
+    {
+      "epoch": 0.6782959705903554,
+      "grad_norm": 0.1282599147552628,
+      "learning_rate": 4.954911902760594e-05,
+      "loss": 1.3498,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6783680530526923,
+      "grad_norm": 0.11017606039944167,
+      "learning_rate": 4.9528962319738304e-05,
+      "loss": 1.2284,
+      "step": 9411
+    },
+    {
+      "epoch": 0.6784401355150292,
+      "grad_norm": 0.12544850509288305,
+      "learning_rate": 4.9508808362993785e-05,
+      "loss": 1.3536,
+      "step": 9412
+    },
+    {
+      "epoch": 0.6785122179773662,
+      "grad_norm": 0.11599225320591447,
+      "learning_rate": 4.948865715847096e-05,
+      "loss": 1.3134,
+      "step": 9413
+    },
+    {
+      "epoch": 0.678584300439703,
+      "grad_norm": 0.14966183533210117,
+      "learning_rate": 4.946850870726817e-05,
+      "loss": 1.3765,
+      "step": 9414
+    },
+    {
+      "epoch": 0.6786563829020399,
+      "grad_norm": 0.13517403487003057,
+      "learning_rate": 4.944836301048368e-05,
+      "loss": 1.2568,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6787284653643768,
+      "grad_norm": 0.11344254855443822,
+      "learning_rate": 4.9428220069215694e-05,
+      "loss": 1.2498,
+      "step": 9416
+    },
+    {
+      "epoch": 0.6788005478267137,
+      "grad_norm": 0.11908505555245534,
+      "learning_rate": 4.940807988456217e-05,
+      "loss": 1.2993,
+      "step": 9417
+    },
+    {
+      "epoch": 0.6788726302890506,
+      "grad_norm": 0.11087555638069922,
+      "learning_rate": 4.938794245762088e-05,
+      "loss": 1.235,
+      "step": 9418
+    },
+    {
+      "epoch": 0.6789447127513876,
+      "grad_norm": 0.10799531210318702,
+      "learning_rate": 4.9367807789489504e-05,
+      "loss": 1.2419,
+      "step": 9419
+    },
+    {
+      "epoch": 0.6790167952137245,
+      "grad_norm": 0.11403572857725483,
+      "learning_rate": 4.934767588126554e-05,
+      "loss": 1.2406,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6790888776760614,
+      "grad_norm": 0.11441539193855763,
+      "learning_rate": 4.932754673404646e-05,
+      "loss": 1.2242,
+      "step": 9421
+    },
+    {
+      "epoch": 0.6791609601383983,
+      "grad_norm": 0.12280691614101277,
+      "learning_rate": 4.930742034892938e-05,
+      "loss": 1.28,
+      "step": 9422
+    },
+    {
+      "epoch": 0.6792330426007352,
+      "grad_norm": 0.10219264922758436,
+      "learning_rate": 4.928729672701139e-05,
+      "loss": 1.2158,
+      "step": 9423
+    },
+    {
+      "epoch": 0.6793051250630722,
+      "grad_norm": 0.1300287790447778,
+      "learning_rate": 4.926717586938938e-05,
+      "loss": 1.3612,
+      "step": 9424
+    },
+    {
+      "epoch": 0.6793772075254091,
+      "grad_norm": 0.10943498308320385,
+      "learning_rate": 4.9247057777160245e-05,
+      "loss": 1.2825,
+      "step": 9425
+    },
+    {
+      "epoch": 0.679449289987746,
+      "grad_norm": 0.11939462796392004,
+      "learning_rate": 4.922694245142045e-05,
+      "loss": 1.2755,
+      "step": 9426
+    },
+    {
+      "epoch": 0.6795213724500829,
+      "grad_norm": 0.12631755103835998,
+      "learning_rate": 4.920682989326654e-05,
+      "loss": 1.294,
+      "step": 9427
+    },
+    {
+      "epoch": 0.6795934549124198,
+      "grad_norm": 0.11938977560779127,
+      "learning_rate": 4.91867201037948e-05,
+      "loss": 1.3019,
+      "step": 9428
+    },
+    {
+      "epoch": 0.6796655373747568,
+      "grad_norm": 0.12953623868618516,
+      "learning_rate": 4.91666130841014e-05,
+      "loss": 1.2959,
+      "step": 9429
+    },
+    {
+      "epoch": 0.6797376198370937,
+      "grad_norm": 0.12374273721384761,
+      "learning_rate": 4.9146508835282354e-05,
+      "loss": 1.363,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6798097022994305,
+      "grad_norm": 0.13031145539426497,
+      "learning_rate": 4.9126407358433535e-05,
+      "loss": 1.365,
+      "step": 9431
+    },
+    {
+      "epoch": 0.6798817847617674,
+      "grad_norm": 0.12815162555148715,
+      "learning_rate": 4.910630865465063e-05,
+      "loss": 1.2162,
+      "step": 9432
+    },
+    {
+      "epoch": 0.6799538672241043,
+      "grad_norm": 0.11024407369836958,
+      "learning_rate": 4.9086212725029226e-05,
+      "loss": 1.1945,
+      "step": 9433
+    },
+    {
+      "epoch": 0.6800259496864413,
+      "grad_norm": 0.12580226209624315,
+      "learning_rate": 4.906611957066472e-05,
+      "loss": 1.2743,
+      "step": 9434
+    },
+    {
+      "epoch": 0.6800980321487782,
+      "grad_norm": 0.12199366828495067,
+      "learning_rate": 4.904602919265234e-05,
+      "loss": 1.3914,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6801701146111151,
+      "grad_norm": 0.28143825312490894,
+      "learning_rate": 4.902594159208723e-05,
+      "loss": 1.2457,
+      "step": 9436
+    },
+    {
+      "epoch": 0.680242197073452,
+      "grad_norm": 0.11796549176997002,
+      "learning_rate": 4.900585677006434e-05,
+      "loss": 1.2537,
+      "step": 9437
+    },
+    {
+      "epoch": 0.6803142795357889,
+      "grad_norm": 0.1800458799578887,
+      "learning_rate": 4.8985774727678446e-05,
+      "loss": 1.3,
+      "step": 9438
+    },
+    {
+      "epoch": 0.6803863619981259,
+      "grad_norm": 0.13510539604591737,
+      "learning_rate": 4.896569546602423e-05,
+      "loss": 1.3367,
+      "step": 9439
+    },
+    {
+      "epoch": 0.6804584444604628,
+      "grad_norm": 0.12364442774786517,
+      "learning_rate": 4.8945618986196175e-05,
+      "loss": 1.3262,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6805305269227997,
+      "grad_norm": 0.12169627193152464,
+      "learning_rate": 4.892554528928862e-05,
+      "loss": 1.1213,
+      "step": 9441
+    },
+    {
+      "epoch": 0.6806026093851366,
+      "grad_norm": 0.1342503683101204,
+      "learning_rate": 4.890547437639578e-05,
+      "loss": 1.2617,
+      "step": 9442
+    },
+    {
+      "epoch": 0.6806746918474735,
+      "grad_norm": 0.1293947557062176,
+      "learning_rate": 4.8885406248611686e-05,
+      "loss": 1.2563,
+      "step": 9443
+    },
+    {
+      "epoch": 0.6807467743098105,
+      "grad_norm": 0.115173515382179,
+      "learning_rate": 4.8865340907030246e-05,
+      "loss": 1.2475,
+      "step": 9444
+    },
+    {
+      "epoch": 0.6808188567721474,
+      "grad_norm": 0.1224454499146634,
+      "learning_rate": 4.8845278352745175e-05,
+      "loss": 1.3177,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6808909392344843,
+      "grad_norm": 0.14584256991733197,
+      "learning_rate": 4.8825218586850085e-05,
+      "loss": 1.2984,
+      "step": 9446
+    },
+    {
+      "epoch": 0.6809630216968212,
+      "grad_norm": 0.12441794930897364,
+      "learning_rate": 4.880516161043841e-05,
+      "loss": 1.2794,
+      "step": 9447
+    },
+    {
+      "epoch": 0.681035104159158,
+      "grad_norm": 0.11746389701352052,
+      "learning_rate": 4.8785107424603414e-05,
+      "loss": 1.3401,
+      "step": 9448
+    },
+    {
+      "epoch": 0.6811071866214949,
+      "grad_norm": 0.17609468963803232,
+      "learning_rate": 4.876505603043826e-05,
+      "loss": 1.281,
+      "step": 9449
+    },
+    {
+      "epoch": 0.6811792690838319,
+      "grad_norm": 0.11828645041813907,
+      "learning_rate": 4.874500742903595e-05,
+      "loss": 1.2066,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6812513515461688,
+      "grad_norm": 0.10958166912284269,
+      "learning_rate": 4.8724961621489194e-05,
+      "loss": 1.3606,
+      "step": 9451
+    },
+    {
+      "epoch": 0.6813234340085057,
+      "grad_norm": 0.11480588485625914,
+      "learning_rate": 4.870491860889079e-05,
+      "loss": 1.2769,
+      "step": 9452
+    },
+    {
+      "epoch": 0.6813955164708426,
+      "grad_norm": 0.12988527866228156,
+      "learning_rate": 4.8684878392333225e-05,
+      "loss": 1.2185,
+      "step": 9453
+    },
+    {
+      "epoch": 0.6814675989331795,
+      "grad_norm": 0.10982679162584495,
+      "learning_rate": 4.8664840972908885e-05,
+      "loss": 1.3045,
+      "step": 9454
+    },
+    {
+      "epoch": 0.6815396813955165,
+      "grad_norm": 0.11837987700441122,
+      "learning_rate": 4.8644806351709915e-05,
+      "loss": 1.238,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6816117638578534,
+      "grad_norm": 0.1295906568860341,
+      "learning_rate": 4.8624774529828464e-05,
+      "loss": 1.1507,
+      "step": 9456
+    },
+    {
+      "epoch": 0.6816838463201903,
+      "grad_norm": 0.11200275743316748,
+      "learning_rate": 4.860474550835642e-05,
+      "loss": 1.2158,
+      "step": 9457
+    },
+    {
+      "epoch": 0.6817559287825272,
+      "grad_norm": 0.1155367861076472,
+      "learning_rate": 4.8584719288385575e-05,
+      "loss": 1.3341,
+      "step": 9458
+    },
+    {
+      "epoch": 0.6818280112448641,
+      "grad_norm": 0.1168146551979583,
+      "learning_rate": 4.856469587100743e-05,
+      "loss": 1.2697,
+      "step": 9459
+    },
+    {
+      "epoch": 0.6819000937072011,
+      "grad_norm": 0.12404451082231237,
+      "learning_rate": 4.8544675257313554e-05,
+      "loss": 1.3155,
+      "step": 9460
+    },
+    {
+      "epoch": 0.681972176169538,
+      "grad_norm": 0.12265859283886786,
+      "learning_rate": 4.85246574483952e-05,
+      "loss": 1.2986,
+      "step": 9461
+    },
+    {
+      "epoch": 0.6820442586318749,
+      "grad_norm": 0.11418656398085873,
+      "learning_rate": 4.8504642445343575e-05,
+      "loss": 1.3458,
+      "step": 9462
+    },
+    {
+      "epoch": 0.6821163410942118,
+      "grad_norm": 0.11320005011792254,
+      "learning_rate": 4.8484630249249584e-05,
+      "loss": 1.2328,
+      "step": 9463
+    },
+    {
+      "epoch": 0.6821884235565487,
+      "grad_norm": 0.10808536870359473,
+      "learning_rate": 4.846462086120407e-05,
+      "loss": 1.2692,
+      "step": 9464
+    },
+    {
+      "epoch": 0.6822605060188857,
+      "grad_norm": 0.12897213133856142,
+      "learning_rate": 4.844461428229782e-05,
+      "loss": 1.2733,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6823325884812225,
+      "grad_norm": 0.11513747478106104,
+      "learning_rate": 4.842461051362135e-05,
+      "loss": 1.3264,
+      "step": 9466
+    },
+    {
+      "epoch": 0.6824046709435594,
+      "grad_norm": 0.11493598333542822,
+      "learning_rate": 4.8404609556264967e-05,
+      "loss": 1.2352,
+      "step": 9467
+    },
+    {
+      "epoch": 0.6824767534058963,
+      "grad_norm": 0.1162656485505921,
+      "learning_rate": 4.838461141131892e-05,
+      "loss": 1.2733,
+      "step": 9468
+    },
+    {
+      "epoch": 0.6825488358682332,
+      "grad_norm": 0.11663130885074001,
+      "learning_rate": 4.836461607987336e-05,
+      "loss": 1.2463,
+      "step": 9469
+    },
+    {
+      "epoch": 0.6826209183305701,
+      "grad_norm": 0.11141666913127345,
+      "learning_rate": 4.8344623563018186e-05,
+      "loss": 1.2762,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6826930007929071,
+      "grad_norm": 0.12751216560079248,
+      "learning_rate": 4.832463386184313e-05,
+      "loss": 1.231,
+      "step": 9471
+    },
+    {
+      "epoch": 0.682765083255244,
+      "grad_norm": 0.13478310237921137,
+      "learning_rate": 4.830464697743783e-05,
+      "loss": 1.2815,
+      "step": 9472
+    },
+    {
+      "epoch": 0.6828371657175809,
+      "grad_norm": 0.11975778381705494,
+      "learning_rate": 4.82846629108917e-05,
+      "loss": 1.1707,
+      "step": 9473
+    },
+    {
+      "epoch": 0.6829092481799178,
+      "grad_norm": 0.11149893810647112,
+      "learning_rate": 4.8264681663294174e-05,
+      "loss": 1.2838,
+      "step": 9474
+    },
+    {
+      "epoch": 0.6829813306422547,
+      "grad_norm": 0.11669133140967529,
+      "learning_rate": 4.8244703235734315e-05,
+      "loss": 1.277,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6830534131045917,
+      "grad_norm": 0.130811909059725,
+      "learning_rate": 4.822472762930112e-05,
+      "loss": 1.2755,
+      "step": 9476
+    },
+    {
+      "epoch": 0.6831254955669286,
+      "grad_norm": 0.1288549455013344,
+      "learning_rate": 4.820475484508345e-05,
+      "loss": 1.235,
+      "step": 9477
+    },
+    {
+      "epoch": 0.6831975780292655,
+      "grad_norm": 0.2074659321796291,
+      "learning_rate": 4.818478488417007e-05,
+      "loss": 1.295,
+      "step": 9478
+    },
+    {
+      "epoch": 0.6832696604916024,
+      "grad_norm": 0.12000153626825541,
+      "learning_rate": 4.816481774764945e-05,
+      "loss": 1.311,
+      "step": 9479
+    },
+    {
+      "epoch": 0.6833417429539393,
+      "grad_norm": 0.11095543342181631,
+      "learning_rate": 4.814485343660998e-05,
+      "loss": 1.2988,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6834138254162763,
+      "grad_norm": 0.13227047576460038,
+      "learning_rate": 4.812489195213987e-05,
+      "loss": 1.2292,
+      "step": 9481
+    },
+    {
+      "epoch": 0.6834859078786132,
+      "grad_norm": 0.11339314199037114,
+      "learning_rate": 4.810493329532731e-05,
+      "loss": 1.2493,
+      "step": 9482
+    },
+    {
+      "epoch": 0.68355799034095,
+      "grad_norm": 0.11406518172047592,
+      "learning_rate": 4.808497746726013e-05,
+      "loss": 1.3309,
+      "step": 9483
+    },
+    {
+      "epoch": 0.6836300728032869,
+      "grad_norm": 0.119393151181573,
+      "learning_rate": 4.80650244690261e-05,
+      "loss": 1.3212,
+      "step": 9484
+    },
+    {
+      "epoch": 0.6837021552656238,
+      "grad_norm": 0.1171718700709399,
+      "learning_rate": 4.804507430171288e-05,
+      "loss": 1.3102,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6837742377279608,
+      "grad_norm": 0.11131379657939079,
+      "learning_rate": 4.8025126966407896e-05,
+      "loss": 1.2514,
+      "step": 9486
+    },
+    {
+      "epoch": 0.6838463201902977,
+      "grad_norm": 0.11828649712751846,
+      "learning_rate": 4.800518246419849e-05,
+      "loss": 1.2434,
+      "step": 9487
+    },
+    {
+      "epoch": 0.6839184026526346,
+      "grad_norm": 0.13539000729143813,
+      "learning_rate": 4.798524079617178e-05,
+      "loss": 1.305,
+      "step": 9488
+    },
+    {
+      "epoch": 0.6839904851149715,
+      "grad_norm": 0.12409306376030309,
+      "learning_rate": 4.79653019634148e-05,
+      "loss": 1.3333,
+      "step": 9489
+    },
+    {
+      "epoch": 0.6840625675773084,
+      "grad_norm": 0.1175849563881839,
+      "learning_rate": 4.7945365967014366e-05,
+      "loss": 1.254,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6841346500396454,
+      "grad_norm": 0.11701813610899829,
+      "learning_rate": 4.792543280805719e-05,
+      "loss": 1.2343,
+      "step": 9491
+    },
+    {
+      "epoch": 0.6842067325019823,
+      "grad_norm": 0.12226684197650149,
+      "learning_rate": 4.7905502487629796e-05,
+      "loss": 1.4152,
+      "step": 9492
+    },
+    {
+      "epoch": 0.6842788149643192,
+      "grad_norm": 0.1283339753180616,
+      "learning_rate": 4.788557500681856e-05,
+      "loss": 1.3131,
+      "step": 9493
+    },
+    {
+      "epoch": 0.6843508974266561,
+      "grad_norm": 0.12028497057991298,
+      "learning_rate": 4.786565036670972e-05,
+      "loss": 1.3265,
+      "step": 9494
+    },
+    {
+      "epoch": 0.684422979888993,
+      "grad_norm": 0.1136454892049601,
+      "learning_rate": 4.784572856838935e-05,
+      "loss": 1.2883,
+      "step": 9495
+    },
+    {
+      "epoch": 0.68449506235133,
+      "grad_norm": 0.12627445135875542,
+      "learning_rate": 4.782580961294335e-05,
+      "loss": 1.3374,
+      "step": 9496
+    },
+    {
+      "epoch": 0.6845671448136669,
+      "grad_norm": 0.11140154937514907,
+      "learning_rate": 4.780589350145749e-05,
+      "loss": 1.2803,
+      "step": 9497
+    },
+    {
+      "epoch": 0.6846392272760038,
+      "grad_norm": 0.11788265545395901,
+      "learning_rate": 4.778598023501738e-05,
+      "loss": 1.2611,
+      "step": 9498
+    },
+    {
+      "epoch": 0.6847113097383407,
+      "grad_norm": 0.11960638336516312,
+      "learning_rate": 4.776606981470846e-05,
+      "loss": 1.297,
+      "step": 9499
+    },
+    {
+      "epoch": 0.6847833922006775,
+      "grad_norm": 0.12003492561308815,
+      "learning_rate": 4.774616224161604e-05,
+      "loss": 1.2406,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6848554746630144,
+      "grad_norm": 0.12584963969544682,
+      "learning_rate": 4.7726257516825243e-05,
+      "loss": 1.336,
+      "step": 9501
+    },
+    {
+      "epoch": 0.6849275571253514,
+      "grad_norm": 0.12315643730525873,
+      "learning_rate": 4.7706355641421084e-05,
+      "loss": 1.332,
+      "step": 9502
+    },
+    {
+      "epoch": 0.6849996395876883,
+      "grad_norm": 0.11789709447961273,
+      "learning_rate": 4.7686456616488365e-05,
+      "loss": 1.3258,
+      "step": 9503
+    },
+    {
+      "epoch": 0.6850717220500252,
+      "grad_norm": 0.12328639768553844,
+      "learning_rate": 4.7666560443111774e-05,
+      "loss": 1.2535,
+      "step": 9504
+    },
+    {
+      "epoch": 0.6851438045123621,
+      "grad_norm": 0.14454943624603467,
+      "learning_rate": 4.764666712237582e-05,
+      "loss": 1.2253,
+      "step": 9505
+    },
+    {
+      "epoch": 0.685215886974699,
+      "grad_norm": 0.1317060082852972,
+      "learning_rate": 4.762677665536488e-05,
+      "loss": 1.3405,
+      "step": 9506
+    },
+    {
+      "epoch": 0.685287969437036,
+      "grad_norm": 0.1087001514342404,
+      "learning_rate": 4.7606889043163196e-05,
+      "loss": 1.3219,
+      "step": 9507
+    },
+    {
+      "epoch": 0.6853600518993729,
+      "grad_norm": 0.12295982271719895,
+      "learning_rate": 4.7587004286854695e-05,
+      "loss": 1.2595,
+      "step": 9508
+    },
+    {
+      "epoch": 0.6854321343617098,
+      "grad_norm": 0.11478943065909321,
+      "learning_rate": 4.756712238752341e-05,
+      "loss": 1.3201,
+      "step": 9509
+    },
+    {
+      "epoch": 0.6855042168240467,
+      "grad_norm": 0.10579038012678023,
+      "learning_rate": 4.7547243346253024e-05,
+      "loss": 1.2421,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6855762992863836,
+      "grad_norm": 0.1217935644182921,
+      "learning_rate": 4.752736716412718e-05,
+      "loss": 1.3479,
+      "step": 9511
+    },
+    {
+      "epoch": 0.6856483817487206,
+      "grad_norm": 0.11196765067406443,
+      "learning_rate": 4.750749384222918e-05,
+      "loss": 1.2947,
+      "step": 9512
+    },
+    {
+      "epoch": 0.6857204642110575,
+      "grad_norm": 0.11472970980275206,
+      "learning_rate": 4.748762338164241e-05,
+      "loss": 1.2175,
+      "step": 9513
+    },
+    {
+      "epoch": 0.6857925466733944,
+      "grad_norm": 0.11807579412180162,
+      "learning_rate": 4.7467755783449955e-05,
+      "loss": 1.2662,
+      "step": 9514
+    },
+    {
+      "epoch": 0.6858646291357313,
+      "grad_norm": 0.11848857711581491,
+      "learning_rate": 4.744789104873482e-05,
+      "loss": 1.1715,
+      "step": 9515
+    },
+    {
+      "epoch": 0.6859367115980682,
+      "grad_norm": 0.11504552026009093,
+      "learning_rate": 4.7428029178579726e-05,
+      "loss": 1.2243,
+      "step": 9516
+    },
+    {
+      "epoch": 0.6860087940604052,
+      "grad_norm": 0.11221339792731133,
+      "learning_rate": 4.7408170174067326e-05,
+      "loss": 1.3016,
+      "step": 9517
+    },
+    {
+      "epoch": 0.686080876522742,
+      "grad_norm": 0.11582124142391428,
+      "learning_rate": 4.7388314036280224e-05,
+      "loss": 1.3629,
+      "step": 9518
+    },
+    {
+      "epoch": 0.6861529589850789,
+      "grad_norm": 0.1267219884759173,
+      "learning_rate": 4.736846076630065e-05,
+      "loss": 1.2398,
+      "step": 9519
+    },
+    {
+      "epoch": 0.6862250414474158,
+      "grad_norm": 0.11964816711775196,
+      "learning_rate": 4.734861036521082e-05,
+      "loss": 1.2149,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6862971239097527,
+      "grad_norm": 0.11389722397558288,
+      "learning_rate": 4.732876283409272e-05,
+      "loss": 1.2457,
+      "step": 9521
+    },
+    {
+      "epoch": 0.6863692063720896,
+      "grad_norm": 0.11896941786321205,
+      "learning_rate": 4.7308918174028326e-05,
+      "loss": 1.2441,
+      "step": 9522
+    },
+    {
+      "epoch": 0.6864412888344266,
+      "grad_norm": 0.11767335299041067,
+      "learning_rate": 4.728907638609925e-05,
+      "loss": 1.1975,
+      "step": 9523
+    },
+    {
+      "epoch": 0.6865133712967635,
+      "grad_norm": 0.11804719750327651,
+      "learning_rate": 4.7269237471387075e-05,
+      "loss": 1.3354,
+      "step": 9524
+    },
+    {
+      "epoch": 0.6865854537591004,
+      "grad_norm": 0.12176160184019959,
+      "learning_rate": 4.724940143097316e-05,
+      "loss": 1.23,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6866575362214373,
+      "grad_norm": 0.11865939439993464,
+      "learning_rate": 4.722956826593886e-05,
+      "loss": 1.3322,
+      "step": 9526
+    },
+    {
+      "epoch": 0.6867296186837742,
+      "grad_norm": 0.11773587964835862,
+      "learning_rate": 4.720973797736515e-05,
+      "loss": 1.2984,
+      "step": 9527
+    },
+    {
+      "epoch": 0.6868017011461112,
+      "grad_norm": 0.11919006180665086,
+      "learning_rate": 4.7189910566333006e-05,
+      "loss": 1.249,
+      "step": 9528
+    },
+    {
+      "epoch": 0.6868737836084481,
+      "grad_norm": 0.12839692138840403,
+      "learning_rate": 4.717008603392318e-05,
+      "loss": 1.2307,
+      "step": 9529
+    },
+    {
+      "epoch": 0.686945866070785,
+      "grad_norm": 0.12277456429089387,
+      "learning_rate": 4.715026438121629e-05,
+      "loss": 1.2106,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6870179485331219,
+      "grad_norm": 0.12219286051655583,
+      "learning_rate": 4.71304456092928e-05,
+      "loss": 1.2769,
+      "step": 9531
+    },
+    {
+      "epoch": 0.6870900309954588,
+      "grad_norm": 0.12048045269669058,
+      "learning_rate": 4.7110629719233025e-05,
+      "loss": 1.304,
+      "step": 9532
+    },
+    {
+      "epoch": 0.6871621134577958,
+      "grad_norm": 0.11819436050477895,
+      "learning_rate": 4.709081671211708e-05,
+      "loss": 1.3238,
+      "step": 9533
+    },
+    {
+      "epoch": 0.6872341959201327,
+      "grad_norm": 0.1108747680719316,
+      "learning_rate": 4.707100658902497e-05,
+      "loss": 1.3204,
+      "step": 9534
+    },
+    {
+      "epoch": 0.6873062783824695,
+      "grad_norm": 0.10424210805548245,
+      "learning_rate": 4.70511993510365e-05,
+      "loss": 1.1092,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6873783608448064,
+      "grad_norm": 0.11785027182875436,
+      "learning_rate": 4.703139499923138e-05,
+      "loss": 1.2105,
+      "step": 9536
+    },
+    {
+      "epoch": 0.6874504433071433,
+      "grad_norm": 0.13511108691716273,
+      "learning_rate": 4.701159353468909e-05,
+      "loss": 1.3647,
+      "step": 9537
+    },
+    {
+      "epoch": 0.6875225257694803,
+      "grad_norm": 0.11368012269116136,
+      "learning_rate": 4.6991794958489e-05,
+      "loss": 1.3662,
+      "step": 9538
+    },
+    {
+      "epoch": 0.6875946082318172,
+      "grad_norm": 0.11631780967758203,
+      "learning_rate": 4.697199927171032e-05,
+      "loss": 1.3441,
+      "step": 9539
+    },
+    {
+      "epoch": 0.6876666906941541,
+      "grad_norm": 0.11736633228416748,
+      "learning_rate": 4.695220647543207e-05,
+      "loss": 1.2401,
+      "step": 9540
+    },
+    {
+      "epoch": 0.687738773156491,
+      "grad_norm": 0.11118211252858111,
+      "learning_rate": 4.693241657073315e-05,
+      "loss": 1.3888,
+      "step": 9541
+    },
+    {
+      "epoch": 0.6878108556188279,
+      "grad_norm": 0.11229313890125078,
+      "learning_rate": 4.691262955869229e-05,
+      "loss": 1.2432,
+      "step": 9542
+    },
+    {
+      "epoch": 0.6878829380811649,
+      "grad_norm": 0.12287294872242457,
+      "learning_rate": 4.6892845440388044e-05,
+      "loss": 1.2374,
+      "step": 9543
+    },
+    {
+      "epoch": 0.6879550205435018,
+      "grad_norm": 0.12385657823210701,
+      "learning_rate": 4.687306421689883e-05,
+      "loss": 1.1887,
+      "step": 9544
+    },
+    {
+      "epoch": 0.6880271030058387,
+      "grad_norm": 0.11255125542614139,
+      "learning_rate": 4.685328588930291e-05,
+      "loss": 1.2316,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6880991854681756,
+      "grad_norm": 0.11730693057534203,
+      "learning_rate": 4.6833510458678375e-05,
+      "loss": 1.3352,
+      "step": 9546
+    },
+    {
+      "epoch": 0.6881712679305125,
+      "grad_norm": 0.12990315478914513,
+      "learning_rate": 4.681373792610315e-05,
+      "loss": 1.2786,
+      "step": 9547
+    },
+    {
+      "epoch": 0.6882433503928495,
+      "grad_norm": 0.11274862744237463,
+      "learning_rate": 4.6793968292655045e-05,
+      "loss": 1.224,
+      "step": 9548
+    },
+    {
+      "epoch": 0.6883154328551864,
+      "grad_norm": 0.11273262497727164,
+      "learning_rate": 4.677420155941165e-05,
+      "loss": 1.248,
+      "step": 9549
+    },
+    {
+      "epoch": 0.6883875153175233,
+      "grad_norm": 0.1298384549483012,
+      "learning_rate": 4.675443772745044e-05,
+      "loss": 1.2593,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6884595977798602,
+      "grad_norm": 0.12219558348852663,
+      "learning_rate": 4.673467679784875e-05,
+      "loss": 1.2688,
+      "step": 9551
+    },
+    {
+      "epoch": 0.688531680242197,
+      "grad_norm": 0.11533663499175749,
+      "learning_rate": 4.6714918771683646e-05,
+      "loss": 1.279,
+      "step": 9552
+    },
+    {
+      "epoch": 0.6886037627045339,
+      "grad_norm": 0.10872639070108381,
+      "learning_rate": 4.6695163650032205e-05,
+      "loss": 1.2871,
+      "step": 9553
+    },
+    {
+      "epoch": 0.6886758451668709,
+      "grad_norm": 0.11391849135944399,
+      "learning_rate": 4.6675411433971205e-05,
+      "loss": 1.2774,
+      "step": 9554
+    },
+    {
+      "epoch": 0.6887479276292078,
+      "grad_norm": 0.11815064647811982,
+      "learning_rate": 4.6655662124577395e-05,
+      "loss": 1.2602,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6888200100915447,
+      "grad_norm": 0.12416543787361906,
+      "learning_rate": 4.663591572292715e-05,
+      "loss": 1.2231,
+      "step": 9556
+    },
+    {
+      "epoch": 0.6888920925538816,
+      "grad_norm": 0.12595090479986942,
+      "learning_rate": 4.661617223009695e-05,
+      "loss": 1.3599,
+      "step": 9557
+    },
+    {
+      "epoch": 0.6889641750162185,
+      "grad_norm": 0.11952698137544016,
+      "learning_rate": 4.659643164716293e-05,
+      "loss": 1.37,
+      "step": 9558
+    },
+    {
+      "epoch": 0.6890362574785555,
+      "grad_norm": 0.11663519453328386,
+      "learning_rate": 4.65766939752012e-05,
+      "loss": 1.225,
+      "step": 9559
+    },
+    {
+      "epoch": 0.6891083399408924,
+      "grad_norm": 0.1122967268205016,
+      "learning_rate": 4.655695921528754e-05,
+      "loss": 1.2961,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6891804224032293,
+      "grad_norm": 0.12206293661953194,
+      "learning_rate": 4.653722736849769e-05,
+      "loss": 1.3148,
+      "step": 9561
+    },
+    {
+      "epoch": 0.6892525048655662,
+      "grad_norm": 0.1308266034438506,
+      "learning_rate": 4.651749843590728e-05,
+      "loss": 1.3184,
+      "step": 9562
+    },
+    {
+      "epoch": 0.6893245873279031,
+      "grad_norm": 0.12226463460777513,
+      "learning_rate": 4.64977724185917e-05,
+      "loss": 1.3262,
+      "step": 9563
+    },
+    {
+      "epoch": 0.6893966697902401,
+      "grad_norm": 0.11478839706181604,
+      "learning_rate": 4.647804931762614e-05,
+      "loss": 1.2968,
+      "step": 9564
+    },
+    {
+      "epoch": 0.689468752252577,
+      "grad_norm": 0.1255614420661547,
+      "learning_rate": 4.645832913408566e-05,
+      "loss": 1.3286,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6895408347149139,
+      "grad_norm": 0.12277906977499052,
+      "learning_rate": 4.6438611869045304e-05,
+      "loss": 1.2583,
+      "step": 9566
+    },
+    {
+      "epoch": 0.6896129171772508,
+      "grad_norm": 0.1323543533373657,
+      "learning_rate": 4.6418897523579785e-05,
+      "loss": 1.3554,
+      "step": 9567
+    },
+    {
+      "epoch": 0.6896849996395877,
+      "grad_norm": 0.12533004237910864,
+      "learning_rate": 4.639918609876368e-05,
+      "loss": 1.2929,
+      "step": 9568
+    },
+    {
+      "epoch": 0.6897570821019247,
+      "grad_norm": 0.1158217890242886,
+      "learning_rate": 4.6379477595671416e-05,
+      "loss": 1.2574,
+      "step": 9569
+    },
+    {
+      "epoch": 0.6898291645642615,
+      "grad_norm": 0.12008740077463838,
+      "learning_rate": 4.635977201537737e-05,
+      "loss": 1.1847,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6899012470265984,
+      "grad_norm": 0.11240725243054861,
+      "learning_rate": 4.634006935895565e-05,
+      "loss": 1.3803,
+      "step": 9571
+    },
+    {
+      "epoch": 0.6899733294889353,
+      "grad_norm": 0.11230480175355995,
+      "learning_rate": 4.6320369627480184e-05,
+      "loss": 1.2238,
+      "step": 9572
+    },
+    {
+      "epoch": 0.6900454119512722,
+      "grad_norm": 0.11321307820016803,
+      "learning_rate": 4.630067282202479e-05,
+      "loss": 1.2015,
+      "step": 9573
+    },
+    {
+      "epoch": 0.6901174944136091,
+      "grad_norm": 0.12832434453300293,
+      "learning_rate": 4.628097894366311e-05,
+      "loss": 1.2994,
+      "step": 9574
+    },
+    {
+      "epoch": 0.6901895768759461,
+      "grad_norm": 0.12644551275928628,
+      "learning_rate": 4.6261287993468737e-05,
+      "loss": 1.2882,
+      "step": 9575
+    },
+    {
+      "epoch": 0.690261659338283,
+      "grad_norm": 0.1246718114400465,
+      "learning_rate": 4.6241599972514893e-05,
+      "loss": 1.3325,
+      "step": 9576
+    },
+    {
+      "epoch": 0.6903337418006199,
+      "grad_norm": 0.11287603495878017,
+      "learning_rate": 4.6221914881874784e-05,
+      "loss": 1.2985,
+      "step": 9577
+    },
+    {
+      "epoch": 0.6904058242629568,
+      "grad_norm": 0.11644117083915703,
+      "learning_rate": 4.620223272262139e-05,
+      "loss": 1.3233,
+      "step": 9578
+    },
+    {
+      "epoch": 0.6904779067252937,
+      "grad_norm": 0.13543065760124617,
+      "learning_rate": 4.618255349582766e-05,
+      "loss": 1.2401,
+      "step": 9579
+    },
+    {
+      "epoch": 0.6905499891876307,
+      "grad_norm": 0.11919544498403474,
+      "learning_rate": 4.616287720256621e-05,
+      "loss": 1.2717,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6906220716499676,
+      "grad_norm": 0.12412470907468312,
+      "learning_rate": 4.614320384390959e-05,
+      "loss": 1.2127,
+      "step": 9581
+    },
+    {
+      "epoch": 0.6906941541123045,
+      "grad_norm": 0.10810157133713244,
+      "learning_rate": 4.612353342093017e-05,
+      "loss": 1.2335,
+      "step": 9582
+    },
+    {
+      "epoch": 0.6907662365746414,
+      "grad_norm": 0.11658901292192209,
+      "learning_rate": 4.6103865934700175e-05,
+      "loss": 1.2765,
+      "step": 9583
+    },
+    {
+      "epoch": 0.6908383190369783,
+      "grad_norm": 0.11414938439382764,
+      "learning_rate": 4.608420138629166e-05,
+      "loss": 1.3127,
+      "step": 9584
+    },
+    {
+      "epoch": 0.6909104014993153,
+      "grad_norm": 0.11586475279758558,
+      "learning_rate": 4.60645397767765e-05,
+      "loss": 1.247,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6909824839616522,
+      "grad_norm": 0.12573988511618972,
+      "learning_rate": 4.6044881107226455e-05,
+      "loss": 1.3733,
+      "step": 9586
+    },
+    {
+      "epoch": 0.691054566423989,
+      "grad_norm": 0.12828623157344193,
+      "learning_rate": 4.602522537871309e-05,
+      "loss": 1.3417,
+      "step": 9587
+    },
+    {
+      "epoch": 0.6911266488863259,
+      "grad_norm": 0.11461041592964943,
+      "learning_rate": 4.60055725923078e-05,
+      "loss": 1.2791,
+      "step": 9588
+    },
+    {
+      "epoch": 0.6911987313486628,
+      "grad_norm": 0.19503206963497652,
+      "learning_rate": 4.5985922749081864e-05,
+      "loss": 1.3436,
+      "step": 9589
+    },
+    {
+      "epoch": 0.6912708138109998,
+      "grad_norm": 0.1258681236300493,
+      "learning_rate": 4.5966275850106354e-05,
+      "loss": 1.2327,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6913428962733367,
+      "grad_norm": 0.12486507562578994,
+      "learning_rate": 4.594663189645222e-05,
+      "loss": 1.2751,
+      "step": 9591
+    },
+    {
+      "epoch": 0.6914149787356736,
+      "grad_norm": 0.15193266356045598,
+      "learning_rate": 4.592699088919021e-05,
+      "loss": 1.3117,
+      "step": 9592
+    },
+    {
+      "epoch": 0.6914870611980105,
+      "grad_norm": 0.10373335144224591,
+      "learning_rate": 4.590735282939096e-05,
+      "loss": 1.227,
+      "step": 9593
+    },
+    {
+      "epoch": 0.6915591436603474,
+      "grad_norm": 0.12351778208434215,
+      "learning_rate": 4.588771771812489e-05,
+      "loss": 1.2965,
+      "step": 9594
+    },
+    {
+      "epoch": 0.6916312261226844,
+      "grad_norm": 0.12887109575978467,
+      "learning_rate": 4.586808555646233e-05,
+      "loss": 1.2659,
+      "step": 9595
+    },
+    {
+      "epoch": 0.6917033085850213,
+      "grad_norm": 0.1256590577064637,
+      "learning_rate": 4.5848456345473365e-05,
+      "loss": 1.3115,
+      "step": 9596
+    },
+    {
+      "epoch": 0.6917753910473582,
+      "grad_norm": 0.12447756814757131,
+      "learning_rate": 4.5828830086227994e-05,
+      "loss": 1.227,
+      "step": 9597
+    },
+    {
+      "epoch": 0.6918474735096951,
+      "grad_norm": 0.11194826394321272,
+      "learning_rate": 4.5809206779796e-05,
+      "loss": 1.2435,
+      "step": 9598
+    },
+    {
+      "epoch": 0.691919555972032,
+      "grad_norm": 0.12446143512932425,
+      "learning_rate": 4.578958642724706e-05,
+      "loss": 1.3173,
+      "step": 9599
+    },
+    {
+      "epoch": 0.691991638434369,
+      "grad_norm": 0.13671823032272312,
+      "learning_rate": 4.5769969029650625e-05,
+      "loss": 1.2867,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6920637208967059,
+      "grad_norm": 0.12140646475404829,
+      "learning_rate": 4.575035458807604e-05,
+      "loss": 1.2383,
+      "step": 9601
+    },
+    {
+      "epoch": 0.6921358033590428,
+      "grad_norm": 0.10866447705801706,
+      "learning_rate": 4.573074310359246e-05,
+      "loss": 1.2223,
+      "step": 9602
+    },
+    {
+      "epoch": 0.6922078858213797,
+      "grad_norm": 0.11238284901469571,
+      "learning_rate": 4.5711134577268874e-05,
+      "loss": 1.2351,
+      "step": 9603
+    },
+    {
+      "epoch": 0.6922799682837165,
+      "grad_norm": 0.11754038747817677,
+      "learning_rate": 4.5691529010174194e-05,
+      "loss": 1.2891,
+      "step": 9604
+    },
+    {
+      "epoch": 0.6923520507460534,
+      "grad_norm": 0.1193621488311641,
+      "learning_rate": 4.5671926403376954e-05,
+      "loss": 1.204,
+      "step": 9605
+    },
+    {
+      "epoch": 0.6924241332083904,
+      "grad_norm": 0.11899630073678477,
+      "learning_rate": 4.5652326757945804e-05,
+      "loss": 1.3312,
+      "step": 9606
+    },
+    {
+      "epoch": 0.6924962156707273,
+      "grad_norm": 0.12413378946233213,
+      "learning_rate": 4.563273007494905e-05,
+      "loss": 1.2135,
+      "step": 9607
+    },
+    {
+      "epoch": 0.6925682981330642,
+      "grad_norm": 0.117524762782378,
+      "learning_rate": 4.561313635545492e-05,
+      "loss": 1.3329,
+      "step": 9608
+    },
+    {
+      "epoch": 0.6926403805954011,
+      "grad_norm": 0.13533720097900181,
+      "learning_rate": 4.559354560053135e-05,
+      "loss": 1.2836,
+      "step": 9609
+    },
+    {
+      "epoch": 0.692712463057738,
+      "grad_norm": 0.1111878783516883,
+      "learning_rate": 4.557395781124632e-05,
+      "loss": 1.3186,
+      "step": 9610
+    },
+    {
+      "epoch": 0.692784545520075,
+      "grad_norm": 0.12623003044702788,
+      "learning_rate": 4.555437298866749e-05,
+      "loss": 1.3789,
+      "step": 9611
+    },
+    {
+      "epoch": 0.6928566279824119,
+      "grad_norm": 0.12089632383972844,
+      "learning_rate": 4.553479113386246e-05,
+      "loss": 1.2784,
+      "step": 9612
+    },
+    {
+      "epoch": 0.6929287104447488,
+      "grad_norm": 0.11752659585340967,
+      "learning_rate": 4.55152122478985e-05,
+      "loss": 1.2613,
+      "step": 9613
+    },
+    {
+      "epoch": 0.6930007929070857,
+      "grad_norm": 0.1254525083809605,
+      "learning_rate": 4.549563633184294e-05,
+      "loss": 1.31,
+      "step": 9614
+    },
+    {
+      "epoch": 0.6930728753694226,
+      "grad_norm": 0.11730831913050635,
+      "learning_rate": 4.54760633867628e-05,
+      "loss": 1.2542,
+      "step": 9615
+    },
+    {
+      "epoch": 0.6931449578317596,
+      "grad_norm": 0.12724271167061038,
+      "learning_rate": 4.5456493413725046e-05,
+      "loss": 1.1985,
+      "step": 9616
+    },
+    {
+      "epoch": 0.6932170402940965,
+      "grad_norm": 0.11329351524600062,
+      "learning_rate": 4.543692641379631e-05,
+      "loss": 1.3279,
+      "step": 9617
+    },
+    {
+      "epoch": 0.6932891227564334,
+      "grad_norm": 0.12083044015828646,
+      "learning_rate": 4.5417362388043194e-05,
+      "loss": 1.2645,
+      "step": 9618
+    },
+    {
+      "epoch": 0.6933612052187703,
+      "grad_norm": 0.1166818815548574,
+      "learning_rate": 4.5397801337532164e-05,
+      "loss": 1.1985,
+      "step": 9619
+    },
+    {
+      "epoch": 0.6934332876811072,
+      "grad_norm": 0.11789853130561545,
+      "learning_rate": 4.537824326332949e-05,
+      "loss": 1.3061,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6935053701434442,
+      "grad_norm": 0.12423899865763428,
+      "learning_rate": 4.535868816650118e-05,
+      "loss": 1.2805,
+      "step": 9621
+    },
+    {
+      "epoch": 0.693577452605781,
+      "grad_norm": 0.10788053838008907,
+      "learning_rate": 4.533913604811317e-05,
+      "loss": 1.3159,
+      "step": 9622
+    },
+    {
+      "epoch": 0.6936495350681179,
+      "grad_norm": 0.12798988156772584,
+      "learning_rate": 4.531958690923128e-05,
+      "loss": 1.2406,
+      "step": 9623
+    },
+    {
+      "epoch": 0.6937216175304548,
+      "grad_norm": 0.12044111746598361,
+      "learning_rate": 4.5300040750921135e-05,
+      "loss": 1.2313,
+      "step": 9624
+    },
+    {
+      "epoch": 0.6937936999927917,
+      "grad_norm": 0.11867648478953219,
+      "learning_rate": 4.5280497574248094e-05,
+      "loss": 1.232,
+      "step": 9625
+    },
+    {
+      "epoch": 0.6938657824551286,
+      "grad_norm": 0.12310043675000999,
+      "learning_rate": 4.5260957380277456e-05,
+      "loss": 1.3575,
+      "step": 9626
+    },
+    {
+      "epoch": 0.6939378649174656,
+      "grad_norm": 0.11731237069882218,
+      "learning_rate": 4.524142017007432e-05,
+      "loss": 1.2715,
+      "step": 9627
+    },
+    {
+      "epoch": 0.6940099473798025,
+      "grad_norm": 0.11614562709255602,
+      "learning_rate": 4.522188594470375e-05,
+      "loss": 1.2035,
+      "step": 9628
+    },
+    {
+      "epoch": 0.6940820298421394,
+      "grad_norm": 0.11766614950235327,
+      "learning_rate": 4.520235470523041e-05,
+      "loss": 1.2844,
+      "step": 9629
+    },
+    {
+      "epoch": 0.6941541123044763,
+      "grad_norm": 0.12200767367619542,
+      "learning_rate": 4.518282645271897e-05,
+      "loss": 1.3574,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6942261947668132,
+      "grad_norm": 0.12105897178831303,
+      "learning_rate": 4.516330118823387e-05,
+      "loss": 1.2015,
+      "step": 9631
+    },
+    {
+      "epoch": 0.6942982772291502,
+      "grad_norm": 0.12435134558894137,
+      "learning_rate": 4.5143778912839506e-05,
+      "loss": 1.3081,
+      "step": 9632
+    },
+    {
+      "epoch": 0.6943703596914871,
+      "grad_norm": 0.125380889566568,
+      "learning_rate": 4.512425962759992e-05,
+      "loss": 1.2858,
+      "step": 9633
+    },
+    {
+      "epoch": 0.694442442153824,
+      "grad_norm": 0.12478048128681507,
+      "learning_rate": 4.5104743333579104e-05,
+      "loss": 1.2367,
+      "step": 9634
+    },
+    {
+      "epoch": 0.6945145246161609,
+      "grad_norm": 0.12781001811053155,
+      "learning_rate": 4.508523003184085e-05,
+      "loss": 1.3055,
+      "step": 9635
+    },
+    {
+      "epoch": 0.6945866070784978,
+      "grad_norm": 0.11365603596545067,
+      "learning_rate": 4.506571972344892e-05,
+      "loss": 1.2218,
+      "step": 9636
+    },
+    {
+      "epoch": 0.6946586895408348,
+      "grad_norm": 0.12563254943816174,
+      "learning_rate": 4.504621240946668e-05,
+      "loss": 1.2076,
+      "step": 9637
+    },
+    {
+      "epoch": 0.6947307720031717,
+      "grad_norm": 0.11540232572421488,
+      "learning_rate": 4.502670809095749e-05,
+      "loss": 1.3077,
+      "step": 9638
+    },
+    {
+      "epoch": 0.6948028544655085,
+      "grad_norm": 0.12873122703604642,
+      "learning_rate": 4.500720676898452e-05,
+      "loss": 1.3183,
+      "step": 9639
+    },
+    {
+      "epoch": 0.6948749369278454,
+      "grad_norm": 0.11020809795621433,
+      "learning_rate": 4.4987708444610763e-05,
+      "loss": 1.2104,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6949470193901823,
+      "grad_norm": 0.11761256409966511,
+      "learning_rate": 4.496821311889905e-05,
+      "loss": 1.2534,
+      "step": 9641
+    },
+    {
+      "epoch": 0.6950191018525192,
+      "grad_norm": 0.12131761777649147,
+      "learning_rate": 4.494872079291205e-05,
+      "loss": 1.305,
+      "step": 9642
+    },
+    {
+      "epoch": 0.6950911843148562,
+      "grad_norm": 0.12485043712517459,
+      "learning_rate": 4.492923146771226e-05,
+      "loss": 1.3123,
+      "step": 9643
+    },
+    {
+      "epoch": 0.6951632667771931,
+      "grad_norm": 0.12852948874826156,
+      "learning_rate": 4.4909745144362046e-05,
+      "loss": 1.2551,
+      "step": 9644
+    },
+    {
+      "epoch": 0.69523534923953,
+      "grad_norm": 0.1153409529134634,
+      "learning_rate": 4.4890261823923564e-05,
+      "loss": 1.2379,
+      "step": 9645
+    },
+    {
+      "epoch": 0.6953074317018669,
+      "grad_norm": 0.11430919046104586,
+      "learning_rate": 4.487078150745883e-05,
+      "loss": 1.2733,
+      "step": 9646
+    },
+    {
+      "epoch": 0.6953795141642038,
+      "grad_norm": 0.11784310473337374,
+      "learning_rate": 4.485130419602972e-05,
+      "loss": 1.262,
+      "step": 9647
+    },
+    {
+      "epoch": 0.6954515966265408,
+      "grad_norm": 0.11898366369811317,
+      "learning_rate": 4.483182989069791e-05,
+      "loss": 1.3177,
+      "step": 9648
+    },
+    {
+      "epoch": 0.6955236790888777,
+      "grad_norm": 0.14176131527771485,
+      "learning_rate": 4.481235859252491e-05,
+      "loss": 1.2761,
+      "step": 9649
+    },
+    {
+      "epoch": 0.6955957615512146,
+      "grad_norm": 0.22936162595299914,
+      "learning_rate": 4.479289030257209e-05,
+      "loss": 1.2203,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6956678440135515,
+      "grad_norm": 0.11416795649475737,
+      "learning_rate": 4.477342502190064e-05,
+      "loss": 1.2311,
+      "step": 9651
+    },
+    {
+      "epoch": 0.6957399264758884,
+      "grad_norm": 0.12917255712099113,
+      "learning_rate": 4.47539627515716e-05,
+      "loss": 1.2586,
+      "step": 9652
+    },
+    {
+      "epoch": 0.6958120089382254,
+      "grad_norm": 0.11438688019721767,
+      "learning_rate": 4.473450349264584e-05,
+      "loss": 1.2422,
+      "step": 9653
+    },
+    {
+      "epoch": 0.6958840914005623,
+      "grad_norm": 0.11371184680089398,
+      "learning_rate": 4.471504724618405e-05,
+      "loss": 1.2302,
+      "step": 9654
+    },
+    {
+      "epoch": 0.6959561738628992,
+      "grad_norm": 0.11633015305697673,
+      "learning_rate": 4.469559401324679e-05,
+      "loss": 1.2852,
+      "step": 9655
+    },
+    {
+      "epoch": 0.696028256325236,
+      "grad_norm": 0.1325860445280333,
+      "learning_rate": 4.467614379489442e-05,
+      "loss": 1.2074,
+      "step": 9656
+    },
+    {
+      "epoch": 0.6961003387875729,
+      "grad_norm": 0.12395201306079706,
+      "learning_rate": 4.465669659218715e-05,
+      "loss": 1.2947,
+      "step": 9657
+    },
+    {
+      "epoch": 0.6961724212499099,
+      "grad_norm": 0.14117652122864177,
+      "learning_rate": 4.463725240618504e-05,
+      "loss": 1.2154,
+      "step": 9658
+    },
+    {
+      "epoch": 0.6962445037122468,
+      "grad_norm": 0.11683783594188023,
+      "learning_rate": 4.461781123794795e-05,
+      "loss": 1.2815,
+      "step": 9659
+    },
+    {
+      "epoch": 0.6963165861745837,
+      "grad_norm": 0.124838948661388,
+      "learning_rate": 4.459837308853561e-05,
+      "loss": 1.3343,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6963886686369206,
+      "grad_norm": 0.12371300042645166,
+      "learning_rate": 4.457893795900763e-05,
+      "loss": 1.3197,
+      "step": 9661
+    },
+    {
+      "epoch": 0.6964607510992575,
+      "grad_norm": 0.11922969980312859,
+      "learning_rate": 4.4559505850423254e-05,
+      "loss": 1.2447,
+      "step": 9662
+    },
+    {
+      "epoch": 0.6965328335615945,
+      "grad_norm": 0.1207148952322425,
+      "learning_rate": 4.454007676384183e-05,
+      "loss": 1.2501,
+      "step": 9663
+    },
+    {
+      "epoch": 0.6966049160239314,
+      "grad_norm": 0.12207792494767107,
+      "learning_rate": 4.452065070032239e-05,
+      "loss": 1.2703,
+      "step": 9664
+    },
+    {
+      "epoch": 0.6966769984862683,
+      "grad_norm": 0.11734564955658551,
+      "learning_rate": 4.450122766092385e-05,
+      "loss": 1.277,
+      "step": 9665
+    },
+    {
+      "epoch": 0.6967490809486052,
+      "grad_norm": 0.1139163566337454,
+      "learning_rate": 4.4481807646704835e-05,
+      "loss": 1.2564,
+      "step": 9666
+    },
+    {
+      "epoch": 0.6968211634109421,
+      "grad_norm": 0.12228362595489527,
+      "learning_rate": 4.446239065872402e-05,
+      "loss": 1.2073,
+      "step": 9667
+    },
+    {
+      "epoch": 0.696893245873279,
+      "grad_norm": 0.1177935156348915,
+      "learning_rate": 4.444297669803981e-05,
+      "loss": 1.2679,
+      "step": 9668
+    },
+    {
+      "epoch": 0.696965328335616,
+      "grad_norm": 0.13610149836413732,
+      "learning_rate": 4.442356576571035e-05,
+      "loss": 1.2386,
+      "step": 9669
+    },
+    {
+      "epoch": 0.6970374107979529,
+      "grad_norm": 0.12422797267875918,
+      "learning_rate": 4.4404157862793764e-05,
+      "loss": 1.2196,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6971094932602898,
+      "grad_norm": 0.1149843109798144,
+      "learning_rate": 4.438475299034791e-05,
+      "loss": 1.278,
+      "step": 9671
+    },
+    {
+      "epoch": 0.6971815757226267,
+      "grad_norm": 0.12399808243809879,
+      "learning_rate": 4.4365351149430645e-05,
+      "loss": 1.4182,
+      "step": 9672
+    },
+    {
+      "epoch": 0.6972536581849635,
+      "grad_norm": 0.11978349810290811,
+      "learning_rate": 4.4345952341099416e-05,
+      "loss": 1.1669,
+      "step": 9673
+    },
+    {
+      "epoch": 0.6973257406473005,
+      "grad_norm": 0.12107520951100215,
+      "learning_rate": 4.4326556566411694e-05,
+      "loss": 1.2232,
+      "step": 9674
+    },
+    {
+      "epoch": 0.6973978231096374,
+      "grad_norm": 0.13038637472177847,
+      "learning_rate": 4.4307163826424656e-05,
+      "loss": 1.3073,
+      "step": 9675
+    },
+    {
+      "epoch": 0.6974699055719743,
+      "grad_norm": 0.13090544393503462,
+      "learning_rate": 4.4287774122195516e-05,
+      "loss": 1.2725,
+      "step": 9676
+    },
+    {
+      "epoch": 0.6975419880343112,
+      "grad_norm": 0.1190887991309252,
+      "learning_rate": 4.4268387454781065e-05,
+      "loss": 1.2642,
+      "step": 9677
+    },
+    {
+      "epoch": 0.6976140704966481,
+      "grad_norm": 0.11442645942626457,
+      "learning_rate": 4.424900382523809e-05,
+      "loss": 1.1562,
+      "step": 9678
+    },
+    {
+      "epoch": 0.6976861529589851,
+      "grad_norm": 0.14256602803040563,
+      "learning_rate": 4.422962323462316e-05,
+      "loss": 1.2487,
+      "step": 9679
+    },
+    {
+      "epoch": 0.697758235421322,
+      "grad_norm": 0.12417775267241424,
+      "learning_rate": 4.42102456839927e-05,
+      "loss": 1.3591,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6978303178836589,
+      "grad_norm": 0.12978466257730423,
+      "learning_rate": 4.419087117440297e-05,
+      "loss": 1.2993,
+      "step": 9681
+    },
+    {
+      "epoch": 0.6979024003459958,
+      "grad_norm": 0.13010794771115905,
+      "learning_rate": 4.4171499706910034e-05,
+      "loss": 1.2775,
+      "step": 9682
+    },
+    {
+      "epoch": 0.6979744828083327,
+      "grad_norm": 0.12361958945681413,
+      "learning_rate": 4.415213128256982e-05,
+      "loss": 1.3048,
+      "step": 9683
+    },
+    {
+      "epoch": 0.6980465652706697,
+      "grad_norm": 0.11167891746429623,
+      "learning_rate": 4.413276590243809e-05,
+      "loss": 1.2791,
+      "step": 9684
+    },
+    {
+      "epoch": 0.6981186477330066,
+      "grad_norm": 0.12384781612062935,
+      "learning_rate": 4.411340356757041e-05,
+      "loss": 1.2591,
+      "step": 9685
+    },
+    {
+      "epoch": 0.6981907301953435,
+      "grad_norm": 0.11662442267559256,
+      "learning_rate": 4.4094044279022216e-05,
+      "loss": 1.31,
+      "step": 9686
+    },
+    {
+      "epoch": 0.6982628126576804,
+      "grad_norm": 0.12079894786331062,
+      "learning_rate": 4.407468803784875e-05,
+      "loss": 1.2417,
+      "step": 9687
+    },
+    {
+      "epoch": 0.6983348951200173,
+      "grad_norm": 0.12210556667814768,
+      "learning_rate": 4.4055334845105125e-05,
+      "loss": 1.2491,
+      "step": 9688
+    },
+    {
+      "epoch": 0.6984069775823543,
+      "grad_norm": 0.11146721003484214,
+      "learning_rate": 4.4035984701846234e-05,
+      "loss": 1.2358,
+      "step": 9689
+    },
+    {
+      "epoch": 0.6984790600446912,
+      "grad_norm": 0.12245380280302927,
+      "learning_rate": 4.4016637609126853e-05,
+      "loss": 1.2101,
+      "step": 9690
+    },
+    {
+      "epoch": 0.698551142507028,
+      "grad_norm": 0.11368850320729466,
+      "learning_rate": 4.399729356800155e-05,
+      "loss": 1.211,
+      "step": 9691
+    },
+    {
+      "epoch": 0.6986232249693649,
+      "grad_norm": 0.11301056937851875,
+      "learning_rate": 4.397795257952477e-05,
+      "loss": 1.4187,
+      "step": 9692
+    },
+    {
+      "epoch": 0.6986953074317018,
+      "grad_norm": 0.12395607540080698,
+      "learning_rate": 4.3958614644750774e-05,
+      "loss": 1.1989,
+      "step": 9693
+    },
+    {
+      "epoch": 0.6987673898940387,
+      "grad_norm": 0.11782225934966178,
+      "learning_rate": 4.3939279764733624e-05,
+      "loss": 1.2884,
+      "step": 9694
+    },
+    {
+      "epoch": 0.6988394723563757,
+      "grad_norm": 0.11876184972860822,
+      "learning_rate": 4.391994794052726e-05,
+      "loss": 1.1757,
+      "step": 9695
+    },
+    {
+      "epoch": 0.6989115548187126,
+      "grad_norm": 0.1169144480846948,
+      "learning_rate": 4.390061917318545e-05,
+      "loss": 1.3822,
+      "step": 9696
+    },
+    {
+      "epoch": 0.6989836372810495,
+      "grad_norm": 0.12432048107602071,
+      "learning_rate": 4.388129346376178e-05,
+      "loss": 1.2487,
+      "step": 9697
+    },
+    {
+      "epoch": 0.6990557197433864,
+      "grad_norm": 0.11938832012563978,
+      "learning_rate": 4.386197081330965e-05,
+      "loss": 1.2719,
+      "step": 9698
+    },
+    {
+      "epoch": 0.6991278022057233,
+      "grad_norm": 0.11374976037654283,
+      "learning_rate": 4.3842651222882346e-05,
+      "loss": 1.2023,
+      "step": 9699
+    },
+    {
+      "epoch": 0.6991998846680603,
+      "grad_norm": 0.11595993278258893,
+      "learning_rate": 4.382333469353295e-05,
+      "loss": 1.2061,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6992719671303972,
+      "grad_norm": 0.1412365654191786,
+      "learning_rate": 4.380402122631441e-05,
+      "loss": 1.1907,
+      "step": 9701
+    },
+    {
+      "epoch": 0.6993440495927341,
+      "grad_norm": 0.1222527990836463,
+      "learning_rate": 4.3784710822279394e-05,
+      "loss": 1.3422,
+      "step": 9702
+    },
+    {
+      "epoch": 0.699416132055071,
+      "grad_norm": 0.13739248102429302,
+      "learning_rate": 4.376540348248058e-05,
+      "loss": 1.2967,
+      "step": 9703
+    },
+    {
+      "epoch": 0.699488214517408,
+      "grad_norm": 0.11847570179774439,
+      "learning_rate": 4.374609920797038e-05,
+      "loss": 1.1876,
+      "step": 9704
+    },
+    {
+      "epoch": 0.6995602969797449,
+      "grad_norm": 0.12478032597036753,
+      "learning_rate": 4.372679799980106e-05,
+      "loss": 1.345,
+      "step": 9705
+    },
+    {
+      "epoch": 0.6996323794420818,
+      "grad_norm": 0.12055085844369406,
+      "learning_rate": 4.3707499859024615e-05,
+      "loss": 1.3143,
+      "step": 9706
+    },
+    {
+      "epoch": 0.6997044619044187,
+      "grad_norm": 0.29647612920375643,
+      "learning_rate": 4.3688204786693075e-05,
+      "loss": 1.251,
+      "step": 9707
+    },
+    {
+      "epoch": 0.6997765443667555,
+      "grad_norm": 0.11994478309423562,
+      "learning_rate": 4.366891278385815e-05,
+      "loss": 1.2579,
+      "step": 9708
+    },
+    {
+      "epoch": 0.6998486268290924,
+      "grad_norm": 0.11385588577583898,
+      "learning_rate": 4.3649623851571464e-05,
+      "loss": 1.1471,
+      "step": 9709
+    },
+    {
+      "epoch": 0.6999207092914294,
+      "grad_norm": 0.11156939765090337,
+      "learning_rate": 4.363033799088433e-05,
+      "loss": 1.2611,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6999927917537663,
+      "grad_norm": 0.12042675177895443,
+      "learning_rate": 4.361105520284812e-05,
+      "loss": 1.1088,
+      "step": 9711
+    },
+    {
+      "epoch": 0.7000648742161032,
+      "grad_norm": 0.12148330071479442,
+      "learning_rate": 4.359177548851386e-05,
+      "loss": 1.2413,
+      "step": 9712
+    },
+    {
+      "epoch": 0.7001369566784401,
+      "grad_norm": 0.12075160854304035,
+      "learning_rate": 4.357249884893252e-05,
+      "loss": 1.3102,
+      "step": 9713
+    },
+    {
+      "epoch": 0.700209039140777,
+      "grad_norm": 0.11480010631939438,
+      "learning_rate": 4.355322528515476e-05,
+      "loss": 1.3291,
+      "step": 9714
+    },
+    {
+      "epoch": 0.700281121603114,
+      "grad_norm": 0.12187618790392919,
+      "learning_rate": 4.353395479823119e-05,
+      "loss": 1.2263,
+      "step": 9715
+    },
+    {
+      "epoch": 0.7003532040654509,
+      "grad_norm": 0.12307883322900623,
+      "learning_rate": 4.351468738921227e-05,
+      "loss": 1.3405,
+      "step": 9716
+    },
+    {
+      "epoch": 0.7004252865277878,
+      "grad_norm": 0.10632515413319332,
+      "learning_rate": 4.349542305914826e-05,
+      "loss": 1.2744,
+      "step": 9717
+    },
+    {
+      "epoch": 0.7004973689901247,
+      "grad_norm": 0.11735159829896592,
+      "learning_rate": 4.347616180908915e-05,
+      "loss": 1.2259,
+      "step": 9718
+    },
+    {
+      "epoch": 0.7005694514524616,
+      "grad_norm": 0.11928654148839846,
+      "learning_rate": 4.345690364008488e-05,
+      "loss": 1.2477,
+      "step": 9719
+    },
+    {
+      "epoch": 0.7006415339147986,
+      "grad_norm": 0.13266809416206363,
+      "learning_rate": 4.343764855318524e-05,
+      "loss": 1.1936,
+      "step": 9720
+    },
+    {
+      "epoch": 0.7007136163771355,
+      "grad_norm": 0.11687914461283543,
+      "learning_rate": 4.3418396549439814e-05,
+      "loss": 1.2332,
+      "step": 9721
+    },
+    {
+      "epoch": 0.7007856988394724,
+      "grad_norm": 0.11441920860523168,
+      "learning_rate": 4.339914762989792e-05,
+      "loss": 1.3229,
+      "step": 9722
+    },
+    {
+      "epoch": 0.7008577813018093,
+      "grad_norm": 0.11810573952475838,
+      "learning_rate": 4.3379901795608856e-05,
+      "loss": 1.2324,
+      "step": 9723
+    },
+    {
+      "epoch": 0.7009298637641462,
+      "grad_norm": 0.1128978368558449,
+      "learning_rate": 4.336065904762164e-05,
+      "loss": 1.3305,
+      "step": 9724
+    },
+    {
+      "epoch": 0.701001946226483,
+      "grad_norm": 0.11033337331773664,
+      "learning_rate": 4.3341419386985295e-05,
+      "loss": 1.2606,
+      "step": 9725
+    },
+    {
+      "epoch": 0.70107402868882,
+      "grad_norm": 0.1285162787838968,
+      "learning_rate": 4.3322182814748436e-05,
+      "loss": 1.2523,
+      "step": 9726
+    },
+    {
+      "epoch": 0.7011461111511569,
+      "grad_norm": 0.11557187312603659,
+      "learning_rate": 4.3302949331959675e-05,
+      "loss": 1.2176,
+      "step": 9727
+    },
+    {
+      "epoch": 0.7012181936134938,
+      "grad_norm": 0.11391207671971633,
+      "learning_rate": 4.328371893966736e-05,
+      "loss": 1.3469,
+      "step": 9728
+    },
+    {
+      "epoch": 0.7012902760758307,
+      "grad_norm": 0.11736542597296518,
+      "learning_rate": 4.3264491638919826e-05,
+      "loss": 1.2892,
+      "step": 9729
+    },
+    {
+      "epoch": 0.7013623585381676,
+      "grad_norm": 0.12012087391204855,
+      "learning_rate": 4.324526743076504e-05,
+      "loss": 1.2191,
+      "step": 9730
+    },
+    {
+      "epoch": 0.7014344410005046,
+      "grad_norm": 0.12273607973908895,
+      "learning_rate": 4.3226046316250927e-05,
+      "loss": 1.2966,
+      "step": 9731
+    },
+    {
+      "epoch": 0.7015065234628415,
+      "grad_norm": 0.1299014259842444,
+      "learning_rate": 4.320682829642515e-05,
+      "loss": 1.2934,
+      "step": 9732
+    },
+    {
+      "epoch": 0.7015786059251784,
+      "grad_norm": 0.11783322679440873,
+      "learning_rate": 4.31876133723354e-05,
+      "loss": 1.2538,
+      "step": 9733
+    },
+    {
+      "epoch": 0.7016506883875153,
+      "grad_norm": 0.12321905828123236,
+      "learning_rate": 4.316840154502895e-05,
+      "loss": 1.2959,
+      "step": 9734
+    },
+    {
+      "epoch": 0.7017227708498522,
+      "grad_norm": 0.1339168807322783,
+      "learning_rate": 4.3149192815553044e-05,
+      "loss": 1.3196,
+      "step": 9735
+    },
+    {
+      "epoch": 0.7017948533121892,
+      "grad_norm": 0.12604597398922393,
+      "learning_rate": 4.3129987184954726e-05,
+      "loss": 1.2494,
+      "step": 9736
+    },
+    {
+      "epoch": 0.7018669357745261,
+      "grad_norm": 0.11244771739574042,
+      "learning_rate": 4.311078465428088e-05,
+      "loss": 1.2247,
+      "step": 9737
+    },
+    {
+      "epoch": 0.701939018236863,
+      "grad_norm": 0.11816687565478633,
+      "learning_rate": 4.3091585224578214e-05,
+      "loss": 1.1923,
+      "step": 9738
+    },
+    {
+      "epoch": 0.7020111006991999,
+      "grad_norm": 0.26732315373982707,
+      "learning_rate": 4.307238889689327e-05,
+      "loss": 1.2098,
+      "step": 9739
+    },
+    {
+      "epoch": 0.7020831831615368,
+      "grad_norm": 0.11867736393733222,
+      "learning_rate": 4.305319567227243e-05,
+      "loss": 1.2676,
+      "step": 9740
+    },
+    {
+      "epoch": 0.7021552656238738,
+      "grad_norm": 0.1169946394267667,
+      "learning_rate": 4.303400555176187e-05,
+      "loss": 1.2349,
+      "step": 9741
+    },
+    {
+      "epoch": 0.7022273480862107,
+      "grad_norm": 0.14832202461820979,
+      "learning_rate": 4.301481853640765e-05,
+      "loss": 1.3013,
+      "step": 9742
+    },
+    {
+      "epoch": 0.7022994305485475,
+      "grad_norm": 0.11364231118039335,
+      "learning_rate": 4.299563462725563e-05,
+      "loss": 1.2513,
+      "step": 9743
+    },
+    {
+      "epoch": 0.7023715130108844,
+      "grad_norm": 0.11826406988906346,
+      "learning_rate": 4.2976453825351495e-05,
+      "loss": 1.369,
+      "step": 9744
+    },
+    {
+      "epoch": 0.7024435954732213,
+      "grad_norm": 0.12009271021539836,
+      "learning_rate": 4.2957276131740766e-05,
+      "loss": 1.3164,
+      "step": 9745
+    },
+    {
+      "epoch": 0.7025156779355582,
+      "grad_norm": 0.11281419013255505,
+      "learning_rate": 4.293810154746881e-05,
+      "loss": 1.2289,
+      "step": 9746
+    },
+    {
+      "epoch": 0.7025877603978952,
+      "grad_norm": 0.11281266156222619,
+      "learning_rate": 4.291893007358081e-05,
+      "loss": 1.2192,
+      "step": 9747
+    },
+    {
+      "epoch": 0.7026598428602321,
+      "grad_norm": 0.1270260598443084,
+      "learning_rate": 4.2899761711121787e-05,
+      "loss": 1.3464,
+      "step": 9748
+    },
+    {
+      "epoch": 0.702731925322569,
+      "grad_norm": 0.131895718663705,
+      "learning_rate": 4.288059646113658e-05,
+      "loss": 1.2118,
+      "step": 9749
+    },
+    {
+      "epoch": 0.7028040077849059,
+      "grad_norm": 0.12552667584473265,
+      "learning_rate": 4.286143432466987e-05,
+      "loss": 1.2394,
+      "step": 9750
+    },
+    {
+      "epoch": 0.7028760902472428,
+      "grad_norm": 0.11704637621175538,
+      "learning_rate": 4.284227530276617e-05,
+      "loss": 1.2448,
+      "step": 9751
+    },
+    {
+      "epoch": 0.7029481727095798,
+      "grad_norm": 0.11836169930321011,
+      "learning_rate": 4.28231193964698e-05,
+      "loss": 1.347,
+      "step": 9752
+    },
+    {
+      "epoch": 0.7030202551719167,
+      "grad_norm": 0.12688278704268194,
+      "learning_rate": 4.280396660682497e-05,
+      "loss": 1.2599,
+      "step": 9753
+    },
+    {
+      "epoch": 0.7030923376342536,
+      "grad_norm": 0.11396852546385768,
+      "learning_rate": 4.2784816934875625e-05,
+      "loss": 1.3295,
+      "step": 9754
+    },
+    {
+      "epoch": 0.7031644200965905,
+      "grad_norm": 0.11663656100282781,
+      "learning_rate": 4.276567038166563e-05,
+      "loss": 1.2,
+      "step": 9755
+    },
+    {
+      "epoch": 0.7032365025589274,
+      "grad_norm": 0.10491284233670786,
+      "learning_rate": 4.274652694823862e-05,
+      "loss": 1.2138,
+      "step": 9756
+    },
+    {
+      "epoch": 0.7033085850212644,
+      "grad_norm": 0.11558367744713083,
+      "learning_rate": 4.272738663563811e-05,
+      "loss": 1.2324,
+      "step": 9757
+    },
+    {
+      "epoch": 0.7033806674836013,
+      "grad_norm": 0.12063281358029776,
+      "learning_rate": 4.2708249444907436e-05,
+      "loss": 1.314,
+      "step": 9758
+    },
+    {
+      "epoch": 0.7034527499459382,
+      "grad_norm": 0.12021919169149133,
+      "learning_rate": 4.268911537708964e-05,
+      "loss": 1.3226,
+      "step": 9759
+    },
+    {
+      "epoch": 0.703524832408275,
+      "grad_norm": 0.11829427151807434,
+      "learning_rate": 4.266998443322781e-05,
+      "loss": 1.1717,
+      "step": 9760
+    },
+    {
+      "epoch": 0.7035969148706119,
+      "grad_norm": 0.11626744027019018,
+      "learning_rate": 4.265085661436471e-05,
+      "loss": 1.2034,
+      "step": 9761
+    },
+    {
+      "epoch": 0.7036689973329489,
+      "grad_norm": 0.11188297925409385,
+      "learning_rate": 4.2631731921543026e-05,
+      "loss": 1.2986,
+      "step": 9762
+    },
+    {
+      "epoch": 0.7037410797952858,
+      "grad_norm": 0.12572895761882416,
+      "learning_rate": 4.2612610355805114e-05,
+      "loss": 1.3414,
+      "step": 9763
+    },
+    {
+      "epoch": 0.7038131622576227,
+      "grad_norm": 0.1297267747879769,
+      "learning_rate": 4.259349191819337e-05,
+      "loss": 1.2446,
+      "step": 9764
+    },
+    {
+      "epoch": 0.7038852447199596,
+      "grad_norm": 0.11080079230876944,
+      "learning_rate": 4.25743766097499e-05,
+      "loss": 1.2181,
+      "step": 9765
+    },
+    {
+      "epoch": 0.7039573271822965,
+      "grad_norm": 0.11920828001273485,
+      "learning_rate": 4.255526443151668e-05,
+      "loss": 1.2582,
+      "step": 9766
+    },
+    {
+      "epoch": 0.7040294096446335,
+      "grad_norm": 0.12464889711395384,
+      "learning_rate": 4.2536155384535426e-05,
+      "loss": 1.3606,
+      "step": 9767
+    },
+    {
+      "epoch": 0.7041014921069704,
+      "grad_norm": 0.10797426642321843,
+      "learning_rate": 4.251704946984776e-05,
+      "loss": 1.2181,
+      "step": 9768
+    },
+    {
+      "epoch": 0.7041735745693073,
+      "grad_norm": 0.12832002429792164,
+      "learning_rate": 4.2497946688495184e-05,
+      "loss": 1.3084,
+      "step": 9769
+    },
+    {
+      "epoch": 0.7042456570316442,
+      "grad_norm": 0.12468742408845232,
+      "learning_rate": 4.247884704151898e-05,
+      "loss": 1.2416,
+      "step": 9770
+    },
+    {
+      "epoch": 0.7043177394939811,
+      "grad_norm": 0.11105935688948274,
+      "learning_rate": 4.2459750529960176e-05,
+      "loss": 1.2785,
+      "step": 9771
+    },
+    {
+      "epoch": 0.704389821956318,
+      "grad_norm": 0.11436921489809243,
+      "learning_rate": 4.24406571548597e-05,
+      "loss": 1.2285,
+      "step": 9772
+    },
+    {
+      "epoch": 0.704461904418655,
+      "grad_norm": 0.10553680749487611,
+      "learning_rate": 4.242156691725838e-05,
+      "loss": 1.165,
+      "step": 9773
+    },
+    {
+      "epoch": 0.7045339868809919,
+      "grad_norm": 0.12062234748106698,
+      "learning_rate": 4.240247981819683e-05,
+      "loss": 1.304,
+      "step": 9774
+    },
+    {
+      "epoch": 0.7046060693433288,
+      "grad_norm": 0.13752014294740722,
+      "learning_rate": 4.238339585871536e-05,
+      "loss": 1.1721,
+      "step": 9775
+    },
+    {
+      "epoch": 0.7046781518056657,
+      "grad_norm": 0.11777097203084382,
+      "learning_rate": 4.236431503985423e-05,
+      "loss": 1.2556,
+      "step": 9776
+    },
+    {
+      "epoch": 0.7047502342680025,
+      "grad_norm": 0.11156718615156104,
+      "learning_rate": 4.23452373626536e-05,
+      "loss": 1.2957,
+      "step": 9777
+    },
+    {
+      "epoch": 0.7048223167303395,
+      "grad_norm": 0.10663411070950686,
+      "learning_rate": 4.232616282815336e-05,
+      "loss": 1.2324,
+      "step": 9778
+    },
+    {
+      "epoch": 0.7048943991926764,
+      "grad_norm": 0.11516980816708348,
+      "learning_rate": 4.230709143739317e-05,
+      "loss": 1.2602,
+      "step": 9779
+    },
+    {
+      "epoch": 0.7049664816550133,
+      "grad_norm": 0.12026441627654522,
+      "learning_rate": 4.228802319141264e-05,
+      "loss": 1.2584,
+      "step": 9780
+    },
+    {
+      "epoch": 0.7050385641173502,
+      "grad_norm": 0.1199863530935817,
+      "learning_rate": 4.2268958091251107e-05,
+      "loss": 1.3175,
+      "step": 9781
+    },
+    {
+      "epoch": 0.7051106465796871,
+      "grad_norm": 0.11550383309913327,
+      "learning_rate": 4.224989613794792e-05,
+      "loss": 1.2277,
+      "step": 9782
+    },
+    {
+      "epoch": 0.7051827290420241,
+      "grad_norm": 0.13073685324166331,
+      "learning_rate": 4.2230837332541994e-05,
+      "loss": 1.2985,
+      "step": 9783
+    },
+    {
+      "epoch": 0.705254811504361,
+      "grad_norm": 0.12118977014096764,
+      "learning_rate": 4.221178167607226e-05,
+      "loss": 1.2367,
+      "step": 9784
+    },
+    {
+      "epoch": 0.7053268939666979,
+      "grad_norm": 0.11591496288185636,
+      "learning_rate": 4.219272916957736e-05,
+      "loss": 1.2778,
+      "step": 9785
+    },
+    {
+      "epoch": 0.7053989764290348,
+      "grad_norm": 0.1149546502624633,
+      "learning_rate": 4.2173679814095964e-05,
+      "loss": 1.2167,
+      "step": 9786
+    },
+    {
+      "epoch": 0.7054710588913717,
+      "grad_norm": 0.114366429253563,
+      "learning_rate": 4.215463361066632e-05,
+      "loss": 1.2573,
+      "step": 9787
+    },
+    {
+      "epoch": 0.7055431413537087,
+      "grad_norm": 0.11257916251608457,
+      "learning_rate": 4.213559056032662e-05,
+      "loss": 1.4716,
+      "step": 9788
+    },
+    {
+      "epoch": 0.7056152238160456,
+      "grad_norm": 0.12380303501656088,
+      "learning_rate": 4.211655066411492e-05,
+      "loss": 1.2775,
+      "step": 9789
+    },
+    {
+      "epoch": 0.7056873062783825,
+      "grad_norm": 0.11430261602398872,
+      "learning_rate": 4.2097513923069045e-05,
+      "loss": 1.1837,
+      "step": 9790
+    },
+    {
+      "epoch": 0.7057593887407194,
+      "grad_norm": 0.12342837608630473,
+      "learning_rate": 4.207848033822668e-05,
+      "loss": 1.3509,
+      "step": 9791
+    },
+    {
+      "epoch": 0.7058314712030563,
+      "grad_norm": 0.1169419246118276,
+      "learning_rate": 4.20594499106253e-05,
+      "loss": 1.2479,
+      "step": 9792
+    },
+    {
+      "epoch": 0.7059035536653933,
+      "grad_norm": 0.11106436022128591,
+      "learning_rate": 4.204042264130227e-05,
+      "loss": 1.3517,
+      "step": 9793
+    },
+    {
+      "epoch": 0.7059756361277302,
+      "grad_norm": 0.121924604439471,
+      "learning_rate": 4.2021398531294706e-05,
+      "loss": 1.3231,
+      "step": 9794
+    },
+    {
+      "epoch": 0.706047718590067,
+      "grad_norm": 0.11319569601296897,
+      "learning_rate": 4.200237758163963e-05,
+      "loss": 1.2508,
+      "step": 9795
+    },
+    {
+      "epoch": 0.7061198010524039,
+      "grad_norm": 0.1218020813574599,
+      "learning_rate": 4.198335979337383e-05,
+      "loss": 1.2372,
+      "step": 9796
+    },
+    {
+      "epoch": 0.7061918835147408,
+      "grad_norm": 0.12832427724785342,
+      "learning_rate": 4.196434516753396e-05,
+      "loss": 1.2457,
+      "step": 9797
+    },
+    {
+      "epoch": 0.7062639659770777,
+      "grad_norm": 0.11558615500405184,
+      "learning_rate": 4.194533370515648e-05,
+      "loss": 1.2311,
+      "step": 9798
+    },
+    {
+      "epoch": 0.7063360484394147,
+      "grad_norm": 0.12339207913770787,
+      "learning_rate": 4.192632540727768e-05,
+      "loss": 1.229,
+      "step": 9799
+    },
+    {
+      "epoch": 0.7064081309017516,
+      "grad_norm": 0.12088463210100131,
+      "learning_rate": 4.190732027493368e-05,
+      "loss": 1.2317,
+      "step": 9800
+    },
+    {
+      "epoch": 0.7064802133640885,
+      "grad_norm": 0.11848447754604721,
+      "learning_rate": 4.1888318309160445e-05,
+      "loss": 1.2176,
+      "step": 9801
+    },
+    {
+      "epoch": 0.7065522958264254,
+      "grad_norm": 0.11538338980497485,
+      "learning_rate": 4.1869319510993755e-05,
+      "loss": 1.2329,
+      "step": 9802
+    },
+    {
+      "epoch": 0.7066243782887623,
+      "grad_norm": 0.12205286603232822,
+      "learning_rate": 4.185032388146919e-05,
+      "loss": 1.2867,
+      "step": 9803
+    },
+    {
+      "epoch": 0.7066964607510993,
+      "grad_norm": 0.11035667154063859,
+      "learning_rate": 4.18313314216222e-05,
+      "loss": 1.2235,
+      "step": 9804
+    },
+    {
+      "epoch": 0.7067685432134362,
+      "grad_norm": 0.14551406492120142,
+      "learning_rate": 4.1812342132488036e-05,
+      "loss": 1.2865,
+      "step": 9805
+    },
+    {
+      "epoch": 0.7068406256757731,
+      "grad_norm": 0.11904442052472254,
+      "learning_rate": 4.179335601510178e-05,
+      "loss": 1.2634,
+      "step": 9806
+    },
+    {
+      "epoch": 0.70691270813811,
+      "grad_norm": 0.13005070102216068,
+      "learning_rate": 4.177437307049836e-05,
+      "loss": 1.2537,
+      "step": 9807
+    },
+    {
+      "epoch": 0.7069847906004469,
+      "grad_norm": 0.11324507386436804,
+      "learning_rate": 4.175539329971252e-05,
+      "loss": 1.2442,
+      "step": 9808
+    },
+    {
+      "epoch": 0.7070568730627839,
+      "grad_norm": 0.12103314645782348,
+      "learning_rate": 4.1736416703778804e-05,
+      "loss": 1.3076,
+      "step": 9809
+    },
+    {
+      "epoch": 0.7071289555251208,
+      "grad_norm": 0.11934281278027215,
+      "learning_rate": 4.1717443283731615e-05,
+      "loss": 1.2439,
+      "step": 9810
+    },
+    {
+      "epoch": 0.7072010379874577,
+      "grad_norm": 0.10831856743739383,
+      "learning_rate": 4.169847304060522e-05,
+      "loss": 1.3043,
+      "step": 9811
+    },
+    {
+      "epoch": 0.7072731204497945,
+      "grad_norm": 0.11178939757671832,
+      "learning_rate": 4.167950597543356e-05,
+      "loss": 1.2584,
+      "step": 9812
+    },
+    {
+      "epoch": 0.7073452029121314,
+      "grad_norm": 0.12320256148621837,
+      "learning_rate": 4.16605420892506e-05,
+      "loss": 1.278,
+      "step": 9813
+    },
+    {
+      "epoch": 0.7074172853744684,
+      "grad_norm": 0.10301553529954983,
+      "learning_rate": 4.164158138309002e-05,
+      "loss": 1.2817,
+      "step": 9814
+    },
+    {
+      "epoch": 0.7074893678368053,
+      "grad_norm": 0.11754412754622574,
+      "learning_rate": 4.1622623857985374e-05,
+      "loss": 1.2509,
+      "step": 9815
+    },
+    {
+      "epoch": 0.7075614502991422,
+      "grad_norm": 0.12298907757787592,
+      "learning_rate": 4.160366951496992e-05,
+      "loss": 1.2448,
+      "step": 9816
+    },
+    {
+      "epoch": 0.7076335327614791,
+      "grad_norm": 0.11716567305610184,
+      "learning_rate": 4.158471835507695e-05,
+      "loss": 1.1869,
+      "step": 9817
+    },
+    {
+      "epoch": 0.707705615223816,
+      "grad_norm": 0.10525661746047178,
+      "learning_rate": 4.156577037933942e-05,
+      "loss": 1.2855,
+      "step": 9818
+    },
+    {
+      "epoch": 0.707777697686153,
+      "grad_norm": 0.12819719547239142,
+      "learning_rate": 4.154682558879021e-05,
+      "loss": 1.2472,
+      "step": 9819
+    },
+    {
+      "epoch": 0.7078497801484899,
+      "grad_norm": 0.11577962511243303,
+      "learning_rate": 4.152788398446186e-05,
+      "loss": 1.2984,
+      "step": 9820
+    },
+    {
+      "epoch": 0.7079218626108268,
+      "grad_norm": 0.12141890229125314,
+      "learning_rate": 4.150894556738698e-05,
+      "loss": 1.1957,
+      "step": 9821
+    },
+    {
+      "epoch": 0.7079939450731637,
+      "grad_norm": 0.11061366033286722,
+      "learning_rate": 4.149001033859788e-05,
+      "loss": 1.3594,
+      "step": 9822
+    },
+    {
+      "epoch": 0.7080660275355006,
+      "grad_norm": 0.12531473827619904,
+      "learning_rate": 4.147107829912663e-05,
+      "loss": 1.2414,
+      "step": 9823
+    },
+    {
+      "epoch": 0.7081381099978376,
+      "grad_norm": 0.12912525216185175,
+      "learning_rate": 4.145214945000523e-05,
+      "loss": 1.234,
+      "step": 9824
+    },
+    {
+      "epoch": 0.7082101924601745,
+      "grad_norm": 0.11377497727646552,
+      "learning_rate": 4.143322379226542e-05,
+      "loss": 1.2611,
+      "step": 9825
+    },
+    {
+      "epoch": 0.7082822749225114,
+      "grad_norm": 0.1336348240168533,
+      "learning_rate": 4.141430132693895e-05,
+      "loss": 1.3119,
+      "step": 9826
+    },
+    {
+      "epoch": 0.7083543573848483,
+      "grad_norm": 0.11760720316943725,
+      "learning_rate": 4.139538205505714e-05,
+      "loss": 1.3065,
+      "step": 9827
+    },
+    {
+      "epoch": 0.7084264398471852,
+      "grad_norm": 0.12118845805520548,
+      "learning_rate": 4.13764659776513e-05,
+      "loss": 1.2249,
+      "step": 9828
+    },
+    {
+      "epoch": 0.708498522309522,
+      "grad_norm": 0.11388276382641224,
+      "learning_rate": 4.135755309575249e-05,
+      "loss": 1.3296,
+      "step": 9829
+    },
+    {
+      "epoch": 0.708570604771859,
+      "grad_norm": 0.12230650108559278,
+      "learning_rate": 4.133864341039174e-05,
+      "loss": 1.2693,
+      "step": 9830
+    },
+    {
+      "epoch": 0.7086426872341959,
+      "grad_norm": 0.12143804674230513,
+      "learning_rate": 4.1319736922599686e-05,
+      "loss": 1.3261,
+      "step": 9831
+    },
+    {
+      "epoch": 0.7087147696965328,
+      "grad_norm": 0.11728322778858775,
+      "learning_rate": 4.1300833633406935e-05,
+      "loss": 1.2947,
+      "step": 9832
+    },
+    {
+      "epoch": 0.7087868521588697,
+      "grad_norm": 0.12358798103529292,
+      "learning_rate": 4.12819335438439e-05,
+      "loss": 1.2374,
+      "step": 9833
+    },
+    {
+      "epoch": 0.7088589346212066,
+      "grad_norm": 0.1307034262786727,
+      "learning_rate": 4.1263036654940775e-05,
+      "loss": 1.2069,
+      "step": 9834
+    },
+    {
+      "epoch": 0.7089310170835436,
+      "grad_norm": 0.11442393177879308,
+      "learning_rate": 4.1244142967727636e-05,
+      "loss": 1.2047,
+      "step": 9835
+    },
+    {
+      "epoch": 0.7090030995458805,
+      "grad_norm": 0.13726518978906385,
+      "learning_rate": 4.122525248323436e-05,
+      "loss": 1.2877,
+      "step": 9836
+    },
+    {
+      "epoch": 0.7090751820082174,
+      "grad_norm": 0.11300289007252114,
+      "learning_rate": 4.120636520249063e-05,
+      "loss": 1.3386,
+      "step": 9837
+    },
+    {
+      "epoch": 0.7091472644705543,
+      "grad_norm": 0.11575714126927186,
+      "learning_rate": 4.1187481126525987e-05,
+      "loss": 1.3442,
+      "step": 9838
+    },
+    {
+      "epoch": 0.7092193469328912,
+      "grad_norm": 0.11132633211043272,
+      "learning_rate": 4.116860025636976e-05,
+      "loss": 1.3023,
+      "step": 9839
+    },
+    {
+      "epoch": 0.7092914293952282,
+      "grad_norm": 0.1137721756804163,
+      "learning_rate": 4.114972259305115e-05,
+      "loss": 1.3154,
+      "step": 9840
+    },
+    {
+      "epoch": 0.7093635118575651,
+      "grad_norm": 0.11644058611232799,
+      "learning_rate": 4.113084813759914e-05,
+      "loss": 1.3793,
+      "step": 9841
+    },
+    {
+      "epoch": 0.709435594319902,
+      "grad_norm": 0.1145814118573321,
+      "learning_rate": 4.111197689104258e-05,
+      "loss": 1.2694,
+      "step": 9842
+    },
+    {
+      "epoch": 0.7095076767822389,
+      "grad_norm": 0.11395782021713652,
+      "learning_rate": 4.109310885441009e-05,
+      "loss": 1.231,
+      "step": 9843
+    },
+    {
+      "epoch": 0.7095797592445758,
+      "grad_norm": 0.11736806295287787,
+      "learning_rate": 4.1074244028730166e-05,
+      "loss": 1.2636,
+      "step": 9844
+    },
+    {
+      "epoch": 0.7096518417069128,
+      "grad_norm": 0.11885602345257634,
+      "learning_rate": 4.105538241503112e-05,
+      "loss": 1.3368,
+      "step": 9845
+    },
+    {
+      "epoch": 0.7097239241692497,
+      "grad_norm": 0.12970155487140692,
+      "learning_rate": 4.103652401434106e-05,
+      "loss": 1.3691,
+      "step": 9846
+    },
+    {
+      "epoch": 0.7097960066315865,
+      "grad_norm": 0.10737149881686855,
+      "learning_rate": 4.101766882768794e-05,
+      "loss": 1.208,
+      "step": 9847
+    },
+    {
+      "epoch": 0.7098680890939234,
+      "grad_norm": 0.1197387071053039,
+      "learning_rate": 4.099881685609954e-05,
+      "loss": 1.3056,
+      "step": 9848
+    },
+    {
+      "epoch": 0.7099401715562603,
+      "grad_norm": 0.11593776794831834,
+      "learning_rate": 4.097996810060346e-05,
+      "loss": 1.3575,
+      "step": 9849
+    },
+    {
+      "epoch": 0.7100122540185972,
+      "grad_norm": 0.11464573145320603,
+      "learning_rate": 4.096112256222713e-05,
+      "loss": 1.3454,
+      "step": 9850
+    },
+    {
+      "epoch": 0.7100843364809342,
+      "grad_norm": 0.12970141322674988,
+      "learning_rate": 4.09422802419978e-05,
+      "loss": 1.2865,
+      "step": 9851
+    },
+    {
+      "epoch": 0.7101564189432711,
+      "grad_norm": 0.11660729650116361,
+      "learning_rate": 4.092344114094254e-05,
+      "loss": 1.2366,
+      "step": 9852
+    },
+    {
+      "epoch": 0.710228501405608,
+      "grad_norm": 0.12537488499566116,
+      "learning_rate": 4.0904605260088244e-05,
+      "loss": 1.1767,
+      "step": 9853
+    },
+    {
+      "epoch": 0.7103005838679449,
+      "grad_norm": 0.12381994014001291,
+      "learning_rate": 4.088577260046166e-05,
+      "loss": 1.2401,
+      "step": 9854
+    },
+    {
+      "epoch": 0.7103726663302818,
+      "grad_norm": 0.11157713782465294,
+      "learning_rate": 4.086694316308934e-05,
+      "loss": 1.2873,
+      "step": 9855
+    },
+    {
+      "epoch": 0.7104447487926188,
+      "grad_norm": 0.15082911145245737,
+      "learning_rate": 4.084811694899757e-05,
+      "loss": 1.268,
+      "step": 9856
+    },
+    {
+      "epoch": 0.7105168312549557,
+      "grad_norm": 0.11795083003735558,
+      "learning_rate": 4.0829293959212656e-05,
+      "loss": 1.2602,
+      "step": 9857
+    },
+    {
+      "epoch": 0.7105889137172926,
+      "grad_norm": 0.10353759461898075,
+      "learning_rate": 4.081047419476057e-05,
+      "loss": 1.1984,
+      "step": 9858
+    },
+    {
+      "epoch": 0.7106609961796295,
+      "grad_norm": 0.11353799581828887,
+      "learning_rate": 4.0791657656667205e-05,
+      "loss": 1.282,
+      "step": 9859
+    },
+    {
+      "epoch": 0.7107330786419664,
+      "grad_norm": 0.11431921461397734,
+      "learning_rate": 4.077284434595813e-05,
+      "loss": 1.2859,
+      "step": 9860
+    },
+    {
+      "epoch": 0.7108051611043034,
+      "grad_norm": 0.12283375576614743,
+      "learning_rate": 4.0754034263658916e-05,
+      "loss": 1.224,
+      "step": 9861
+    },
+    {
+      "epoch": 0.7108772435666403,
+      "grad_norm": 0.14894922183316756,
+      "learning_rate": 4.0735227410794875e-05,
+      "loss": 1.2558,
+      "step": 9862
+    },
+    {
+      "epoch": 0.7109493260289772,
+      "grad_norm": 0.11417111987987096,
+      "learning_rate": 4.071642378839117e-05,
+      "loss": 1.3112,
+      "step": 9863
+    },
+    {
+      "epoch": 0.711021408491314,
+      "grad_norm": 0.1162291994876297,
+      "learning_rate": 4.069762339747266e-05,
+      "loss": 1.2164,
+      "step": 9864
+    },
+    {
+      "epoch": 0.7110934909536509,
+      "grad_norm": 0.12814503473140995,
+      "learning_rate": 4.067882623906425e-05,
+      "loss": 1.4001,
+      "step": 9865
+    },
+    {
+      "epoch": 0.7111655734159879,
+      "grad_norm": 0.11349031061593318,
+      "learning_rate": 4.066003231419051e-05,
+      "loss": 1.1961,
+      "step": 9866
+    },
+    {
+      "epoch": 0.7112376558783248,
+      "grad_norm": 0.10734025210204173,
+      "learning_rate": 4.0641241623875916e-05,
+      "loss": 1.22,
+      "step": 9867
+    },
+    {
+      "epoch": 0.7113097383406617,
+      "grad_norm": 0.11993577534373923,
+      "learning_rate": 4.062245416914465e-05,
+      "loss": 1.3358,
+      "step": 9868
+    },
+    {
+      "epoch": 0.7113818208029986,
+      "grad_norm": 0.11613070287300448,
+      "learning_rate": 4.060366995102082e-05,
+      "loss": 1.3039,
+      "step": 9869
+    },
+    {
+      "epoch": 0.7114539032653355,
+      "grad_norm": 0.11291252607125007,
+      "learning_rate": 4.0584888970528364e-05,
+      "loss": 1.2678,
+      "step": 9870
+    },
+    {
+      "epoch": 0.7115259857276724,
+      "grad_norm": 0.11568153797547549,
+      "learning_rate": 4.0566111228691064e-05,
+      "loss": 1.2505,
+      "step": 9871
+    },
+    {
+      "epoch": 0.7115980681900094,
+      "grad_norm": 0.1224059051940209,
+      "learning_rate": 4.054733672653236e-05,
+      "loss": 1.2629,
+      "step": 9872
+    },
+    {
+      "epoch": 0.7116701506523463,
+      "grad_norm": 0.11215208966261708,
+      "learning_rate": 4.052856546507565e-05,
+      "loss": 1.2163,
+      "step": 9873
+    },
+    {
+      "epoch": 0.7117422331146832,
+      "grad_norm": 0.13873882358708708,
+      "learning_rate": 4.050979744534421e-05,
+      "loss": 1.3635,
+      "step": 9874
+    },
+    {
+      "epoch": 0.7118143155770201,
+      "grad_norm": 0.11090805429683591,
+      "learning_rate": 4.0491032668361054e-05,
+      "loss": 1.3301,
+      "step": 9875
+    },
+    {
+      "epoch": 0.711886398039357,
+      "grad_norm": 0.11583062297470205,
+      "learning_rate": 4.047227113514897e-05,
+      "loss": 1.183,
+      "step": 9876
+    },
+    {
+      "epoch": 0.711958480501694,
+      "grad_norm": 0.10986619194046066,
+      "learning_rate": 4.0453512846730646e-05,
+      "loss": 1.1827,
+      "step": 9877
+    },
+    {
+      "epoch": 0.7120305629640309,
+      "grad_norm": 0.2729044781157872,
+      "learning_rate": 4.0434757804128564e-05,
+      "loss": 1.2444,
+      "step": 9878
+    },
+    {
+      "epoch": 0.7121026454263678,
+      "grad_norm": 0.11903235402050652,
+      "learning_rate": 4.041600600836514e-05,
+      "loss": 1.2531,
+      "step": 9879
+    },
+    {
+      "epoch": 0.7121747278887047,
+      "grad_norm": 0.1197957865623074,
+      "learning_rate": 4.03972574604624e-05,
+      "loss": 1.215,
+      "step": 9880
+    },
+    {
+      "epoch": 0.7122468103510415,
+      "grad_norm": 0.11977690320463867,
+      "learning_rate": 4.037851216144236e-05,
+      "loss": 1.2229,
+      "step": 9881
+    },
+    {
+      "epoch": 0.7123188928133785,
+      "grad_norm": 0.10952854366378775,
+      "learning_rate": 4.035977011232676e-05,
+      "loss": 1.2687,
+      "step": 9882
+    },
+    {
+      "epoch": 0.7123909752757154,
+      "grad_norm": 0.11512529091408771,
+      "learning_rate": 4.0341031314137304e-05,
+      "loss": 1.2899,
+      "step": 9883
+    },
+    {
+      "epoch": 0.7124630577380523,
+      "grad_norm": 0.12038979621460248,
+      "learning_rate": 4.0322295767895335e-05,
+      "loss": 1.3671,
+      "step": 9884
+    },
+    {
+      "epoch": 0.7125351402003892,
+      "grad_norm": 0.12001522425782381,
+      "learning_rate": 4.030356347462214e-05,
+      "loss": 1.2858,
+      "step": 9885
+    },
+    {
+      "epoch": 0.7126072226627261,
+      "grad_norm": 0.12125590055803352,
+      "learning_rate": 4.028483443533877e-05,
+      "loss": 1.3158,
+      "step": 9886
+    },
+    {
+      "epoch": 0.7126793051250631,
+      "grad_norm": 0.10985523379540882,
+      "learning_rate": 4.0266108651066216e-05,
+      "loss": 1.2231,
+      "step": 9887
+    },
+    {
+      "epoch": 0.7127513875874,
+      "grad_norm": 0.1249653097239802,
+      "learning_rate": 4.0247386122825095e-05,
+      "loss": 1.2291,
+      "step": 9888
+    },
+    {
+      "epoch": 0.7128234700497369,
+      "grad_norm": 0.1150483819621081,
+      "learning_rate": 4.0228666851635985e-05,
+      "loss": 1.1416,
+      "step": 9889
+    },
+    {
+      "epoch": 0.7128955525120738,
+      "grad_norm": 0.11724711536465567,
+      "learning_rate": 4.020995083851927e-05,
+      "loss": 1.2473,
+      "step": 9890
+    },
+    {
+      "epoch": 0.7129676349744107,
+      "grad_norm": 0.1192025261106227,
+      "learning_rate": 4.019123808449514e-05,
+      "loss": 1.3272,
+      "step": 9891
+    },
+    {
+      "epoch": 0.7130397174367477,
+      "grad_norm": 0.11661968571001355,
+      "learning_rate": 4.017252859058359e-05,
+      "loss": 1.3754,
+      "step": 9892
+    },
+    {
+      "epoch": 0.7131117998990846,
+      "grad_norm": 0.11856689342469161,
+      "learning_rate": 4.015382235780446e-05,
+      "loss": 1.2649,
+      "step": 9893
+    },
+    {
+      "epoch": 0.7131838823614215,
+      "grad_norm": 0.11189310860565789,
+      "learning_rate": 4.0135119387177425e-05,
+      "loss": 1.2149,
+      "step": 9894
+    },
+    {
+      "epoch": 0.7132559648237584,
+      "grad_norm": 0.12935491606579336,
+      "learning_rate": 4.011641967972193e-05,
+      "loss": 1.1569,
+      "step": 9895
+    },
+    {
+      "epoch": 0.7133280472860953,
+      "grad_norm": 0.12276726711346952,
+      "learning_rate": 4.009772323645731e-05,
+      "loss": 1.3087,
+      "step": 9896
+    },
+    {
+      "epoch": 0.7134001297484323,
+      "grad_norm": 0.10705784366199357,
+      "learning_rate": 4.007903005840267e-05,
+      "loss": 1.3169,
+      "step": 9897
+    },
+    {
+      "epoch": 0.7134722122107692,
+      "grad_norm": 0.1153826438967347,
+      "learning_rate": 4.006034014657696e-05,
+      "loss": 1.2983,
+      "step": 9898
+    },
+    {
+      "epoch": 0.713544294673106,
+      "grad_norm": 0.1356229305771396,
+      "learning_rate": 4.0041653501998946e-05,
+      "loss": 1.1081,
+      "step": 9899
+    },
+    {
+      "epoch": 0.7136163771354429,
+      "grad_norm": 0.11521262400697668,
+      "learning_rate": 4.002297012568722e-05,
+      "loss": 1.2505,
+      "step": 9900
+    },
+    {
+      "epoch": 0.7136884595977798,
+      "grad_norm": 0.12378024845414123,
+      "learning_rate": 4.0004290018660195e-05,
+      "loss": 1.2564,
+      "step": 9901
+    },
+    {
+      "epoch": 0.7137605420601167,
+      "grad_norm": 0.13162204056060847,
+      "learning_rate": 3.99856131819361e-05,
+      "loss": 1.3255,
+      "step": 9902
+    },
+    {
+      "epoch": 0.7138326245224537,
+      "grad_norm": 0.11381450942760797,
+      "learning_rate": 3.9966939616532994e-05,
+      "loss": 1.171,
+      "step": 9903
+    },
+    {
+      "epoch": 0.7139047069847906,
+      "grad_norm": 0.11508841640874928,
+      "learning_rate": 3.9948269323468744e-05,
+      "loss": 1.1979,
+      "step": 9904
+    },
+    {
+      "epoch": 0.7139767894471275,
+      "grad_norm": 0.11079828207791684,
+      "learning_rate": 3.992960230376105e-05,
+      "loss": 1.2358,
+      "step": 9905
+    },
+    {
+      "epoch": 0.7140488719094644,
+      "grad_norm": 0.1311584255012269,
+      "learning_rate": 3.9910938558427445e-05,
+      "loss": 1.2648,
+      "step": 9906
+    },
+    {
+      "epoch": 0.7141209543718013,
+      "grad_norm": 0.11969575956067796,
+      "learning_rate": 3.989227808848526e-05,
+      "loss": 1.1791,
+      "step": 9907
+    },
+    {
+      "epoch": 0.7141930368341383,
+      "grad_norm": 0.11484211790125005,
+      "learning_rate": 3.987362089495166e-05,
+      "loss": 1.1988,
+      "step": 9908
+    },
+    {
+      "epoch": 0.7142651192964752,
+      "grad_norm": 0.10942215507810746,
+      "learning_rate": 3.985496697884361e-05,
+      "loss": 1.2599,
+      "step": 9909
+    },
+    {
+      "epoch": 0.7143372017588121,
+      "grad_norm": 0.11469339457614154,
+      "learning_rate": 3.9836316341177946e-05,
+      "loss": 1.2205,
+      "step": 9910
+    },
+    {
+      "epoch": 0.714409284221149,
+      "grad_norm": 0.13234805529579707,
+      "learning_rate": 3.981766898297128e-05,
+      "loss": 1.2901,
+      "step": 9911
+    },
+    {
+      "epoch": 0.7144813666834859,
+      "grad_norm": 0.11416139901557405,
+      "learning_rate": 3.97990249052401e-05,
+      "loss": 1.3019,
+      "step": 9912
+    },
+    {
+      "epoch": 0.7145534491458229,
+      "grad_norm": 0.1206878577438669,
+      "learning_rate": 3.9780384109000556e-05,
+      "loss": 1.2696,
+      "step": 9913
+    },
+    {
+      "epoch": 0.7146255316081598,
+      "grad_norm": 0.11950383592997506,
+      "learning_rate": 3.9761746595268865e-05,
+      "loss": 1.2419,
+      "step": 9914
+    },
+    {
+      "epoch": 0.7146976140704967,
+      "grad_norm": 0.12615065268001818,
+      "learning_rate": 3.974311236506089e-05,
+      "loss": 1.2914,
+      "step": 9915
+    },
+    {
+      "epoch": 0.7147696965328335,
+      "grad_norm": 0.10651914170815302,
+      "learning_rate": 3.972448141939239e-05,
+      "loss": 1.2277,
+      "step": 9916
+    },
+    {
+      "epoch": 0.7148417789951704,
+      "grad_norm": 0.11278349026249855,
+      "learning_rate": 3.970585375927883e-05,
+      "loss": 1.2336,
+      "step": 9917
+    },
+    {
+      "epoch": 0.7149138614575073,
+      "grad_norm": 0.1195032737482814,
+      "learning_rate": 3.968722938573568e-05,
+      "loss": 1.2147,
+      "step": 9918
+    },
+    {
+      "epoch": 0.7149859439198443,
+      "grad_norm": 0.11533511933272339,
+      "learning_rate": 3.9668608299778096e-05,
+      "loss": 1.3075,
+      "step": 9919
+    },
+    {
+      "epoch": 0.7150580263821812,
+      "grad_norm": 0.10873229748562227,
+      "learning_rate": 3.9649990502421144e-05,
+      "loss": 1.2648,
+      "step": 9920
+    },
+    {
+      "epoch": 0.7151301088445181,
+      "grad_norm": 0.12042129759194034,
+      "learning_rate": 3.963137599467958e-05,
+      "loss": 1.22,
+      "step": 9921
+    },
+    {
+      "epoch": 0.715202191306855,
+      "grad_norm": 0.1126525125295973,
+      "learning_rate": 3.9612764777568056e-05,
+      "loss": 1.2207,
+      "step": 9922
+    },
+    {
+      "epoch": 0.715274273769192,
+      "grad_norm": 0.1146901783366612,
+      "learning_rate": 3.959415685210114e-05,
+      "loss": 1.3654,
+      "step": 9923
+    },
+    {
+      "epoch": 0.7153463562315289,
+      "grad_norm": 0.12329256148854144,
+      "learning_rate": 3.95755522192931e-05,
+      "loss": 1.355,
+      "step": 9924
+    },
+    {
+      "epoch": 0.7154184386938658,
+      "grad_norm": 0.1128975801270478,
+      "learning_rate": 3.9556950880158004e-05,
+      "loss": 1.3185,
+      "step": 9925
+    },
+    {
+      "epoch": 0.7154905211562027,
+      "grad_norm": 0.11706670734086254,
+      "learning_rate": 3.95383528357098e-05,
+      "loss": 1.3458,
+      "step": 9926
+    },
+    {
+      "epoch": 0.7155626036185396,
+      "grad_norm": 0.1196302340358214,
+      "learning_rate": 3.95197580869623e-05,
+      "loss": 1.2084,
+      "step": 9927
+    },
+    {
+      "epoch": 0.7156346860808765,
+      "grad_norm": 0.2087018757002518,
+      "learning_rate": 3.950116663492909e-05,
+      "loss": 1.2942,
+      "step": 9928
+    },
+    {
+      "epoch": 0.7157067685432135,
+      "grad_norm": 0.12632339290255948,
+      "learning_rate": 3.948257848062351e-05,
+      "loss": 1.2694,
+      "step": 9929
+    },
+    {
+      "epoch": 0.7157788510055504,
+      "grad_norm": 0.11699816984890132,
+      "learning_rate": 3.946399362505876e-05,
+      "loss": 1.2037,
+      "step": 9930
+    },
+    {
+      "epoch": 0.7158509334678873,
+      "grad_norm": 0.11589053872250153,
+      "learning_rate": 3.9445412069247965e-05,
+      "loss": 1.2786,
+      "step": 9931
+    },
+    {
+      "epoch": 0.7159230159302242,
+      "grad_norm": 0.11587614299628114,
+      "learning_rate": 3.942683381420399e-05,
+      "loss": 1.3073,
+      "step": 9932
+    },
+    {
+      "epoch": 0.715995098392561,
+      "grad_norm": 0.11211197973538767,
+      "learning_rate": 3.940825886093944e-05,
+      "loss": 1.2389,
+      "step": 9933
+    },
+    {
+      "epoch": 0.716067180854898,
+      "grad_norm": 0.115439200670348,
+      "learning_rate": 3.9389687210466855e-05,
+      "loss": 1.2285,
+      "step": 9934
+    },
+    {
+      "epoch": 0.7161392633172349,
+      "grad_norm": 0.11803999387905909,
+      "learning_rate": 3.93711188637985e-05,
+      "loss": 1.2592,
+      "step": 9935
+    },
+    {
+      "epoch": 0.7162113457795718,
+      "grad_norm": 0.12275865496868549,
+      "learning_rate": 3.935255382194665e-05,
+      "loss": 1.415,
+      "step": 9936
+    },
+    {
+      "epoch": 0.7162834282419087,
+      "grad_norm": 0.11048899499752797,
+      "learning_rate": 3.9333992085923156e-05,
+      "loss": 1.2772,
+      "step": 9937
+    },
+    {
+      "epoch": 0.7163555107042456,
+      "grad_norm": 0.11884454026724131,
+      "learning_rate": 3.931543365673982e-05,
+      "loss": 1.2186,
+      "step": 9938
+    },
+    {
+      "epoch": 0.7164275931665826,
+      "grad_norm": 0.12465173619380429,
+      "learning_rate": 3.929687853540821e-05,
+      "loss": 1.2865,
+      "step": 9939
+    },
+    {
+      "epoch": 0.7164996756289195,
+      "grad_norm": 0.10839513669767376,
+      "learning_rate": 3.927832672293986e-05,
+      "loss": 1.2356,
+      "step": 9940
+    },
+    {
+      "epoch": 0.7165717580912564,
+      "grad_norm": 0.12570976538309037,
+      "learning_rate": 3.92597782203459e-05,
+      "loss": 1.2875,
+      "step": 9941
+    },
+    {
+      "epoch": 0.7166438405535933,
+      "grad_norm": 0.12196361394513214,
+      "learning_rate": 3.924123302863743e-05,
+      "loss": 1.2733,
+      "step": 9942
+    },
+    {
+      "epoch": 0.7167159230159302,
+      "grad_norm": 0.10897402042263227,
+      "learning_rate": 3.922269114882531e-05,
+      "loss": 1.2519,
+      "step": 9943
+    },
+    {
+      "epoch": 0.7167880054782672,
+      "grad_norm": 0.11308856757677882,
+      "learning_rate": 3.920415258192026e-05,
+      "loss": 1.287,
+      "step": 9944
+    },
+    {
+      "epoch": 0.7168600879406041,
+      "grad_norm": 0.1250755074129482,
+      "learning_rate": 3.918561732893279e-05,
+      "loss": 1.227,
+      "step": 9945
+    },
+    {
+      "epoch": 0.716932170402941,
+      "grad_norm": 0.12253448244297858,
+      "learning_rate": 3.916708539087324e-05,
+      "loss": 1.2859,
+      "step": 9946
+    },
+    {
+      "epoch": 0.7170042528652779,
+      "grad_norm": 0.11777755293940936,
+      "learning_rate": 3.9148556768751754e-05,
+      "loss": 1.2314,
+      "step": 9947
+    },
+    {
+      "epoch": 0.7170763353276148,
+      "grad_norm": 0.12369090203209138,
+      "learning_rate": 3.913003146357832e-05,
+      "loss": 1.3601,
+      "step": 9948
+    },
+    {
+      "epoch": 0.7171484177899518,
+      "grad_norm": 0.12729037766453177,
+      "learning_rate": 3.911150947636273e-05,
+      "loss": 1.3136,
+      "step": 9949
+    },
+    {
+      "epoch": 0.7172205002522887,
+      "grad_norm": 0.12104873926830868,
+      "learning_rate": 3.9092990808114595e-05,
+      "loss": 1.3165,
+      "step": 9950
+    },
+    {
+      "epoch": 0.7172925827146255,
+      "grad_norm": 0.11288018432264978,
+      "learning_rate": 3.907447545984335e-05,
+      "loss": 1.3282,
+      "step": 9951
+    },
+    {
+      "epoch": 0.7173646651769624,
+      "grad_norm": 0.11037041725957898,
+      "learning_rate": 3.9055963432558265e-05,
+      "loss": 1.2802,
+      "step": 9952
+    },
+    {
+      "epoch": 0.7174367476392993,
+      "grad_norm": 0.12564363046873175,
+      "learning_rate": 3.903745472726837e-05,
+      "loss": 1.2933,
+      "step": 9953
+    },
+    {
+      "epoch": 0.7175088301016362,
+      "grad_norm": 0.11138350929192936,
+      "learning_rate": 3.901894934498259e-05,
+      "loss": 1.1955,
+      "step": 9954
+    },
+    {
+      "epoch": 0.7175809125639732,
+      "grad_norm": 0.10899551083069764,
+      "learning_rate": 3.900044728670963e-05,
+      "loss": 1.2436,
+      "step": 9955
+    },
+    {
+      "epoch": 0.7176529950263101,
+      "grad_norm": 0.11183395140873689,
+      "learning_rate": 3.8981948553457994e-05,
+      "loss": 1.2916,
+      "step": 9956
+    },
+    {
+      "epoch": 0.717725077488647,
+      "grad_norm": 0.11937744065065878,
+      "learning_rate": 3.896345314623605e-05,
+      "loss": 1.2779,
+      "step": 9957
+    },
+    {
+      "epoch": 0.7177971599509839,
+      "grad_norm": 0.12365763645512076,
+      "learning_rate": 3.894496106605197e-05,
+      "loss": 1.3286,
+      "step": 9958
+    },
+    {
+      "epoch": 0.7178692424133208,
+      "grad_norm": 0.12049752106125544,
+      "learning_rate": 3.892647231391372e-05,
+      "loss": 1.2995,
+      "step": 9959
+    },
+    {
+      "epoch": 0.7179413248756578,
+      "grad_norm": 0.30072778371294984,
+      "learning_rate": 3.890798689082909e-05,
+      "loss": 1.2227,
+      "step": 9960
+    },
+    {
+      "epoch": 0.7180134073379947,
+      "grad_norm": 0.11538897010045543,
+      "learning_rate": 3.888950479780574e-05,
+      "loss": 1.2471,
+      "step": 9961
+    },
+    {
+      "epoch": 0.7180854898003316,
+      "grad_norm": 0.12009371569090353,
+      "learning_rate": 3.887102603585109e-05,
+      "loss": 1.1948,
+      "step": 9962
+    },
+    {
+      "epoch": 0.7181575722626685,
+      "grad_norm": 0.11139050256362312,
+      "learning_rate": 3.8852550605972396e-05,
+      "loss": 1.311,
+      "step": 9963
+    },
+    {
+      "epoch": 0.7182296547250054,
+      "grad_norm": 0.11625300721253155,
+      "learning_rate": 3.883407850917672e-05,
+      "loss": 1.3469,
+      "step": 9964
+    },
+    {
+      "epoch": 0.7183017371873424,
+      "grad_norm": 0.11515367855057795,
+      "learning_rate": 3.881560974647101e-05,
+      "loss": 1.306,
+      "step": 9965
+    },
+    {
+      "epoch": 0.7183738196496793,
+      "grad_norm": 0.12581248513652607,
+      "learning_rate": 3.8797144318861875e-05,
+      "loss": 1.3022,
+      "step": 9966
+    },
+    {
+      "epoch": 0.7184459021120162,
+      "grad_norm": 0.12415388563188345,
+      "learning_rate": 3.8778682227355955e-05,
+      "loss": 1.3047,
+      "step": 9967
+    },
+    {
+      "epoch": 0.718517984574353,
+      "grad_norm": 0.12151087004404326,
+      "learning_rate": 3.876022347295955e-05,
+      "loss": 1.3256,
+      "step": 9968
+    },
+    {
+      "epoch": 0.7185900670366899,
+      "grad_norm": 0.11590121388300509,
+      "learning_rate": 3.874176805667886e-05,
+      "loss": 1.2584,
+      "step": 9969
+    },
+    {
+      "epoch": 0.7186621494990268,
+      "grad_norm": 0.11981678831697656,
+      "learning_rate": 3.872331597951977e-05,
+      "loss": 1.309,
+      "step": 9970
+    },
+    {
+      "epoch": 0.7187342319613638,
+      "grad_norm": 0.11890802412768124,
+      "learning_rate": 3.87048672424882e-05,
+      "loss": 1.297,
+      "step": 9971
+    },
+    {
+      "epoch": 0.7188063144237007,
+      "grad_norm": 0.23161352595907356,
+      "learning_rate": 3.8686421846589726e-05,
+      "loss": 1.2218,
+      "step": 9972
+    },
+    {
+      "epoch": 0.7188783968860376,
+      "grad_norm": 0.13626796078553102,
+      "learning_rate": 3.8667979792829824e-05,
+      "loss": 1.2573,
+      "step": 9973
+    },
+    {
+      "epoch": 0.7189504793483745,
+      "grad_norm": 0.10724719702531017,
+      "learning_rate": 3.864954108221367e-05,
+      "loss": 1.2448,
+      "step": 9974
+    },
+    {
+      "epoch": 0.7190225618107114,
+      "grad_norm": 0.12652481273207186,
+      "learning_rate": 3.863110571574636e-05,
+      "loss": 1.288,
+      "step": 9975
+    },
+    {
+      "epoch": 0.7190946442730484,
+      "grad_norm": 0.11061228540094983,
+      "learning_rate": 3.861267369443288e-05,
+      "loss": 1.3241,
+      "step": 9976
+    },
+    {
+      "epoch": 0.7191667267353853,
+      "grad_norm": 0.1103778471191747,
+      "learning_rate": 3.859424501927783e-05,
+      "loss": 1.238,
+      "step": 9977
+    },
+    {
+      "epoch": 0.7192388091977222,
+      "grad_norm": 0.12071182048856985,
+      "learning_rate": 3.8575819691285783e-05,
+      "loss": 1.2603,
+      "step": 9978
+    },
+    {
+      "epoch": 0.7193108916600591,
+      "grad_norm": 0.12172423030869255,
+      "learning_rate": 3.855739771146104e-05,
+      "loss": 1.1978,
+      "step": 9979
+    },
+    {
+      "epoch": 0.719382974122396,
+      "grad_norm": 0.11561489453544553,
+      "learning_rate": 3.853897908080788e-05,
+      "loss": 1.2213,
+      "step": 9980
+    },
+    {
+      "epoch": 0.719455056584733,
+      "grad_norm": 0.11370601812702054,
+      "learning_rate": 3.8520563800330157e-05,
+      "loss": 1.2985,
+      "step": 9981
+    },
+    {
+      "epoch": 0.7195271390470699,
+      "grad_norm": 0.10777948404723704,
+      "learning_rate": 3.850215187103173e-05,
+      "loss": 1.2326,
+      "step": 9982
+    },
+    {
+      "epoch": 0.7195992215094068,
+      "grad_norm": 0.12590712868121712,
+      "learning_rate": 3.848374329391617e-05,
+      "loss": 1.336,
+      "step": 9983
+    },
+    {
+      "epoch": 0.7196713039717437,
+      "grad_norm": 0.12287134568409655,
+      "learning_rate": 3.8465338069987e-05,
+      "loss": 1.2968,
+      "step": 9984
+    },
+    {
+      "epoch": 0.7197433864340805,
+      "grad_norm": 0.11503620042469245,
+      "learning_rate": 3.8446936200247375e-05,
+      "loss": 1.329,
+      "step": 9985
+    },
+    {
+      "epoch": 0.7198154688964175,
+      "grad_norm": 0.12391034944435367,
+      "learning_rate": 3.8428537685700406e-05,
+      "loss": 1.1414,
+      "step": 9986
+    },
+    {
+      "epoch": 0.7198875513587544,
+      "grad_norm": 0.11531269407969111,
+      "learning_rate": 3.841014252734896e-05,
+      "loss": 1.3088,
+      "step": 9987
+    },
+    {
+      "epoch": 0.7199596338210913,
+      "grad_norm": 0.11503121584718809,
+      "learning_rate": 3.839175072619574e-05,
+      "loss": 1.325,
+      "step": 9988
+    },
+    {
+      "epoch": 0.7200317162834282,
+      "grad_norm": 0.10605462785211804,
+      "learning_rate": 3.8373362283243274e-05,
+      "loss": 1.2725,
+      "step": 9989
+    },
+    {
+      "epoch": 0.7201037987457651,
+      "grad_norm": 0.1463625245543884,
+      "learning_rate": 3.835497719949389e-05,
+      "loss": 1.2916,
+      "step": 9990
+    },
+    {
+      "epoch": 0.720175881208102,
+      "grad_norm": 0.11987544277158382,
+      "learning_rate": 3.8336595475949745e-05,
+      "loss": 1.3161,
+      "step": 9991
+    },
+    {
+      "epoch": 0.720247963670439,
+      "grad_norm": 0.4581381847142387,
+      "learning_rate": 3.831821711361281e-05,
+      "loss": 1.1981,
+      "step": 9992
+    },
+    {
+      "epoch": 0.7203200461327759,
+      "grad_norm": 0.1203771168120353,
+      "learning_rate": 3.8299842113484855e-05,
+      "loss": 1.2439,
+      "step": 9993
+    },
+    {
+      "epoch": 0.7203921285951128,
+      "grad_norm": 0.12008791981231419,
+      "learning_rate": 3.828147047656748e-05,
+      "loss": 1.2633,
+      "step": 9994
+    },
+    {
+      "epoch": 0.7204642110574497,
+      "grad_norm": 0.1153095283275913,
+      "learning_rate": 3.826310220386212e-05,
+      "loss": 1.3248,
+      "step": 9995
+    },
+    {
+      "epoch": 0.7205362935197867,
+      "grad_norm": 0.12186889112564615,
+      "learning_rate": 3.8244737296370016e-05,
+      "loss": 1.341,
+      "step": 9996
+    },
+    {
+      "epoch": 0.7206083759821236,
+      "grad_norm": 0.1394448296511925,
+      "learning_rate": 3.822637575509219e-05,
+      "loss": 1.2611,
+      "step": 9997
+    },
+    {
+      "epoch": 0.7206804584444605,
+      "grad_norm": 0.12208920090319463,
+      "learning_rate": 3.8208017581029545e-05,
+      "loss": 1.2909,
+      "step": 9998
+    },
+    {
+      "epoch": 0.7207525409067974,
+      "grad_norm": 0.1184659242621817,
+      "learning_rate": 3.8189662775182734e-05,
+      "loss": 1.2386,
+      "step": 9999
+    },
+    {
+      "epoch": 0.7208246233691343,
+      "grad_norm": 0.11301846448387365,
+      "learning_rate": 3.817131133855229e-05,
+      "loss": 1.226,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7208967058314713,
+      "grad_norm": 0.10316409499042724,
+      "learning_rate": 3.81529632721385e-05,
+      "loss": 1.2648,
+      "step": 10001
+    },
+    {
+      "epoch": 0.7209687882938082,
+      "grad_norm": 0.11738144501488268,
+      "learning_rate": 3.8134618576941505e-05,
+      "loss": 1.2212,
+      "step": 10002
+    },
+    {
+      "epoch": 0.721040870756145,
+      "grad_norm": 0.11255204510157432,
+      "learning_rate": 3.811627725396128e-05,
+      "loss": 1.2031,
+      "step": 10003
+    },
+    {
+      "epoch": 0.7211129532184819,
+      "grad_norm": 0.12064853417550256,
+      "learning_rate": 3.809793930419756e-05,
+      "loss": 1.304,
+      "step": 10004
+    },
+    {
+      "epoch": 0.7211850356808188,
+      "grad_norm": 0.11573899026360206,
+      "learning_rate": 3.807960472864994e-05,
+      "loss": 1.233,
+      "step": 10005
+    },
+    {
+      "epoch": 0.7212571181431557,
+      "grad_norm": 0.1066848403250684,
+      "learning_rate": 3.806127352831782e-05,
+      "loss": 1.3201,
+      "step": 10006
+    },
+    {
+      "epoch": 0.7213292006054927,
+      "grad_norm": 0.12592572401304655,
+      "learning_rate": 3.804294570420041e-05,
+      "loss": 1.2479,
+      "step": 10007
+    },
+    {
+      "epoch": 0.7214012830678296,
+      "grad_norm": 0.11828375880682254,
+      "learning_rate": 3.8024621257296747e-05,
+      "loss": 1.3239,
+      "step": 10008
+    },
+    {
+      "epoch": 0.7214733655301665,
+      "grad_norm": 0.11690770907423967,
+      "learning_rate": 3.8006300188605705e-05,
+      "loss": 1.3136,
+      "step": 10009
+    },
+    {
+      "epoch": 0.7215454479925034,
+      "grad_norm": 0.12774051144196677,
+      "learning_rate": 3.798798249912584e-05,
+      "loss": 1.2629,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7216175304548403,
+      "grad_norm": 0.20615093858308467,
+      "learning_rate": 3.796966818985574e-05,
+      "loss": 1.3008,
+      "step": 10011
+    },
+    {
+      "epoch": 0.7216896129171773,
+      "grad_norm": 0.10255402836531625,
+      "learning_rate": 3.795135726179367e-05,
+      "loss": 1.2191,
+      "step": 10012
+    },
+    {
+      "epoch": 0.7217616953795142,
+      "grad_norm": 0.1152674248977096,
+      "learning_rate": 3.793304971593777e-05,
+      "loss": 1.2965,
+      "step": 10013
+    },
+    {
+      "epoch": 0.7218337778418511,
+      "grad_norm": 0.12277410051141674,
+      "learning_rate": 3.7914745553285846e-05,
+      "loss": 1.2525,
+      "step": 10014
+    },
+    {
+      "epoch": 0.721905860304188,
+      "grad_norm": 0.1306410709232309,
+      "learning_rate": 3.789644477483576e-05,
+      "loss": 1.178,
+      "step": 10015
+    },
+    {
+      "epoch": 0.7219779427665249,
+      "grad_norm": 0.1296259700564312,
+      "learning_rate": 3.787814738158504e-05,
+      "loss": 1.2474,
+      "step": 10016
+    },
+    {
+      "epoch": 0.7220500252288619,
+      "grad_norm": 0.11206728369790696,
+      "learning_rate": 3.7859853374531065e-05,
+      "loss": 1.3683,
+      "step": 10017
+    },
+    {
+      "epoch": 0.7221221076911988,
+      "grad_norm": 0.1179861400021065,
+      "learning_rate": 3.784156275467097e-05,
+      "loss": 1.3187,
+      "step": 10018
+    },
+    {
+      "epoch": 0.7221941901535357,
+      "grad_norm": 0.12543809576019818,
+      "learning_rate": 3.782327552300175e-05,
+      "loss": 1.3295,
+      "step": 10019
+    },
+    {
+      "epoch": 0.7222662726158725,
+      "grad_norm": 0.11861988952091375,
+      "learning_rate": 3.780499168052031e-05,
+      "loss": 1.3388,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7223383550782094,
+      "grad_norm": 0.11021512209146395,
+      "learning_rate": 3.778671122822327e-05,
+      "loss": 1.2614,
+      "step": 10021
+    },
+    {
+      "epoch": 0.7224104375405463,
+      "grad_norm": 0.12608485164157346,
+      "learning_rate": 3.7768434167107006e-05,
+      "loss": 1.2349,
+      "step": 10022
+    },
+    {
+      "epoch": 0.7224825200028833,
+      "grad_norm": 0.12145601214197514,
+      "learning_rate": 3.775016049816778e-05,
+      "loss": 1.3313,
+      "step": 10023
+    },
+    {
+      "epoch": 0.7225546024652202,
+      "grad_norm": 0.11843193990023698,
+      "learning_rate": 3.7731890222401744e-05,
+      "loss": 1.1866,
+      "step": 10024
+    },
+    {
+      "epoch": 0.7226266849275571,
+      "grad_norm": 0.1246696249338573,
+      "learning_rate": 3.7713623340804805e-05,
+      "loss": 1.2752,
+      "step": 10025
+    },
+    {
+      "epoch": 0.722698767389894,
+      "grad_norm": 0.1141621065209545,
+      "learning_rate": 3.769535985437257e-05,
+      "loss": 1.2696,
+      "step": 10026
+    },
+    {
+      "epoch": 0.722770849852231,
+      "grad_norm": 0.12398481852124754,
+      "learning_rate": 3.76770997641006e-05,
+      "loss": 1.247,
+      "step": 10027
+    },
+    {
+      "epoch": 0.7228429323145679,
+      "grad_norm": 0.135410631128964,
+      "learning_rate": 3.765884307098427e-05,
+      "loss": 1.1651,
+      "step": 10028
+    },
+    {
+      "epoch": 0.7229150147769048,
+      "grad_norm": 0.13695952367600625,
+      "learning_rate": 3.764058977601875e-05,
+      "loss": 1.2051,
+      "step": 10029
+    },
+    {
+      "epoch": 0.7229870972392417,
+      "grad_norm": 0.12856537455526534,
+      "learning_rate": 3.762233988019894e-05,
+      "loss": 1.2969,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7230591797015786,
+      "grad_norm": 0.10706369927464907,
+      "learning_rate": 3.7604093384519655e-05,
+      "loss": 1.2467,
+      "step": 10031
+    },
+    {
+      "epoch": 0.7231312621639155,
+      "grad_norm": 0.11515158695448112,
+      "learning_rate": 3.758585028997545e-05,
+      "loss": 1.2644,
+      "step": 10032
+    },
+    {
+      "epoch": 0.7232033446262525,
+      "grad_norm": 0.11433413671708653,
+      "learning_rate": 3.7567610597560856e-05,
+      "loss": 1.2309,
+      "step": 10033
+    },
+    {
+      "epoch": 0.7232754270885894,
+      "grad_norm": 0.11322489494778105,
+      "learning_rate": 3.754937430826998e-05,
+      "loss": 1.2619,
+      "step": 10034
+    },
+    {
+      "epoch": 0.7233475095509263,
+      "grad_norm": 0.1252082708473652,
+      "learning_rate": 3.75311414230969e-05,
+      "loss": 1.2861,
+      "step": 10035
+    },
+    {
+      "epoch": 0.7234195920132632,
+      "grad_norm": 0.1205308705009557,
+      "learning_rate": 3.7512911943035446e-05,
+      "loss": 1.3061,
+      "step": 10036
+    },
+    {
+      "epoch": 0.7234916744756,
+      "grad_norm": 0.14061383350954615,
+      "learning_rate": 3.749468586907937e-05,
+      "loss": 1.2832,
+      "step": 10037
+    },
+    {
+      "epoch": 0.723563756937937,
+      "grad_norm": 0.11491934663030388,
+      "learning_rate": 3.747646320222209e-05,
+      "loss": 1.3468,
+      "step": 10038
+    },
+    {
+      "epoch": 0.7236358394002739,
+      "grad_norm": 0.11688834351953062,
+      "learning_rate": 3.7458243943456905e-05,
+      "loss": 1.3028,
+      "step": 10039
+    },
+    {
+      "epoch": 0.7237079218626108,
+      "grad_norm": 0.12112485699333252,
+      "learning_rate": 3.744002809377693e-05,
+      "loss": 1.2761,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7237800043249477,
+      "grad_norm": 0.13279357685479126,
+      "learning_rate": 3.742181565417512e-05,
+      "loss": 1.3268,
+      "step": 10041
+    },
+    {
+      "epoch": 0.7238520867872846,
+      "grad_norm": 0.11050450302066542,
+      "learning_rate": 3.7403606625644195e-05,
+      "loss": 1.2632,
+      "step": 10042
+    },
+    {
+      "epoch": 0.7239241692496216,
+      "grad_norm": 0.1287718711500916,
+      "learning_rate": 3.738540100917671e-05,
+      "loss": 1.2615,
+      "step": 10043
+    },
+    {
+      "epoch": 0.7239962517119585,
+      "grad_norm": 0.11915186072540442,
+      "learning_rate": 3.736719880576503e-05,
+      "loss": 1.2774,
+      "step": 10044
+    },
+    {
+      "epoch": 0.7240683341742954,
+      "grad_norm": 0.1344857368990193,
+      "learning_rate": 3.734900001640135e-05,
+      "loss": 1.2835,
+      "step": 10045
+    },
+    {
+      "epoch": 0.7241404166366323,
+      "grad_norm": 0.14978982356626458,
+      "learning_rate": 3.7330804642077655e-05,
+      "loss": 1.2683,
+      "step": 10046
+    },
+    {
+      "epoch": 0.7242124990989692,
+      "grad_norm": 0.11395953845254461,
+      "learning_rate": 3.731261268378576e-05,
+      "loss": 1.3665,
+      "step": 10047
+    },
+    {
+      "epoch": 0.7242845815613062,
+      "grad_norm": 0.12902379987606832,
+      "learning_rate": 3.7294424142517304e-05,
+      "loss": 1.3108,
+      "step": 10048
+    },
+    {
+      "epoch": 0.7243566640236431,
+      "grad_norm": 0.13330036514595467,
+      "learning_rate": 3.7276239019263705e-05,
+      "loss": 1.212,
+      "step": 10049
+    },
+    {
+      "epoch": 0.72442874648598,
+      "grad_norm": 0.12205532935370388,
+      "learning_rate": 3.7258057315016226e-05,
+      "loss": 1.2208,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7245008289483169,
+      "grad_norm": 0.11663755610760486,
+      "learning_rate": 3.723987903076592e-05,
+      "loss": 1.2429,
+      "step": 10051
+    },
+    {
+      "epoch": 0.7245729114106538,
+      "grad_norm": 0.13866813828194335,
+      "learning_rate": 3.722170416750368e-05,
+      "loss": 1.1495,
+      "step": 10052
+    },
+    {
+      "epoch": 0.7246449938729908,
+      "grad_norm": 0.11549687825833513,
+      "learning_rate": 3.72035327262202e-05,
+      "loss": 1.277,
+      "step": 10053
+    },
+    {
+      "epoch": 0.7247170763353277,
+      "grad_norm": 0.11318162224026385,
+      "learning_rate": 3.718536470790597e-05,
+      "loss": 1.2618,
+      "step": 10054
+    },
+    {
+      "epoch": 0.7247891587976645,
+      "grad_norm": 0.11415523922170685,
+      "learning_rate": 3.7167200113551325e-05,
+      "loss": 1.2661,
+      "step": 10055
+    },
+    {
+      "epoch": 0.7248612412600014,
+      "grad_norm": 0.11044886072916553,
+      "learning_rate": 3.714903894414638e-05,
+      "loss": 1.2567,
+      "step": 10056
+    },
+    {
+      "epoch": 0.7249333237223383,
+      "grad_norm": 0.11571595898651968,
+      "learning_rate": 3.7130881200681115e-05,
+      "loss": 1.3418,
+      "step": 10057
+    },
+    {
+      "epoch": 0.7250054061846752,
+      "grad_norm": 0.11373328121360649,
+      "learning_rate": 3.7112726884145254e-05,
+      "loss": 1.1737,
+      "step": 10058
+    },
+    {
+      "epoch": 0.7250774886470122,
+      "grad_norm": 0.11923792684341439,
+      "learning_rate": 3.709457599552838e-05,
+      "loss": 1.2666,
+      "step": 10059
+    },
+    {
+      "epoch": 0.7251495711093491,
+      "grad_norm": 0.1197093997088651,
+      "learning_rate": 3.7076428535819896e-05,
+      "loss": 1.2197,
+      "step": 10060
+    },
+    {
+      "epoch": 0.725221653571686,
+      "grad_norm": 0.12285775988673493,
+      "learning_rate": 3.705828450600898e-05,
+      "loss": 1.3098,
+      "step": 10061
+    },
+    {
+      "epoch": 0.7252937360340229,
+      "grad_norm": 0.1129354956574019,
+      "learning_rate": 3.704014390708468e-05,
+      "loss": 1.2473,
+      "step": 10062
+    },
+    {
+      "epoch": 0.7253658184963598,
+      "grad_norm": 0.11796125219733031,
+      "learning_rate": 3.7022006740035734e-05,
+      "loss": 1.3139,
+      "step": 10063
+    },
+    {
+      "epoch": 0.7254379009586968,
+      "grad_norm": 0.11794831356832233,
+      "learning_rate": 3.7003873005850856e-05,
+      "loss": 1.1908,
+      "step": 10064
+    },
+    {
+      "epoch": 0.7255099834210337,
+      "grad_norm": 0.109314697197058,
+      "learning_rate": 3.698574270551849e-05,
+      "loss": 1.2419,
+      "step": 10065
+    },
+    {
+      "epoch": 0.7255820658833706,
+      "grad_norm": 0.1187766378437471,
+      "learning_rate": 3.6967615840026924e-05,
+      "loss": 1.2863,
+      "step": 10066
+    },
+    {
+      "epoch": 0.7256541483457075,
+      "grad_norm": 0.11894094367628542,
+      "learning_rate": 3.6949492410364126e-05,
+      "loss": 1.2784,
+      "step": 10067
+    },
+    {
+      "epoch": 0.7257262308080444,
+      "grad_norm": 0.11427986762925108,
+      "learning_rate": 3.69313724175181e-05,
+      "loss": 1.3369,
+      "step": 10068
+    },
+    {
+      "epoch": 0.7257983132703814,
+      "grad_norm": 0.115279931825956,
+      "learning_rate": 3.6913255862476494e-05,
+      "loss": 1.1536,
+      "step": 10069
+    },
+    {
+      "epoch": 0.7258703957327183,
+      "grad_norm": 0.11640392414783222,
+      "learning_rate": 3.6895142746226865e-05,
+      "loss": 1.3667,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7259424781950552,
+      "grad_norm": 0.12034809004226489,
+      "learning_rate": 3.687703306975645e-05,
+      "loss": 1.2023,
+      "step": 10071
+    },
+    {
+      "epoch": 0.726014560657392,
+      "grad_norm": 0.11046564733310704,
+      "learning_rate": 3.685892683405248e-05,
+      "loss": 1.3093,
+      "step": 10072
+    },
+    {
+      "epoch": 0.7260866431197289,
+      "grad_norm": 0.12754710080970827,
+      "learning_rate": 3.684082404010186e-05,
+      "loss": 1.3727,
+      "step": 10073
+    },
+    {
+      "epoch": 0.7261587255820658,
+      "grad_norm": 0.11097331497274861,
+      "learning_rate": 3.6822724688891416e-05,
+      "loss": 1.2594,
+      "step": 10074
+    },
+    {
+      "epoch": 0.7262308080444028,
+      "grad_norm": 0.1329454462383232,
+      "learning_rate": 3.6804628781407647e-05,
+      "loss": 1.2573,
+      "step": 10075
+    },
+    {
+      "epoch": 0.7263028905067397,
+      "grad_norm": 0.12208164248672172,
+      "learning_rate": 3.678653631863693e-05,
+      "loss": 1.2658,
+      "step": 10076
+    },
+    {
+      "epoch": 0.7263749729690766,
+      "grad_norm": 0.13123549427872164,
+      "learning_rate": 3.676844730156554e-05,
+      "loss": 1.3052,
+      "step": 10077
+    },
+    {
+      "epoch": 0.7264470554314135,
+      "grad_norm": 0.11990774868974292,
+      "learning_rate": 3.6750361731179494e-05,
+      "loss": 1.2981,
+      "step": 10078
+    },
+    {
+      "epoch": 0.7265191378937504,
+      "grad_norm": 0.12983396681467996,
+      "learning_rate": 3.673227960846455e-05,
+      "loss": 1.2593,
+      "step": 10079
+    },
+    {
+      "epoch": 0.7265912203560874,
+      "grad_norm": 0.12222009882485582,
+      "learning_rate": 3.671420093440634e-05,
+      "loss": 1.2778,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7266633028184243,
+      "grad_norm": 0.12111613908368318,
+      "learning_rate": 3.669612570999038e-05,
+      "loss": 1.2448,
+      "step": 10081
+    },
+    {
+      "epoch": 0.7267353852807612,
+      "grad_norm": 0.12319337742038852,
+      "learning_rate": 3.667805393620194e-05,
+      "loss": 1.3551,
+      "step": 10082
+    },
+    {
+      "epoch": 0.7268074677430981,
+      "grad_norm": 0.12128526587557896,
+      "learning_rate": 3.665998561402602e-05,
+      "loss": 1.3285,
+      "step": 10083
+    },
+    {
+      "epoch": 0.726879550205435,
+      "grad_norm": 0.11288301995835087,
+      "learning_rate": 3.6641920744447534e-05,
+      "loss": 1.2657,
+      "step": 10084
+    },
+    {
+      "epoch": 0.726951632667772,
+      "grad_norm": 0.12901736530182636,
+      "learning_rate": 3.662385932845116e-05,
+      "loss": 1.249,
+      "step": 10085
+    },
+    {
+      "epoch": 0.7270237151301089,
+      "grad_norm": 0.13025772159376933,
+      "learning_rate": 3.6605801367021494e-05,
+      "loss": 1.2495,
+      "step": 10086
+    },
+    {
+      "epoch": 0.7270957975924458,
+      "grad_norm": 0.13198985851073497,
+      "learning_rate": 3.658774686114277e-05,
+      "loss": 1.3405,
+      "step": 10087
+    },
+    {
+      "epoch": 0.7271678800547827,
+      "grad_norm": 0.11969729659039545,
+      "learning_rate": 3.656969581179913e-05,
+      "loss": 1.1988,
+      "step": 10088
+    },
+    {
+      "epoch": 0.7272399625171195,
+      "grad_norm": 0.12979288951293758,
+      "learning_rate": 3.655164821997451e-05,
+      "loss": 1.2077,
+      "step": 10089
+    },
+    {
+      "epoch": 0.7273120449794565,
+      "grad_norm": 0.13456175577304597,
+      "learning_rate": 3.653360408665275e-05,
+      "loss": 1.2167,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7273841274417934,
+      "grad_norm": 0.129505966819648,
+      "learning_rate": 3.651556341281732e-05,
+      "loss": 1.2214,
+      "step": 10091
+    },
+    {
+      "epoch": 0.7274562099041303,
+      "grad_norm": 0.10612328445517576,
+      "learning_rate": 3.6497526199451626e-05,
+      "loss": 1.23,
+      "step": 10092
+    },
+    {
+      "epoch": 0.7275282923664672,
+      "grad_norm": 0.11932745828187447,
+      "learning_rate": 3.647949244753884e-05,
+      "loss": 1.19,
+      "step": 10093
+    },
+    {
+      "epoch": 0.7276003748288041,
+      "grad_norm": 0.11652843175682785,
+      "learning_rate": 3.6461462158062046e-05,
+      "loss": 1.2957,
+      "step": 10094
+    },
+    {
+      "epoch": 0.727672457291141,
+      "grad_norm": 0.10875674150492302,
+      "learning_rate": 3.644343533200397e-05,
+      "loss": 1.2839,
+      "step": 10095
+    },
+    {
+      "epoch": 0.727744539753478,
+      "grad_norm": 0.11342580620846661,
+      "learning_rate": 3.642541197034726e-05,
+      "loss": 1.1886,
+      "step": 10096
+    },
+    {
+      "epoch": 0.7278166222158149,
+      "grad_norm": 0.12393392276437036,
+      "learning_rate": 3.6407392074074344e-05,
+      "loss": 1.2741,
+      "step": 10097
+    },
+    {
+      "epoch": 0.7278887046781518,
+      "grad_norm": 0.114108062415651,
+      "learning_rate": 3.638937564416748e-05,
+      "loss": 1.3203,
+      "step": 10098
+    },
+    {
+      "epoch": 0.7279607871404887,
+      "grad_norm": 0.11309163589624574,
+      "learning_rate": 3.6371362681608725e-05,
+      "loss": 1.2794,
+      "step": 10099
+    },
+    {
+      "epoch": 0.7280328696028256,
+      "grad_norm": 0.1261034921456826,
+      "learning_rate": 3.635335318737993e-05,
+      "loss": 1.3094,
+      "step": 10100
+    },
+    {
+      "epoch": 0.7281049520651626,
+      "grad_norm": 0.11665976151026859,
+      "learning_rate": 3.633534716246281e-05,
+      "loss": 1.2819,
+      "step": 10101
+    },
+    {
+      "epoch": 0.7281770345274995,
+      "grad_norm": 0.13160947558713534,
+      "learning_rate": 3.631734460783881e-05,
+      "loss": 1.2444,
+      "step": 10102
+    },
+    {
+      "epoch": 0.7282491169898364,
+      "grad_norm": 0.12478200322813748,
+      "learning_rate": 3.629934552448925e-05,
+      "loss": 1.2909,
+      "step": 10103
+    },
+    {
+      "epoch": 0.7283211994521733,
+      "grad_norm": 0.11264532979819253,
+      "learning_rate": 3.628134991339524e-05,
+      "loss": 1.1825,
+      "step": 10104
+    },
+    {
+      "epoch": 0.7283932819145102,
+      "grad_norm": 0.11396267817176904,
+      "learning_rate": 3.626335777553772e-05,
+      "loss": 1.2907,
+      "step": 10105
+    },
+    {
+      "epoch": 0.7284653643768471,
+      "grad_norm": 0.11872119584999846,
+      "learning_rate": 3.624536911189739e-05,
+      "loss": 1.2126,
+      "step": 10106
+    },
+    {
+      "epoch": 0.728537446839184,
+      "grad_norm": 0.12181660439104253,
+      "learning_rate": 3.622738392345482e-05,
+      "loss": 1.2938,
+      "step": 10107
+    },
+    {
+      "epoch": 0.7286095293015209,
+      "grad_norm": 0.11465487893629804,
+      "learning_rate": 3.6209402211190337e-05,
+      "loss": 1.2449,
+      "step": 10108
+    },
+    {
+      "epoch": 0.7286816117638578,
+      "grad_norm": 0.12302185319104286,
+      "learning_rate": 3.6191423976084126e-05,
+      "loss": 1.3231,
+      "step": 10109
+    },
+    {
+      "epoch": 0.7287536942261947,
+      "grad_norm": 0.11692131252168204,
+      "learning_rate": 3.6173449219116164e-05,
+      "loss": 1.4177,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7288257766885317,
+      "grad_norm": 0.12208670443602491,
+      "learning_rate": 3.615547794126622e-05,
+      "loss": 1.2577,
+      "step": 10111
+    },
+    {
+      "epoch": 0.7288978591508686,
+      "grad_norm": 0.12362136609456721,
+      "learning_rate": 3.6137510143513895e-05,
+      "loss": 1.2487,
+      "step": 10112
+    },
+    {
+      "epoch": 0.7289699416132055,
+      "grad_norm": 0.12005505280257378,
+      "learning_rate": 3.6119545826838606e-05,
+      "loss": 1.2041,
+      "step": 10113
+    },
+    {
+      "epoch": 0.7290420240755424,
+      "grad_norm": 0.11419803362552602,
+      "learning_rate": 3.610158499221956e-05,
+      "loss": 1.2785,
+      "step": 10114
+    },
+    {
+      "epoch": 0.7291141065378793,
+      "grad_norm": 0.11512942031782608,
+      "learning_rate": 3.6083627640635773e-05,
+      "loss": 1.2312,
+      "step": 10115
+    },
+    {
+      "epoch": 0.7291861890002163,
+      "grad_norm": 0.1135310679986204,
+      "learning_rate": 3.606567377306611e-05,
+      "loss": 1.2834,
+      "step": 10116
+    },
+    {
+      "epoch": 0.7292582714625532,
+      "grad_norm": 0.15248802160016503,
+      "learning_rate": 3.604772339048919e-05,
+      "loss": 1.2215,
+      "step": 10117
+    },
+    {
+      "epoch": 0.7293303539248901,
+      "grad_norm": 0.13056620515789874,
+      "learning_rate": 3.602977649388347e-05,
+      "loss": 1.2778,
+      "step": 10118
+    },
+    {
+      "epoch": 0.729402436387227,
+      "grad_norm": 0.11995892903934445,
+      "learning_rate": 3.6011833084227266e-05,
+      "loss": 1.2589,
+      "step": 10119
+    },
+    {
+      "epoch": 0.7294745188495639,
+      "grad_norm": 0.11539582310705057,
+      "learning_rate": 3.599389316249855e-05,
+      "loss": 1.2656,
+      "step": 10120
+    },
+    {
+      "epoch": 0.7295466013119009,
+      "grad_norm": 0.12378602524597138,
+      "learning_rate": 3.5975956729675294e-05,
+      "loss": 1.3062,
+      "step": 10121
+    },
+    {
+      "epoch": 0.7296186837742378,
+      "grad_norm": 0.1388288816148396,
+      "learning_rate": 3.5958023786735186e-05,
+      "loss": 1.219,
+      "step": 10122
+    },
+    {
+      "epoch": 0.7296907662365747,
+      "grad_norm": 0.1145514880277898,
+      "learning_rate": 3.594009433465575e-05,
+      "loss": 1.356,
+      "step": 10123
+    },
+    {
+      "epoch": 0.7297628486989115,
+      "grad_norm": 0.11994166993286905,
+      "learning_rate": 3.592216837441419e-05,
+      "loss": 1.2291,
+      "step": 10124
+    },
+    {
+      "epoch": 0.7298349311612484,
+      "grad_norm": 0.1280877122114852,
+      "learning_rate": 3.590424590698775e-05,
+      "loss": 1.2532,
+      "step": 10125
+    },
+    {
+      "epoch": 0.7299070136235853,
+      "grad_norm": 0.11684370601020012,
+      "learning_rate": 3.588632693335332e-05,
+      "loss": 1.2863,
+      "step": 10126
+    },
+    {
+      "epoch": 0.7299790960859223,
+      "grad_norm": 0.12954307211488864,
+      "learning_rate": 3.586841145448769e-05,
+      "loss": 1.4051,
+      "step": 10127
+    },
+    {
+      "epoch": 0.7300511785482592,
+      "grad_norm": 0.1107758537282004,
+      "learning_rate": 3.585049947136734e-05,
+      "loss": 1.3061,
+      "step": 10128
+    },
+    {
+      "epoch": 0.7301232610105961,
+      "grad_norm": 0.11115461883527536,
+      "learning_rate": 3.583259098496863e-05,
+      "loss": 1.2775,
+      "step": 10129
+    },
+    {
+      "epoch": 0.730195343472933,
+      "grad_norm": 0.13282658552703103,
+      "learning_rate": 3.581468599626784e-05,
+      "loss": 1.2833,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7302674259352699,
+      "grad_norm": 0.12720983284913961,
+      "learning_rate": 3.579678450624084e-05,
+      "loss": 1.2506,
+      "step": 10131
+    },
+    {
+      "epoch": 0.7303395083976069,
+      "grad_norm": 0.12064178245362774,
+      "learning_rate": 3.5778886515863474e-05,
+      "loss": 1.2887,
+      "step": 10132
+    },
+    {
+      "epoch": 0.7304115908599438,
+      "grad_norm": 0.11232701168677116,
+      "learning_rate": 3.5760992026111285e-05,
+      "loss": 1.2704,
+      "step": 10133
+    },
+    {
+      "epoch": 0.7304836733222807,
+      "grad_norm": 0.11001555754962658,
+      "learning_rate": 3.5743101037959805e-05,
+      "loss": 1.3287,
+      "step": 10134
+    },
+    {
+      "epoch": 0.7305557557846176,
+      "grad_norm": 0.1162333822941189,
+      "learning_rate": 3.572521355238414e-05,
+      "loss": 1.2468,
+      "step": 10135
+    },
+    {
+      "epoch": 0.7306278382469545,
+      "grad_norm": 0.12013392077554404,
+      "learning_rate": 3.570732957035936e-05,
+      "loss": 1.3122,
+      "step": 10136
+    },
+    {
+      "epoch": 0.7306999207092915,
+      "grad_norm": 0.11569735326562021,
+      "learning_rate": 3.568944909286026e-05,
+      "loss": 1.2755,
+      "step": 10137
+    },
+    {
+      "epoch": 0.7307720031716284,
+      "grad_norm": 0.11483049127997715,
+      "learning_rate": 3.5671572120861596e-05,
+      "loss": 1.4369,
+      "step": 10138
+    },
+    {
+      "epoch": 0.7308440856339653,
+      "grad_norm": 0.12342118717602864,
+      "learning_rate": 3.565369865533771e-05,
+      "loss": 1.3343,
+      "step": 10139
+    },
+    {
+      "epoch": 0.7309161680963022,
+      "grad_norm": 0.10920954500624136,
+      "learning_rate": 3.5635828697262904e-05,
+      "loss": 1.2596,
+      "step": 10140
+    },
+    {
+      "epoch": 0.730988250558639,
+      "grad_norm": 0.11690798202465426,
+      "learning_rate": 3.561796224761126e-05,
+      "loss": 1.2866,
+      "step": 10141
+    },
+    {
+      "epoch": 0.731060333020976,
+      "grad_norm": 0.13650014813573186,
+      "learning_rate": 3.560009930735664e-05,
+      "loss": 1.3089,
+      "step": 10142
+    },
+    {
+      "epoch": 0.7311324154833129,
+      "grad_norm": 0.11927624105106856,
+      "learning_rate": 3.5582239877472756e-05,
+      "loss": 1.264,
+      "step": 10143
+    },
+    {
+      "epoch": 0.7312044979456498,
+      "grad_norm": 0.12219507728175054,
+      "learning_rate": 3.556438395893309e-05,
+      "loss": 1.3122,
+      "step": 10144
+    },
+    {
+      "epoch": 0.7312765804079867,
+      "grad_norm": 0.11201727427714982,
+      "learning_rate": 3.554653155271096e-05,
+      "loss": 1.2243,
+      "step": 10145
+    },
+    {
+      "epoch": 0.7313486628703236,
+      "grad_norm": 0.11870704513551232,
+      "learning_rate": 3.552868265977948e-05,
+      "loss": 1.2186,
+      "step": 10146
+    },
+    {
+      "epoch": 0.7314207453326605,
+      "grad_norm": 0.1582768114986185,
+      "learning_rate": 3.551083728111157e-05,
+      "loss": 1.3198,
+      "step": 10147
+    },
+    {
+      "epoch": 0.7314928277949975,
+      "grad_norm": 0.13280655869003202,
+      "learning_rate": 3.5492995417679956e-05,
+      "loss": 1.2898,
+      "step": 10148
+    },
+    {
+      "epoch": 0.7315649102573344,
+      "grad_norm": 0.12997030219540312,
+      "learning_rate": 3.547515707045719e-05,
+      "loss": 1.2959,
+      "step": 10149
+    },
+    {
+      "epoch": 0.7316369927196713,
+      "grad_norm": 0.12775820839926832,
+      "learning_rate": 3.545732224041562e-05,
+      "loss": 1.3415,
+      "step": 10150
+    },
+    {
+      "epoch": 0.7317090751820082,
+      "grad_norm": 0.12480393990351338,
+      "learning_rate": 3.543949092852741e-05,
+      "loss": 1.3048,
+      "step": 10151
+    },
+    {
+      "epoch": 0.7317811576443451,
+      "grad_norm": 0.12182052989868579,
+      "learning_rate": 3.542166313576451e-05,
+      "loss": 1.328,
+      "step": 10152
+    },
+    {
+      "epoch": 0.7318532401066821,
+      "grad_norm": 0.12906930041376843,
+      "learning_rate": 3.5403838863098715e-05,
+      "loss": 1.1734,
+      "step": 10153
+    },
+    {
+      "epoch": 0.731925322569019,
+      "grad_norm": 0.1280383743930584,
+      "learning_rate": 3.538601811150159e-05,
+      "loss": 1.2623,
+      "step": 10154
+    },
+    {
+      "epoch": 0.7319974050313559,
+      "grad_norm": 0.11873796037985658,
+      "learning_rate": 3.536820088194453e-05,
+      "loss": 1.3675,
+      "step": 10155
+    },
+    {
+      "epoch": 0.7320694874936928,
+      "grad_norm": 0.118121683506864,
+      "learning_rate": 3.535038717539875e-05,
+      "loss": 1.3067,
+      "step": 10156
+    },
+    {
+      "epoch": 0.7321415699560297,
+      "grad_norm": 0.11453647214051219,
+      "learning_rate": 3.533257699283522e-05,
+      "loss": 1.1917,
+      "step": 10157
+    },
+    {
+      "epoch": 0.7322136524183666,
+      "grad_norm": 0.11142123440380823,
+      "learning_rate": 3.5314770335224775e-05,
+      "loss": 1.3416,
+      "step": 10158
+    },
+    {
+      "epoch": 0.7322857348807035,
+      "grad_norm": 0.12157017384487886,
+      "learning_rate": 3.529696720353805e-05,
+      "loss": 1.2766,
+      "step": 10159
+    },
+    {
+      "epoch": 0.7323578173430404,
+      "grad_norm": 0.12727316297812186,
+      "learning_rate": 3.5279167598745455e-05,
+      "loss": 1.1594,
+      "step": 10160
+    },
+    {
+      "epoch": 0.7324298998053773,
+      "grad_norm": 0.22804584213441176,
+      "learning_rate": 3.5261371521817244e-05,
+      "loss": 1.2382,
+      "step": 10161
+    },
+    {
+      "epoch": 0.7325019822677142,
+      "grad_norm": 0.10955651769822018,
+      "learning_rate": 3.5243578973723444e-05,
+      "loss": 1.3312,
+      "step": 10162
+    },
+    {
+      "epoch": 0.7325740647300512,
+      "grad_norm": 0.11669936432748307,
+      "learning_rate": 3.5225789955433954e-05,
+      "loss": 1.3548,
+      "step": 10163
+    },
+    {
+      "epoch": 0.7326461471923881,
+      "grad_norm": 0.12156920827824484,
+      "learning_rate": 3.5208004467918324e-05,
+      "loss": 1.281,
+      "step": 10164
+    },
+    {
+      "epoch": 0.732718229654725,
+      "grad_norm": 0.11409500876383663,
+      "learning_rate": 3.519022251214614e-05,
+      "loss": 1.2363,
+      "step": 10165
+    },
+    {
+      "epoch": 0.7327903121170619,
+      "grad_norm": 0.11522700498076091,
+      "learning_rate": 3.517244408908663e-05,
+      "loss": 1.2401,
+      "step": 10166
+    },
+    {
+      "epoch": 0.7328623945793988,
+      "grad_norm": 0.1024783595863152,
+      "learning_rate": 3.5154669199708915e-05,
+      "loss": 1.2246,
+      "step": 10167
+    },
+    {
+      "epoch": 0.7329344770417358,
+      "grad_norm": 0.1326119873644706,
+      "learning_rate": 3.513689784498178e-05,
+      "loss": 1.2916,
+      "step": 10168
+    },
+    {
+      "epoch": 0.7330065595040727,
+      "grad_norm": 0.12200771315677203,
+      "learning_rate": 3.5119130025874034e-05,
+      "loss": 1.2874,
+      "step": 10169
+    },
+    {
+      "epoch": 0.7330786419664096,
+      "grad_norm": 0.12627800007700132,
+      "learning_rate": 3.510136574335414e-05,
+      "loss": 1.3242,
+      "step": 10170
+    },
+    {
+      "epoch": 0.7331507244287465,
+      "grad_norm": 0.13169947011693342,
+      "learning_rate": 3.508360499839044e-05,
+      "loss": 1.4002,
+      "step": 10171
+    },
+    {
+      "epoch": 0.7332228068910834,
+      "grad_norm": 0.11188985686099796,
+      "learning_rate": 3.5065847791951e-05,
+      "loss": 1.3069,
+      "step": 10172
+    },
+    {
+      "epoch": 0.7332948893534204,
+      "grad_norm": 0.11720013895575908,
+      "learning_rate": 3.504809412500374e-05,
+      "loss": 1.2629,
+      "step": 10173
+    },
+    {
+      "epoch": 0.7333669718157573,
+      "grad_norm": 0.3741100503787521,
+      "learning_rate": 3.5030343998516455e-05,
+      "loss": 1.2189,
+      "step": 10174
+    },
+    {
+      "epoch": 0.7334390542780942,
+      "grad_norm": 0.11589171231746417,
+      "learning_rate": 3.5012597413456706e-05,
+      "loss": 1.1819,
+      "step": 10175
+    },
+    {
+      "epoch": 0.733511136740431,
+      "grad_norm": 0.12478032642262214,
+      "learning_rate": 3.499485437079174e-05,
+      "loss": 1.246,
+      "step": 10176
+    },
+    {
+      "epoch": 0.7335832192027679,
+      "grad_norm": 0.12474298735747358,
+      "learning_rate": 3.497711487148874e-05,
+      "loss": 1.2908,
+      "step": 10177
+    },
+    {
+      "epoch": 0.7336553016651048,
+      "grad_norm": 0.10435812242338857,
+      "learning_rate": 3.495937891651472e-05,
+      "loss": 1.322,
+      "step": 10178
+    },
+    {
+      "epoch": 0.7337273841274418,
+      "grad_norm": 0.1275438774014604,
+      "learning_rate": 3.494164650683646e-05,
+      "loss": 1.2047,
+      "step": 10179
+    },
+    {
+      "epoch": 0.7337994665897787,
+      "grad_norm": 0.11682901042700064,
+      "learning_rate": 3.492391764342044e-05,
+      "loss": 1.2722,
+      "step": 10180
+    },
+    {
+      "epoch": 0.7338715490521156,
+      "grad_norm": 0.11089652912126846,
+      "learning_rate": 3.490619232723308e-05,
+      "loss": 1.3453,
+      "step": 10181
+    },
+    {
+      "epoch": 0.7339436315144525,
+      "grad_norm": 0.12261145315627486,
+      "learning_rate": 3.4888470559240616e-05,
+      "loss": 1.3266,
+      "step": 10182
+    },
+    {
+      "epoch": 0.7340157139767894,
+      "grad_norm": 0.11642374456952582,
+      "learning_rate": 3.4870752340409043e-05,
+      "loss": 1.2239,
+      "step": 10183
+    },
+    {
+      "epoch": 0.7340877964391264,
+      "grad_norm": 0.11500010942750928,
+      "learning_rate": 3.4853037671704094e-05,
+      "loss": 1.1751,
+      "step": 10184
+    },
+    {
+      "epoch": 0.7341598789014633,
+      "grad_norm": 0.1303123463962188,
+      "learning_rate": 3.483532655409142e-05,
+      "loss": 1.2378,
+      "step": 10185
+    },
+    {
+      "epoch": 0.7342319613638002,
+      "grad_norm": 0.11891604189671692,
+      "learning_rate": 3.4817618988536394e-05,
+      "loss": 1.2774,
+      "step": 10186
+    },
+    {
+      "epoch": 0.7343040438261371,
+      "grad_norm": 0.1518837344394934,
+      "learning_rate": 3.4799914976004336e-05,
+      "loss": 1.2641,
+      "step": 10187
+    },
+    {
+      "epoch": 0.734376126288474,
+      "grad_norm": 0.12585621947193174,
+      "learning_rate": 3.4782214517460165e-05,
+      "loss": 1.2145,
+      "step": 10188
+    },
+    {
+      "epoch": 0.734448208750811,
+      "grad_norm": 0.12283955732515434,
+      "learning_rate": 3.476451761386876e-05,
+      "loss": 1.2826,
+      "step": 10189
+    },
+    {
+      "epoch": 0.7345202912131479,
+      "grad_norm": 0.1240612432763927,
+      "learning_rate": 3.4746824266194744e-05,
+      "loss": 1.2482,
+      "step": 10190
+    },
+    {
+      "epoch": 0.7345923736754848,
+      "grad_norm": 0.11909292193823579,
+      "learning_rate": 3.472913447540263e-05,
+      "loss": 1.2252,
+      "step": 10191
+    },
+    {
+      "epoch": 0.7346644561378217,
+      "grad_norm": 0.11791712773673244,
+      "learning_rate": 3.471144824245658e-05,
+      "loss": 1.1828,
+      "step": 10192
+    },
+    {
+      "epoch": 0.7347365386001585,
+      "grad_norm": 0.11285981139235736,
+      "learning_rate": 3.4693765568320694e-05,
+      "loss": 1.2444,
+      "step": 10193
+    },
+    {
+      "epoch": 0.7348086210624954,
+      "grad_norm": 0.11962556329596429,
+      "learning_rate": 3.467608645395883e-05,
+      "loss": 1.2759,
+      "step": 10194
+    },
+    {
+      "epoch": 0.7348807035248324,
+      "grad_norm": 0.11356276921961665,
+      "learning_rate": 3.465841090033466e-05,
+      "loss": 1.2648,
+      "step": 10195
+    },
+    {
+      "epoch": 0.7349527859871693,
+      "grad_norm": 0.12024894189723741,
+      "learning_rate": 3.4640738908411664e-05,
+      "loss": 1.3195,
+      "step": 10196
+    },
+    {
+      "epoch": 0.7350248684495062,
+      "grad_norm": 0.12929160132351714,
+      "learning_rate": 3.462307047915311e-05,
+      "loss": 1.2988,
+      "step": 10197
+    },
+    {
+      "epoch": 0.7350969509118431,
+      "grad_norm": 0.12307520445544792,
+      "learning_rate": 3.460540561352209e-05,
+      "loss": 1.2141,
+      "step": 10198
+    },
+    {
+      "epoch": 0.73516903337418,
+      "grad_norm": 0.14550893524459352,
+      "learning_rate": 3.458774431248151e-05,
+      "loss": 1.3346,
+      "step": 10199
+    },
+    {
+      "epoch": 0.735241115836517,
+      "grad_norm": 0.1151506664145388,
+      "learning_rate": 3.457008657699405e-05,
+      "loss": 1.3063,
+      "step": 10200
+    },
+    {
+      "epoch": 0.7353131982988539,
+      "grad_norm": 0.1234355010639715,
+      "learning_rate": 3.455243240802224e-05,
+      "loss": 1.2764,
+      "step": 10201
+    },
+    {
+      "epoch": 0.7353852807611908,
+      "grad_norm": 0.11680814523396486,
+      "learning_rate": 3.453478180652836e-05,
+      "loss": 1.2707,
+      "step": 10202
+    },
+    {
+      "epoch": 0.7354573632235277,
+      "grad_norm": 0.10424706950151474,
+      "learning_rate": 3.451713477347455e-05,
+      "loss": 1.4097,
+      "step": 10203
+    },
+    {
+      "epoch": 0.7355294456858646,
+      "grad_norm": 0.1064633267342327,
+      "learning_rate": 3.449949130982271e-05,
+      "loss": 1.2904,
+      "step": 10204
+    },
+    {
+      "epoch": 0.7356015281482016,
+      "grad_norm": 0.1217691640028133,
+      "learning_rate": 3.448185141653458e-05,
+      "loss": 1.3359,
+      "step": 10205
+    },
+    {
+      "epoch": 0.7356736106105385,
+      "grad_norm": 0.13258204001726218,
+      "learning_rate": 3.446421509457169e-05,
+      "loss": 1.3105,
+      "step": 10206
+    },
+    {
+      "epoch": 0.7357456930728754,
+      "grad_norm": 0.1149383474173665,
+      "learning_rate": 3.44465823448954e-05,
+      "loss": 1.2214,
+      "step": 10207
+    },
+    {
+      "epoch": 0.7358177755352123,
+      "grad_norm": 0.11558426062363598,
+      "learning_rate": 3.4428953168466806e-05,
+      "loss": 1.2445,
+      "step": 10208
+    },
+    {
+      "epoch": 0.7358898579975492,
+      "grad_norm": 0.12268621168061072,
+      "learning_rate": 3.441132756624687e-05,
+      "loss": 1.2873,
+      "step": 10209
+    },
+    {
+      "epoch": 0.735961940459886,
+      "grad_norm": 0.11170599356965737,
+      "learning_rate": 3.4393705539196376e-05,
+      "loss": 1.2667,
+      "step": 10210
+    },
+    {
+      "epoch": 0.736034022922223,
+      "grad_norm": 0.11553646791191985,
+      "learning_rate": 3.437608708827585e-05,
+      "loss": 1.3365,
+      "step": 10211
+    },
+    {
+      "epoch": 0.7361061053845599,
+      "grad_norm": 0.11789155972293763,
+      "learning_rate": 3.435847221444568e-05,
+      "loss": 1.2805,
+      "step": 10212
+    },
+    {
+      "epoch": 0.7361781878468968,
+      "grad_norm": 0.12354538640111905,
+      "learning_rate": 3.434086091866601e-05,
+      "loss": 1.3339,
+      "step": 10213
+    },
+    {
+      "epoch": 0.7362502703092337,
+      "grad_norm": 0.11052360435524937,
+      "learning_rate": 3.4323253201896824e-05,
+      "loss": 1.2293,
+      "step": 10214
+    },
+    {
+      "epoch": 0.7363223527715707,
+      "grad_norm": 0.11423208782312765,
+      "learning_rate": 3.43056490650979e-05,
+      "loss": 1.3525,
+      "step": 10215
+    },
+    {
+      "epoch": 0.7363944352339076,
+      "grad_norm": 0.12023353910725298,
+      "learning_rate": 3.4288048509228855e-05,
+      "loss": 1.3078,
+      "step": 10216
+    },
+    {
+      "epoch": 0.7364665176962445,
+      "grad_norm": 0.1336213733385316,
+      "learning_rate": 3.427045153524897e-05,
+      "loss": 1.3262,
+      "step": 10217
+    },
+    {
+      "epoch": 0.7365386001585814,
+      "grad_norm": 0.10700050008679449,
+      "learning_rate": 3.425285814411755e-05,
+      "loss": 1.2683,
+      "step": 10218
+    },
+    {
+      "epoch": 0.7366106826209183,
+      "grad_norm": 0.13997046668807744,
+      "learning_rate": 3.423526833679355e-05,
+      "loss": 1.309,
+      "step": 10219
+    },
+    {
+      "epoch": 0.7366827650832553,
+      "grad_norm": 0.1090485756209805,
+      "learning_rate": 3.42176821142358e-05,
+      "loss": 1.2278,
+      "step": 10220
+    },
+    {
+      "epoch": 0.7367548475455922,
+      "grad_norm": 0.11390334796362941,
+      "learning_rate": 3.4200099477402814e-05,
+      "loss": 1.2154,
+      "step": 10221
+    },
+    {
+      "epoch": 0.7368269300079291,
+      "grad_norm": 0.12396025468049338,
+      "learning_rate": 3.418252042725312e-05,
+      "loss": 1.1983,
+      "step": 10222
+    },
+    {
+      "epoch": 0.736899012470266,
+      "grad_norm": 0.11929232059660572,
+      "learning_rate": 3.416494496474486e-05,
+      "loss": 1.2994,
+      "step": 10223
+    },
+    {
+      "epoch": 0.7369710949326029,
+      "grad_norm": 0.12062094305231622,
+      "learning_rate": 3.414737309083612e-05,
+      "loss": 1.1998,
+      "step": 10224
+    },
+    {
+      "epoch": 0.7370431773949399,
+      "grad_norm": 0.12152658032031027,
+      "learning_rate": 3.4129804806484624e-05,
+      "loss": 1.3115,
+      "step": 10225
+    },
+    {
+      "epoch": 0.7371152598572768,
+      "grad_norm": 0.13570353672661245,
+      "learning_rate": 3.4112240112648075e-05,
+      "loss": 1.2611,
+      "step": 10226
+    },
+    {
+      "epoch": 0.7371873423196137,
+      "grad_norm": 0.10392515149933611,
+      "learning_rate": 3.4094679010283895e-05,
+      "loss": 1.3453,
+      "step": 10227
+    },
+    {
+      "epoch": 0.7372594247819505,
+      "grad_norm": 0.13476802949957678,
+      "learning_rate": 3.407712150034935e-05,
+      "loss": 1.2744,
+      "step": 10228
+    },
+    {
+      "epoch": 0.7373315072442874,
+      "grad_norm": 0.11862063717097843,
+      "learning_rate": 3.405956758380141e-05,
+      "loss": 1.1945,
+      "step": 10229
+    },
+    {
+      "epoch": 0.7374035897066243,
+      "grad_norm": 0.12419206764577345,
+      "learning_rate": 3.404201726159693e-05,
+      "loss": 1.1829,
+      "step": 10230
+    },
+    {
+      "epoch": 0.7374756721689613,
+      "grad_norm": 0.13139706060738623,
+      "learning_rate": 3.402447053469261e-05,
+      "loss": 1.2707,
+      "step": 10231
+    },
+    {
+      "epoch": 0.7375477546312982,
+      "grad_norm": 0.1246294868399954,
+      "learning_rate": 3.4006927404044916e-05,
+      "loss": 1.2488,
+      "step": 10232
+    },
+    {
+      "epoch": 0.7376198370936351,
+      "grad_norm": 0.12366386974824167,
+      "learning_rate": 3.3989387870610036e-05,
+      "loss": 1.2935,
+      "step": 10233
+    },
+    {
+      "epoch": 0.737691919555972,
+      "grad_norm": 0.11955249551785524,
+      "learning_rate": 3.397185193534403e-05,
+      "loss": 1.2445,
+      "step": 10234
+    },
+    {
+      "epoch": 0.7377640020183089,
+      "grad_norm": 0.13745840426829276,
+      "learning_rate": 3.395431959920283e-05,
+      "loss": 1.2142,
+      "step": 10235
+    },
+    {
+      "epoch": 0.7378360844806459,
+      "grad_norm": 0.117181922801554,
+      "learning_rate": 3.393679086314212e-05,
+      "loss": 1.2546,
+      "step": 10236
+    },
+    {
+      "epoch": 0.7379081669429828,
+      "grad_norm": 0.11836419766207076,
+      "learning_rate": 3.391926572811728e-05,
+      "loss": 1.2869,
+      "step": 10237
+    },
+    {
+      "epoch": 0.7379802494053197,
+      "grad_norm": 0.12915813700913606,
+      "learning_rate": 3.390174419508364e-05,
+      "loss": 1.3398,
+      "step": 10238
+    },
+    {
+      "epoch": 0.7380523318676566,
+      "grad_norm": 0.11541981425234342,
+      "learning_rate": 3.388422626499623e-05,
+      "loss": 1.2932,
+      "step": 10239
+    },
+    {
+      "epoch": 0.7381244143299935,
+      "grad_norm": 0.12091737635533661,
+      "learning_rate": 3.386671193881006e-05,
+      "loss": 1.2867,
+      "step": 10240
+    },
+    {
+      "epoch": 0.7381964967923305,
+      "grad_norm": 0.1474572483501177,
+      "learning_rate": 3.3849201217479684e-05,
+      "loss": 1.3428,
+      "step": 10241
+    },
+    {
+      "epoch": 0.7382685792546674,
+      "grad_norm": 0.11608261111242402,
+      "learning_rate": 3.383169410195965e-05,
+      "loss": 1.1727,
+      "step": 10242
+    },
+    {
+      "epoch": 0.7383406617170043,
+      "grad_norm": 0.1205912968448111,
+      "learning_rate": 3.381419059320422e-05,
+      "loss": 1.3043,
+      "step": 10243
+    },
+    {
+      "epoch": 0.7384127441793412,
+      "grad_norm": 0.12530046454930377,
+      "learning_rate": 3.3796690692167576e-05,
+      "loss": 1.3225,
+      "step": 10244
+    },
+    {
+      "epoch": 0.738484826641678,
+      "grad_norm": 0.12024131936372102,
+      "learning_rate": 3.3779194399803536e-05,
+      "loss": 1.3323,
+      "step": 10245
+    },
+    {
+      "epoch": 0.738556909104015,
+      "grad_norm": 0.1229200706946679,
+      "learning_rate": 3.376170171706582e-05,
+      "loss": 1.2132,
+      "step": 10246
+    },
+    {
+      "epoch": 0.7386289915663519,
+      "grad_norm": 0.1175712276684676,
+      "learning_rate": 3.374421264490795e-05,
+      "loss": 1.2564,
+      "step": 10247
+    },
+    {
+      "epoch": 0.7387010740286888,
+      "grad_norm": 0.12601702780765742,
+      "learning_rate": 3.3726727184283236e-05,
+      "loss": 1.2514,
+      "step": 10248
+    },
+    {
+      "epoch": 0.7387731564910257,
+      "grad_norm": 0.121043523737243,
+      "learning_rate": 3.3709245336144777e-05,
+      "loss": 1.2436,
+      "step": 10249
+    },
+    {
+      "epoch": 0.7388452389533626,
+      "grad_norm": 0.11599152730239065,
+      "learning_rate": 3.36917671014455e-05,
+      "loss": 1.3246,
+      "step": 10250
+    },
+    {
+      "epoch": 0.7389173214156995,
+      "grad_norm": 0.12387831982900487,
+      "learning_rate": 3.367429248113815e-05,
+      "loss": 1.2907,
+      "step": 10251
+    },
+    {
+      "epoch": 0.7389894038780365,
+      "grad_norm": 0.12695599883898195,
+      "learning_rate": 3.365682147617521e-05,
+      "loss": 1.3404,
+      "step": 10252
+    },
+    {
+      "epoch": 0.7390614863403734,
+      "grad_norm": 0.12953777287975612,
+      "learning_rate": 3.363935408750901e-05,
+      "loss": 1.2594,
+      "step": 10253
+    },
+    {
+      "epoch": 0.7391335688027103,
+      "grad_norm": 0.11619377396098522,
+      "learning_rate": 3.3621890316091706e-05,
+      "loss": 1.2411,
+      "step": 10254
+    },
+    {
+      "epoch": 0.7392056512650472,
+      "grad_norm": 0.1322650618122938,
+      "learning_rate": 3.360443016287522e-05,
+      "loss": 1.2422,
+      "step": 10255
+    },
+    {
+      "epoch": 0.7392777337273841,
+      "grad_norm": 0.14377582674616246,
+      "learning_rate": 3.358697362881127e-05,
+      "loss": 1.2436,
+      "step": 10256
+    },
+    {
+      "epoch": 0.7393498161897211,
+      "grad_norm": 0.12526948009004196,
+      "learning_rate": 3.356952071485141e-05,
+      "loss": 1.2477,
+      "step": 10257
+    },
+    {
+      "epoch": 0.739421898652058,
+      "grad_norm": 0.13331185231149345,
+      "learning_rate": 3.3552071421946983e-05,
+      "loss": 1.2231,
+      "step": 10258
+    },
+    {
+      "epoch": 0.7394939811143949,
+      "grad_norm": 0.12282529860871995,
+      "learning_rate": 3.353462575104911e-05,
+      "loss": 1.2465,
+      "step": 10259
+    },
+    {
+      "epoch": 0.7395660635767318,
+      "grad_norm": 0.13329532766310967,
+      "learning_rate": 3.351718370310876e-05,
+      "loss": 1.2416,
+      "step": 10260
+    },
+    {
+      "epoch": 0.7396381460390687,
+      "grad_norm": 0.11292161535122519,
+      "learning_rate": 3.3499745279076665e-05,
+      "loss": 1.3156,
+      "step": 10261
+    },
+    {
+      "epoch": 0.7397102285014056,
+      "grad_norm": 0.18365766057571375,
+      "learning_rate": 3.3482310479903386e-05,
+      "loss": 1.2483,
+      "step": 10262
+    },
+    {
+      "epoch": 0.7397823109637425,
+      "grad_norm": 0.1275848621633368,
+      "learning_rate": 3.3464879306539267e-05,
+      "loss": 1.3184,
+      "step": 10263
+    },
+    {
+      "epoch": 0.7398543934260794,
+      "grad_norm": 0.11692002196661426,
+      "learning_rate": 3.344745175993447e-05,
+      "loss": 1.1381,
+      "step": 10264
+    },
+    {
+      "epoch": 0.7399264758884163,
+      "grad_norm": 0.12319649729336321,
+      "learning_rate": 3.343002784103893e-05,
+      "loss": 1.3433,
+      "step": 10265
+    },
+    {
+      "epoch": 0.7399985583507532,
+      "grad_norm": 0.12153873659784488,
+      "learning_rate": 3.3412607550802444e-05,
+      "loss": 1.3333,
+      "step": 10266
+    },
+    {
+      "epoch": 0.7400706408130902,
+      "grad_norm": 0.13488244204003189,
+      "learning_rate": 3.339519089017455e-05,
+      "loss": 1.2498,
+      "step": 10267
+    },
+    {
+      "epoch": 0.7401427232754271,
+      "grad_norm": 0.11332383826147462,
+      "learning_rate": 3.337777786010461e-05,
+      "loss": 1.2418,
+      "step": 10268
+    },
+    {
+      "epoch": 0.740214805737764,
+      "grad_norm": 0.11419551985994382,
+      "learning_rate": 3.3360368461541835e-05,
+      "loss": 1.2854,
+      "step": 10269
+    },
+    {
+      "epoch": 0.7402868882001009,
+      "grad_norm": 0.10473104115113564,
+      "learning_rate": 3.334296269543509e-05,
+      "loss": 1.161,
+      "step": 10270
+    },
+    {
+      "epoch": 0.7403589706624378,
+      "grad_norm": 0.12746195592232615,
+      "learning_rate": 3.332556056273324e-05,
+      "loss": 1.2792,
+      "step": 10271
+    },
+    {
+      "epoch": 0.7404310531247748,
+      "grad_norm": 0.12575860067503836,
+      "learning_rate": 3.330816206438482e-05,
+      "loss": 1.3011,
+      "step": 10272
+    },
+    {
+      "epoch": 0.7405031355871117,
+      "grad_norm": 0.12402246917576613,
+      "learning_rate": 3.329076720133825e-05,
+      "loss": 1.1525,
+      "step": 10273
+    },
+    {
+      "epoch": 0.7405752180494486,
+      "grad_norm": 0.11488422616973427,
+      "learning_rate": 3.32733759745416e-05,
+      "loss": 1.2573,
+      "step": 10274
+    },
+    {
+      "epoch": 0.7406473005117855,
+      "grad_norm": 0.12664569726433514,
+      "learning_rate": 3.325598838494294e-05,
+      "loss": 1.2877,
+      "step": 10275
+    },
+    {
+      "epoch": 0.7407193829741224,
+      "grad_norm": 0.13346778197084472,
+      "learning_rate": 3.323860443349003e-05,
+      "loss": 1.3713,
+      "step": 10276
+    },
+    {
+      "epoch": 0.7407914654364594,
+      "grad_norm": 0.12084695404143699,
+      "learning_rate": 3.322122412113047e-05,
+      "loss": 1.3817,
+      "step": 10277
+    },
+    {
+      "epoch": 0.7408635478987963,
+      "grad_norm": 0.1285022948630132,
+      "learning_rate": 3.3203847448811545e-05,
+      "loss": 1.3191,
+      "step": 10278
+    },
+    {
+      "epoch": 0.7409356303611332,
+      "grad_norm": 0.11430330751348423,
+      "learning_rate": 3.318647441748055e-05,
+      "loss": 1.2228,
+      "step": 10279
+    },
+    {
+      "epoch": 0.74100771282347,
+      "grad_norm": 0.112493036304887,
+      "learning_rate": 3.3169105028084456e-05,
+      "loss": 1.3191,
+      "step": 10280
+    },
+    {
+      "epoch": 0.7410797952858069,
+      "grad_norm": 0.11056050982821806,
+      "learning_rate": 3.315173928156999e-05,
+      "loss": 1.3392,
+      "step": 10281
+    },
+    {
+      "epoch": 0.7411518777481438,
+      "grad_norm": 0.1156296067424169,
+      "learning_rate": 3.313437717888379e-05,
+      "loss": 1.2894,
+      "step": 10282
+    },
+    {
+      "epoch": 0.7412239602104808,
+      "grad_norm": 0.11523056898494678,
+      "learning_rate": 3.311701872097218e-05,
+      "loss": 1.3348,
+      "step": 10283
+    },
+    {
+      "epoch": 0.7412960426728177,
+      "grad_norm": 0.16460709787183953,
+      "learning_rate": 3.309966390878149e-05,
+      "loss": 1.2047,
+      "step": 10284
+    },
+    {
+      "epoch": 0.7413681251351546,
+      "grad_norm": 0.11313430803771082,
+      "learning_rate": 3.3082312743257574e-05,
+      "loss": 1.3036,
+      "step": 10285
+    },
+    {
+      "epoch": 0.7414402075974915,
+      "grad_norm": 0.10681996224880701,
+      "learning_rate": 3.306496522534628e-05,
+      "loss": 1.2875,
+      "step": 10286
+    },
+    {
+      "epoch": 0.7415122900598284,
+      "grad_norm": 0.12181199164777676,
+      "learning_rate": 3.3047621355993165e-05,
+      "loss": 1.2117,
+      "step": 10287
+    },
+    {
+      "epoch": 0.7415843725221654,
+      "grad_norm": 0.10940484129657485,
+      "learning_rate": 3.3030281136143723e-05,
+      "loss": 1.2437,
+      "step": 10288
+    },
+    {
+      "epoch": 0.7416564549845023,
+      "grad_norm": 0.12854350138615792,
+      "learning_rate": 3.3012944566743054e-05,
+      "loss": 1.1697,
+      "step": 10289
+    },
+    {
+      "epoch": 0.7417285374468392,
+      "grad_norm": 0.12023658981704645,
+      "learning_rate": 3.29956116487362e-05,
+      "loss": 1.2706,
+      "step": 10290
+    },
+    {
+      "epoch": 0.7418006199091761,
+      "grad_norm": 0.12750017250763837,
+      "learning_rate": 3.297828238306794e-05,
+      "loss": 1.2808,
+      "step": 10291
+    },
+    {
+      "epoch": 0.741872702371513,
+      "grad_norm": 0.12130359079985571,
+      "learning_rate": 3.296095677068288e-05,
+      "loss": 1.28,
+      "step": 10292
+    },
+    {
+      "epoch": 0.74194478483385,
+      "grad_norm": 0.20778752490055657,
+      "learning_rate": 3.2943634812525435e-05,
+      "loss": 1.2695,
+      "step": 10293
+    },
+    {
+      "epoch": 0.7420168672961869,
+      "grad_norm": 0.10672009003208684,
+      "learning_rate": 3.292631650953979e-05,
+      "loss": 1.198,
+      "step": 10294
+    },
+    {
+      "epoch": 0.7420889497585238,
+      "grad_norm": 0.11402193456565403,
+      "learning_rate": 3.290900186266995e-05,
+      "loss": 1.3314,
+      "step": 10295
+    },
+    {
+      "epoch": 0.7421610322208607,
+      "grad_norm": 0.13262483011262066,
+      "learning_rate": 3.2891690872859736e-05,
+      "loss": 1.3185,
+      "step": 10296
+    },
+    {
+      "epoch": 0.7422331146831975,
+      "grad_norm": 0.12376133848662653,
+      "learning_rate": 3.287438354105272e-05,
+      "loss": 1.1889,
+      "step": 10297
+    },
+    {
+      "epoch": 0.7423051971455344,
+      "grad_norm": 0.12485788067755076,
+      "learning_rate": 3.285707986819234e-05,
+      "loss": 1.305,
+      "step": 10298
+    },
+    {
+      "epoch": 0.7423772796078714,
+      "grad_norm": 0.11349815086722033,
+      "learning_rate": 3.283977985522178e-05,
+      "loss": 1.2425,
+      "step": 10299
+    },
+    {
+      "epoch": 0.7424493620702083,
+      "grad_norm": 0.12151902299356453,
+      "learning_rate": 3.2822483503084045e-05,
+      "loss": 1.2948,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7425214445325452,
+      "grad_norm": 0.11840196300842637,
+      "learning_rate": 3.2805190812721955e-05,
+      "loss": 1.192,
+      "step": 10301
+    },
+    {
+      "epoch": 0.7425935269948821,
+      "grad_norm": 0.12458541044182049,
+      "learning_rate": 3.278790178507809e-05,
+      "loss": 1.3054,
+      "step": 10302
+    },
+    {
+      "epoch": 0.742665609457219,
+      "grad_norm": 0.11073897332048557,
+      "learning_rate": 3.277061642109489e-05,
+      "loss": 1.2038,
+      "step": 10303
+    },
+    {
+      "epoch": 0.742737691919556,
+      "grad_norm": 0.10822688813713859,
+      "learning_rate": 3.275333472171455e-05,
+      "loss": 1.2291,
+      "step": 10304
+    },
+    {
+      "epoch": 0.7428097743818929,
+      "grad_norm": 0.11254555579286965,
+      "learning_rate": 3.273605668787906e-05,
+      "loss": 1.3505,
+      "step": 10305
+    },
+    {
+      "epoch": 0.7428818568442298,
+      "grad_norm": 0.11484540163999457,
+      "learning_rate": 3.271878232053025e-05,
+      "loss": 1.2995,
+      "step": 10306
+    },
+    {
+      "epoch": 0.7429539393065667,
+      "grad_norm": 0.11797992198361118,
+      "learning_rate": 3.270151162060971e-05,
+      "loss": 1.2432,
+      "step": 10307
+    },
+    {
+      "epoch": 0.7430260217689036,
+      "grad_norm": 0.10813345217086046,
+      "learning_rate": 3.2684244589058867e-05,
+      "loss": 1.222,
+      "step": 10308
+    },
+    {
+      "epoch": 0.7430981042312406,
+      "grad_norm": 0.11773026699954592,
+      "learning_rate": 3.26669812268189e-05,
+      "loss": 1.2973,
+      "step": 10309
+    },
+    {
+      "epoch": 0.7431701866935775,
+      "grad_norm": 0.12652538187043857,
+      "learning_rate": 3.264972153483086e-05,
+      "loss": 1.2005,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7432422691559144,
+      "grad_norm": 0.25324992533876545,
+      "learning_rate": 3.263246551403552e-05,
+      "loss": 1.2376,
+      "step": 10311
+    },
+    {
+      "epoch": 0.7433143516182513,
+      "grad_norm": 0.1374195131852353,
+      "learning_rate": 3.261521316537348e-05,
+      "loss": 1.2327,
+      "step": 10312
+    },
+    {
+      "epoch": 0.7433864340805882,
+      "grad_norm": 0.1257576968850496,
+      "learning_rate": 3.259796448978522e-05,
+      "loss": 1.2661,
+      "step": 10313
+    },
+    {
+      "epoch": 0.743458516542925,
+      "grad_norm": 0.11690769348745522,
+      "learning_rate": 3.258071948821081e-05,
+      "loss": 1.2661,
+      "step": 10314
+    },
+    {
+      "epoch": 0.743530599005262,
+      "grad_norm": 0.1161871981626156,
+      "learning_rate": 3.256347816159038e-05,
+      "loss": 1.2947,
+      "step": 10315
+    },
+    {
+      "epoch": 0.7436026814675989,
+      "grad_norm": 0.11393334405690005,
+      "learning_rate": 3.2546240510863676e-05,
+      "loss": 1.2422,
+      "step": 10316
+    },
+    {
+      "epoch": 0.7436747639299358,
+      "grad_norm": 0.1244561713919257,
+      "learning_rate": 3.2529006536970373e-05,
+      "loss": 1.2395,
+      "step": 10317
+    },
+    {
+      "epoch": 0.7437468463922727,
+      "grad_norm": 0.11089679942088587,
+      "learning_rate": 3.2511776240849757e-05,
+      "loss": 1.265,
+      "step": 10318
+    },
+    {
+      "epoch": 0.7438189288546097,
+      "grad_norm": 0.12058796411159403,
+      "learning_rate": 3.249454962344114e-05,
+      "loss": 1.3214,
+      "step": 10319
+    },
+    {
+      "epoch": 0.7438910113169466,
+      "grad_norm": 0.11335342571339657,
+      "learning_rate": 3.2477326685683484e-05,
+      "loss": 1.2574,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7439630937792835,
+      "grad_norm": 0.14912630088960316,
+      "learning_rate": 3.246010742851564e-05,
+      "loss": 1.2788,
+      "step": 10321
+    },
+    {
+      "epoch": 0.7440351762416204,
+      "grad_norm": 0.11186652648848348,
+      "learning_rate": 3.244289185287609e-05,
+      "loss": 1.246,
+      "step": 10322
+    },
+    {
+      "epoch": 0.7441072587039573,
+      "grad_norm": 0.12067663652880713,
+      "learning_rate": 3.242567995970336e-05,
+      "loss": 1.2654,
+      "step": 10323
+    },
+    {
+      "epoch": 0.7441793411662943,
+      "grad_norm": 0.11584760857771663,
+      "learning_rate": 3.240847174993561e-05,
+      "loss": 1.273,
+      "step": 10324
+    },
+    {
+      "epoch": 0.7442514236286312,
+      "grad_norm": 0.12085009393652593,
+      "learning_rate": 3.2391267224510866e-05,
+      "loss": 1.27,
+      "step": 10325
+    },
+    {
+      "epoch": 0.7443235060909681,
+      "grad_norm": 0.1143292761317842,
+      "learning_rate": 3.2374066384366874e-05,
+      "loss": 1.2451,
+      "step": 10326
+    },
+    {
+      "epoch": 0.744395588553305,
+      "grad_norm": 0.1209158931849235,
+      "learning_rate": 3.235686923044123e-05,
+      "loss": 1.2352,
+      "step": 10327
+    },
+    {
+      "epoch": 0.7444676710156419,
+      "grad_norm": 0.14471818610813375,
+      "learning_rate": 3.23396757636714e-05,
+      "loss": 1.3042,
+      "step": 10328
+    },
+    {
+      "epoch": 0.7445397534779788,
+      "grad_norm": 0.13334390808256216,
+      "learning_rate": 3.232248598499459e-05,
+      "loss": 1.2923,
+      "step": 10329
+    },
+    {
+      "epoch": 0.7446118359403158,
+      "grad_norm": 0.12611237209508327,
+      "learning_rate": 3.230529989534771e-05,
+      "loss": 1.2279,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7446839184026527,
+      "grad_norm": 0.11844032930015555,
+      "learning_rate": 3.228811749566757e-05,
+      "loss": 1.2161,
+      "step": 10331
+    },
+    {
+      "epoch": 0.7447560008649895,
+      "grad_norm": 0.121584226346611,
+      "learning_rate": 3.227093878689082e-05,
+      "loss": 1.24,
+      "step": 10332
+    },
+    {
+      "epoch": 0.7448280833273264,
+      "grad_norm": 0.11080041947964687,
+      "learning_rate": 3.225376376995387e-05,
+      "loss": 1.1822,
+      "step": 10333
+    },
+    {
+      "epoch": 0.7449001657896633,
+      "grad_norm": 0.11730919914910119,
+      "learning_rate": 3.223659244579284e-05,
+      "loss": 1.3049,
+      "step": 10334
+    },
+    {
+      "epoch": 0.7449722482520003,
+      "grad_norm": 0.1188269769686752,
+      "learning_rate": 3.2219424815343735e-05,
+      "loss": 1.1943,
+      "step": 10335
+    },
+    {
+      "epoch": 0.7450443307143372,
+      "grad_norm": 0.12814309963030543,
+      "learning_rate": 3.2202260879542335e-05,
+      "loss": 1.2716,
+      "step": 10336
+    },
+    {
+      "epoch": 0.7451164131766741,
+      "grad_norm": 0.11446617832473267,
+      "learning_rate": 3.218510063932432e-05,
+      "loss": 1.2989,
+      "step": 10337
+    },
+    {
+      "epoch": 0.745188495639011,
+      "grad_norm": 0.11267254250707433,
+      "learning_rate": 3.216794409562498e-05,
+      "loss": 1.2721,
+      "step": 10338
+    },
+    {
+      "epoch": 0.7452605781013479,
+      "grad_norm": 0.12386406041494925,
+      "learning_rate": 3.2150791249379516e-05,
+      "loss": 1.371,
+      "step": 10339
+    },
+    {
+      "epoch": 0.7453326605636849,
+      "grad_norm": 0.11374021496115634,
+      "learning_rate": 3.213364210152291e-05,
+      "loss": 1.2572,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7454047430260218,
+      "grad_norm": 0.11825505581907465,
+      "learning_rate": 3.211649665299001e-05,
+      "loss": 1.2796,
+      "step": 10341
+    },
+    {
+      "epoch": 0.7454768254883587,
+      "grad_norm": 0.11451218019527164,
+      "learning_rate": 3.209935490471533e-05,
+      "loss": 1.3544,
+      "step": 10342
+    },
+    {
+      "epoch": 0.7455489079506956,
+      "grad_norm": 0.13461732519554265,
+      "learning_rate": 3.208221685763325e-05,
+      "loss": 1.2308,
+      "step": 10343
+    },
+    {
+      "epoch": 0.7456209904130325,
+      "grad_norm": 0.11589983643406418,
+      "learning_rate": 3.2065082512677936e-05,
+      "loss": 1.2046,
+      "step": 10344
+    },
+    {
+      "epoch": 0.7456930728753695,
+      "grad_norm": 0.11710821708824071,
+      "learning_rate": 3.204795187078344e-05,
+      "loss": 1.2949,
+      "step": 10345
+    },
+    {
+      "epoch": 0.7457651553377064,
+      "grad_norm": 0.14048449987922068,
+      "learning_rate": 3.203082493288346e-05,
+      "loss": 1.3549,
+      "step": 10346
+    },
+    {
+      "epoch": 0.7458372378000433,
+      "grad_norm": 0.12829186602246315,
+      "learning_rate": 3.201370169991159e-05,
+      "loss": 1.2696,
+      "step": 10347
+    },
+    {
+      "epoch": 0.7459093202623802,
+      "grad_norm": 0.11043709155731032,
+      "learning_rate": 3.19965821728012e-05,
+      "loss": 1.2476,
+      "step": 10348
+    },
+    {
+      "epoch": 0.745981402724717,
+      "grad_norm": 0.1267296764756946,
+      "learning_rate": 3.1979466352485464e-05,
+      "loss": 1.3592,
+      "step": 10349
+    },
+    {
+      "epoch": 0.7460534851870539,
+      "grad_norm": 0.11705276248734682,
+      "learning_rate": 3.196235423989734e-05,
+      "loss": 1.2102,
+      "step": 10350
+    },
+    {
+      "epoch": 0.7461255676493909,
+      "grad_norm": 0.12860502962232942,
+      "learning_rate": 3.194524583596958e-05,
+      "loss": 1.3199,
+      "step": 10351
+    },
+    {
+      "epoch": 0.7461976501117278,
+      "grad_norm": 0.12373153510482782,
+      "learning_rate": 3.1928141141634773e-05,
+      "loss": 1.3328,
+      "step": 10352
+    },
+    {
+      "epoch": 0.7462697325740647,
+      "grad_norm": 0.12316854700476547,
+      "learning_rate": 3.191104015782525e-05,
+      "loss": 1.2132,
+      "step": 10353
+    },
+    {
+      "epoch": 0.7463418150364016,
+      "grad_norm": 0.11631213892080618,
+      "learning_rate": 3.1893942885473184e-05,
+      "loss": 1.3091,
+      "step": 10354
+    },
+    {
+      "epoch": 0.7464138974987385,
+      "grad_norm": 0.12663029303598566,
+      "learning_rate": 3.187684932551053e-05,
+      "loss": 1.2409,
+      "step": 10355
+    },
+    {
+      "epoch": 0.7464859799610755,
+      "grad_norm": 0.11644257213762528,
+      "learning_rate": 3.1859759478869024e-05,
+      "loss": 1.255,
+      "step": 10356
+    },
+    {
+      "epoch": 0.7465580624234124,
+      "grad_norm": 0.1084474005228095,
+      "learning_rate": 3.1842673346480235e-05,
+      "loss": 1.2159,
+      "step": 10357
+    },
+    {
+      "epoch": 0.7466301448857493,
+      "grad_norm": 0.12856446014136905,
+      "learning_rate": 3.18255909292755e-05,
+      "loss": 1.3336,
+      "step": 10358
+    },
+    {
+      "epoch": 0.7467022273480862,
+      "grad_norm": 0.10828477845308679,
+      "learning_rate": 3.180851222818596e-05,
+      "loss": 1.2179,
+      "step": 10359
+    },
+    {
+      "epoch": 0.7467743098104231,
+      "grad_norm": 0.11981747327606233,
+      "learning_rate": 3.179143724414257e-05,
+      "loss": 1.2524,
+      "step": 10360
+    },
+    {
+      "epoch": 0.7468463922727601,
+      "grad_norm": 0.11464122145000846,
+      "learning_rate": 3.177436597807606e-05,
+      "loss": 1.2791,
+      "step": 10361
+    },
+    {
+      "epoch": 0.746918474735097,
+      "grad_norm": 0.11948189898042096,
+      "learning_rate": 3.175729843091696e-05,
+      "loss": 1.2973,
+      "step": 10362
+    },
+    {
+      "epoch": 0.7469905571974339,
+      "grad_norm": 0.11585002229030072,
+      "learning_rate": 3.174023460359562e-05,
+      "loss": 1.2801,
+      "step": 10363
+    },
+    {
+      "epoch": 0.7470626396597708,
+      "grad_norm": 0.1282893728853761,
+      "learning_rate": 3.172317449704216e-05,
+      "loss": 1.2838,
+      "step": 10364
+    },
+    {
+      "epoch": 0.7471347221221077,
+      "grad_norm": 0.12026566512082669,
+      "learning_rate": 3.170611811218651e-05,
+      "loss": 1.2533,
+      "step": 10365
+    },
+    {
+      "epoch": 0.7472068045844446,
+      "grad_norm": 0.11942873758819407,
+      "learning_rate": 3.16890654499584e-05,
+      "loss": 1.2938,
+      "step": 10366
+    },
+    {
+      "epoch": 0.7472788870467815,
+      "grad_norm": 0.11579534395132698,
+      "learning_rate": 3.167201651128735e-05,
+      "loss": 1.1608,
+      "step": 10367
+    },
+    {
+      "epoch": 0.7473509695091184,
+      "grad_norm": 0.12521253620541925,
+      "learning_rate": 3.165497129710269e-05,
+      "loss": 1.2672,
+      "step": 10368
+    },
+    {
+      "epoch": 0.7474230519714553,
+      "grad_norm": 0.13180855365353267,
+      "learning_rate": 3.163792980833351e-05,
+      "loss": 1.281,
+      "step": 10369
+    },
+    {
+      "epoch": 0.7474951344337922,
+      "grad_norm": 0.1116653215865103,
+      "learning_rate": 3.162089204590878e-05,
+      "loss": 1.259,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7475672168961291,
+      "grad_norm": 0.12323795883169414,
+      "learning_rate": 3.1603858010757104e-05,
+      "loss": 1.3058,
+      "step": 10371
+    },
+    {
+      "epoch": 0.7476392993584661,
+      "grad_norm": 0.1524599357737869,
+      "learning_rate": 3.158682770380709e-05,
+      "loss": 1.2719,
+      "step": 10372
+    },
+    {
+      "epoch": 0.747711381820803,
+      "grad_norm": 0.1404989527935165,
+      "learning_rate": 3.156980112598701e-05,
+      "loss": 1.1987,
+      "step": 10373
+    },
+    {
+      "epoch": 0.7477834642831399,
+      "grad_norm": 0.12465013669591961,
+      "learning_rate": 3.155277827822498e-05,
+      "loss": 1.2977,
+      "step": 10374
+    },
+    {
+      "epoch": 0.7478555467454768,
+      "grad_norm": 0.1261580034554153,
+      "learning_rate": 3.153575916144883e-05,
+      "loss": 1.2699,
+      "step": 10375
+    },
+    {
+      "epoch": 0.7479276292078137,
+      "grad_norm": 0.11856752012585758,
+      "learning_rate": 3.151874377658633e-05,
+      "loss": 1.2902,
+      "step": 10376
+    },
+    {
+      "epoch": 0.7479997116701507,
+      "grad_norm": 0.11676994279314871,
+      "learning_rate": 3.1501732124564954e-05,
+      "loss": 1.1977,
+      "step": 10377
+    },
+    {
+      "epoch": 0.7480717941324876,
+      "grad_norm": 0.12463183395800827,
+      "learning_rate": 3.148472420631201e-05,
+      "loss": 1.3842,
+      "step": 10378
+    },
+    {
+      "epoch": 0.7481438765948245,
+      "grad_norm": 0.12866178294892713,
+      "learning_rate": 3.146772002275452e-05,
+      "loss": 1.1575,
+      "step": 10379
+    },
+    {
+      "epoch": 0.7482159590571614,
+      "grad_norm": 0.13123199693097515,
+      "learning_rate": 3.145071957481935e-05,
+      "loss": 1.3875,
+      "step": 10380
+    },
+    {
+      "epoch": 0.7482880415194983,
+      "grad_norm": 0.12155190403673263,
+      "learning_rate": 3.143372286343327e-05,
+      "loss": 1.2392,
+      "step": 10381
+    },
+    {
+      "epoch": 0.7483601239818353,
+      "grad_norm": 0.12360984877065305,
+      "learning_rate": 3.1416729889522735e-05,
+      "loss": 1.3053,
+      "step": 10382
+    },
+    {
+      "epoch": 0.7484322064441722,
+      "grad_norm": 0.11752332368426252,
+      "learning_rate": 3.139974065401395e-05,
+      "loss": 1.2023,
+      "step": 10383
+    },
+    {
+      "epoch": 0.748504288906509,
+      "grad_norm": 0.11835848302006537,
+      "learning_rate": 3.1382755157832986e-05,
+      "loss": 1.3864,
+      "step": 10384
+    },
+    {
+      "epoch": 0.7485763713688459,
+      "grad_norm": 0.11467928877404406,
+      "learning_rate": 3.136577340190576e-05,
+      "loss": 1.2056,
+      "step": 10385
+    },
+    {
+      "epoch": 0.7486484538311828,
+      "grad_norm": 0.124445130248809,
+      "learning_rate": 3.134879538715794e-05,
+      "loss": 1.2147,
+      "step": 10386
+    },
+    {
+      "epoch": 0.7487205362935198,
+      "grad_norm": 0.11975659153551176,
+      "learning_rate": 3.1331821114514905e-05,
+      "loss": 1.2143,
+      "step": 10387
+    },
+    {
+      "epoch": 0.7487926187558567,
+      "grad_norm": 0.1483853335652019,
+      "learning_rate": 3.131485058490191e-05,
+      "loss": 1.2286,
+      "step": 10388
+    },
+    {
+      "epoch": 0.7488647012181936,
+      "grad_norm": 0.12303765116701582,
+      "learning_rate": 3.129788379924408e-05,
+      "loss": 1.2545,
+      "step": 10389
+    },
+    {
+      "epoch": 0.7489367836805305,
+      "grad_norm": 0.1209357660672735,
+      "learning_rate": 3.128092075846623e-05,
+      "loss": 1.2487,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7490088661428674,
+      "grad_norm": 0.11508201993789575,
+      "learning_rate": 3.126396146349294e-05,
+      "loss": 1.2798,
+      "step": 10391
+    },
+    {
+      "epoch": 0.7490809486052044,
+      "grad_norm": 0.12182145461432715,
+      "learning_rate": 3.1247005915248686e-05,
+      "loss": 1.223,
+      "step": 10392
+    },
+    {
+      "epoch": 0.7491530310675413,
+      "grad_norm": 0.13143355246798386,
+      "learning_rate": 3.123005411465766e-05,
+      "loss": 1.2477,
+      "step": 10393
+    },
+    {
+      "epoch": 0.7492251135298782,
+      "grad_norm": 0.11652717999402826,
+      "learning_rate": 3.121310606264398e-05,
+      "loss": 1.3029,
+      "step": 10394
+    },
+    {
+      "epoch": 0.7492971959922151,
+      "grad_norm": 0.11352816257204369,
+      "learning_rate": 3.119616176013137e-05,
+      "loss": 1.2887,
+      "step": 10395
+    },
+    {
+      "epoch": 0.749369278454552,
+      "grad_norm": 0.11220701058686422,
+      "learning_rate": 3.117922120804349e-05,
+      "loss": 1.2056,
+      "step": 10396
+    },
+    {
+      "epoch": 0.749441360916889,
+      "grad_norm": 0.12457410268866137,
+      "learning_rate": 3.11622844073037e-05,
+      "loss": 1.313,
+      "step": 10397
+    },
+    {
+      "epoch": 0.7495134433792259,
+      "grad_norm": 0.11984508340997746,
+      "learning_rate": 3.114535135883532e-05,
+      "loss": 1.3066,
+      "step": 10398
+    },
+    {
+      "epoch": 0.7495855258415628,
+      "grad_norm": 0.11483682334308022,
+      "learning_rate": 3.112842206356125e-05,
+      "loss": 1.3204,
+      "step": 10399
+    },
+    {
+      "epoch": 0.7496576083038997,
+      "grad_norm": 0.11886005416298533,
+      "learning_rate": 3.1111496522404315e-05,
+      "loss": 1.2681,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7497296907662365,
+      "grad_norm": 0.13407066603550907,
+      "learning_rate": 3.109457473628712e-05,
+      "loss": 1.189,
+      "step": 10401
+    },
+    {
+      "epoch": 0.7498017732285734,
+      "grad_norm": 0.11672230976319305,
+      "learning_rate": 3.1077656706132054e-05,
+      "loss": 1.2806,
+      "step": 10402
+    },
+    {
+      "epoch": 0.7498738556909104,
+      "grad_norm": 0.1221230274991943,
+      "learning_rate": 3.106074243286128e-05,
+      "loss": 1.2339,
+      "step": 10403
+    },
+    {
+      "epoch": 0.7499459381532473,
+      "grad_norm": 0.12354305249601422,
+      "learning_rate": 3.1043831917396795e-05,
+      "loss": 1.2176,
+      "step": 10404
+    },
+    {
+      "epoch": 0.7500180206155842,
+      "grad_norm": 0.11744545570054171,
+      "learning_rate": 3.102692516066037e-05,
+      "loss": 1.3796,
+      "step": 10405
+    },
+    {
+      "epoch": 0.7500901030779211,
+      "grad_norm": 0.12599283814510132,
+      "learning_rate": 3.1010022163573585e-05,
+      "loss": 1.2685,
+      "step": 10406
+    },
+    {
+      "epoch": 0.750162185540258,
+      "grad_norm": 0.14694463156506093,
+      "learning_rate": 3.099312292705778e-05,
+      "loss": 1.3183,
+      "step": 10407
+    },
+    {
+      "epoch": 0.750234268002595,
+      "grad_norm": 0.118607413216093,
+      "learning_rate": 3.097622745203413e-05,
+      "loss": 1.2695,
+      "step": 10408
+    },
+    {
+      "epoch": 0.7503063504649319,
+      "grad_norm": 0.12013238521083584,
+      "learning_rate": 3.09593357394236e-05,
+      "loss": 1.2769,
+      "step": 10409
+    },
+    {
+      "epoch": 0.7503784329272688,
+      "grad_norm": 0.10742266080494221,
+      "learning_rate": 3.0942447790146925e-05,
+      "loss": 1.2807,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7504505153896057,
+      "grad_norm": 0.11131654248330301,
+      "learning_rate": 3.092556360512464e-05,
+      "loss": 1.268,
+      "step": 10411
+    },
+    {
+      "epoch": 0.7505225978519426,
+      "grad_norm": 0.1132478666767687,
+      "learning_rate": 3.090868318527711e-05,
+      "loss": 1.1758,
+      "step": 10412
+    },
+    {
+      "epoch": 0.7505946803142796,
+      "grad_norm": 0.11154194976305341,
+      "learning_rate": 3.089180653152445e-05,
+      "loss": 1.2289,
+      "step": 10413
+    },
+    {
+      "epoch": 0.7506667627766165,
+      "grad_norm": 0.11068919204584095,
+      "learning_rate": 3.087493364478659e-05,
+      "loss": 1.2099,
+      "step": 10414
+    },
+    {
+      "epoch": 0.7507388452389534,
+      "grad_norm": 0.11864861857661776,
+      "learning_rate": 3.085806452598325e-05,
+      "loss": 1.3059,
+      "step": 10415
+    },
+    {
+      "epoch": 0.7508109277012903,
+      "grad_norm": 0.11025351523715934,
+      "learning_rate": 3.0841199176033974e-05,
+      "loss": 1.2945,
+      "step": 10416
+    },
+    {
+      "epoch": 0.7508830101636272,
+      "grad_norm": 0.11444539462125432,
+      "learning_rate": 3.082433759585804e-05,
+      "loss": 1.199,
+      "step": 10417
+    },
+    {
+      "epoch": 0.750955092625964,
+      "grad_norm": 0.11638361707238688,
+      "learning_rate": 3.080747978637457e-05,
+      "loss": 1.2653,
+      "step": 10418
+    },
+    {
+      "epoch": 0.751027175088301,
+      "grad_norm": 0.11127582833677284,
+      "learning_rate": 3.0790625748502475e-05,
+      "loss": 1.2394,
+      "step": 10419
+    },
+    {
+      "epoch": 0.7510992575506379,
+      "grad_norm": 0.1176703825280969,
+      "learning_rate": 3.077377548316044e-05,
+      "loss": 1.2132,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7511713400129748,
+      "grad_norm": 0.14004418935872592,
+      "learning_rate": 3.075692899126695e-05,
+      "loss": 1.1803,
+      "step": 10421
+    },
+    {
+      "epoch": 0.7512434224753117,
+      "grad_norm": 0.11399214204154622,
+      "learning_rate": 3.0740086273740295e-05,
+      "loss": 1.2458,
+      "step": 10422
+    },
+    {
+      "epoch": 0.7513155049376486,
+      "grad_norm": 0.11955948018151515,
+      "learning_rate": 3.072324733149861e-05,
+      "loss": 1.2471,
+      "step": 10423
+    },
+    {
+      "epoch": 0.7513875873999856,
+      "grad_norm": 0.11047164427818966,
+      "learning_rate": 3.070641216545962e-05,
+      "loss": 1.2892,
+      "step": 10424
+    },
+    {
+      "epoch": 0.7514596698623225,
+      "grad_norm": 0.11030319045483193,
+      "learning_rate": 3.068958077654113e-05,
+      "loss": 1.2587,
+      "step": 10425
+    },
+    {
+      "epoch": 0.7515317523246594,
+      "grad_norm": 0.10804263989416381,
+      "learning_rate": 3.067275316566055e-05,
+      "loss": 1.2769,
+      "step": 10426
+    },
+    {
+      "epoch": 0.7516038347869963,
+      "grad_norm": 0.1228444898615077,
+      "learning_rate": 3.065592933373518e-05,
+      "loss": 1.3418,
+      "step": 10427
+    },
+    {
+      "epoch": 0.7516759172493332,
+      "grad_norm": 0.13119859694646852,
+      "learning_rate": 3.0639109281681966e-05,
+      "loss": 1.2788,
+      "step": 10428
+    },
+    {
+      "epoch": 0.7517479997116702,
+      "grad_norm": 0.1266357957041589,
+      "learning_rate": 3.062229301041785e-05,
+      "loss": 1.301,
+      "step": 10429
+    },
+    {
+      "epoch": 0.7518200821740071,
+      "grad_norm": 0.12968856298711445,
+      "learning_rate": 3.0605480520859433e-05,
+      "loss": 1.3247,
+      "step": 10430
+    },
+    {
+      "epoch": 0.751892164636344,
+      "grad_norm": 0.12960817772503058,
+      "learning_rate": 3.058867181392319e-05,
+      "loss": 1.3025,
+      "step": 10431
+    },
+    {
+      "epoch": 0.7519642470986809,
+      "grad_norm": 0.12175508586375933,
+      "learning_rate": 3.057186689052523e-05,
+      "loss": 1.1548,
+      "step": 10432
+    },
+    {
+      "epoch": 0.7520363295610178,
+      "grad_norm": 0.1181752777038379,
+      "learning_rate": 3.055506575158168e-05,
+      "loss": 1.2611,
+      "step": 10433
+    },
+    {
+      "epoch": 0.7521084120233548,
+      "grad_norm": 0.11191657427544618,
+      "learning_rate": 3.053826839800834e-05,
+      "loss": 1.1591,
+      "step": 10434
+    },
+    {
+      "epoch": 0.7521804944856917,
+      "grad_norm": 0.1340680859600284,
+      "learning_rate": 3.052147483072078e-05,
+      "loss": 1.2933,
+      "step": 10435
+    },
+    {
+      "epoch": 0.7522525769480285,
+      "grad_norm": 0.1106907965223477,
+      "learning_rate": 3.050468505063441e-05,
+      "loss": 1.1875,
+      "step": 10436
+    },
+    {
+      "epoch": 0.7523246594103654,
+      "grad_norm": 0.11817845035526947,
+      "learning_rate": 3.0487899058664395e-05,
+      "loss": 1.3038,
+      "step": 10437
+    },
+    {
+      "epoch": 0.7523967418727023,
+      "grad_norm": 0.1535269381572771,
+      "learning_rate": 3.0471116855725813e-05,
+      "loss": 1.2179,
+      "step": 10438
+    },
+    {
+      "epoch": 0.7524688243350393,
+      "grad_norm": 0.1150267571664176,
+      "learning_rate": 3.0454338442733355e-05,
+      "loss": 1.3313,
+      "step": 10439
+    },
+    {
+      "epoch": 0.7525409067973762,
+      "grad_norm": 0.12511451019811237,
+      "learning_rate": 3.0437563820601633e-05,
+      "loss": 1.2262,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7526129892597131,
+      "grad_norm": 0.12512894572887379,
+      "learning_rate": 3.042079299024496e-05,
+      "loss": 1.1896,
+      "step": 10441
+    },
+    {
+      "epoch": 0.75268507172205,
+      "grad_norm": 0.12186966016271107,
+      "learning_rate": 3.040402595257762e-05,
+      "loss": 1.274,
+      "step": 10442
+    },
+    {
+      "epoch": 0.7527571541843869,
+      "grad_norm": 0.12089038905149929,
+      "learning_rate": 3.0387262708513454e-05,
+      "loss": 1.1879,
+      "step": 10443
+    },
+    {
+      "epoch": 0.7528292366467239,
+      "grad_norm": 0.11840590666448862,
+      "learning_rate": 3.0370503258966244e-05,
+      "loss": 1.202,
+      "step": 10444
+    },
+    {
+      "epoch": 0.7529013191090608,
+      "grad_norm": 0.12067442829612401,
+      "learning_rate": 3.0353747604849527e-05,
+      "loss": 1.2465,
+      "step": 10445
+    },
+    {
+      "epoch": 0.7529734015713977,
+      "grad_norm": 0.185624212230686,
+      "learning_rate": 3.0336995747076636e-05,
+      "loss": 1.2159,
+      "step": 10446
+    },
+    {
+      "epoch": 0.7530454840337346,
+      "grad_norm": 0.10289613143245693,
+      "learning_rate": 3.03202476865607e-05,
+      "loss": 1.2803,
+      "step": 10447
+    },
+    {
+      "epoch": 0.7531175664960715,
+      "grad_norm": 0.12017591133951692,
+      "learning_rate": 3.0303503424214643e-05,
+      "loss": 1.2208,
+      "step": 10448
+    },
+    {
+      "epoch": 0.7531896489584085,
+      "grad_norm": 0.1208402898735493,
+      "learning_rate": 3.0286762960951166e-05,
+      "loss": 1.3034,
+      "step": 10449
+    },
+    {
+      "epoch": 0.7532617314207454,
+      "grad_norm": 0.11767530024441607,
+      "learning_rate": 3.0270026297682785e-05,
+      "loss": 1.1742,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7533338138830823,
+      "grad_norm": 0.13442722034902085,
+      "learning_rate": 3.0253293435321793e-05,
+      "loss": 1.2462,
+      "step": 10451
+    },
+    {
+      "epoch": 0.7534058963454192,
+      "grad_norm": 0.1155231229446613,
+      "learning_rate": 3.0236564374780284e-05,
+      "loss": 1.2441,
+      "step": 10452
+    },
+    {
+      "epoch": 0.753477978807756,
+      "grad_norm": 0.10532950674119541,
+      "learning_rate": 3.021983911697013e-05,
+      "loss": 1.1965,
+      "step": 10453
+    },
+    {
+      "epoch": 0.7535500612700929,
+      "grad_norm": 0.11391352350602109,
+      "learning_rate": 3.0203117662803016e-05,
+      "loss": 1.2089,
+      "step": 10454
+    },
+    {
+      "epoch": 0.7536221437324299,
+      "grad_norm": 0.14674074263569228,
+      "learning_rate": 3.018640001319042e-05,
+      "loss": 1.1805,
+      "step": 10455
+    },
+    {
+      "epoch": 0.7536942261947668,
+      "grad_norm": 0.11969036120428096,
+      "learning_rate": 3.0169686169043587e-05,
+      "loss": 1.2862,
+      "step": 10456
+    },
+    {
+      "epoch": 0.7537663086571037,
+      "grad_norm": 0.12500911908215645,
+      "learning_rate": 3.0152976131273568e-05,
+      "loss": 1.1539,
+      "step": 10457
+    },
+    {
+      "epoch": 0.7538383911194406,
+      "grad_norm": 0.23827514239856065,
+      "learning_rate": 3.0136269900791235e-05,
+      "loss": 1.2986,
+      "step": 10458
+    },
+    {
+      "epoch": 0.7539104735817775,
+      "grad_norm": 0.12484955164860832,
+      "learning_rate": 3.0119567478507206e-05,
+      "loss": 1.2653,
+      "step": 10459
+    },
+    {
+      "epoch": 0.7539825560441145,
+      "grad_norm": 0.11695523392680528,
+      "learning_rate": 3.0102868865331914e-05,
+      "loss": 1.2199,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7540546385064514,
+      "grad_norm": 0.12087979701704878,
+      "learning_rate": 3.008617406217559e-05,
+      "loss": 1.2247,
+      "step": 10461
+    },
+    {
+      "epoch": 0.7541267209687883,
+      "grad_norm": 0.11761115731835646,
+      "learning_rate": 3.0069483069948246e-05,
+      "loss": 1.2878,
+      "step": 10462
+    },
+    {
+      "epoch": 0.7541988034311252,
+      "grad_norm": 0.11223890953914417,
+      "learning_rate": 3.0052795889559694e-05,
+      "loss": 1.2101,
+      "step": 10463
+    },
+    {
+      "epoch": 0.7542708858934621,
+      "grad_norm": 0.12442941166922876,
+      "learning_rate": 3.0036112521919525e-05,
+      "loss": 1.2573,
+      "step": 10464
+    },
+    {
+      "epoch": 0.7543429683557991,
+      "grad_norm": 0.11372644640697349,
+      "learning_rate": 3.0019432967937143e-05,
+      "loss": 1.2745,
+      "step": 10465
+    },
+    {
+      "epoch": 0.754415050818136,
+      "grad_norm": 0.11840755128571843,
+      "learning_rate": 3.0002757228521726e-05,
+      "loss": 1.2954,
+      "step": 10466
+    },
+    {
+      "epoch": 0.7544871332804729,
+      "grad_norm": 0.11775769646531722,
+      "learning_rate": 2.998608530458228e-05,
+      "loss": 1.2872,
+      "step": 10467
+    },
+    {
+      "epoch": 0.7545592157428098,
+      "grad_norm": 0.11513617846685036,
+      "learning_rate": 2.9969417197027495e-05,
+      "loss": 1.2752,
+      "step": 10468
+    },
+    {
+      "epoch": 0.7546312982051467,
+      "grad_norm": 0.12211321756514945,
+      "learning_rate": 2.9952752906766013e-05,
+      "loss": 1.2658,
+      "step": 10469
+    },
+    {
+      "epoch": 0.7547033806674835,
+      "grad_norm": 0.11869696209017304,
+      "learning_rate": 2.9936092434706153e-05,
+      "loss": 1.0949,
+      "step": 10470
+    },
+    {
+      "epoch": 0.7547754631298205,
+      "grad_norm": 0.11440724684338935,
+      "learning_rate": 2.99194357817561e-05,
+      "loss": 1.1924,
+      "step": 10471
+    },
+    {
+      "epoch": 0.7548475455921574,
+      "grad_norm": 0.1560928828131219,
+      "learning_rate": 2.9902782948823694e-05,
+      "loss": 1.2464,
+      "step": 10472
+    },
+    {
+      "epoch": 0.7549196280544943,
+      "grad_norm": 0.12255214826187488,
+      "learning_rate": 2.988613393681675e-05,
+      "loss": 1.3081,
+      "step": 10473
+    },
+    {
+      "epoch": 0.7549917105168312,
+      "grad_norm": 0.1111977253499952,
+      "learning_rate": 2.986948874664276e-05,
+      "loss": 1.1972,
+      "step": 10474
+    },
+    {
+      "epoch": 0.7550637929791681,
+      "grad_norm": 0.13536180015434007,
+      "learning_rate": 2.985284737920908e-05,
+      "loss": 1.2482,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7551358754415051,
+      "grad_norm": 0.11117458374673317,
+      "learning_rate": 2.9836209835422713e-05,
+      "loss": 1.3984,
+      "step": 10476
+    },
+    {
+      "epoch": 0.755207957903842,
+      "grad_norm": 0.11920024190677067,
+      "learning_rate": 2.981957611619063e-05,
+      "loss": 1.3334,
+      "step": 10477
+    },
+    {
+      "epoch": 0.7552800403661789,
+      "grad_norm": 0.10806082106177858,
+      "learning_rate": 2.9802946222419513e-05,
+      "loss": 1.2649,
+      "step": 10478
+    },
+    {
+      "epoch": 0.7553521228285158,
+      "grad_norm": 0.11112294502057257,
+      "learning_rate": 2.9786320155015857e-05,
+      "loss": 1.2761,
+      "step": 10479
+    },
+    {
+      "epoch": 0.7554242052908527,
+      "grad_norm": 0.12503881171216166,
+      "learning_rate": 2.9769697914885862e-05,
+      "loss": 1.2384,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7554962877531897,
+      "grad_norm": 0.12006599796425284,
+      "learning_rate": 2.975307950293561e-05,
+      "loss": 1.2989,
+      "step": 10481
+    },
+    {
+      "epoch": 0.7555683702155266,
+      "grad_norm": 0.13717822802900614,
+      "learning_rate": 2.9736464920070993e-05,
+      "loss": 1.2005,
+      "step": 10482
+    },
+    {
+      "epoch": 0.7556404526778635,
+      "grad_norm": 0.10920774743943465,
+      "learning_rate": 2.971985416719767e-05,
+      "loss": 1.2496,
+      "step": 10483
+    },
+    {
+      "epoch": 0.7557125351402004,
+      "grad_norm": 0.11354162698326335,
+      "learning_rate": 2.970324724522101e-05,
+      "loss": 1.3329,
+      "step": 10484
+    },
+    {
+      "epoch": 0.7557846176025373,
+      "grad_norm": 0.12624195630750193,
+      "learning_rate": 2.968664415504623e-05,
+      "loss": 1.285,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7558567000648743,
+      "grad_norm": 0.11401199167141232,
+      "learning_rate": 2.9670044897578418e-05,
+      "loss": 1.3317,
+      "step": 10486
+    },
+    {
+      "epoch": 0.7559287825272112,
+      "grad_norm": 0.11397951361435692,
+      "learning_rate": 2.9653449473722383e-05,
+      "loss": 1.2287,
+      "step": 10487
+    },
+    {
+      "epoch": 0.756000864989548,
+      "grad_norm": 0.11349995896815283,
+      "learning_rate": 2.963685788438266e-05,
+      "loss": 1.3084,
+      "step": 10488
+    },
+    {
+      "epoch": 0.7560729474518849,
+      "grad_norm": 0.12352659000730208,
+      "learning_rate": 2.9620270130463667e-05,
+      "loss": 1.2904,
+      "step": 10489
+    },
+    {
+      "epoch": 0.7561450299142218,
+      "grad_norm": 0.12413936049533815,
+      "learning_rate": 2.9603686212869552e-05,
+      "loss": 1.3444,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7562171123765588,
+      "grad_norm": 0.1379207721882299,
+      "learning_rate": 2.9587106132504393e-05,
+      "loss": 1.2539,
+      "step": 10491
+    },
+    {
+      "epoch": 0.7562891948388957,
+      "grad_norm": 0.12068362399929242,
+      "learning_rate": 2.9570529890271835e-05,
+      "loss": 1.3248,
+      "step": 10492
+    },
+    {
+      "epoch": 0.7563612773012326,
+      "grad_norm": 0.11336613597945243,
+      "learning_rate": 2.955395748707549e-05,
+      "loss": 1.2602,
+      "step": 10493
+    },
+    {
+      "epoch": 0.7564333597635695,
+      "grad_norm": 0.11545475346121978,
+      "learning_rate": 2.9537388923818656e-05,
+      "loss": 1.275,
+      "step": 10494
+    },
+    {
+      "epoch": 0.7565054422259064,
+      "grad_norm": 0.1103325010272704,
+      "learning_rate": 2.9520824201404573e-05,
+      "loss": 1.3297,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7565775246882434,
+      "grad_norm": 0.11959815549334278,
+      "learning_rate": 2.9504263320736046e-05,
+      "loss": 1.3502,
+      "step": 10496
+    },
+    {
+      "epoch": 0.7566496071505803,
+      "grad_norm": 0.11939317046171469,
+      "learning_rate": 2.9487706282715856e-05,
+      "loss": 1.3307,
+      "step": 10497
+    },
+    {
+      "epoch": 0.7567216896129172,
+      "grad_norm": 0.125498439341664,
+      "learning_rate": 2.9471153088246462e-05,
+      "loss": 1.4084,
+      "step": 10498
+    },
+    {
+      "epoch": 0.7567937720752541,
+      "grad_norm": 0.11341261631859123,
+      "learning_rate": 2.9454603738230256e-05,
+      "loss": 1.2153,
+      "step": 10499
+    },
+    {
+      "epoch": 0.756865854537591,
+      "grad_norm": 0.11483768294175102,
+      "learning_rate": 2.943805823356922e-05,
+      "loss": 1.2287,
+      "step": 10500
+    },
+    {
+      "epoch": 0.756937936999928,
+      "grad_norm": 0.1137700074378297,
+      "learning_rate": 2.9421516575165288e-05,
+      "loss": 1.2327,
+      "step": 10501
+    },
+    {
+      "epoch": 0.7570100194622649,
+      "grad_norm": 0.11397218103333982,
+      "learning_rate": 2.940497876392011e-05,
+      "loss": 1.2113,
+      "step": 10502
+    },
+    {
+      "epoch": 0.7570821019246018,
+      "grad_norm": 0.12290312581435264,
+      "learning_rate": 2.938844480073516e-05,
+      "loss": 1.3105,
+      "step": 10503
+    },
+    {
+      "epoch": 0.7571541843869387,
+      "grad_norm": 0.12440801174069012,
+      "learning_rate": 2.937191468651167e-05,
+      "loss": 1.2984,
+      "step": 10504
+    },
+    {
+      "epoch": 0.7572262668492755,
+      "grad_norm": 0.11244632939467106,
+      "learning_rate": 2.9355388422150688e-05,
+      "loss": 1.2238,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7572983493116124,
+      "grad_norm": 0.11852605373807315,
+      "learning_rate": 2.9338866008553035e-05,
+      "loss": 1.1989,
+      "step": 10506
+    },
+    {
+      "epoch": 0.7573704317739494,
+      "grad_norm": 0.12512846445034603,
+      "learning_rate": 2.9322347446619345e-05,
+      "loss": 1.2663,
+      "step": 10507
+    },
+    {
+      "epoch": 0.7574425142362863,
+      "grad_norm": 0.12675516939987938,
+      "learning_rate": 2.9305832737250017e-05,
+      "loss": 1.2223,
+      "step": 10508
+    },
+    {
+      "epoch": 0.7575145966986232,
+      "grad_norm": 0.12289422126083027,
+      "learning_rate": 2.9289321881345254e-05,
+      "loss": 1.2676,
+      "step": 10509
+    },
+    {
+      "epoch": 0.7575866791609601,
+      "grad_norm": 0.12363892173778984,
+      "learning_rate": 2.9272814879805045e-05,
+      "loss": 1.2788,
+      "step": 10510
+    },
+    {
+      "epoch": 0.757658761623297,
+      "grad_norm": 0.12532648941768967,
+      "learning_rate": 2.9256311733529163e-05,
+      "loss": 1.3196,
+      "step": 10511
+    },
+    {
+      "epoch": 0.757730844085634,
+      "grad_norm": 0.14608746686430402,
+      "learning_rate": 2.923981244341719e-05,
+      "loss": 1.3353,
+      "step": 10512
+    },
+    {
+      "epoch": 0.7578029265479709,
+      "grad_norm": 0.11991668671564884,
+      "learning_rate": 2.9223317010368478e-05,
+      "loss": 1.2796,
+      "step": 10513
+    },
+    {
+      "epoch": 0.7578750090103078,
+      "grad_norm": 0.12123225471503221,
+      "learning_rate": 2.920682543528217e-05,
+      "loss": 1.2627,
+      "step": 10514
+    },
+    {
+      "epoch": 0.7579470914726447,
+      "grad_norm": 0.1224896115197127,
+      "learning_rate": 2.9190337719057215e-05,
+      "loss": 1.2804,
+      "step": 10515
+    },
+    {
+      "epoch": 0.7580191739349816,
+      "grad_norm": 0.12830200671212827,
+      "learning_rate": 2.9173853862592326e-05,
+      "loss": 1.281,
+      "step": 10516
+    },
+    {
+      "epoch": 0.7580912563973186,
+      "grad_norm": 0.12451070733502229,
+      "learning_rate": 2.9157373866786043e-05,
+      "loss": 1.2911,
+      "step": 10517
+    },
+    {
+      "epoch": 0.7581633388596555,
+      "grad_norm": 0.11836567299328801,
+      "learning_rate": 2.9140897732536643e-05,
+      "loss": 1.3571,
+      "step": 10518
+    },
+    {
+      "epoch": 0.7582354213219924,
+      "grad_norm": 0.1368791833559967,
+      "learning_rate": 2.9124425460742254e-05,
+      "loss": 1.2699,
+      "step": 10519
+    },
+    {
+      "epoch": 0.7583075037843293,
+      "grad_norm": 0.11829039549209128,
+      "learning_rate": 2.9107957052300783e-05,
+      "loss": 1.2622,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7583795862466662,
+      "grad_norm": 0.11772034835444904,
+      "learning_rate": 2.9091492508109795e-05,
+      "loss": 1.2513,
+      "step": 10521
+    },
+    {
+      "epoch": 0.758451668709003,
+      "grad_norm": 0.12314989532736216,
+      "learning_rate": 2.9075031829066867e-05,
+      "loss": 1.3068,
+      "step": 10522
+    },
+    {
+      "epoch": 0.75852375117134,
+      "grad_norm": 0.1124880314048645,
+      "learning_rate": 2.9058575016069223e-05,
+      "loss": 1.3433,
+      "step": 10523
+    },
+    {
+      "epoch": 0.7585958336336769,
+      "grad_norm": 0.136947894800872,
+      "learning_rate": 2.9042122070013932e-05,
+      "loss": 1.3078,
+      "step": 10524
+    },
+    {
+      "epoch": 0.7586679160960138,
+      "grad_norm": 0.12029915955913807,
+      "learning_rate": 2.9025672991797724e-05,
+      "loss": 1.2633,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7587399985583507,
+      "grad_norm": 0.12140434456697287,
+      "learning_rate": 2.9009227782317327e-05,
+      "loss": 1.2778,
+      "step": 10526
+    },
+    {
+      "epoch": 0.7588120810206876,
+      "grad_norm": 0.11969080130818249,
+      "learning_rate": 2.8992786442469123e-05,
+      "loss": 1.2429,
+      "step": 10527
+    },
+    {
+      "epoch": 0.7588841634830246,
+      "grad_norm": 0.11294279260164564,
+      "learning_rate": 2.8976348973149337e-05,
+      "loss": 1.2799,
+      "step": 10528
+    },
+    {
+      "epoch": 0.7589562459453615,
+      "grad_norm": 0.12310889548783494,
+      "learning_rate": 2.8959915375253866e-05,
+      "loss": 1.3616,
+      "step": 10529
+    },
+    {
+      "epoch": 0.7590283284076984,
+      "grad_norm": 0.14048296418422584,
+      "learning_rate": 2.894348564967857e-05,
+      "loss": 1.2878,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7591004108700353,
+      "grad_norm": 0.11786450083940546,
+      "learning_rate": 2.8927059797319002e-05,
+      "loss": 1.2243,
+      "step": 10531
+    },
+    {
+      "epoch": 0.7591724933323722,
+      "grad_norm": 0.1454429155167535,
+      "learning_rate": 2.8910637819070552e-05,
+      "loss": 1.1872,
+      "step": 10532
+    },
+    {
+      "epoch": 0.7592445757947092,
+      "grad_norm": 0.1157152584566582,
+      "learning_rate": 2.8894219715828286e-05,
+      "loss": 1.3334,
+      "step": 10533
+    },
+    {
+      "epoch": 0.7593166582570461,
+      "grad_norm": 0.12264205254645491,
+      "learning_rate": 2.8877805488487143e-05,
+      "loss": 1.1858,
+      "step": 10534
+    },
+    {
+      "epoch": 0.759388740719383,
+      "grad_norm": 0.12037908867007201,
+      "learning_rate": 2.8861395137941905e-05,
+      "loss": 1.2363,
+      "step": 10535
+    },
+    {
+      "epoch": 0.7594608231817199,
+      "grad_norm": 0.11423901701056603,
+      "learning_rate": 2.8844988665087092e-05,
+      "loss": 1.3039,
+      "step": 10536
+    },
+    {
+      "epoch": 0.7595329056440568,
+      "grad_norm": 0.1163975600324938,
+      "learning_rate": 2.8828586070816944e-05,
+      "loss": 1.3676,
+      "step": 10537
+    },
+    {
+      "epoch": 0.7596049881063938,
+      "grad_norm": 0.12218117962253121,
+      "learning_rate": 2.881218735602553e-05,
+      "loss": 1.3297,
+      "step": 10538
+    },
+    {
+      "epoch": 0.7596770705687306,
+      "grad_norm": 0.1275380557505943,
+      "learning_rate": 2.8795792521606814e-05,
+      "loss": 1.2526,
+      "step": 10539
+    },
+    {
+      "epoch": 0.7597491530310675,
+      "grad_norm": 0.1395722241314583,
+      "learning_rate": 2.8779401568454433e-05,
+      "loss": 1.3809,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7598212354934044,
+      "grad_norm": 0.11798779404706315,
+      "learning_rate": 2.8763014497461803e-05,
+      "loss": 1.3738,
+      "step": 10541
+    },
+    {
+      "epoch": 0.7598933179557413,
+      "grad_norm": 0.11693377890811452,
+      "learning_rate": 2.8746631309522175e-05,
+      "loss": 1.2922,
+      "step": 10542
+    },
+    {
+      "epoch": 0.7599654004180783,
+      "grad_norm": 0.11774872736598724,
+      "learning_rate": 2.873025200552858e-05,
+      "loss": 1.1954,
+      "step": 10543
+    },
+    {
+      "epoch": 0.7600374828804152,
+      "grad_norm": 0.11288930859318269,
+      "learning_rate": 2.87138765863739e-05,
+      "loss": 1.3478,
+      "step": 10544
+    },
+    {
+      "epoch": 0.7601095653427521,
+      "grad_norm": 0.11025166939526966,
+      "learning_rate": 2.8697505052950657e-05,
+      "loss": 1.2213,
+      "step": 10545
+    },
+    {
+      "epoch": 0.760181647805089,
+      "grad_norm": 0.12648544195755124,
+      "learning_rate": 2.8681137406151282e-05,
+      "loss": 1.2991,
+      "step": 10546
+    },
+    {
+      "epoch": 0.7602537302674259,
+      "grad_norm": 0.12352945413837385,
+      "learning_rate": 2.8664773646867915e-05,
+      "loss": 1.2601,
+      "step": 10547
+    },
+    {
+      "epoch": 0.7603258127297629,
+      "grad_norm": 0.12388246408632345,
+      "learning_rate": 2.8648413775992644e-05,
+      "loss": 1.2683,
+      "step": 10548
+    },
+    {
+      "epoch": 0.7603978951920998,
+      "grad_norm": 0.12178640246797837,
+      "learning_rate": 2.863205779441711e-05,
+      "loss": 1.238,
+      "step": 10549
+    },
+    {
+      "epoch": 0.7604699776544367,
+      "grad_norm": 0.1209138005845628,
+      "learning_rate": 2.8615705703032902e-05,
+      "loss": 1.2702,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7605420601167736,
+      "grad_norm": 0.12273078450856986,
+      "learning_rate": 2.859935750273133e-05,
+      "loss": 1.3148,
+      "step": 10551
+    },
+    {
+      "epoch": 0.7606141425791105,
+      "grad_norm": 0.1190869567653472,
+      "learning_rate": 2.858301319440361e-05,
+      "loss": 1.234,
+      "step": 10552
+    },
+    {
+      "epoch": 0.7606862250414475,
+      "grad_norm": 0.12934663297505025,
+      "learning_rate": 2.856667277894054e-05,
+      "loss": 1.328,
+      "step": 10553
+    },
+    {
+      "epoch": 0.7607583075037844,
+      "grad_norm": 0.13380993326375099,
+      "learning_rate": 2.855033625723287e-05,
+      "loss": 1.24,
+      "step": 10554
+    },
+    {
+      "epoch": 0.7608303899661213,
+      "grad_norm": 0.12080180740511905,
+      "learning_rate": 2.853400363017108e-05,
+      "loss": 1.2869,
+      "step": 10555
+    },
+    {
+      "epoch": 0.7609024724284582,
+      "grad_norm": 0.11604147841060636,
+      "learning_rate": 2.8517674898645452e-05,
+      "loss": 1.3804,
+      "step": 10556
+    },
+    {
+      "epoch": 0.760974554890795,
+      "grad_norm": 0.13539530659070514,
+      "learning_rate": 2.850135006354604e-05,
+      "loss": 1.3325,
+      "step": 10557
+    },
+    {
+      "epoch": 0.7610466373531319,
+      "grad_norm": 0.117647152425111,
+      "learning_rate": 2.8485029125762687e-05,
+      "loss": 1.3645,
+      "step": 10558
+    },
+    {
+      "epoch": 0.7611187198154689,
+      "grad_norm": 0.10909317400870568,
+      "learning_rate": 2.8468712086185047e-05,
+      "loss": 1.199,
+      "step": 10559
+    },
+    {
+      "epoch": 0.7611908022778058,
+      "grad_norm": 0.11143887132564535,
+      "learning_rate": 2.845239894570253e-05,
+      "loss": 1.339,
+      "step": 10560
+    },
+    {
+      "epoch": 0.7612628847401427,
+      "grad_norm": 0.13440185381528277,
+      "learning_rate": 2.8436089705204348e-05,
+      "loss": 1.2524,
+      "step": 10561
+    },
+    {
+      "epoch": 0.7613349672024796,
+      "grad_norm": 0.11338572451227955,
+      "learning_rate": 2.8419784365579516e-05,
+      "loss": 1.3211,
+      "step": 10562
+    },
+    {
+      "epoch": 0.7614070496648165,
+      "grad_norm": 0.11898274013948397,
+      "learning_rate": 2.8403482927716807e-05,
+      "loss": 1.1797,
+      "step": 10563
+    },
+    {
+      "epoch": 0.7614791321271535,
+      "grad_norm": 0.12995612688672414,
+      "learning_rate": 2.83871853925048e-05,
+      "loss": 1.3529,
+      "step": 10564
+    },
+    {
+      "epoch": 0.7615512145894904,
+      "grad_norm": 0.18023814848425507,
+      "learning_rate": 2.8370891760831853e-05,
+      "loss": 1.242,
+      "step": 10565
+    },
+    {
+      "epoch": 0.7616232970518273,
+      "grad_norm": 0.12134288308672514,
+      "learning_rate": 2.835460203358612e-05,
+      "loss": 1.2217,
+      "step": 10566
+    },
+    {
+      "epoch": 0.7616953795141642,
+      "grad_norm": 0.11431926815128293,
+      "learning_rate": 2.8338316211655536e-05,
+      "loss": 1.1739,
+      "step": 10567
+    },
+    {
+      "epoch": 0.7617674619765011,
+      "grad_norm": 0.14039308319246688,
+      "learning_rate": 2.8322034295927823e-05,
+      "loss": 1.3177,
+      "step": 10568
+    },
+    {
+      "epoch": 0.7618395444388381,
+      "grad_norm": 0.12244805782046046,
+      "learning_rate": 2.8305756287290484e-05,
+      "loss": 1.3019,
+      "step": 10569
+    },
+    {
+      "epoch": 0.761911626901175,
+      "grad_norm": 0.11471175946204185,
+      "learning_rate": 2.8289482186630823e-05,
+      "loss": 1.305,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7619837093635119,
+      "grad_norm": 0.12053331271983175,
+      "learning_rate": 2.8273211994835923e-05,
+      "loss": 1.264,
+      "step": 10571
+    },
+    {
+      "epoch": 0.7620557918258488,
+      "grad_norm": 0.115452861808211,
+      "learning_rate": 2.825694571279266e-05,
+      "loss": 1.215,
+      "step": 10572
+    },
+    {
+      "epoch": 0.7621278742881857,
+      "grad_norm": 0.18071732488005793,
+      "learning_rate": 2.8240683341387687e-05,
+      "loss": 1.278,
+      "step": 10573
+    },
+    {
+      "epoch": 0.7621999567505225,
+      "grad_norm": 0.11398591418551207,
+      "learning_rate": 2.822442488150745e-05,
+      "loss": 1.3069,
+      "step": 10574
+    },
+    {
+      "epoch": 0.7622720392128595,
+      "grad_norm": 0.11938321927333505,
+      "learning_rate": 2.8208170334038176e-05,
+      "loss": 1.2162,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7623441216751964,
+      "grad_norm": 0.13461597365560526,
+      "learning_rate": 2.8191919699865888e-05,
+      "loss": 1.2219,
+      "step": 10576
+    },
+    {
+      "epoch": 0.7624162041375333,
+      "grad_norm": 0.11794436517428238,
+      "learning_rate": 2.8175672979876423e-05,
+      "loss": 1.1807,
+      "step": 10577
+    },
+    {
+      "epoch": 0.7624882865998702,
+      "grad_norm": 0.11887822889708816,
+      "learning_rate": 2.815943017495528e-05,
+      "loss": 1.2143,
+      "step": 10578
+    },
+    {
+      "epoch": 0.7625603690622071,
+      "grad_norm": 0.1313025667437936,
+      "learning_rate": 2.8143191285987926e-05,
+      "loss": 1.2049,
+      "step": 10579
+    },
+    {
+      "epoch": 0.7626324515245441,
+      "grad_norm": 0.11989373936459762,
+      "learning_rate": 2.8126956313859486e-05,
+      "loss": 1.3533,
+      "step": 10580
+    },
+    {
+      "epoch": 0.762704533986881,
+      "grad_norm": 0.11666314247367417,
+      "learning_rate": 2.811072525945496e-05,
+      "loss": 1.2779,
+      "step": 10581
+    },
+    {
+      "epoch": 0.7627766164492179,
+      "grad_norm": 0.12885477528901884,
+      "learning_rate": 2.8094498123658985e-05,
+      "loss": 1.3246,
+      "step": 10582
+    },
+    {
+      "epoch": 0.7628486989115548,
+      "grad_norm": 0.12076148281484075,
+      "learning_rate": 2.8078274907356185e-05,
+      "loss": 1.3116,
+      "step": 10583
+    },
+    {
+      "epoch": 0.7629207813738917,
+      "grad_norm": 0.11116733709417131,
+      "learning_rate": 2.8062055611430825e-05,
+      "loss": 1.1733,
+      "step": 10584
+    },
+    {
+      "epoch": 0.7629928638362287,
+      "grad_norm": 0.12003961604524803,
+      "learning_rate": 2.804584023676704e-05,
+      "loss": 1.1901,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7630649462985656,
+      "grad_norm": 0.12338658000180128,
+      "learning_rate": 2.8029628784248653e-05,
+      "loss": 1.2504,
+      "step": 10586
+    },
+    {
+      "epoch": 0.7631370287609025,
+      "grad_norm": 0.1178601308125348,
+      "learning_rate": 2.801342125475933e-05,
+      "loss": 1.2987,
+      "step": 10587
+    },
+    {
+      "epoch": 0.7632091112232394,
+      "grad_norm": 0.11008604883972184,
+      "learning_rate": 2.799721764918263e-05,
+      "loss": 1.2148,
+      "step": 10588
+    },
+    {
+      "epoch": 0.7632811936855763,
+      "grad_norm": 0.10232350587849366,
+      "learning_rate": 2.7981017968401678e-05,
+      "loss": 1.228,
+      "step": 10589
+    },
+    {
+      "epoch": 0.7633532761479133,
+      "grad_norm": 0.13017611989975483,
+      "learning_rate": 2.7964822213299556e-05,
+      "loss": 1.2451,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7634253586102501,
+      "grad_norm": 0.12847950144338144,
+      "learning_rate": 2.7948630384759035e-05,
+      "loss": 1.2999,
+      "step": 10591
+    },
+    {
+      "epoch": 0.763497441072587,
+      "grad_norm": 0.11903983674787035,
+      "learning_rate": 2.793244248366281e-05,
+      "loss": 1.2873,
+      "step": 10592
+    },
+    {
+      "epoch": 0.7635695235349239,
+      "grad_norm": 0.11947650929662203,
+      "learning_rate": 2.7916258510893168e-05,
+      "loss": 1.2751,
+      "step": 10593
+    },
+    {
+      "epoch": 0.7636416059972608,
+      "grad_norm": 0.1307285256417351,
+      "learning_rate": 2.7900078467332323e-05,
+      "loss": 1.2587,
+      "step": 10594
+    },
+    {
+      "epoch": 0.7637136884595978,
+      "grad_norm": 0.1255709053114885,
+      "learning_rate": 2.78839023538622e-05,
+      "loss": 1.3476,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7637857709219347,
+      "grad_norm": 0.137973338214756,
+      "learning_rate": 2.7867730171364624e-05,
+      "loss": 1.1975,
+      "step": 10596
+    },
+    {
+      "epoch": 0.7638578533842716,
+      "grad_norm": 0.10764384572688082,
+      "learning_rate": 2.7851561920721035e-05,
+      "loss": 1.3124,
+      "step": 10597
+    },
+    {
+      "epoch": 0.7639299358466085,
+      "grad_norm": 0.1295354005491379,
+      "learning_rate": 2.7835397602812786e-05,
+      "loss": 1.3187,
+      "step": 10598
+    },
+    {
+      "epoch": 0.7640020183089454,
+      "grad_norm": 0.11905851831994466,
+      "learning_rate": 2.7819237218520966e-05,
+      "loss": 1.1595,
+      "step": 10599
+    },
+    {
+      "epoch": 0.7640741007712823,
+      "grad_norm": 0.11190827844841084,
+      "learning_rate": 2.7803080768726462e-05,
+      "loss": 1.1065,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7641461832336193,
+      "grad_norm": 0.11587603953531542,
+      "learning_rate": 2.7786928254309953e-05,
+      "loss": 1.2035,
+      "step": 10601
+    },
+    {
+      "epoch": 0.7642182656959562,
+      "grad_norm": 0.1102022373442963,
+      "learning_rate": 2.777077967615188e-05,
+      "loss": 1.1812,
+      "step": 10602
+    },
+    {
+      "epoch": 0.7642903481582931,
+      "grad_norm": 0.10747221284978344,
+      "learning_rate": 2.7754635035132514e-05,
+      "loss": 1.0683,
+      "step": 10603
+    },
+    {
+      "epoch": 0.76436243062063,
+      "grad_norm": 0.1222399112776011,
+      "learning_rate": 2.7738494332131846e-05,
+      "loss": 1.2789,
+      "step": 10604
+    },
+    {
+      "epoch": 0.764434513082967,
+      "grad_norm": 0.11639571303113029,
+      "learning_rate": 2.7722357568029723e-05,
+      "loss": 1.2585,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7645065955453039,
+      "grad_norm": 0.1205569331882626,
+      "learning_rate": 2.7706224743705724e-05,
+      "loss": 1.2508,
+      "step": 10606
+    },
+    {
+      "epoch": 0.7645786780076408,
+      "grad_norm": 0.12514776307391712,
+      "learning_rate": 2.7690095860039224e-05,
+      "loss": 1.382,
+      "step": 10607
+    },
+    {
+      "epoch": 0.7646507604699777,
+      "grad_norm": 0.11919131906953483,
+      "learning_rate": 2.767397091790941e-05,
+      "loss": 1.3564,
+      "step": 10608
+    },
+    {
+      "epoch": 0.7647228429323145,
+      "grad_norm": 0.11791735263619448,
+      "learning_rate": 2.7657849918195223e-05,
+      "loss": 1.2589,
+      "step": 10609
+    },
+    {
+      "epoch": 0.7647949253946514,
+      "grad_norm": 0.1130515823774802,
+      "learning_rate": 2.7641732861775414e-05,
+      "loss": 1.2403,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7648670078569884,
+      "grad_norm": 0.13540935524720932,
+      "learning_rate": 2.76256197495285e-05,
+      "loss": 1.2845,
+      "step": 10611
+    },
+    {
+      "epoch": 0.7649390903193253,
+      "grad_norm": 0.11912948808507708,
+      "learning_rate": 2.7609510582332788e-05,
+      "loss": 1.2505,
+      "step": 10612
+    },
+    {
+      "epoch": 0.7650111727816622,
+      "grad_norm": 0.12243522824880564,
+      "learning_rate": 2.759340536106637e-05,
+      "loss": 1.2789,
+      "step": 10613
+    },
+    {
+      "epoch": 0.7650832552439991,
+      "grad_norm": 0.11361923910053215,
+      "learning_rate": 2.7577304086607125e-05,
+      "loss": 1.268,
+      "step": 10614
+    },
+    {
+      "epoch": 0.765155337706336,
+      "grad_norm": 0.11199241022974349,
+      "learning_rate": 2.7561206759832724e-05,
+      "loss": 1.3581,
+      "step": 10615
+    },
+    {
+      "epoch": 0.765227420168673,
+      "grad_norm": 0.11479885285441112,
+      "learning_rate": 2.7545113381620612e-05,
+      "loss": 1.2361,
+      "step": 10616
+    },
+    {
+      "epoch": 0.7652995026310099,
+      "grad_norm": 0.13083880707186626,
+      "learning_rate": 2.752902395284801e-05,
+      "loss": 1.3366,
+      "step": 10617
+    },
+    {
+      "epoch": 0.7653715850933468,
+      "grad_norm": 0.12403582176513556,
+      "learning_rate": 2.751293847439196e-05,
+      "loss": 1.3454,
+      "step": 10618
+    },
+    {
+      "epoch": 0.7654436675556837,
+      "grad_norm": 0.1171620166053329,
+      "learning_rate": 2.7496856947129247e-05,
+      "loss": 1.2585,
+      "step": 10619
+    },
+    {
+      "epoch": 0.7655157500180206,
+      "grad_norm": 0.2638655367954201,
+      "learning_rate": 2.7480779371936462e-05,
+      "loss": 1.2435,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7655878324803576,
+      "grad_norm": 0.11845967855674334,
+      "learning_rate": 2.7464705749690002e-05,
+      "loss": 1.2251,
+      "step": 10621
+    },
+    {
+      "epoch": 0.7656599149426945,
+      "grad_norm": 0.11795703281983413,
+      "learning_rate": 2.7448636081265942e-05,
+      "loss": 1.256,
+      "step": 10622
+    },
+    {
+      "epoch": 0.7657319974050314,
+      "grad_norm": 0.12290825881703792,
+      "learning_rate": 2.743257036754031e-05,
+      "loss": 1.2429,
+      "step": 10623
+    },
+    {
+      "epoch": 0.7658040798673683,
+      "grad_norm": 0.11935776929989446,
+      "learning_rate": 2.741650860938879e-05,
+      "loss": 1.1855,
+      "step": 10624
+    },
+    {
+      "epoch": 0.7658761623297052,
+      "grad_norm": 0.12303999986194486,
+      "learning_rate": 2.7400450807686938e-05,
+      "loss": 1.275,
+      "step": 10625
+    },
+    {
+      "epoch": 0.765948244792042,
+      "grad_norm": 0.1264932882012706,
+      "learning_rate": 2.7384396963309955e-05,
+      "loss": 1.1642,
+      "step": 10626
+    },
+    {
+      "epoch": 0.766020327254379,
+      "grad_norm": 0.1170065965288473,
+      "learning_rate": 2.7368347077133004e-05,
+      "loss": 1.2072,
+      "step": 10627
+    },
+    {
+      "epoch": 0.7660924097167159,
+      "grad_norm": 0.12241509512708705,
+      "learning_rate": 2.7352301150030912e-05,
+      "loss": 1.3838,
+      "step": 10628
+    },
+    {
+      "epoch": 0.7661644921790528,
+      "grad_norm": 0.12072356313069461,
+      "learning_rate": 2.7336259182878376e-05,
+      "loss": 1.2987,
+      "step": 10629
+    },
+    {
+      "epoch": 0.7662365746413897,
+      "grad_norm": 0.14128711193671367,
+      "learning_rate": 2.7320221176549744e-05,
+      "loss": 1.2763,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7663086571037266,
+      "grad_norm": 0.11614133742457923,
+      "learning_rate": 2.7304187131919235e-05,
+      "loss": 1.2651,
+      "step": 10631
+    },
+    {
+      "epoch": 0.7663807395660636,
+      "grad_norm": 0.12896031065047864,
+      "learning_rate": 2.728815704986092e-05,
+      "loss": 1.383,
+      "step": 10632
+    },
+    {
+      "epoch": 0.7664528220284005,
+      "grad_norm": 0.10707521398487113,
+      "learning_rate": 2.7272130931248573e-05,
+      "loss": 1.2065,
+      "step": 10633
+    },
+    {
+      "epoch": 0.7665249044907374,
+      "grad_norm": 0.13543729457013717,
+      "learning_rate": 2.7256108776955702e-05,
+      "loss": 1.2337,
+      "step": 10634
+    },
+    {
+      "epoch": 0.7665969869530743,
+      "grad_norm": 0.13320516648907543,
+      "learning_rate": 2.7240090587855647e-05,
+      "loss": 1.2751,
+      "step": 10635
+    },
+    {
+      "epoch": 0.7666690694154112,
+      "grad_norm": 0.12737338441218726,
+      "learning_rate": 2.7224076364821627e-05,
+      "loss": 1.3414,
+      "step": 10636
+    },
+    {
+      "epoch": 0.7667411518777482,
+      "grad_norm": 0.11905409330360285,
+      "learning_rate": 2.720806610872655e-05,
+      "loss": 1.2947,
+      "step": 10637
+    },
+    {
+      "epoch": 0.7668132343400851,
+      "grad_norm": 0.11344307103933605,
+      "learning_rate": 2.719205982044305e-05,
+      "loss": 1.2639,
+      "step": 10638
+    },
+    {
+      "epoch": 0.766885316802422,
+      "grad_norm": 0.13302636614314972,
+      "learning_rate": 2.717605750084361e-05,
+      "loss": 1.3317,
+      "step": 10639
+    },
+    {
+      "epoch": 0.7669573992647589,
+      "grad_norm": 0.1347027467465626,
+      "learning_rate": 2.7160059150800578e-05,
+      "loss": 1.1875,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7670294817270958,
+      "grad_norm": 0.23523856618253847,
+      "learning_rate": 2.7144064771185997e-05,
+      "loss": 1.296,
+      "step": 10641
+    },
+    {
+      "epoch": 0.7671015641894328,
+      "grad_norm": 0.11919531909385891,
+      "learning_rate": 2.7128074362871646e-05,
+      "loss": 1.2295,
+      "step": 10642
+    },
+    {
+      "epoch": 0.7671736466517696,
+      "grad_norm": 0.12464187931098013,
+      "learning_rate": 2.7112087926729168e-05,
+      "loss": 1.3181,
+      "step": 10643
+    },
+    {
+      "epoch": 0.7672457291141065,
+      "grad_norm": 0.12189409996890184,
+      "learning_rate": 2.7096105463629938e-05,
+      "loss": 1.3377,
+      "step": 10644
+    },
+    {
+      "epoch": 0.7673178115764434,
+      "grad_norm": 0.12849194290909116,
+      "learning_rate": 2.7080126974445242e-05,
+      "loss": 1.1519,
+      "step": 10645
+    },
+    {
+      "epoch": 0.7673898940387803,
+      "grad_norm": 0.11766140155785773,
+      "learning_rate": 2.7064152460045955e-05,
+      "loss": 1.1919,
+      "step": 10646
+    },
+    {
+      "epoch": 0.7674619765011172,
+      "grad_norm": 0.13151408882056007,
+      "learning_rate": 2.704818192130286e-05,
+      "loss": 1.2918,
+      "step": 10647
+    },
+    {
+      "epoch": 0.7675340589634542,
+      "grad_norm": 0.12503243704180295,
+      "learning_rate": 2.7032215359086465e-05,
+      "loss": 1.2803,
+      "step": 10648
+    },
+    {
+      "epoch": 0.7676061414257911,
+      "grad_norm": 0.1201623594008619,
+      "learning_rate": 2.7016252774267192e-05,
+      "loss": 1.2434,
+      "step": 10649
+    },
+    {
+      "epoch": 0.767678223888128,
+      "grad_norm": 0.11584325696908594,
+      "learning_rate": 2.7000294167715034e-05,
+      "loss": 1.2611,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7677503063504649,
+      "grad_norm": 0.1202078420351592,
+      "learning_rate": 2.698433954029992e-05,
+      "loss": 1.3073,
+      "step": 10651
+    },
+    {
+      "epoch": 0.7678223888128018,
+      "grad_norm": 0.12913610840769787,
+      "learning_rate": 2.6968388892891516e-05,
+      "loss": 1.2603,
+      "step": 10652
+    },
+    {
+      "epoch": 0.7678944712751388,
+      "grad_norm": 0.16929396185109533,
+      "learning_rate": 2.6952442226359277e-05,
+      "loss": 1.1884,
+      "step": 10653
+    },
+    {
+      "epoch": 0.7679665537374757,
+      "grad_norm": 0.11711411881747376,
+      "learning_rate": 2.6936499541572445e-05,
+      "loss": 1.273,
+      "step": 10654
+    },
+    {
+      "epoch": 0.7680386361998126,
+      "grad_norm": 0.1110808958870647,
+      "learning_rate": 2.6920560839400023e-05,
+      "loss": 1.1856,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7681107186621495,
+      "grad_norm": 0.1228233711546265,
+      "learning_rate": 2.690462612071083e-05,
+      "loss": 1.3852,
+      "step": 10656
+    },
+    {
+      "epoch": 0.7681828011244864,
+      "grad_norm": 0.11585453831465439,
+      "learning_rate": 2.688869538637344e-05,
+      "loss": 1.2832,
+      "step": 10657
+    },
+    {
+      "epoch": 0.7682548835868234,
+      "grad_norm": 0.12009393442687595,
+      "learning_rate": 2.687276863725623e-05,
+      "loss": 1.2703,
+      "step": 10658
+    },
+    {
+      "epoch": 0.7683269660491603,
+      "grad_norm": 0.11495124339196071,
+      "learning_rate": 2.6856845874227333e-05,
+      "loss": 1.2907,
+      "step": 10659
+    },
+    {
+      "epoch": 0.7683990485114972,
+      "grad_norm": 0.11498131504847452,
+      "learning_rate": 2.68409270981547e-05,
+      "loss": 1.2427,
+      "step": 10660
+    },
+    {
+      "epoch": 0.768471130973834,
+      "grad_norm": 0.10977940371815172,
+      "learning_rate": 2.682501230990604e-05,
+      "loss": 1.2555,
+      "step": 10661
+    },
+    {
+      "epoch": 0.7685432134361709,
+      "grad_norm": 0.11724060047568986,
+      "learning_rate": 2.6809101510348857e-05,
+      "loss": 1.2614,
+      "step": 10662
+    },
+    {
+      "epoch": 0.7686152958985079,
+      "grad_norm": 0.12560178885380904,
+      "learning_rate": 2.6793194700350422e-05,
+      "loss": 1.2831,
+      "step": 10663
+    },
+    {
+      "epoch": 0.7686873783608448,
+      "grad_norm": 0.13313677618470768,
+      "learning_rate": 2.67772918807778e-05,
+      "loss": 1.1705,
+      "step": 10664
+    },
+    {
+      "epoch": 0.7687594608231817,
+      "grad_norm": 0.10746157779844713,
+      "learning_rate": 2.676139305249783e-05,
+      "loss": 1.3462,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7688315432855186,
+      "grad_norm": 0.11957094179569983,
+      "learning_rate": 2.6745498216377163e-05,
+      "loss": 1.3786,
+      "step": 10666
+    },
+    {
+      "epoch": 0.7689036257478555,
+      "grad_norm": 0.13005843668775421,
+      "learning_rate": 2.6729607373282193e-05,
+      "loss": 1.3133,
+      "step": 10667
+    },
+    {
+      "epoch": 0.7689757082101925,
+      "grad_norm": 0.10877995978984253,
+      "learning_rate": 2.671372052407912e-05,
+      "loss": 1.3079,
+      "step": 10668
+    },
+    {
+      "epoch": 0.7690477906725294,
+      "grad_norm": 0.1302747916772236,
+      "learning_rate": 2.669783766963392e-05,
+      "loss": 1.2161,
+      "step": 10669
+    },
+    {
+      "epoch": 0.7691198731348663,
+      "grad_norm": 0.12462889949740111,
+      "learning_rate": 2.6681958810812336e-05,
+      "loss": 1.3703,
+      "step": 10670
+    },
+    {
+      "epoch": 0.7691919555972032,
+      "grad_norm": 0.10944366131297921,
+      "learning_rate": 2.6666083948479924e-05,
+      "loss": 1.3091,
+      "step": 10671
+    },
+    {
+      "epoch": 0.7692640380595401,
+      "grad_norm": 0.12082931128100183,
+      "learning_rate": 2.6650213083501995e-05,
+      "loss": 1.2656,
+      "step": 10672
+    },
+    {
+      "epoch": 0.769336120521877,
+      "grad_norm": 0.12627677297194914,
+      "learning_rate": 2.6634346216743665e-05,
+      "loss": 1.2495,
+      "step": 10673
+    },
+    {
+      "epoch": 0.769408202984214,
+      "grad_norm": 0.11461770304180116,
+      "learning_rate": 2.6618483349069846e-05,
+      "loss": 1.3176,
+      "step": 10674
+    },
+    {
+      "epoch": 0.7694802854465509,
+      "grad_norm": 0.11349473946031881,
+      "learning_rate": 2.6602624481345108e-05,
+      "loss": 1.2671,
+      "step": 10675
+    },
+    {
+      "epoch": 0.7695523679088878,
+      "grad_norm": 0.1084861580902203,
+      "learning_rate": 2.6586769614434003e-05,
+      "loss": 1.2981,
+      "step": 10676
+    },
+    {
+      "epoch": 0.7696244503712247,
+      "grad_norm": 0.12392551687917495,
+      "learning_rate": 2.6570918749200722e-05,
+      "loss": 1.3292,
+      "step": 10677
+    },
+    {
+      "epoch": 0.7696965328335615,
+      "grad_norm": 0.12255154840895018,
+      "learning_rate": 2.655507188650932e-05,
+      "loss": 1.2229,
+      "step": 10678
+    },
+    {
+      "epoch": 0.7697686152958985,
+      "grad_norm": 0.11543974472318705,
+      "learning_rate": 2.65392290272235e-05,
+      "loss": 1.2722,
+      "step": 10679
+    },
+    {
+      "epoch": 0.7698406977582354,
+      "grad_norm": 0.1244313953358607,
+      "learning_rate": 2.6523390172206918e-05,
+      "loss": 1.2572,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7699127802205723,
+      "grad_norm": 0.1251168631889096,
+      "learning_rate": 2.6507555322322918e-05,
+      "loss": 1.2473,
+      "step": 10681
+    },
+    {
+      "epoch": 0.7699848626829092,
+      "grad_norm": 0.1096664707422631,
+      "learning_rate": 2.6491724478434666e-05,
+      "loss": 1.3635,
+      "step": 10682
+    },
+    {
+      "epoch": 0.7700569451452461,
+      "grad_norm": 0.11553020301943019,
+      "learning_rate": 2.647589764140499e-05,
+      "loss": 1.2233,
+      "step": 10683
+    },
+    {
+      "epoch": 0.7701290276075831,
+      "grad_norm": 0.13648843951445402,
+      "learning_rate": 2.646007481209668e-05,
+      "loss": 1.2426,
+      "step": 10684
+    },
+    {
+      "epoch": 0.77020111006992,
+      "grad_norm": 0.11887076362383206,
+      "learning_rate": 2.6444255991372214e-05,
+      "loss": 1.2578,
+      "step": 10685
+    },
+    {
+      "epoch": 0.7702731925322569,
+      "grad_norm": 0.129735436010245,
+      "learning_rate": 2.6428441180093866e-05,
+      "loss": 1.3219,
+      "step": 10686
+    },
+    {
+      "epoch": 0.7703452749945938,
+      "grad_norm": 0.11045849895433275,
+      "learning_rate": 2.6412630379123637e-05,
+      "loss": 1.2241,
+      "step": 10687
+    },
+    {
+      "epoch": 0.7704173574569307,
+      "grad_norm": 0.1175780882272209,
+      "learning_rate": 2.639682358932336e-05,
+      "loss": 1.3959,
+      "step": 10688
+    },
+    {
+      "epoch": 0.7704894399192677,
+      "grad_norm": 0.11992831807028648,
+      "learning_rate": 2.638102081155469e-05,
+      "loss": 1.3568,
+      "step": 10689
+    },
+    {
+      "epoch": 0.7705615223816046,
+      "grad_norm": 0.11794461157578777,
+      "learning_rate": 2.636522204667904e-05,
+      "loss": 1.2386,
+      "step": 10690
+    },
+    {
+      "epoch": 0.7706336048439415,
+      "grad_norm": 0.11514397306816684,
+      "learning_rate": 2.6349427295557506e-05,
+      "loss": 1.306,
+      "step": 10691
+    },
+    {
+      "epoch": 0.7707056873062784,
+      "grad_norm": 0.13040265418110628,
+      "learning_rate": 2.6333636559051046e-05,
+      "loss": 1.2797,
+      "step": 10692
+    },
+    {
+      "epoch": 0.7707777697686153,
+      "grad_norm": 0.12042868771003777,
+      "learning_rate": 2.631784983802047e-05,
+      "loss": 1.2015,
+      "step": 10693
+    },
+    {
+      "epoch": 0.7708498522309523,
+      "grad_norm": 0.12170870763218167,
+      "learning_rate": 2.6302067133326293e-05,
+      "loss": 1.2736,
+      "step": 10694
+    },
+    {
+      "epoch": 0.7709219346932891,
+      "grad_norm": 0.11960007111660902,
+      "learning_rate": 2.6286288445828734e-05,
+      "loss": 1.2001,
+      "step": 10695
+    },
+    {
+      "epoch": 0.770994017155626,
+      "grad_norm": 0.12446014220687163,
+      "learning_rate": 2.6270513776387927e-05,
+      "loss": 1.1454,
+      "step": 10696
+    },
+    {
+      "epoch": 0.7710660996179629,
+      "grad_norm": 0.11804551598203612,
+      "learning_rate": 2.625474312586369e-05,
+      "loss": 1.319,
+      "step": 10697
+    },
+    {
+      "epoch": 0.7711381820802998,
+      "grad_norm": 0.11359705128678844,
+      "learning_rate": 2.6238976495115754e-05,
+      "loss": 1.2879,
+      "step": 10698
+    },
+    {
+      "epoch": 0.7712102645426367,
+      "grad_norm": 0.11895707092050203,
+      "learning_rate": 2.622321388500345e-05,
+      "loss": 1.2388,
+      "step": 10699
+    },
+    {
+      "epoch": 0.7712823470049737,
+      "grad_norm": 0.12199948244751543,
+      "learning_rate": 2.620745529638603e-05,
+      "loss": 1.3639,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7713544294673106,
+      "grad_norm": 0.11951841610787213,
+      "learning_rate": 2.6191700730122416e-05,
+      "loss": 1.2805,
+      "step": 10701
+    },
+    {
+      "epoch": 0.7714265119296475,
+      "grad_norm": 0.11675254296898098,
+      "learning_rate": 2.6175950187071487e-05,
+      "loss": 1.2968,
+      "step": 10702
+    },
+    {
+      "epoch": 0.7714985943919844,
+      "grad_norm": 0.15164875289714022,
+      "learning_rate": 2.61602036680917e-05,
+      "loss": 1.275,
+      "step": 10703
+    },
+    {
+      "epoch": 0.7715706768543213,
+      "grad_norm": 0.12083310983397394,
+      "learning_rate": 2.614446117404139e-05,
+      "loss": 1.2599,
+      "step": 10704
+    },
+    {
+      "epoch": 0.7716427593166583,
+      "grad_norm": 0.1250465686937132,
+      "learning_rate": 2.6128722705778653e-05,
+      "loss": 1.3148,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7717148417789952,
+      "grad_norm": 0.14784988979685085,
+      "learning_rate": 2.6112988264161454e-05,
+      "loss": 1.1916,
+      "step": 10706
+    },
+    {
+      "epoch": 0.7717869242413321,
+      "grad_norm": 0.1265707969374425,
+      "learning_rate": 2.6097257850047386e-05,
+      "loss": 1.2242,
+      "step": 10707
+    },
+    {
+      "epoch": 0.771859006703669,
+      "grad_norm": 0.11380887917895588,
+      "learning_rate": 2.6081531464293906e-05,
+      "loss": 1.2718,
+      "step": 10708
+    },
+    {
+      "epoch": 0.771931089166006,
+      "grad_norm": 0.11459918770422861,
+      "learning_rate": 2.6065809107758266e-05,
+      "loss": 1.2644,
+      "step": 10709
+    },
+    {
+      "epoch": 0.7720031716283429,
+      "grad_norm": 0.1131548109581303,
+      "learning_rate": 2.6050090781297464e-05,
+      "loss": 1.2746,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7720752540906798,
+      "grad_norm": 0.12696948037059794,
+      "learning_rate": 2.6034376485768286e-05,
+      "loss": 1.2754,
+      "step": 10711
+    },
+    {
+      "epoch": 0.7721473365530167,
+      "grad_norm": 0.1224542596359772,
+      "learning_rate": 2.60186662220273e-05,
+      "loss": 1.2459,
+      "step": 10712
+    },
+    {
+      "epoch": 0.7722194190153535,
+      "grad_norm": 0.11715740048306948,
+      "learning_rate": 2.6002959990930864e-05,
+      "loss": 1.2914,
+      "step": 10713
+    },
+    {
+      "epoch": 0.7722915014776904,
+      "grad_norm": 0.12495597890260005,
+      "learning_rate": 2.5987257793335107e-05,
+      "loss": 1.3379,
+      "step": 10714
+    },
+    {
+      "epoch": 0.7723635839400274,
+      "grad_norm": 0.11377792783775371,
+      "learning_rate": 2.5971559630095932e-05,
+      "loss": 1.2801,
+      "step": 10715
+    },
+    {
+      "epoch": 0.7724356664023643,
+      "grad_norm": 0.1290766509169426,
+      "learning_rate": 2.5955865502069043e-05,
+      "loss": 1.2292,
+      "step": 10716
+    },
+    {
+      "epoch": 0.7725077488647012,
+      "grad_norm": 0.12885989757493765,
+      "learning_rate": 2.59401754101099e-05,
+      "loss": 1.2773,
+      "step": 10717
+    },
+    {
+      "epoch": 0.7725798313270381,
+      "grad_norm": 0.1195027757773923,
+      "learning_rate": 2.5924489355073744e-05,
+      "loss": 1.2097,
+      "step": 10718
+    },
+    {
+      "epoch": 0.772651913789375,
+      "grad_norm": 0.12079041129372477,
+      "learning_rate": 2.5908807337815632e-05,
+      "loss": 1.2718,
+      "step": 10719
+    },
+    {
+      "epoch": 0.772723996251712,
+      "grad_norm": 0.11236683965038416,
+      "learning_rate": 2.589312935919035e-05,
+      "loss": 1.1817,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7727960787140489,
+      "grad_norm": 0.12692193508017588,
+      "learning_rate": 2.58774554200525e-05,
+      "loss": 1.2799,
+      "step": 10721
+    },
+    {
+      "epoch": 0.7728681611763858,
+      "grad_norm": 0.11681971765750876,
+      "learning_rate": 2.586178552125644e-05,
+      "loss": 1.2662,
+      "step": 10722
+    },
+    {
+      "epoch": 0.7729402436387227,
+      "grad_norm": 0.12594282313023458,
+      "learning_rate": 2.5846119663656343e-05,
+      "loss": 1.2995,
+      "step": 10723
+    },
+    {
+      "epoch": 0.7730123261010596,
+      "grad_norm": 0.12074944830233875,
+      "learning_rate": 2.5830457848106116e-05,
+      "loss": 1.3565,
+      "step": 10724
+    },
+    {
+      "epoch": 0.7730844085633966,
+      "grad_norm": 0.11685976519603412,
+      "learning_rate": 2.5814800075459477e-05,
+      "loss": 1.1875,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7731564910257335,
+      "grad_norm": 0.12235994973557777,
+      "learning_rate": 2.5799146346569914e-05,
+      "loss": 1.2679,
+      "step": 10726
+    },
+    {
+      "epoch": 0.7732285734880704,
+      "grad_norm": 0.12401506839394404,
+      "learning_rate": 2.5783496662290695e-05,
+      "loss": 1.3477,
+      "step": 10727
+    },
+    {
+      "epoch": 0.7733006559504073,
+      "grad_norm": 0.1258022957609564,
+      "learning_rate": 2.5767851023474865e-05,
+      "loss": 1.2469,
+      "step": 10728
+    },
+    {
+      "epoch": 0.7733727384127442,
+      "grad_norm": 0.12964394141098454,
+      "learning_rate": 2.575220943097526e-05,
+      "loss": 1.334,
+      "step": 10729
+    },
+    {
+      "epoch": 0.773444820875081,
+      "grad_norm": 0.13300568017854691,
+      "learning_rate": 2.5736571885644477e-05,
+      "loss": 1.1995,
+      "step": 10730
+    },
+    {
+      "epoch": 0.773516903337418,
+      "grad_norm": 0.12240779647095455,
+      "learning_rate": 2.5720938388334937e-05,
+      "loss": 1.4374,
+      "step": 10731
+    },
+    {
+      "epoch": 0.7735889857997549,
+      "grad_norm": 0.20323870218775195,
+      "learning_rate": 2.570530893989871e-05,
+      "loss": 1.2907,
+      "step": 10732
+    },
+    {
+      "epoch": 0.7736610682620918,
+      "grad_norm": 0.13085281058293274,
+      "learning_rate": 2.5689683541187838e-05,
+      "loss": 1.1858,
+      "step": 10733
+    },
+    {
+      "epoch": 0.7737331507244287,
+      "grad_norm": 0.1256633873178836,
+      "learning_rate": 2.5674062193054016e-05,
+      "loss": 1.254,
+      "step": 10734
+    },
+    {
+      "epoch": 0.7738052331867656,
+      "grad_norm": 0.12702743641445657,
+      "learning_rate": 2.5658444896348766e-05,
+      "loss": 1.2117,
+      "step": 10735
+    },
+    {
+      "epoch": 0.7738773156491026,
+      "grad_norm": 0.11937119902513627,
+      "learning_rate": 2.5642831651923295e-05,
+      "loss": 1.279,
+      "step": 10736
+    },
+    {
+      "epoch": 0.7739493981114395,
+      "grad_norm": 0.11272325460840954,
+      "learning_rate": 2.5627222460628742e-05,
+      "loss": 1.2801,
+      "step": 10737
+    },
+    {
+      "epoch": 0.7740214805737764,
+      "grad_norm": 0.11771367239366239,
+      "learning_rate": 2.561161732331593e-05,
+      "loss": 1.2892,
+      "step": 10738
+    },
+    {
+      "epoch": 0.7740935630361133,
+      "grad_norm": 0.12452454407215953,
+      "learning_rate": 2.5596016240835486e-05,
+      "loss": 1.3532,
+      "step": 10739
+    },
+    {
+      "epoch": 0.7741656454984502,
+      "grad_norm": 0.13547757376716274,
+      "learning_rate": 2.5580419214037777e-05,
+      "loss": 1.1991,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7742377279607872,
+      "grad_norm": 0.11288713078209416,
+      "learning_rate": 2.5564826243772966e-05,
+      "loss": 1.2359,
+      "step": 10741
+    },
+    {
+      "epoch": 0.7743098104231241,
+      "grad_norm": 0.12233819683939329,
+      "learning_rate": 2.5549237330891097e-05,
+      "loss": 1.2177,
+      "step": 10742
+    },
+    {
+      "epoch": 0.774381892885461,
+      "grad_norm": 0.10689593030355946,
+      "learning_rate": 2.553365247624182e-05,
+      "loss": 1.2994,
+      "step": 10743
+    },
+    {
+      "epoch": 0.7744539753477979,
+      "grad_norm": 0.12109131424679269,
+      "learning_rate": 2.5518071680674683e-05,
+      "loss": 1.2836,
+      "step": 10744
+    },
+    {
+      "epoch": 0.7745260578101348,
+      "grad_norm": 0.11171087289040044,
+      "learning_rate": 2.5502494945038934e-05,
+      "loss": 1.2944,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7745981402724718,
+      "grad_norm": 0.11389510871915488,
+      "learning_rate": 2.5486922270183756e-05,
+      "loss": 1.2452,
+      "step": 10746
+    },
+    {
+      "epoch": 0.7746702227348086,
+      "grad_norm": 0.12190906220441665,
+      "learning_rate": 2.5471353656957886e-05,
+      "loss": 1.2618,
+      "step": 10747
+    },
+    {
+      "epoch": 0.7747423051971455,
+      "grad_norm": 0.1249859302124325,
+      "learning_rate": 2.5455789106210004e-05,
+      "loss": 1.247,
+      "step": 10748
+    },
+    {
+      "epoch": 0.7748143876594824,
+      "grad_norm": 0.11128629226082631,
+      "learning_rate": 2.5440228618788465e-05,
+      "loss": 1.2549,
+      "step": 10749
+    },
+    {
+      "epoch": 0.7748864701218193,
+      "grad_norm": 0.12720510965637744,
+      "learning_rate": 2.5424672195541565e-05,
+      "loss": 1.3421,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7749585525841562,
+      "grad_norm": 0.12060560160201629,
+      "learning_rate": 2.5409119837317173e-05,
+      "loss": 1.3362,
+      "step": 10751
+    },
+    {
+      "epoch": 0.7750306350464932,
+      "grad_norm": 0.12427688676244966,
+      "learning_rate": 2.5393571544963056e-05,
+      "loss": 1.2038,
+      "step": 10752
+    },
+    {
+      "epoch": 0.7751027175088301,
+      "grad_norm": 0.12352988368960069,
+      "learning_rate": 2.5378027319326736e-05,
+      "loss": 1.1828,
+      "step": 10753
+    },
+    {
+      "epoch": 0.775174799971167,
+      "grad_norm": 0.13065474340371264,
+      "learning_rate": 2.536248716125552e-05,
+      "loss": 1.3909,
+      "step": 10754
+    },
+    {
+      "epoch": 0.7752468824335039,
+      "grad_norm": 0.12740310135421665,
+      "learning_rate": 2.534695107159647e-05,
+      "loss": 1.3207,
+      "step": 10755
+    },
+    {
+      "epoch": 0.7753189648958408,
+      "grad_norm": 0.12265835852903949,
+      "learning_rate": 2.5331419051196458e-05,
+      "loss": 1.2837,
+      "step": 10756
+    },
+    {
+      "epoch": 0.7753910473581778,
+      "grad_norm": 0.13463329460698203,
+      "learning_rate": 2.531589110090211e-05,
+      "loss": 1.2146,
+      "step": 10757
+    },
+    {
+      "epoch": 0.7754631298205147,
+      "grad_norm": 0.10827136156256627,
+      "learning_rate": 2.530036722155985e-05,
+      "loss": 1.291,
+      "step": 10758
+    },
+    {
+      "epoch": 0.7755352122828516,
+      "grad_norm": 0.1242171808074373,
+      "learning_rate": 2.5284847414015845e-05,
+      "loss": 1.2642,
+      "step": 10759
+    },
+    {
+      "epoch": 0.7756072947451885,
+      "grad_norm": 0.12597351692644929,
+      "learning_rate": 2.5269331679116082e-05,
+      "loss": 1.2253,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7756793772075254,
+      "grad_norm": 0.1219722184375612,
+      "learning_rate": 2.525382001770631e-05,
+      "loss": 1.2421,
+      "step": 10761
+    },
+    {
+      "epoch": 0.7757514596698624,
+      "grad_norm": 0.11104773948716799,
+      "learning_rate": 2.523831243063204e-05,
+      "loss": 1.2449,
+      "step": 10762
+    },
+    {
+      "epoch": 0.7758235421321993,
+      "grad_norm": 0.13072250242196365,
+      "learning_rate": 2.5222808918738573e-05,
+      "loss": 1.1994,
+      "step": 10763
+    },
+    {
+      "epoch": 0.7758956245945362,
+      "grad_norm": 0.12239941531256639,
+      "learning_rate": 2.5207309482871e-05,
+      "loss": 1.2616,
+      "step": 10764
+    },
+    {
+      "epoch": 0.775967707056873,
+      "grad_norm": 0.11981739660078516,
+      "learning_rate": 2.519181412387418e-05,
+      "loss": 1.2722,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7760397895192099,
+      "grad_norm": 0.11856102295758768,
+      "learning_rate": 2.5176322842592738e-05,
+      "loss": 1.3761,
+      "step": 10766
+    },
+    {
+      "epoch": 0.7761118719815469,
+      "grad_norm": 0.12707191312883914,
+      "learning_rate": 2.5160835639871084e-05,
+      "loss": 1.2904,
+      "step": 10767
+    },
+    {
+      "epoch": 0.7761839544438838,
+      "grad_norm": 0.11364746396341499,
+      "learning_rate": 2.5145352516553434e-05,
+      "loss": 1.2724,
+      "step": 10768
+    },
+    {
+      "epoch": 0.7762560369062207,
+      "grad_norm": 0.11500284288782633,
+      "learning_rate": 2.5129873473483734e-05,
+      "loss": 1.2639,
+      "step": 10769
+    },
+    {
+      "epoch": 0.7763281193685576,
+      "grad_norm": 0.1255294864250013,
+      "learning_rate": 2.5114398511505734e-05,
+      "loss": 1.3485,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7764002018308945,
+      "grad_norm": 0.1248006218068425,
+      "learning_rate": 2.5098927631462966e-05,
+      "loss": 1.2708,
+      "step": 10771
+    },
+    {
+      "epoch": 0.7764722842932315,
+      "grad_norm": 0.12521393409067633,
+      "learning_rate": 2.5083460834198723e-05,
+      "loss": 1.2888,
+      "step": 10772
+    },
+    {
+      "epoch": 0.7765443667555684,
+      "grad_norm": 0.1227124545024407,
+      "learning_rate": 2.506799812055609e-05,
+      "loss": 1.2363,
+      "step": 10773
+    },
+    {
+      "epoch": 0.7766164492179053,
+      "grad_norm": 0.11170844749551967,
+      "learning_rate": 2.5052539491377912e-05,
+      "loss": 1.1847,
+      "step": 10774
+    },
+    {
+      "epoch": 0.7766885316802422,
+      "grad_norm": 0.10040598137658464,
+      "learning_rate": 2.503708494750687e-05,
+      "loss": 1.1599,
+      "step": 10775
+    },
+    {
+      "epoch": 0.7767606141425791,
+      "grad_norm": 0.11581555612760168,
+      "learning_rate": 2.5021634489785273e-05,
+      "loss": 1.3424,
+      "step": 10776
+    },
+    {
+      "epoch": 0.776832696604916,
+      "grad_norm": 0.1180371207370496,
+      "learning_rate": 2.5006188119055405e-05,
+      "loss": 1.3678,
+      "step": 10777
+    },
+    {
+      "epoch": 0.776904779067253,
+      "grad_norm": 0.11127483723887473,
+      "learning_rate": 2.499074583615919e-05,
+      "loss": 1.2868,
+      "step": 10778
+    },
+    {
+      "epoch": 0.7769768615295899,
+      "grad_norm": 0.11478922091077043,
+      "learning_rate": 2.4975307641938417e-05,
+      "loss": 1.1794,
+      "step": 10779
+    },
+    {
+      "epoch": 0.7770489439919268,
+      "grad_norm": 0.12461146447876172,
+      "learning_rate": 2.49598735372345e-05,
+      "loss": 1.3713,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7771210264542637,
+      "grad_norm": 0.11786612478514119,
+      "learning_rate": 2.494444352288884e-05,
+      "loss": 1.3395,
+      "step": 10781
+    },
+    {
+      "epoch": 0.7771931089166005,
+      "grad_norm": 0.14089842545547007,
+      "learning_rate": 2.4929017599742476e-05,
+      "loss": 1.3827,
+      "step": 10782
+    },
+    {
+      "epoch": 0.7772651913789375,
+      "grad_norm": 0.11613913621081054,
+      "learning_rate": 2.4913595768636278e-05,
+      "loss": 1.2682,
+      "step": 10783
+    },
+    {
+      "epoch": 0.7773372738412744,
+      "grad_norm": 0.1309828765935969,
+      "learning_rate": 2.4898178030410836e-05,
+      "loss": 1.226,
+      "step": 10784
+    },
+    {
+      "epoch": 0.7774093563036113,
+      "grad_norm": 0.11613531671036209,
+      "learning_rate": 2.488276438590653e-05,
+      "loss": 1.2054,
+      "step": 10785
+    },
+    {
+      "epoch": 0.7774814387659482,
+      "grad_norm": 0.12054580472178102,
+      "learning_rate": 2.486735483596362e-05,
+      "loss": 1.3054,
+      "step": 10786
+    },
+    {
+      "epoch": 0.7775535212282851,
+      "grad_norm": 0.1278574007103607,
+      "learning_rate": 2.4851949381422046e-05,
+      "loss": 1.325,
+      "step": 10787
+    },
+    {
+      "epoch": 0.7776256036906221,
+      "grad_norm": 0.11841438279346668,
+      "learning_rate": 2.483654802312151e-05,
+      "loss": 1.2631,
+      "step": 10788
+    },
+    {
+      "epoch": 0.777697686152959,
+      "grad_norm": 0.11019815279278183,
+      "learning_rate": 2.4821150761901503e-05,
+      "loss": 1.2714,
+      "step": 10789
+    },
+    {
+      "epoch": 0.7777697686152959,
+      "grad_norm": 0.11907107175133515,
+      "learning_rate": 2.4805757598601363e-05,
+      "loss": 1.2775,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7778418510776328,
+      "grad_norm": 0.12122776076131374,
+      "learning_rate": 2.4790368534060194e-05,
+      "loss": 1.3033,
+      "step": 10791
+    },
+    {
+      "epoch": 0.7779139335399697,
+      "grad_norm": 0.11872847712193234,
+      "learning_rate": 2.4774983569116737e-05,
+      "loss": 1.2682,
+      "step": 10792
+    },
+    {
+      "epoch": 0.7779860160023067,
+      "grad_norm": 0.13786732531599086,
+      "learning_rate": 2.4759602704609664e-05,
+      "loss": 1.2466,
+      "step": 10793
+    },
+    {
+      "epoch": 0.7780580984646436,
+      "grad_norm": 0.1292589913207786,
+      "learning_rate": 2.474422594137734e-05,
+      "loss": 1.2637,
+      "step": 10794
+    },
+    {
+      "epoch": 0.7781301809269805,
+      "grad_norm": 0.11911590149063644,
+      "learning_rate": 2.4728853280258013e-05,
+      "loss": 1.3142,
+      "step": 10795
+    },
+    {
+      "epoch": 0.7782022633893174,
+      "grad_norm": 0.12612726246048225,
+      "learning_rate": 2.4713484722089553e-05,
+      "loss": 1.2883,
+      "step": 10796
+    },
+    {
+      "epoch": 0.7782743458516543,
+      "grad_norm": 0.11789112922981637,
+      "learning_rate": 2.46981202677097e-05,
+      "loss": 1.3637,
+      "step": 10797
+    },
+    {
+      "epoch": 0.7783464283139913,
+      "grad_norm": 0.1152861827895048,
+      "learning_rate": 2.4682759917955932e-05,
+      "loss": 1.2493,
+      "step": 10798
+    },
+    {
+      "epoch": 0.7784185107763281,
+      "grad_norm": 0.12147464745662787,
+      "learning_rate": 2.4667403673665623e-05,
+      "loss": 1.2461,
+      "step": 10799
+    },
+    {
+      "epoch": 0.778490593238665,
+      "grad_norm": 0.1285150648143667,
+      "learning_rate": 2.4652051535675725e-05,
+      "loss": 1.2465,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7785626757010019,
+      "grad_norm": 0.11681861340411466,
+      "learning_rate": 2.46367035048231e-05,
+      "loss": 1.2732,
+      "step": 10801
+    },
+    {
+      "epoch": 0.7786347581633388,
+      "grad_norm": 0.12673181396657734,
+      "learning_rate": 2.4621359581944326e-05,
+      "loss": 1.2564,
+      "step": 10802
+    },
+    {
+      "epoch": 0.7787068406256757,
+      "grad_norm": 0.11780044659587571,
+      "learning_rate": 2.4606019767875865e-05,
+      "loss": 1.3269,
+      "step": 10803
+    },
+    {
+      "epoch": 0.7787789230880127,
+      "grad_norm": 0.12708010329342578,
+      "learning_rate": 2.4590684063453794e-05,
+      "loss": 1.2853,
+      "step": 10804
+    },
+    {
+      "epoch": 0.7788510055503496,
+      "grad_norm": 0.11331141642595445,
+      "learning_rate": 2.4575352469514066e-05,
+      "loss": 1.2434,
+      "step": 10805
+    },
+    {
+      "epoch": 0.7789230880126865,
+      "grad_norm": 0.11928231523185201,
+      "learning_rate": 2.4560024986892406e-05,
+      "loss": 1.1757,
+      "step": 10806
+    },
+    {
+      "epoch": 0.7789951704750234,
+      "grad_norm": 0.11773580052312155,
+      "learning_rate": 2.4544701616424292e-05,
+      "loss": 1.2377,
+      "step": 10807
+    },
+    {
+      "epoch": 0.7790672529373603,
+      "grad_norm": 0.12411596518839503,
+      "learning_rate": 2.4529382358944974e-05,
+      "loss": 1.275,
+      "step": 10808
+    },
+    {
+      "epoch": 0.7791393353996973,
+      "grad_norm": 0.1186747474212209,
+      "learning_rate": 2.4514067215289503e-05,
+      "loss": 1.183,
+      "step": 10809
+    },
+    {
+      "epoch": 0.7792114178620342,
+      "grad_norm": 0.11649414928343968,
+      "learning_rate": 2.4498756186292682e-05,
+      "loss": 1.2926,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7792835003243711,
+      "grad_norm": 0.12277464028317903,
+      "learning_rate": 2.4483449272789106e-05,
+      "loss": 1.3691,
+      "step": 10811
+    },
+    {
+      "epoch": 0.779355582786708,
+      "grad_norm": 0.11745765504278025,
+      "learning_rate": 2.446814647561313e-05,
+      "loss": 1.2372,
+      "step": 10812
+    },
+    {
+      "epoch": 0.7794276652490449,
+      "grad_norm": 0.11967205143477576,
+      "learning_rate": 2.44528477955989e-05,
+      "loss": 1.2692,
+      "step": 10813
+    },
+    {
+      "epoch": 0.7794997477113819,
+      "grad_norm": 0.12405485619351019,
+      "learning_rate": 2.4437553233580322e-05,
+      "loss": 1.2859,
+      "step": 10814
+    },
+    {
+      "epoch": 0.7795718301737188,
+      "grad_norm": 0.11915100491931124,
+      "learning_rate": 2.4422262790391103e-05,
+      "loss": 1.2201,
+      "step": 10815
+    },
+    {
+      "epoch": 0.7796439126360557,
+      "grad_norm": 0.11151672595555216,
+      "learning_rate": 2.4406976466864696e-05,
+      "loss": 1.3046,
+      "step": 10816
+    },
+    {
+      "epoch": 0.7797159950983925,
+      "grad_norm": 0.12686989224242323,
+      "learning_rate": 2.4391694263834354e-05,
+      "loss": 1.2334,
+      "step": 10817
+    },
+    {
+      "epoch": 0.7797880775607294,
+      "grad_norm": 0.12538331479478337,
+      "learning_rate": 2.4376416182133076e-05,
+      "loss": 1.2565,
+      "step": 10818
+    },
+    {
+      "epoch": 0.7798601600230664,
+      "grad_norm": 0.11831553627605151,
+      "learning_rate": 2.436114222259367e-05,
+      "loss": 1.2744,
+      "step": 10819
+    },
+    {
+      "epoch": 0.7799322424854033,
+      "grad_norm": 0.11444815205688447,
+      "learning_rate": 2.4345872386048696e-05,
+      "loss": 1.2935,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7800043249477402,
+      "grad_norm": 0.12772949897209498,
+      "learning_rate": 2.433060667333049e-05,
+      "loss": 1.3378,
+      "step": 10821
+    },
+    {
+      "epoch": 0.7800764074100771,
+      "grad_norm": 0.12124671896678262,
+      "learning_rate": 2.431534508527119e-05,
+      "loss": 1.2565,
+      "step": 10822
+    },
+    {
+      "epoch": 0.780148489872414,
+      "grad_norm": 0.12409353994751275,
+      "learning_rate": 2.430008762270267e-05,
+      "loss": 1.345,
+      "step": 10823
+    },
+    {
+      "epoch": 0.780220572334751,
+      "grad_norm": 0.12756592383785922,
+      "learning_rate": 2.4284834286456603e-05,
+      "loss": 1.3353,
+      "step": 10824
+    },
+    {
+      "epoch": 0.7802926547970879,
+      "grad_norm": 0.1224740611765861,
+      "learning_rate": 2.4269585077364443e-05,
+      "loss": 1.2523,
+      "step": 10825
+    },
+    {
+      "epoch": 0.7803647372594248,
+      "grad_norm": 0.11595570315557767,
+      "learning_rate": 2.425433999625739e-05,
+      "loss": 1.2977,
+      "step": 10826
+    },
+    {
+      "epoch": 0.7804368197217617,
+      "grad_norm": 0.11182425969488739,
+      "learning_rate": 2.4239099043966462e-05,
+      "loss": 1.2729,
+      "step": 10827
+    },
+    {
+      "epoch": 0.7805089021840986,
+      "grad_norm": 0.19631323272045845,
+      "learning_rate": 2.4223862221322424e-05,
+      "loss": 1.2957,
+      "step": 10828
+    },
+    {
+      "epoch": 0.7805809846464355,
+      "grad_norm": 0.11380949616256636,
+      "learning_rate": 2.420862952915577e-05,
+      "loss": 1.3335,
+      "step": 10829
+    },
+    {
+      "epoch": 0.7806530671087725,
+      "grad_norm": 0.12929836468958097,
+      "learning_rate": 2.4193400968296865e-05,
+      "loss": 1.2274,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7807251495711094,
+      "grad_norm": 0.11848025065232566,
+      "learning_rate": 2.4178176539575803e-05,
+      "loss": 1.2867,
+      "step": 10831
+    },
+    {
+      "epoch": 0.7807972320334463,
+      "grad_norm": 0.12280240233995465,
+      "learning_rate": 2.4162956243822464e-05,
+      "loss": 1.2408,
+      "step": 10832
+    },
+    {
+      "epoch": 0.7808693144957832,
+      "grad_norm": 0.12845578152521434,
+      "learning_rate": 2.414774008186642e-05,
+      "loss": 1.338,
+      "step": 10833
+    },
+    {
+      "epoch": 0.78094139695812,
+      "grad_norm": 0.1288723938413484,
+      "learning_rate": 2.413252805453715e-05,
+      "loss": 1.2275,
+      "step": 10834
+    },
+    {
+      "epoch": 0.781013479420457,
+      "grad_norm": 0.1219676131944716,
+      "learning_rate": 2.4117320162663836e-05,
+      "loss": 1.2812,
+      "step": 10835
+    },
+    {
+      "epoch": 0.7810855618827939,
+      "grad_norm": 0.11613408088167178,
+      "learning_rate": 2.410211640707547e-05,
+      "loss": 1.331,
+      "step": 10836
+    },
+    {
+      "epoch": 0.7811576443451308,
+      "grad_norm": 0.11594793048273885,
+      "learning_rate": 2.4086916788600734e-05,
+      "loss": 1.2487,
+      "step": 10837
+    },
+    {
+      "epoch": 0.7812297268074677,
+      "grad_norm": 0.12350004387749318,
+      "learning_rate": 2.407172130806814e-05,
+      "loss": 1.2901,
+      "step": 10838
+    },
+    {
+      "epoch": 0.7813018092698046,
+      "grad_norm": 0.11401800297338159,
+      "learning_rate": 2.405652996630604e-05,
+      "loss": 1.2317,
+      "step": 10839
+    },
+    {
+      "epoch": 0.7813738917321416,
+      "grad_norm": 0.1248612907775712,
+      "learning_rate": 2.4041342764142494e-05,
+      "loss": 1.3116,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7814459741944785,
+      "grad_norm": 0.12409961308974805,
+      "learning_rate": 2.4026159702405282e-05,
+      "loss": 1.2184,
+      "step": 10841
+    },
+    {
+      "epoch": 0.7815180566568154,
+      "grad_norm": 0.12450493477022334,
+      "learning_rate": 2.4010980781922022e-05,
+      "loss": 1.3295,
+      "step": 10842
+    },
+    {
+      "epoch": 0.7815901391191523,
+      "grad_norm": 0.11189474262628403,
+      "learning_rate": 2.3995806003520148e-05,
+      "loss": 1.269,
+      "step": 10843
+    },
+    {
+      "epoch": 0.7816622215814892,
+      "grad_norm": 0.1259783729475738,
+      "learning_rate": 2.3980635368026838e-05,
+      "loss": 1.1708,
+      "step": 10844
+    },
+    {
+      "epoch": 0.7817343040438262,
+      "grad_norm": 0.1197241525135815,
+      "learning_rate": 2.396546887626896e-05,
+      "loss": 1.3286,
+      "step": 10845
+    },
+    {
+      "epoch": 0.7818063865061631,
+      "grad_norm": 0.11853715222895551,
+      "learning_rate": 2.3950306529073218e-05,
+      "loss": 1.2769,
+      "step": 10846
+    },
+    {
+      "epoch": 0.7818784689685,
+      "grad_norm": 0.11566004849649097,
+      "learning_rate": 2.3935148327266167e-05,
+      "loss": 1.2835,
+      "step": 10847
+    },
+    {
+      "epoch": 0.7819505514308369,
+      "grad_norm": 0.1195930204087446,
+      "learning_rate": 2.391999427167404e-05,
+      "loss": 1.3399,
+      "step": 10848
+    },
+    {
+      "epoch": 0.7820226338931738,
+      "grad_norm": 0.12481933383328724,
+      "learning_rate": 2.3904844363122847e-05,
+      "loss": 1.2851,
+      "step": 10849
+    },
+    {
+      "epoch": 0.7820947163555108,
+      "grad_norm": 0.13225534465643424,
+      "learning_rate": 2.388969860243839e-05,
+      "loss": 1.2596,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7821667988178476,
+      "grad_norm": 0.12261764177266168,
+      "learning_rate": 2.387455699044624e-05,
+      "loss": 1.3419,
+      "step": 10851
+    },
+    {
+      "epoch": 0.7822388812801845,
+      "grad_norm": 0.13313876537721378,
+      "learning_rate": 2.3859419527971827e-05,
+      "loss": 1.2712,
+      "step": 10852
+    },
+    {
+      "epoch": 0.7823109637425214,
+      "grad_norm": 0.1138743589281662,
+      "learning_rate": 2.3844286215840193e-05,
+      "loss": 1.1899,
+      "step": 10853
+    },
+    {
+      "epoch": 0.7823830462048583,
+      "grad_norm": 0.11446979759463699,
+      "learning_rate": 2.382915705487626e-05,
+      "loss": 1.2363,
+      "step": 10854
+    },
+    {
+      "epoch": 0.7824551286671952,
+      "grad_norm": 0.11410816706879494,
+      "learning_rate": 2.38140320459047e-05,
+      "loss": 1.2111,
+      "step": 10855
+    },
+    {
+      "epoch": 0.7825272111295322,
+      "grad_norm": 0.11463562759084527,
+      "learning_rate": 2.3798911189750017e-05,
+      "loss": 1.3104,
+      "step": 10856
+    },
+    {
+      "epoch": 0.7825992935918691,
+      "grad_norm": 0.126937060092846,
+      "learning_rate": 2.3783794487236365e-05,
+      "loss": 1.2933,
+      "step": 10857
+    },
+    {
+      "epoch": 0.782671376054206,
+      "grad_norm": 0.11318126715745887,
+      "learning_rate": 2.3768681939187763e-05,
+      "loss": 1.2402,
+      "step": 10858
+    },
+    {
+      "epoch": 0.7827434585165429,
+      "grad_norm": 0.11569392083781684,
+      "learning_rate": 2.3753573546427976e-05,
+      "loss": 1.2014,
+      "step": 10859
+    },
+    {
+      "epoch": 0.7828155409788798,
+      "grad_norm": 0.11278513001722819,
+      "learning_rate": 2.3738469309780554e-05,
+      "loss": 1.2859,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7828876234412168,
+      "grad_norm": 0.11792613047625651,
+      "learning_rate": 2.3723369230068815e-05,
+      "loss": 1.245,
+      "step": 10861
+    },
+    {
+      "epoch": 0.7829597059035537,
+      "grad_norm": 0.11249764877473212,
+      "learning_rate": 2.3708273308115835e-05,
+      "loss": 1.1235,
+      "step": 10862
+    },
+    {
+      "epoch": 0.7830317883658906,
+      "grad_norm": 0.10951856282116525,
+      "learning_rate": 2.3693181544744493e-05,
+      "loss": 1.2642,
+      "step": 10863
+    },
+    {
+      "epoch": 0.7831038708282275,
+      "grad_norm": 0.1180366653476412,
+      "learning_rate": 2.367809394077741e-05,
+      "loss": 1.1988,
+      "step": 10864
+    },
+    {
+      "epoch": 0.7831759532905644,
+      "grad_norm": 0.11261419417077069,
+      "learning_rate": 2.3663010497037007e-05,
+      "loss": 1.2823,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7832480357529014,
+      "grad_norm": 0.12108062773929652,
+      "learning_rate": 2.3647931214345464e-05,
+      "loss": 1.2347,
+      "step": 10866
+    },
+    {
+      "epoch": 0.7833201182152383,
+      "grad_norm": 0.11161367196554679,
+      "learning_rate": 2.3632856093524724e-05,
+      "loss": 1.3357,
+      "step": 10867
+    },
+    {
+      "epoch": 0.7833922006775752,
+      "grad_norm": 0.11404327045734917,
+      "learning_rate": 2.361778513539654e-05,
+      "loss": 1.3243,
+      "step": 10868
+    },
+    {
+      "epoch": 0.783464283139912,
+      "grad_norm": 0.12005155630029404,
+      "learning_rate": 2.360271834078239e-05,
+      "loss": 1.196,
+      "step": 10869
+    },
+    {
+      "epoch": 0.7835363656022489,
+      "grad_norm": 0.11381617080896426,
+      "learning_rate": 2.3587655710503564e-05,
+      "loss": 1.3419,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7836084480645858,
+      "grad_norm": 0.13590013811224524,
+      "learning_rate": 2.3572597245381112e-05,
+      "loss": 1.2939,
+      "step": 10871
+    },
+    {
+      "epoch": 0.7836805305269228,
+      "grad_norm": 0.12156021767564992,
+      "learning_rate": 2.3557542946235844e-05,
+      "loss": 1.21,
+      "step": 10872
+    },
+    {
+      "epoch": 0.7837526129892597,
+      "grad_norm": 0.12193692551256605,
+      "learning_rate": 2.354249281388836e-05,
+      "loss": 1.3708,
+      "step": 10873
+    },
+    {
+      "epoch": 0.7838246954515966,
+      "grad_norm": 0.10981484991835433,
+      "learning_rate": 2.3527446849159017e-05,
+      "loss": 1.2582,
+      "step": 10874
+    },
+    {
+      "epoch": 0.7838967779139335,
+      "grad_norm": 0.1156092298841586,
+      "learning_rate": 2.351240505286798e-05,
+      "loss": 1.2466,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7839688603762704,
+      "grad_norm": 0.13658999584651782,
+      "learning_rate": 2.3497367425835138e-05,
+      "loss": 1.3012,
+      "step": 10876
+    },
+    {
+      "epoch": 0.7840409428386074,
+      "grad_norm": 0.10711795738942266,
+      "learning_rate": 2.3482333968880176e-05,
+      "loss": 1.2118,
+      "step": 10877
+    },
+    {
+      "epoch": 0.7841130253009443,
+      "grad_norm": 0.11332282331607077,
+      "learning_rate": 2.3467304682822565e-05,
+      "loss": 1.2709,
+      "step": 10878
+    },
+    {
+      "epoch": 0.7841851077632812,
+      "grad_norm": 0.11477183278504717,
+      "learning_rate": 2.345227956848154e-05,
+      "loss": 1.2931,
+      "step": 10879
+    },
+    {
+      "epoch": 0.7842571902256181,
+      "grad_norm": 0.12044344502172663,
+      "learning_rate": 2.3437258626676084e-05,
+      "loss": 1.2516,
+      "step": 10880
+    },
+    {
+      "epoch": 0.784329272687955,
+      "grad_norm": 0.11806636297052343,
+      "learning_rate": 2.3422241858225013e-05,
+      "loss": 1.2039,
+      "step": 10881
+    },
+    {
+      "epoch": 0.784401355150292,
+      "grad_norm": 0.11730670089185635,
+      "learning_rate": 2.3407229263946784e-05,
+      "loss": 1.2877,
+      "step": 10882
+    },
+    {
+      "epoch": 0.7844734376126289,
+      "grad_norm": 0.1071209489322999,
+      "learning_rate": 2.3392220844659806e-05,
+      "loss": 1.1088,
+      "step": 10883
+    },
+    {
+      "epoch": 0.7845455200749658,
+      "grad_norm": 0.12526715480143688,
+      "learning_rate": 2.337721660118215e-05,
+      "loss": 1.2981,
+      "step": 10884
+    },
+    {
+      "epoch": 0.7846176025373027,
+      "grad_norm": 0.1277908794312501,
+      "learning_rate": 2.3362216534331705e-05,
+      "loss": 1.4175,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7846896849996395,
+      "grad_norm": 0.11796481088037093,
+      "learning_rate": 2.3347220644926028e-05,
+      "loss": 1.2022,
+      "step": 10886
+    },
+    {
+      "epoch": 0.7847617674619765,
+      "grad_norm": 0.12516731021254332,
+      "learning_rate": 2.3332228933782596e-05,
+      "loss": 1.2095,
+      "step": 10887
+    },
+    {
+      "epoch": 0.7848338499243134,
+      "grad_norm": 0.1389957184221273,
+      "learning_rate": 2.3317241401718592e-05,
+      "loss": 1.2286,
+      "step": 10888
+    },
+    {
+      "epoch": 0.7849059323866503,
+      "grad_norm": 0.1272383866712457,
+      "learning_rate": 2.3302258049550974e-05,
+      "loss": 1.1856,
+      "step": 10889
+    },
+    {
+      "epoch": 0.7849780148489872,
+      "grad_norm": 0.12832912103100222,
+      "learning_rate": 2.3287278878096396e-05,
+      "loss": 1.3745,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7850500973113241,
+      "grad_norm": 0.13444384514632904,
+      "learning_rate": 2.3272303888171432e-05,
+      "loss": 1.1764,
+      "step": 10891
+    },
+    {
+      "epoch": 0.785122179773661,
+      "grad_norm": 0.1290391867710194,
+      "learning_rate": 2.3257333080592368e-05,
+      "loss": 1.2932,
+      "step": 10892
+    },
+    {
+      "epoch": 0.785194262235998,
+      "grad_norm": 0.10986537893145615,
+      "learning_rate": 2.324236645617518e-05,
+      "loss": 1.2868,
+      "step": 10893
+    },
+    {
+      "epoch": 0.7852663446983349,
+      "grad_norm": 0.22454213188372638,
+      "learning_rate": 2.322740401573571e-05,
+      "loss": 1.349,
+      "step": 10894
+    },
+    {
+      "epoch": 0.7853384271606718,
+      "grad_norm": 0.12452213964495694,
+      "learning_rate": 2.321244576008953e-05,
+      "loss": 1.1966,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7854105096230087,
+      "grad_norm": 0.12922281765699256,
+      "learning_rate": 2.3197491690052065e-05,
+      "loss": 1.2673,
+      "step": 10896
+    },
+    {
+      "epoch": 0.7854825920853457,
+      "grad_norm": 0.1146352208405936,
+      "learning_rate": 2.3182541806438374e-05,
+      "loss": 1.3622,
+      "step": 10897
+    },
+    {
+      "epoch": 0.7855546745476826,
+      "grad_norm": 0.11515699389585003,
+      "learning_rate": 2.3167596110063384e-05,
+      "loss": 1.3072,
+      "step": 10898
+    },
+    {
+      "epoch": 0.7856267570100195,
+      "grad_norm": 0.1241751715371345,
+      "learning_rate": 2.315265460174173e-05,
+      "loss": 1.3298,
+      "step": 10899
+    },
+    {
+      "epoch": 0.7856988394723564,
+      "grad_norm": 0.11583142456202433,
+      "learning_rate": 2.3137717282287963e-05,
+      "loss": 1.3041,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7857709219346933,
+      "grad_norm": 0.12541568977558054,
+      "learning_rate": 2.3122784152516208e-05,
+      "loss": 1.2154,
+      "step": 10901
+    },
+    {
+      "epoch": 0.7858430043970303,
+      "grad_norm": 0.11181417011613505,
+      "learning_rate": 2.310785521324047e-05,
+      "loss": 1.343,
+      "step": 10902
+    },
+    {
+      "epoch": 0.7859150868593671,
+      "grad_norm": 0.11699226337748506,
+      "learning_rate": 2.309293046527452e-05,
+      "loss": 1.2736,
+      "step": 10903
+    },
+    {
+      "epoch": 0.785987169321704,
+      "grad_norm": 0.12341167077131664,
+      "learning_rate": 2.3078009909431896e-05,
+      "loss": 1.2581,
+      "step": 10904
+    },
+    {
+      "epoch": 0.7860592517840409,
+      "grad_norm": 0.10823657657260206,
+      "learning_rate": 2.3063093546525892e-05,
+      "loss": 1.3435,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7861313342463778,
+      "grad_norm": 0.1377021019544597,
+      "learning_rate": 2.3048181377369592e-05,
+      "loss": 1.2582,
+      "step": 10906
+    },
+    {
+      "epoch": 0.7862034167087147,
+      "grad_norm": 0.11876577492002947,
+      "learning_rate": 2.3033273402775833e-05,
+      "loss": 1.2864,
+      "step": 10907
+    },
+    {
+      "epoch": 0.7862754991710517,
+      "grad_norm": 0.11917659541889951,
+      "learning_rate": 2.3018369623557244e-05,
+      "loss": 1.264,
+      "step": 10908
+    },
+    {
+      "epoch": 0.7863475816333886,
+      "grad_norm": 0.11789097887837885,
+      "learning_rate": 2.30034700405262e-05,
+      "loss": 1.3392,
+      "step": 10909
+    },
+    {
+      "epoch": 0.7864196640957255,
+      "grad_norm": 0.1201410711905354,
+      "learning_rate": 2.2988574654494875e-05,
+      "loss": 1.2494,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7864917465580624,
+      "grad_norm": 0.1163623183665073,
+      "learning_rate": 2.2973683466275188e-05,
+      "loss": 1.2006,
+      "step": 10911
+    },
+    {
+      "epoch": 0.7865638290203993,
+      "grad_norm": 0.11463055520527457,
+      "learning_rate": 2.2958796476678857e-05,
+      "loss": 1.3032,
+      "step": 10912
+    },
+    {
+      "epoch": 0.7866359114827363,
+      "grad_norm": 0.12958034045574057,
+      "learning_rate": 2.294391368651735e-05,
+      "loss": 1.1798,
+      "step": 10913
+    },
+    {
+      "epoch": 0.7867079939450732,
+      "grad_norm": 0.11435127846139166,
+      "learning_rate": 2.2929035096601905e-05,
+      "loss": 1.2928,
+      "step": 10914
+    },
+    {
+      "epoch": 0.7867800764074101,
+      "grad_norm": 0.1152469744510159,
+      "learning_rate": 2.2914160707743538e-05,
+      "loss": 1.3356,
+      "step": 10915
+    },
+    {
+      "epoch": 0.786852158869747,
+      "grad_norm": 0.1238639685966433,
+      "learning_rate": 2.2899290520753046e-05,
+      "loss": 1.3426,
+      "step": 10916
+    },
+    {
+      "epoch": 0.7869242413320839,
+      "grad_norm": 0.11769678600229816,
+      "learning_rate": 2.2884424536440985e-05,
+      "loss": 1.3145,
+      "step": 10917
+    },
+    {
+      "epoch": 0.7869963237944209,
+      "grad_norm": 0.10988428572943564,
+      "learning_rate": 2.286956275561769e-05,
+      "loss": 1.3132,
+      "step": 10918
+    },
+    {
+      "epoch": 0.7870684062567578,
+      "grad_norm": 0.13218312167309737,
+      "learning_rate": 2.2854705179093238e-05,
+      "loss": 1.3601,
+      "step": 10919
+    },
+    {
+      "epoch": 0.7871404887190947,
+      "grad_norm": 0.13845334536786302,
+      "learning_rate": 2.2839851807677527e-05,
+      "loss": 1.2969,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7872125711814315,
+      "grad_norm": 0.13487894899917893,
+      "learning_rate": 2.2825002642180182e-05,
+      "loss": 1.3052,
+      "step": 10921
+    },
+    {
+      "epoch": 0.7872846536437684,
+      "grad_norm": 0.1255699265853035,
+      "learning_rate": 2.2810157683410627e-05,
+      "loss": 1.3059,
+      "step": 10922
+    },
+    {
+      "epoch": 0.7873567361061053,
+      "grad_norm": 0.13258174184316127,
+      "learning_rate": 2.2795316932178036e-05,
+      "loss": 1.3229,
+      "step": 10923
+    },
+    {
+      "epoch": 0.7874288185684423,
+      "grad_norm": 0.1209545796862831,
+      "learning_rate": 2.278048038929136e-05,
+      "loss": 1.2682,
+      "step": 10924
+    },
+    {
+      "epoch": 0.7875009010307792,
+      "grad_norm": 0.13093639345027402,
+      "learning_rate": 2.2765648055559353e-05,
+      "loss": 1.2153,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7875729834931161,
+      "grad_norm": 0.14088485940734388,
+      "learning_rate": 2.2750819931790436e-05,
+      "loss": 1.2865,
+      "step": 10926
+    },
+    {
+      "epoch": 0.787645065955453,
+      "grad_norm": 0.13362436553115584,
+      "learning_rate": 2.273599601879295e-05,
+      "loss": 1.2522,
+      "step": 10927
+    },
+    {
+      "epoch": 0.78771714841779,
+      "grad_norm": 0.10795709048049527,
+      "learning_rate": 2.2721176317374903e-05,
+      "loss": 1.1901,
+      "step": 10928
+    },
+    {
+      "epoch": 0.7877892308801269,
+      "grad_norm": 0.11331620635175235,
+      "learning_rate": 2.270636082834414e-05,
+      "loss": 1.3162,
+      "step": 10929
+    },
+    {
+      "epoch": 0.7878613133424638,
+      "grad_norm": 0.10796823883324799,
+      "learning_rate": 2.2691549552508128e-05,
+      "loss": 1.2425,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7879333958048007,
+      "grad_norm": 0.11811888302385984,
+      "learning_rate": 2.267674249067433e-05,
+      "loss": 1.2947,
+      "step": 10931
+    },
+    {
+      "epoch": 0.7880054782671376,
+      "grad_norm": 0.12505473780981502,
+      "learning_rate": 2.266193964364981e-05,
+      "loss": 1.2919,
+      "step": 10932
+    },
+    {
+      "epoch": 0.7880775607294745,
+      "grad_norm": 0.13254347867226438,
+      "learning_rate": 2.26471410122415e-05,
+      "loss": 1.3002,
+      "step": 10933
+    },
+    {
+      "epoch": 0.7881496431918115,
+      "grad_norm": 0.11352766008561452,
+      "learning_rate": 2.263234659725596e-05,
+      "loss": 1.3207,
+      "step": 10934
+    },
+    {
+      "epoch": 0.7882217256541484,
+      "grad_norm": 0.11682559630407302,
+      "learning_rate": 2.2617556399499716e-05,
+      "loss": 1.3044,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7882938081164853,
+      "grad_norm": 0.12721409989060567,
+      "learning_rate": 2.260277041977893e-05,
+      "loss": 1.1734,
+      "step": 10936
+    },
+    {
+      "epoch": 0.7883658905788222,
+      "grad_norm": 0.12252656989105955,
+      "learning_rate": 2.2587988658899606e-05,
+      "loss": 1.2862,
+      "step": 10937
+    },
+    {
+      "epoch": 0.788437973041159,
+      "grad_norm": 0.12111897957576188,
+      "learning_rate": 2.2573211117667404e-05,
+      "loss": 1.238,
+      "step": 10938
+    },
+    {
+      "epoch": 0.788510055503496,
+      "grad_norm": 0.11193694284684144,
+      "learning_rate": 2.2558437796887853e-05,
+      "loss": 1.3002,
+      "step": 10939
+    },
+    {
+      "epoch": 0.7885821379658329,
+      "grad_norm": 0.10608600685817032,
+      "learning_rate": 2.2543668697366293e-05,
+      "loss": 1.1762,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7886542204281698,
+      "grad_norm": 0.11869961585484892,
+      "learning_rate": 2.2528903819907752e-05,
+      "loss": 1.2179,
+      "step": 10941
+    },
+    {
+      "epoch": 0.7887263028905067,
+      "grad_norm": 0.12724472413275997,
+      "learning_rate": 2.2514143165317002e-05,
+      "loss": 1.2295,
+      "step": 10942
+    },
+    {
+      "epoch": 0.7887983853528436,
+      "grad_norm": 0.13358046630655393,
+      "learning_rate": 2.249938673439863e-05,
+      "loss": 1.252,
+      "step": 10943
+    },
+    {
+      "epoch": 0.7888704678151806,
+      "grad_norm": 0.11567778808612254,
+      "learning_rate": 2.248463452795705e-05,
+      "loss": 1.3067,
+      "step": 10944
+    },
+    {
+      "epoch": 0.7889425502775175,
+      "grad_norm": 0.11276329321342002,
+      "learning_rate": 2.246988654679639e-05,
+      "loss": 1.2791,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7890146327398544,
+      "grad_norm": 0.11537855889988859,
+      "learning_rate": 2.2455142791720486e-05,
+      "loss": 1.2506,
+      "step": 10946
+    },
+    {
+      "epoch": 0.7890867152021913,
+      "grad_norm": 0.11094466701872742,
+      "learning_rate": 2.244040326353304e-05,
+      "loss": 1.1922,
+      "step": 10947
+    },
+    {
+      "epoch": 0.7891587976645282,
+      "grad_norm": 0.12383954880093201,
+      "learning_rate": 2.2425667963037454e-05,
+      "loss": 1.2648,
+      "step": 10948
+    },
+    {
+      "epoch": 0.7892308801268652,
+      "grad_norm": 0.1211186139564945,
+      "learning_rate": 2.2410936891037015e-05,
+      "loss": 1.2777,
+      "step": 10949
+    },
+    {
+      "epoch": 0.7893029625892021,
+      "grad_norm": 0.10712145612686162,
+      "learning_rate": 2.2396210048334622e-05,
+      "loss": 1.2557,
+      "step": 10950
+    },
+    {
+      "epoch": 0.789375045051539,
+      "grad_norm": 0.11752204344350783,
+      "learning_rate": 2.238148743573305e-05,
+      "loss": 1.2735,
+      "step": 10951
+    },
+    {
+      "epoch": 0.7894471275138759,
+      "grad_norm": 0.12460951384845138,
+      "learning_rate": 2.2366769054034775e-05,
+      "loss": 1.2985,
+      "step": 10952
+    },
+    {
+      "epoch": 0.7895192099762128,
+      "grad_norm": 0.10450340427437022,
+      "learning_rate": 2.2352054904042174e-05,
+      "loss": 1.2378,
+      "step": 10953
+    },
+    {
+      "epoch": 0.7895912924385498,
+      "grad_norm": 0.11869988184133964,
+      "learning_rate": 2.2337344986557208e-05,
+      "loss": 1.2958,
+      "step": 10954
+    },
+    {
+      "epoch": 0.7896633749008866,
+      "grad_norm": 0.11627280398176813,
+      "learning_rate": 2.232263930238172e-05,
+      "loss": 1.297,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7897354573632235,
+      "grad_norm": 0.10614631101446728,
+      "learning_rate": 2.2307937852317296e-05,
+      "loss": 1.2173,
+      "step": 10956
+    },
+    {
+      "epoch": 0.7898075398255604,
+      "grad_norm": 0.1188511198293265,
+      "learning_rate": 2.229324063716537e-05,
+      "loss": 1.2874,
+      "step": 10957
+    },
+    {
+      "epoch": 0.7898796222878973,
+      "grad_norm": 0.105435326376427,
+      "learning_rate": 2.2278547657726978e-05,
+      "loss": 1.3026,
+      "step": 10958
+    },
+    {
+      "epoch": 0.7899517047502342,
+      "grad_norm": 0.10639662487081032,
+      "learning_rate": 2.226385891480306e-05,
+      "loss": 1.28,
+      "step": 10959
+    },
+    {
+      "epoch": 0.7900237872125712,
+      "grad_norm": 0.12057785025479442,
+      "learning_rate": 2.2249174409194274e-05,
+      "loss": 1.2903,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7900958696749081,
+      "grad_norm": 0.11912395941507196,
+      "learning_rate": 2.223449414170107e-05,
+      "loss": 1.2738,
+      "step": 10961
+    },
+    {
+      "epoch": 0.790167952137245,
+      "grad_norm": 0.11831037002674719,
+      "learning_rate": 2.2219818113123635e-05,
+      "loss": 1.1807,
+      "step": 10962
+    },
+    {
+      "epoch": 0.7902400345995819,
+      "grad_norm": 0.21274751833843691,
+      "learning_rate": 2.220514632426196e-05,
+      "loss": 1.1898,
+      "step": 10963
+    },
+    {
+      "epoch": 0.7903121170619188,
+      "grad_norm": 0.1253422102522692,
+      "learning_rate": 2.219047877591578e-05,
+      "loss": 1.3344,
+      "step": 10964
+    },
+    {
+      "epoch": 0.7903841995242558,
+      "grad_norm": 0.16528242993028966,
+      "learning_rate": 2.217581546888461e-05,
+      "loss": 1.364,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7904562819865927,
+      "grad_norm": 0.1273408540445682,
+      "learning_rate": 2.2161156403967733e-05,
+      "loss": 1.1966,
+      "step": 10966
+    },
+    {
+      "epoch": 0.7905283644489296,
+      "grad_norm": 0.12044055256114213,
+      "learning_rate": 2.2146501581964196e-05,
+      "loss": 1.2797,
+      "step": 10967
+    },
+    {
+      "epoch": 0.7906004469112665,
+      "grad_norm": 0.11666970586878586,
+      "learning_rate": 2.213185100367282e-05,
+      "loss": 1.2375,
+      "step": 10968
+    },
+    {
+      "epoch": 0.7906725293736034,
+      "grad_norm": 0.1248448154809287,
+      "learning_rate": 2.2117204669892198e-05,
+      "loss": 1.214,
+      "step": 10969
+    },
+    {
+      "epoch": 0.7907446118359404,
+      "grad_norm": 0.12845867048204818,
+      "learning_rate": 2.210256258142068e-05,
+      "loss": 1.3025,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7908166942982773,
+      "grad_norm": 0.12278993039095402,
+      "learning_rate": 2.2087924739056387e-05,
+      "loss": 1.2708,
+      "step": 10971
+    },
+    {
+      "epoch": 0.7908887767606141,
+      "grad_norm": 0.1184045879727166,
+      "learning_rate": 2.207329114359723e-05,
+      "loss": 1.2177,
+      "step": 10972
+    },
+    {
+      "epoch": 0.790960859222951,
+      "grad_norm": 0.11738197269798098,
+      "learning_rate": 2.205866179584084e-05,
+      "loss": 1.2079,
+      "step": 10973
+    },
+    {
+      "epoch": 0.7910329416852879,
+      "grad_norm": 0.12977177086118669,
+      "learning_rate": 2.204403669658468e-05,
+      "loss": 1.3442,
+      "step": 10974
+    },
+    {
+      "epoch": 0.7911050241476248,
+      "grad_norm": 0.12425349455044513,
+      "learning_rate": 2.202941584662592e-05,
+      "loss": 1.3282,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7911771066099618,
+      "grad_norm": 0.12473149067706724,
+      "learning_rate": 2.2014799246761553e-05,
+      "loss": 1.2953,
+      "step": 10976
+    },
+    {
+      "epoch": 0.7912491890722987,
+      "grad_norm": 0.12227666870882288,
+      "learning_rate": 2.2000186897788298e-05,
+      "loss": 1.3106,
+      "step": 10977
+    },
+    {
+      "epoch": 0.7913212715346356,
+      "grad_norm": 0.12202472110464055,
+      "learning_rate": 2.198557880050266e-05,
+      "loss": 1.304,
+      "step": 10978
+    },
+    {
+      "epoch": 0.7913933539969725,
+      "grad_norm": 0.12697698846156125,
+      "learning_rate": 2.1970974955700917e-05,
+      "loss": 1.3477,
+      "step": 10979
+    },
+    {
+      "epoch": 0.7914654364593094,
+      "grad_norm": 0.11755746242342899,
+      "learning_rate": 2.1956375364179116e-05,
+      "loss": 1.2551,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7915375189216464,
+      "grad_norm": 0.12987193687150594,
+      "learning_rate": 2.194178002673304e-05,
+      "loss": 1.2894,
+      "step": 10981
+    },
+    {
+      "epoch": 0.7916096013839833,
+      "grad_norm": 0.12552644506594324,
+      "learning_rate": 2.1927188944158315e-05,
+      "loss": 1.1982,
+      "step": 10982
+    },
+    {
+      "epoch": 0.7916816838463202,
+      "grad_norm": 0.1258935054309892,
+      "learning_rate": 2.1912602117250203e-05,
+      "loss": 1.2822,
+      "step": 10983
+    },
+    {
+      "epoch": 0.7917537663086571,
+      "grad_norm": 0.11817012066348649,
+      "learning_rate": 2.1898019546803882e-05,
+      "loss": 1.2483,
+      "step": 10984
+    },
+    {
+      "epoch": 0.791825848770994,
+      "grad_norm": 0.11583370028047803,
+      "learning_rate": 2.188344123361423e-05,
+      "loss": 1.2333,
+      "step": 10985
+    },
+    {
+      "epoch": 0.791897931233331,
+      "grad_norm": 0.13962309075303433,
+      "learning_rate": 2.1868867178475893e-05,
+      "loss": 1.3218,
+      "step": 10986
+    },
+    {
+      "epoch": 0.7919700136956679,
+      "grad_norm": 0.1089358910443372,
+      "learning_rate": 2.185429738218323e-05,
+      "loss": 1.2378,
+      "step": 10987
+    },
+    {
+      "epoch": 0.7920420961580048,
+      "grad_norm": 0.11365131624964386,
+      "learning_rate": 2.18397318455305e-05,
+      "loss": 1.3429,
+      "step": 10988
+    },
+    {
+      "epoch": 0.7921141786203417,
+      "grad_norm": 0.1317828699734514,
+      "learning_rate": 2.1825170569311627e-05,
+      "loss": 1.351,
+      "step": 10989
+    },
+    {
+      "epoch": 0.7921862610826785,
+      "grad_norm": 0.10449360786966574,
+      "learning_rate": 2.181061355432036e-05,
+      "loss": 1.1998,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7922583435450155,
+      "grad_norm": 0.12431287178167837,
+      "learning_rate": 2.1796060801350116e-05,
+      "loss": 1.2965,
+      "step": 10991
+    },
+    {
+      "epoch": 0.7923304260073524,
+      "grad_norm": 0.12850978965821705,
+      "learning_rate": 2.1781512311194173e-05,
+      "loss": 1.4412,
+      "step": 10992
+    },
+    {
+      "epoch": 0.7924025084696893,
+      "grad_norm": 0.1200644904987094,
+      "learning_rate": 2.176696808464559e-05,
+      "loss": 1.3064,
+      "step": 10993
+    },
+    {
+      "epoch": 0.7924745909320262,
+      "grad_norm": 0.20356009891197932,
+      "learning_rate": 2.1752428122497182e-05,
+      "loss": 1.2369,
+      "step": 10994
+    },
+    {
+      "epoch": 0.7925466733943631,
+      "grad_norm": 0.12999287123455697,
+      "learning_rate": 2.1737892425541416e-05,
+      "loss": 1.3697,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7926187558567,
+      "grad_norm": 0.12191280841183841,
+      "learning_rate": 2.172336099457065e-05,
+      "loss": 1.2734,
+      "step": 10996
+    },
+    {
+      "epoch": 0.792690838319037,
+      "grad_norm": 0.11099596768164563,
+      "learning_rate": 2.170883383037703e-05,
+      "loss": 1.3356,
+      "step": 10997
+    },
+    {
+      "epoch": 0.7927629207813739,
+      "grad_norm": 0.12659429060240507,
+      "learning_rate": 2.1694310933752395e-05,
+      "loss": 1.2838,
+      "step": 10998
+    },
+    {
+      "epoch": 0.7928350032437108,
+      "grad_norm": 0.12319927942268069,
+      "learning_rate": 2.167979230548832e-05,
+      "loss": 1.2245,
+      "step": 10999
+    },
+    {
+      "epoch": 0.7929070857060477,
+      "grad_norm": 0.12218443438988617,
+      "learning_rate": 2.166527794637623e-05,
+      "loss": 1.1389,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7929791681683847,
+      "grad_norm": 0.11597930433270762,
+      "learning_rate": 2.165076785720731e-05,
+      "loss": 1.2143,
+      "step": 11001
+    },
+    {
+      "epoch": 0.7930512506307216,
+      "grad_norm": 0.11616167330428127,
+      "learning_rate": 2.1636262038772504e-05,
+      "loss": 1.248,
+      "step": 11002
+    },
+    {
+      "epoch": 0.7931233330930585,
+      "grad_norm": 0.13666806229480125,
+      "learning_rate": 2.162176049186245e-05,
+      "loss": 1.2463,
+      "step": 11003
+    },
+    {
+      "epoch": 0.7931954155553954,
+      "grad_norm": 0.11027181513990768,
+      "learning_rate": 2.160726321726764e-05,
+      "loss": 1.3523,
+      "step": 11004
+    },
+    {
+      "epoch": 0.7932674980177323,
+      "grad_norm": 0.1265521259903243,
+      "learning_rate": 2.159277021577828e-05,
+      "loss": 1.2571,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7933395804800693,
+      "grad_norm": 0.1267487554885954,
+      "learning_rate": 2.157828148818445e-05,
+      "loss": 1.2925,
+      "step": 11006
+    },
+    {
+      "epoch": 0.7934116629424061,
+      "grad_norm": 0.13087434162550804,
+      "learning_rate": 2.1563797035275836e-05,
+      "loss": 1.304,
+      "step": 11007
+    },
+    {
+      "epoch": 0.793483745404743,
+      "grad_norm": 0.1122253256242175,
+      "learning_rate": 2.1549316857841993e-05,
+      "loss": 1.2591,
+      "step": 11008
+    },
+    {
+      "epoch": 0.7935558278670799,
+      "grad_norm": 0.11911743026212981,
+      "learning_rate": 2.1534840956672186e-05,
+      "loss": 1.2605,
+      "step": 11009
+    },
+    {
+      "epoch": 0.7936279103294168,
+      "grad_norm": 0.1352820633579962,
+      "learning_rate": 2.1520369332555577e-05,
+      "loss": 1.3097,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7936999927917537,
+      "grad_norm": 0.12017456985797988,
+      "learning_rate": 2.1505901986280908e-05,
+      "loss": 1.1805,
+      "step": 11011
+    },
+    {
+      "epoch": 0.7937720752540907,
+      "grad_norm": 0.11582799865538884,
+      "learning_rate": 2.149143891863681e-05,
+      "loss": 1.3585,
+      "step": 11012
+    },
+    {
+      "epoch": 0.7938441577164276,
+      "grad_norm": 0.1241352718166538,
+      "learning_rate": 2.147698013041165e-05,
+      "loss": 1.2445,
+      "step": 11013
+    },
+    {
+      "epoch": 0.7939162401787645,
+      "grad_norm": 0.11522775980807158,
+      "learning_rate": 2.146252562239356e-05,
+      "loss": 1.2576,
+      "step": 11014
+    },
+    {
+      "epoch": 0.7939883226411014,
+      "grad_norm": 0.12141706945282324,
+      "learning_rate": 2.1448075395370437e-05,
+      "loss": 1.2319,
+      "step": 11015
+    },
+    {
+      "epoch": 0.7940604051034383,
+      "grad_norm": 0.1454668913010386,
+      "learning_rate": 2.1433629450129954e-05,
+      "loss": 1.354,
+      "step": 11016
+    },
+    {
+      "epoch": 0.7941324875657753,
+      "grad_norm": 0.10607012173671518,
+      "learning_rate": 2.141918778745954e-05,
+      "loss": 1.2987,
+      "step": 11017
+    },
+    {
+      "epoch": 0.7942045700281122,
+      "grad_norm": 0.1111605441377713,
+      "learning_rate": 2.140475040814639e-05,
+      "loss": 1.3351,
+      "step": 11018
+    },
+    {
+      "epoch": 0.7942766524904491,
+      "grad_norm": 0.1230297623455713,
+      "learning_rate": 2.1390317312977482e-05,
+      "loss": 1.3665,
+      "step": 11019
+    },
+    {
+      "epoch": 0.794348734952786,
+      "grad_norm": 0.11885981473659597,
+      "learning_rate": 2.137588850273955e-05,
+      "loss": 1.2661,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7944208174151229,
+      "grad_norm": 0.1220098667717008,
+      "learning_rate": 2.1361463978219077e-05,
+      "loss": 1.2895,
+      "step": 11021
+    },
+    {
+      "epoch": 0.7944928998774599,
+      "grad_norm": 0.11318447336081398,
+      "learning_rate": 2.134704374020233e-05,
+      "loss": 1.2731,
+      "step": 11022
+    },
+    {
+      "epoch": 0.7945649823397968,
+      "grad_norm": 0.12680548821204385,
+      "learning_rate": 2.1332627789475358e-05,
+      "loss": 1.277,
+      "step": 11023
+    },
+    {
+      "epoch": 0.7946370648021336,
+      "grad_norm": 0.11413882043713558,
+      "learning_rate": 2.1318216126823952e-05,
+      "loss": 1.3981,
+      "step": 11024
+    },
+    {
+      "epoch": 0.7947091472644705,
+      "grad_norm": 0.12475461642801021,
+      "learning_rate": 2.1303808753033672e-05,
+      "loss": 1.1855,
+      "step": 11025
+    },
+    {
+      "epoch": 0.7947812297268074,
+      "grad_norm": 0.12919023701645763,
+      "learning_rate": 2.128940566888985e-05,
+      "loss": 1.1528,
+      "step": 11026
+    },
+    {
+      "epoch": 0.7948533121891443,
+      "grad_norm": 0.11574318158137177,
+      "learning_rate": 2.1275006875177584e-05,
+      "loss": 1.228,
+      "step": 11027
+    },
+    {
+      "epoch": 0.7949253946514813,
+      "grad_norm": 0.11859171874043026,
+      "learning_rate": 2.1260612372681744e-05,
+      "loss": 1.3297,
+      "step": 11028
+    },
+    {
+      "epoch": 0.7949974771138182,
+      "grad_norm": 0.12340413016066595,
+      "learning_rate": 2.1246222162186935e-05,
+      "loss": 1.254,
+      "step": 11029
+    },
+    {
+      "epoch": 0.7950695595761551,
+      "grad_norm": 0.11434480896539527,
+      "learning_rate": 2.123183624447759e-05,
+      "loss": 1.2127,
+      "step": 11030
+    },
+    {
+      "epoch": 0.795141642038492,
+      "grad_norm": 0.11442908594564848,
+      "learning_rate": 2.121745462033784e-05,
+      "loss": 1.2559,
+      "step": 11031
+    },
+    {
+      "epoch": 0.795213724500829,
+      "grad_norm": 0.1263806274729562,
+      "learning_rate": 2.120307729055162e-05,
+      "loss": 1.2731,
+      "step": 11032
+    },
+    {
+      "epoch": 0.7952858069631659,
+      "grad_norm": 0.11331731991853808,
+      "learning_rate": 2.1188704255902626e-05,
+      "loss": 1.261,
+      "step": 11033
+    },
+    {
+      "epoch": 0.7953578894255028,
+      "grad_norm": 0.1329344174411992,
+      "learning_rate": 2.1174335517174317e-05,
+      "loss": 1.2797,
+      "step": 11034
+    },
+    {
+      "epoch": 0.7954299718878397,
+      "grad_norm": 0.1158110567857461,
+      "learning_rate": 2.1159971075149944e-05,
+      "loss": 1.3677,
+      "step": 11035
+    },
+    {
+      "epoch": 0.7955020543501766,
+      "grad_norm": 0.12079472760525874,
+      "learning_rate": 2.1145610930612413e-05,
+      "loss": 1.2345,
+      "step": 11036
+    },
+    {
+      "epoch": 0.7955741368125135,
+      "grad_norm": 0.11697072843146639,
+      "learning_rate": 2.113125508434457e-05,
+      "loss": 1.2253,
+      "step": 11037
+    },
+    {
+      "epoch": 0.7956462192748505,
+      "grad_norm": 0.12198579278787247,
+      "learning_rate": 2.1116903537128907e-05,
+      "loss": 1.3219,
+      "step": 11038
+    },
+    {
+      "epoch": 0.7957183017371874,
+      "grad_norm": 0.1312199063764888,
+      "learning_rate": 2.110255628974773e-05,
+      "loss": 1.3052,
+      "step": 11039
+    },
+    {
+      "epoch": 0.7957903841995243,
+      "grad_norm": 0.11352806409562045,
+      "learning_rate": 2.1088213342983022e-05,
+      "loss": 1.2199,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7958624666618612,
+      "grad_norm": 0.10520317539988411,
+      "learning_rate": 2.107387469761668e-05,
+      "loss": 1.1658,
+      "step": 11041
+    },
+    {
+      "epoch": 0.795934549124198,
+      "grad_norm": 0.11842367169433518,
+      "learning_rate": 2.1059540354430264e-05,
+      "loss": 1.3771,
+      "step": 11042
+    },
+    {
+      "epoch": 0.796006631586535,
+      "grad_norm": 0.1274798719780482,
+      "learning_rate": 2.1045210314205134e-05,
+      "loss": 1.3189,
+      "step": 11043
+    },
+    {
+      "epoch": 0.7960787140488719,
+      "grad_norm": 0.11936296570392183,
+      "learning_rate": 2.1030884577722344e-05,
+      "loss": 1.3011,
+      "step": 11044
+    },
+    {
+      "epoch": 0.7961507965112088,
+      "grad_norm": 0.12885393167498746,
+      "learning_rate": 2.1016563145762857e-05,
+      "loss": 1.2301,
+      "step": 11045
+    },
+    {
+      "epoch": 0.7962228789735457,
+      "grad_norm": 0.11923453173739229,
+      "learning_rate": 2.1002246019107308e-05,
+      "loss": 1.1999,
+      "step": 11046
+    },
+    {
+      "epoch": 0.7962949614358826,
+      "grad_norm": 0.1443963147780714,
+      "learning_rate": 2.098793319853605e-05,
+      "loss": 1.2159,
+      "step": 11047
+    },
+    {
+      "epoch": 0.7963670438982196,
+      "grad_norm": 0.12560582887945138,
+      "learning_rate": 2.0973624684829296e-05,
+      "loss": 1.1941,
+      "step": 11048
+    },
+    {
+      "epoch": 0.7964391263605565,
+      "grad_norm": 0.13288079360254823,
+      "learning_rate": 2.0959320478766965e-05,
+      "loss": 1.1741,
+      "step": 11049
+    },
+    {
+      "epoch": 0.7965112088228934,
+      "grad_norm": 0.11892106676586349,
+      "learning_rate": 2.094502058112884e-05,
+      "loss": 1.2915,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7965832912852303,
+      "grad_norm": 0.1113774525816408,
+      "learning_rate": 2.0930724992694318e-05,
+      "loss": 1.2398,
+      "step": 11051
+    },
+    {
+      "epoch": 0.7966553737475672,
+      "grad_norm": 0.1280198248268832,
+      "learning_rate": 2.0916433714242646e-05,
+      "loss": 1.3119,
+      "step": 11052
+    },
+    {
+      "epoch": 0.7967274562099042,
+      "grad_norm": 0.12298635686453699,
+      "learning_rate": 2.090214674655282e-05,
+      "loss": 1.1293,
+      "step": 11053
+    },
+    {
+      "epoch": 0.7967995386722411,
+      "grad_norm": 0.12038626829551567,
+      "learning_rate": 2.0887864090403676e-05,
+      "loss": 1.3319,
+      "step": 11054
+    },
+    {
+      "epoch": 0.796871621134578,
+      "grad_norm": 0.11948760372963513,
+      "learning_rate": 2.0873585746573666e-05,
+      "loss": 1.2565,
+      "step": 11055
+    },
+    {
+      "epoch": 0.7969437035969149,
+      "grad_norm": 0.12610439679248825,
+      "learning_rate": 2.0859311715841122e-05,
+      "loss": 1.2185,
+      "step": 11056
+    },
+    {
+      "epoch": 0.7970157860592518,
+      "grad_norm": 0.12520550922762605,
+      "learning_rate": 2.0845041998984105e-05,
+      "loss": 1.1864,
+      "step": 11057
+    },
+    {
+      "epoch": 0.7970878685215887,
+      "grad_norm": 0.10719879405745106,
+      "learning_rate": 2.0830776596780432e-05,
+      "loss": 1.2509,
+      "step": 11058
+    },
+    {
+      "epoch": 0.7971599509839256,
+      "grad_norm": 0.11690482676311004,
+      "learning_rate": 2.0816515510007707e-05,
+      "loss": 1.3394,
+      "step": 11059
+    },
+    {
+      "epoch": 0.7972320334462625,
+      "grad_norm": 0.12404337948721526,
+      "learning_rate": 2.080225873944328e-05,
+      "loss": 1.3373,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7973041159085994,
+      "grad_norm": 0.1289844430176528,
+      "learning_rate": 2.078800628586428e-05,
+      "loss": 1.3095,
+      "step": 11061
+    },
+    {
+      "epoch": 0.7973761983709363,
+      "grad_norm": 0.11481648530026388,
+      "learning_rate": 2.0773758150047583e-05,
+      "loss": 1.2605,
+      "step": 11062
+    },
+    {
+      "epoch": 0.7974482808332732,
+      "grad_norm": 0.12439942715703525,
+      "learning_rate": 2.0759514332769848e-05,
+      "loss": 1.3529,
+      "step": 11063
+    },
+    {
+      "epoch": 0.7975203632956102,
+      "grad_norm": 0.11300742277143426,
+      "learning_rate": 2.0745274834807492e-05,
+      "loss": 1.1924,
+      "step": 11064
+    },
+    {
+      "epoch": 0.7975924457579471,
+      "grad_norm": 0.1301192916739168,
+      "learning_rate": 2.073103965693668e-05,
+      "loss": 1.2652,
+      "step": 11065
+    },
+    {
+      "epoch": 0.797664528220284,
+      "grad_norm": 0.12853455266984812,
+      "learning_rate": 2.0716808799933375e-05,
+      "loss": 1.2191,
+      "step": 11066
+    },
+    {
+      "epoch": 0.7977366106826209,
+      "grad_norm": 0.2801231922554782,
+      "learning_rate": 2.0702582264573277e-05,
+      "loss": 1.3685,
+      "step": 11067
+    },
+    {
+      "epoch": 0.7978086931449578,
+      "grad_norm": 0.12183258481577179,
+      "learning_rate": 2.068836005163186e-05,
+      "loss": 1.3012,
+      "step": 11068
+    },
+    {
+      "epoch": 0.7978807756072948,
+      "grad_norm": 0.11941198168878452,
+      "learning_rate": 2.0674142161884358e-05,
+      "loss": 1.3116,
+      "step": 11069
+    },
+    {
+      "epoch": 0.7979528580696317,
+      "grad_norm": 0.19520997930955988,
+      "learning_rate": 2.065992859610578e-05,
+      "loss": 1.2683,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7980249405319686,
+      "grad_norm": 0.12448799413136118,
+      "learning_rate": 2.0645719355070902e-05,
+      "loss": 1.297,
+      "step": 11071
+    },
+    {
+      "epoch": 0.7980970229943055,
+      "grad_norm": 0.12801301529382733,
+      "learning_rate": 2.063151443955422e-05,
+      "loss": 1.2453,
+      "step": 11072
+    },
+    {
+      "epoch": 0.7981691054566424,
+      "grad_norm": 0.1346676619228079,
+      "learning_rate": 2.0617313850330066e-05,
+      "loss": 1.3018,
+      "step": 11073
+    },
+    {
+      "epoch": 0.7982411879189794,
+      "grad_norm": 0.11489249232198992,
+      "learning_rate": 2.0603117588172495e-05,
+      "loss": 1.2978,
+      "step": 11074
+    },
+    {
+      "epoch": 0.7983132703813163,
+      "grad_norm": 0.13940400036046038,
+      "learning_rate": 2.0588925653855307e-05,
+      "loss": 1.2484,
+      "step": 11075
+    },
+    {
+      "epoch": 0.7983853528436531,
+      "grad_norm": 0.12326677348786252,
+      "learning_rate": 2.0574738048152097e-05,
+      "loss": 1.2154,
+      "step": 11076
+    },
+    {
+      "epoch": 0.79845743530599,
+      "grad_norm": 0.1338465316591459,
+      "learning_rate": 2.0560554771836227e-05,
+      "loss": 1.3375,
+      "step": 11077
+    },
+    {
+      "epoch": 0.7985295177683269,
+      "grad_norm": 0.1085877844522942,
+      "learning_rate": 2.054637582568082e-05,
+      "loss": 1.2723,
+      "step": 11078
+    },
+    {
+      "epoch": 0.7986016002306638,
+      "grad_norm": 0.1409865453382635,
+      "learning_rate": 2.0532201210458755e-05,
+      "loss": 1.3139,
+      "step": 11079
+    },
+    {
+      "epoch": 0.7986736826930008,
+      "grad_norm": 0.11945858132435633,
+      "learning_rate": 2.051803092694261e-05,
+      "loss": 1.3044,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7987457651553377,
+      "grad_norm": 0.11727723228174143,
+      "learning_rate": 2.0503864975904873e-05,
+      "loss": 1.3287,
+      "step": 11081
+    },
+    {
+      "epoch": 0.7988178476176746,
+      "grad_norm": 0.13032834073489008,
+      "learning_rate": 2.0489703358117696e-05,
+      "loss": 1.1849,
+      "step": 11082
+    },
+    {
+      "epoch": 0.7988899300800115,
+      "grad_norm": 0.1104638335618379,
+      "learning_rate": 2.0475546074353026e-05,
+      "loss": 1.2856,
+      "step": 11083
+    },
+    {
+      "epoch": 0.7989620125423484,
+      "grad_norm": 0.12033156563122052,
+      "learning_rate": 2.046139312538249e-05,
+      "loss": 1.1957,
+      "step": 11084
+    },
+    {
+      "epoch": 0.7990340950046854,
+      "grad_norm": 0.12200245321668576,
+      "learning_rate": 2.0447244511977627e-05,
+      "loss": 1.3413,
+      "step": 11085
+    },
+    {
+      "epoch": 0.7991061774670223,
+      "grad_norm": 0.12556161629848145,
+      "learning_rate": 2.0433100234909642e-05,
+      "loss": 1.2623,
+      "step": 11086
+    },
+    {
+      "epoch": 0.7991782599293592,
+      "grad_norm": 0.11895971534399023,
+      "learning_rate": 2.041896029494954e-05,
+      "loss": 1.2097,
+      "step": 11087
+    },
+    {
+      "epoch": 0.7992503423916961,
+      "grad_norm": 0.1259582161498347,
+      "learning_rate": 2.040482469286803e-05,
+      "loss": 1.2561,
+      "step": 11088
+    },
+    {
+      "epoch": 0.799322424854033,
+      "grad_norm": 0.12916442289963392,
+      "learning_rate": 2.0390693429435627e-05,
+      "loss": 1.3127,
+      "step": 11089
+    },
+    {
+      "epoch": 0.79939450731637,
+      "grad_norm": 0.12247338679874241,
+      "learning_rate": 2.0376566505422657e-05,
+      "loss": 1.1916,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7994665897787069,
+      "grad_norm": 0.10902006971555493,
+      "learning_rate": 2.0362443921599173e-05,
+      "loss": 1.1873,
+      "step": 11091
+    },
+    {
+      "epoch": 0.7995386722410438,
+      "grad_norm": 0.13402794340854182,
+      "learning_rate": 2.0348325678734936e-05,
+      "loss": 1.3175,
+      "step": 11092
+    },
+    {
+      "epoch": 0.7996107547033807,
+      "grad_norm": 0.19780624243324707,
+      "learning_rate": 2.0334211777599498e-05,
+      "loss": 1.2786,
+      "step": 11093
+    },
+    {
+      "epoch": 0.7996828371657175,
+      "grad_norm": 0.11535477955314034,
+      "learning_rate": 2.0320102218962256e-05,
+      "loss": 1.283,
+      "step": 11094
+    },
+    {
+      "epoch": 0.7997549196280545,
+      "grad_norm": 0.11435762593526294,
+      "learning_rate": 2.0305997003592314e-05,
+      "loss": 1.3074,
+      "step": 11095
+    },
+    {
+      "epoch": 0.7998270020903914,
+      "grad_norm": 0.12317044951288714,
+      "learning_rate": 2.029189613225847e-05,
+      "loss": 1.338,
+      "step": 11096
+    },
+    {
+      "epoch": 0.7998990845527283,
+      "grad_norm": 0.12168738997697408,
+      "learning_rate": 2.0277799605729342e-05,
+      "loss": 1.3206,
+      "step": 11097
+    },
+    {
+      "epoch": 0.7999711670150652,
+      "grad_norm": 0.12247258303171096,
+      "learning_rate": 2.0263707424773382e-05,
+      "loss": 1.2641,
+      "step": 11098
+    },
+    {
+      "epoch": 0.8000432494774021,
+      "grad_norm": 0.1156854239837751,
+      "learning_rate": 2.024961959015873e-05,
+      "loss": 1.1975,
+      "step": 11099
+    },
+    {
+      "epoch": 0.800115331939739,
+      "grad_norm": 0.1175610217945022,
+      "learning_rate": 2.0235536102653253e-05,
+      "loss": 1.252,
+      "step": 11100
+    },
+    {
+      "epoch": 0.800187414402076,
+      "grad_norm": 0.13083512509612807,
+      "learning_rate": 2.0221456963024656e-05,
+      "loss": 1.1919,
+      "step": 11101
+    },
+    {
+      "epoch": 0.8002594968644129,
+      "grad_norm": 0.1697666583499286,
+      "learning_rate": 2.0207382172040347e-05,
+      "loss": 1.2265,
+      "step": 11102
+    },
+    {
+      "epoch": 0.8003315793267498,
+      "grad_norm": 0.11367482174355772,
+      "learning_rate": 2.0193311730467602e-05,
+      "loss": 1.2758,
+      "step": 11103
+    },
+    {
+      "epoch": 0.8004036617890867,
+      "grad_norm": 0.12404132492503667,
+      "learning_rate": 2.017924563907332e-05,
+      "loss": 1.2399,
+      "step": 11104
+    },
+    {
+      "epoch": 0.8004757442514236,
+      "grad_norm": 0.12590924210375667,
+      "learning_rate": 2.016518389862424e-05,
+      "loss": 1.1854,
+      "step": 11105
+    },
+    {
+      "epoch": 0.8005478267137606,
+      "grad_norm": 0.12680831790324867,
+      "learning_rate": 2.0151126509886843e-05,
+      "loss": 1.1888,
+      "step": 11106
+    },
+    {
+      "epoch": 0.8006199091760975,
+      "grad_norm": 0.12061613995451527,
+      "learning_rate": 2.0137073473627444e-05,
+      "loss": 1.225,
+      "step": 11107
+    },
+    {
+      "epoch": 0.8006919916384344,
+      "grad_norm": 0.13264791874878254,
+      "learning_rate": 2.0123024790611987e-05,
+      "loss": 1.1886,
+      "step": 11108
+    },
+    {
+      "epoch": 0.8007640741007713,
+      "grad_norm": 0.14702886914236074,
+      "learning_rate": 2.0108980461606287e-05,
+      "loss": 1.3641,
+      "step": 11109
+    },
+    {
+      "epoch": 0.8008361565631082,
+      "grad_norm": 0.11258067415320419,
+      "learning_rate": 2.009494048737587e-05,
+      "loss": 1.2672,
+      "step": 11110
+    },
+    {
+      "epoch": 0.8009082390254451,
+      "grad_norm": 0.12307941206204764,
+      "learning_rate": 2.0080904868686057e-05,
+      "loss": 1.2281,
+      "step": 11111
+    },
+    {
+      "epoch": 0.800980321487782,
+      "grad_norm": 0.11053410108904321,
+      "learning_rate": 2.0066873606301896e-05,
+      "loss": 1.2761,
+      "step": 11112
+    },
+    {
+      "epoch": 0.8010524039501189,
+      "grad_norm": 0.11902428308571059,
+      "learning_rate": 2.0052846700988226e-05,
+      "loss": 1.3233,
+      "step": 11113
+    },
+    {
+      "epoch": 0.8011244864124558,
+      "grad_norm": 0.13497804415568318,
+      "learning_rate": 2.003882415350965e-05,
+      "loss": 1.3387,
+      "step": 11114
+    },
+    {
+      "epoch": 0.8011965688747927,
+      "grad_norm": 0.11000451172243117,
+      "learning_rate": 2.00248059646305e-05,
+      "loss": 1.3096,
+      "step": 11115
+    },
+    {
+      "epoch": 0.8012686513371297,
+      "grad_norm": 0.11604877702667576,
+      "learning_rate": 2.0010792135114897e-05,
+      "loss": 1.2663,
+      "step": 11116
+    },
+    {
+      "epoch": 0.8013407337994666,
+      "grad_norm": 0.11194793589909156,
+      "learning_rate": 1.9996782665726743e-05,
+      "loss": 1.282,
+      "step": 11117
+    },
+    {
+      "epoch": 0.8014128162618035,
+      "grad_norm": 0.12144491084292874,
+      "learning_rate": 1.998277755722965e-05,
+      "loss": 1.2587,
+      "step": 11118
+    },
+    {
+      "epoch": 0.8014848987241404,
+      "grad_norm": 0.1321592496855657,
+      "learning_rate": 1.9968776810387048e-05,
+      "loss": 1.2996,
+      "step": 11119
+    },
+    {
+      "epoch": 0.8015569811864773,
+      "grad_norm": 0.11957269368732867,
+      "learning_rate": 1.995478042596207e-05,
+      "loss": 1.3241,
+      "step": 11120
+    },
+    {
+      "epoch": 0.8016290636488143,
+      "grad_norm": 0.10732163048710186,
+      "learning_rate": 1.994078840471768e-05,
+      "loss": 1.2663,
+      "step": 11121
+    },
+    {
+      "epoch": 0.8017011461111512,
+      "grad_norm": 0.12544794321556385,
+      "learning_rate": 1.992680074741655e-05,
+      "loss": 1.2895,
+      "step": 11122
+    },
+    {
+      "epoch": 0.8017732285734881,
+      "grad_norm": 0.11401439139112093,
+      "learning_rate": 1.991281745482113e-05,
+      "loss": 1.3321,
+      "step": 11123
+    },
+    {
+      "epoch": 0.801845311035825,
+      "grad_norm": 0.12309441095968304,
+      "learning_rate": 1.9898838527693643e-05,
+      "loss": 1.2335,
+      "step": 11124
+    },
+    {
+      "epoch": 0.8019173934981619,
+      "grad_norm": 0.1450014831874713,
+      "learning_rate": 1.9884863966796052e-05,
+      "loss": 1.2054,
+      "step": 11125
+    },
+    {
+      "epoch": 0.8019894759604989,
+      "grad_norm": 0.11881176548026133,
+      "learning_rate": 1.987089377289012e-05,
+      "loss": 1.2664,
+      "step": 11126
+    },
+    {
+      "epoch": 0.8020615584228358,
+      "grad_norm": 0.1240698111310702,
+      "learning_rate": 1.9856927946737324e-05,
+      "loss": 1.2824,
+      "step": 11127
+    },
+    {
+      "epoch": 0.8021336408851726,
+      "grad_norm": 0.11178592193535845,
+      "learning_rate": 1.9842966489098935e-05,
+      "loss": 1.2538,
+      "step": 11128
+    },
+    {
+      "epoch": 0.8022057233475095,
+      "grad_norm": 0.1102730714054707,
+      "learning_rate": 1.9829009400735977e-05,
+      "loss": 1.2776,
+      "step": 11129
+    },
+    {
+      "epoch": 0.8022778058098464,
+      "grad_norm": 0.16210804241783833,
+      "learning_rate": 1.981505668240924e-05,
+      "loss": 1.1488,
+      "step": 11130
+    },
+    {
+      "epoch": 0.8023498882721833,
+      "grad_norm": 0.12039574066849718,
+      "learning_rate": 1.980110833487927e-05,
+      "loss": 1.2633,
+      "step": 11131
+    },
+    {
+      "epoch": 0.8024219707345203,
+      "grad_norm": 0.11790861933072454,
+      "learning_rate": 1.9787164358906407e-05,
+      "loss": 1.2343,
+      "step": 11132
+    },
+    {
+      "epoch": 0.8024940531968572,
+      "grad_norm": 0.1210206861436659,
+      "learning_rate": 1.977322475525063e-05,
+      "loss": 1.2454,
+      "step": 11133
+    },
+    {
+      "epoch": 0.8025661356591941,
+      "grad_norm": 0.13210766810589514,
+      "learning_rate": 1.9759289524671854e-05,
+      "loss": 1.2333,
+      "step": 11134
+    },
+    {
+      "epoch": 0.802638218121531,
+      "grad_norm": 0.1168268192135467,
+      "learning_rate": 1.9745358667929658e-05,
+      "loss": 1.1744,
+      "step": 11135
+    },
+    {
+      "epoch": 0.8027103005838679,
+      "grad_norm": 0.12267993730382395,
+      "learning_rate": 1.9731432185783426e-05,
+      "loss": 1.1774,
+      "step": 11136
+    },
+    {
+      "epoch": 0.8027823830462049,
+      "grad_norm": 0.109345843619595,
+      "learning_rate": 1.9717510078992185e-05,
+      "loss": 1.1997,
+      "step": 11137
+    },
+    {
+      "epoch": 0.8028544655085418,
+      "grad_norm": 0.12114956186246949,
+      "learning_rate": 1.9703592348314903e-05,
+      "loss": 1.2763,
+      "step": 11138
+    },
+    {
+      "epoch": 0.8029265479708787,
+      "grad_norm": 0.12048737784477223,
+      "learning_rate": 1.9689678994510184e-05,
+      "loss": 1.3264,
+      "step": 11139
+    },
+    {
+      "epoch": 0.8029986304332156,
+      "grad_norm": 0.12126991516733925,
+      "learning_rate": 1.967577001833647e-05,
+      "loss": 1.3057,
+      "step": 11140
+    },
+    {
+      "epoch": 0.8030707128955525,
+      "grad_norm": 0.1180527344021625,
+      "learning_rate": 1.9661865420551828e-05,
+      "loss": 1.3148,
+      "step": 11141
+    },
+    {
+      "epoch": 0.8031427953578895,
+      "grad_norm": 0.11206523520985939,
+      "learning_rate": 1.9647965201914277e-05,
+      "loss": 1.3092,
+      "step": 11142
+    },
+    {
+      "epoch": 0.8032148778202264,
+      "grad_norm": 0.12237404632039808,
+      "learning_rate": 1.9634069363181474e-05,
+      "loss": 1.2703,
+      "step": 11143
+    },
+    {
+      "epoch": 0.8032869602825633,
+      "grad_norm": 0.11880981809839283,
+      "learning_rate": 1.9620177905110892e-05,
+      "loss": 1.2585,
+      "step": 11144
+    },
+    {
+      "epoch": 0.8033590427449002,
+      "grad_norm": 0.12662073015262718,
+      "learning_rate": 1.9606290828459685e-05,
+      "loss": 1.2164,
+      "step": 11145
+    },
+    {
+      "epoch": 0.803431125207237,
+      "grad_norm": 0.12574434760663894,
+      "learning_rate": 1.959240813398483e-05,
+      "loss": 1.322,
+      "step": 11146
+    },
+    {
+      "epoch": 0.803503207669574,
+      "grad_norm": 0.1359940915809025,
+      "learning_rate": 1.957852982244309e-05,
+      "loss": 1.2566,
+      "step": 11147
+    },
+    {
+      "epoch": 0.8035752901319109,
+      "grad_norm": 0.12407592415073053,
+      "learning_rate": 1.9564655894590987e-05,
+      "loss": 1.2251,
+      "step": 11148
+    },
+    {
+      "epoch": 0.8036473725942478,
+      "grad_norm": 0.12025251384551386,
+      "learning_rate": 1.9550786351184703e-05,
+      "loss": 1.2171,
+      "step": 11149
+    },
+    {
+      "epoch": 0.8037194550565847,
+      "grad_norm": 0.1314550674930621,
+      "learning_rate": 1.953692119298025e-05,
+      "loss": 1.2056,
+      "step": 11150
+    },
+    {
+      "epoch": 0.8037915375189216,
+      "grad_norm": 0.12641837967808423,
+      "learning_rate": 1.9523060420733462e-05,
+      "loss": 1.2358,
+      "step": 11151
+    },
+    {
+      "epoch": 0.8038636199812585,
+      "grad_norm": 0.1282561298193267,
+      "learning_rate": 1.950920403519988e-05,
+      "loss": 1.2367,
+      "step": 11152
+    },
+    {
+      "epoch": 0.8039357024435955,
+      "grad_norm": 0.11691046914290645,
+      "learning_rate": 1.949535203713474e-05,
+      "loss": 1.2508,
+      "step": 11153
+    },
+    {
+      "epoch": 0.8040077849059324,
+      "grad_norm": 0.11508605268191902,
+      "learning_rate": 1.948150442729313e-05,
+      "loss": 1.2896,
+      "step": 11154
+    },
+    {
+      "epoch": 0.8040798673682693,
+      "grad_norm": 0.11969291560557563,
+      "learning_rate": 1.9467661206429844e-05,
+      "loss": 1.2108,
+      "step": 11155
+    },
+    {
+      "epoch": 0.8041519498306062,
+      "grad_norm": 0.11789070172065659,
+      "learning_rate": 1.9453822375299537e-05,
+      "loss": 1.2851,
+      "step": 11156
+    },
+    {
+      "epoch": 0.8042240322929431,
+      "grad_norm": 0.11190390806076714,
+      "learning_rate": 1.9439987934656467e-05,
+      "loss": 1.1959,
+      "step": 11157
+    },
+    {
+      "epoch": 0.8042961147552801,
+      "grad_norm": 0.12044522330914832,
+      "learning_rate": 1.942615788525476e-05,
+      "loss": 1.3076,
+      "step": 11158
+    },
+    {
+      "epoch": 0.804368197217617,
+      "grad_norm": 0.12469916772838566,
+      "learning_rate": 1.9412332227848263e-05,
+      "loss": 1.2587,
+      "step": 11159
+    },
+    {
+      "epoch": 0.8044402796799539,
+      "grad_norm": 0.24295393919493907,
+      "learning_rate": 1.9398510963190676e-05,
+      "loss": 1.1634,
+      "step": 11160
+    },
+    {
+      "epoch": 0.8045123621422908,
+      "grad_norm": 0.13401656397579964,
+      "learning_rate": 1.9384694092035294e-05,
+      "loss": 1.2664,
+      "step": 11161
+    },
+    {
+      "epoch": 0.8045844446046277,
+      "grad_norm": 0.12248993861313037,
+      "learning_rate": 1.9370881615135285e-05,
+      "loss": 1.2222,
+      "step": 11162
+    },
+    {
+      "epoch": 0.8046565270669646,
+      "grad_norm": 0.15136914609739416,
+      "learning_rate": 1.935707353324353e-05,
+      "loss": 1.2186,
+      "step": 11163
+    },
+    {
+      "epoch": 0.8047286095293015,
+      "grad_norm": 0.12379092493499601,
+      "learning_rate": 1.9343269847112765e-05,
+      "loss": 1.2613,
+      "step": 11164
+    },
+    {
+      "epoch": 0.8048006919916384,
+      "grad_norm": 0.12705675657598295,
+      "learning_rate": 1.9329470557495354e-05,
+      "loss": 1.3695,
+      "step": 11165
+    },
+    {
+      "epoch": 0.8048727744539753,
+      "grad_norm": 0.11691740273068543,
+      "learning_rate": 1.9315675665143495e-05,
+      "loss": 1.2497,
+      "step": 11166
+    },
+    {
+      "epoch": 0.8049448569163122,
+      "grad_norm": 0.12499873486481441,
+      "learning_rate": 1.930188517080913e-05,
+      "loss": 1.2521,
+      "step": 11167
+    },
+    {
+      "epoch": 0.8050169393786492,
+      "grad_norm": 0.12424536182940876,
+      "learning_rate": 1.9288099075243983e-05,
+      "loss": 1.2437,
+      "step": 11168
+    },
+    {
+      "epoch": 0.8050890218409861,
+      "grad_norm": 0.12056364659578199,
+      "learning_rate": 1.9274317379199492e-05,
+      "loss": 1.2448,
+      "step": 11169
+    },
+    {
+      "epoch": 0.805161104303323,
+      "grad_norm": 0.11070754724530987,
+      "learning_rate": 1.9260540083426903e-05,
+      "loss": 1.2452,
+      "step": 11170
+    },
+    {
+      "epoch": 0.8052331867656599,
+      "grad_norm": 0.12644273244245902,
+      "learning_rate": 1.9246767188677195e-05,
+      "loss": 1.2788,
+      "step": 11171
+    },
+    {
+      "epoch": 0.8053052692279968,
+      "grad_norm": 0.1124013058587513,
+      "learning_rate": 1.923299869570111e-05,
+      "loss": 1.3523,
+      "step": 11172
+    },
+    {
+      "epoch": 0.8053773516903338,
+      "grad_norm": 0.11411631637274151,
+      "learning_rate": 1.9219234605249158e-05,
+      "loss": 1.3046,
+      "step": 11173
+    },
+    {
+      "epoch": 0.8054494341526707,
+      "grad_norm": 0.11611951045887145,
+      "learning_rate": 1.9205474918071608e-05,
+      "loss": 1.2341,
+      "step": 11174
+    },
+    {
+      "epoch": 0.8055215166150076,
+      "grad_norm": 0.11726876629034759,
+      "learning_rate": 1.9191719634918482e-05,
+      "loss": 1.3347,
+      "step": 11175
+    },
+    {
+      "epoch": 0.8055935990773445,
+      "grad_norm": 0.13037472561064012,
+      "learning_rate": 1.9177968756539567e-05,
+      "loss": 1.3545,
+      "step": 11176
+    },
+    {
+      "epoch": 0.8056656815396814,
+      "grad_norm": 0.11930047944307011,
+      "learning_rate": 1.9164222283684408e-05,
+      "loss": 1.305,
+      "step": 11177
+    },
+    {
+      "epoch": 0.8057377640020184,
+      "grad_norm": 0.10819069488919897,
+      "learning_rate": 1.9150480217102318e-05,
+      "loss": 1.2345,
+      "step": 11178
+    },
+    {
+      "epoch": 0.8058098464643553,
+      "grad_norm": 0.11426502698163897,
+      "learning_rate": 1.9136742557542352e-05,
+      "loss": 1.2238,
+      "step": 11179
+    },
+    {
+      "epoch": 0.8058819289266921,
+      "grad_norm": 0.1165182705104336,
+      "learning_rate": 1.912300930575335e-05,
+      "loss": 1.2583,
+      "step": 11180
+    },
+    {
+      "epoch": 0.805954011389029,
+      "grad_norm": 0.12261229575632743,
+      "learning_rate": 1.910928046248387e-05,
+      "loss": 1.296,
+      "step": 11181
+    },
+    {
+      "epoch": 0.8060260938513659,
+      "grad_norm": 0.11442173742181351,
+      "learning_rate": 1.9095556028482286e-05,
+      "loss": 1.2703,
+      "step": 11182
+    },
+    {
+      "epoch": 0.8060981763137028,
+      "grad_norm": 0.1199426793188914,
+      "learning_rate": 1.908183600449669e-05,
+      "loss": 1.2637,
+      "step": 11183
+    },
+    {
+      "epoch": 0.8061702587760398,
+      "grad_norm": 0.11204867968171427,
+      "learning_rate": 1.9068120391274936e-05,
+      "loss": 1.2755,
+      "step": 11184
+    },
+    {
+      "epoch": 0.8062423412383767,
+      "grad_norm": 0.12324769456715964,
+      "learning_rate": 1.905440918956467e-05,
+      "loss": 1.4239,
+      "step": 11185
+    },
+    {
+      "epoch": 0.8063144237007136,
+      "grad_norm": 0.11429108259044789,
+      "learning_rate": 1.9040702400113252e-05,
+      "loss": 1.276,
+      "step": 11186
+    },
+    {
+      "epoch": 0.8063865061630505,
+      "grad_norm": 0.13284719694500516,
+      "learning_rate": 1.902700002366784e-05,
+      "loss": 1.25,
+      "step": 11187
+    },
+    {
+      "epoch": 0.8064585886253874,
+      "grad_norm": 0.11312854057973751,
+      "learning_rate": 1.9013302060975336e-05,
+      "loss": 1.2425,
+      "step": 11188
+    },
+    {
+      "epoch": 0.8065306710877244,
+      "grad_norm": 0.12227281130184182,
+      "learning_rate": 1.899960851278243e-05,
+      "loss": 1.3101,
+      "step": 11189
+    },
+    {
+      "epoch": 0.8066027535500613,
+      "grad_norm": 0.12595769987256197,
+      "learning_rate": 1.8985919379835448e-05,
+      "loss": 1.3361,
+      "step": 11190
+    },
+    {
+      "epoch": 0.8066748360123982,
+      "grad_norm": 0.12457866602300578,
+      "learning_rate": 1.897223466288065e-05,
+      "loss": 1.2863,
+      "step": 11191
+    },
+    {
+      "epoch": 0.8067469184747351,
+      "grad_norm": 0.114334790773291,
+      "learning_rate": 1.8958554362663972e-05,
+      "loss": 1.2611,
+      "step": 11192
+    },
+    {
+      "epoch": 0.806819000937072,
+      "grad_norm": 0.10948426239419018,
+      "learning_rate": 1.8944878479931126e-05,
+      "loss": 1.2472,
+      "step": 11193
+    },
+    {
+      "epoch": 0.806891083399409,
+      "grad_norm": 0.11802459470696844,
+      "learning_rate": 1.8931207015427477e-05,
+      "loss": 1.2829,
+      "step": 11194
+    },
+    {
+      "epoch": 0.8069631658617459,
+      "grad_norm": 0.11254305204191972,
+      "learning_rate": 1.8917539969898346e-05,
+      "loss": 1.34,
+      "step": 11195
+    },
+    {
+      "epoch": 0.8070352483240828,
+      "grad_norm": 0.11919386675655547,
+      "learning_rate": 1.890387734408867e-05,
+      "loss": 1.2725,
+      "step": 11196
+    },
+    {
+      "epoch": 0.8071073307864197,
+      "grad_norm": 0.12776797804991216,
+      "learning_rate": 1.8890219138743203e-05,
+      "loss": 1.1469,
+      "step": 11197
+    },
+    {
+      "epoch": 0.8071794132487565,
+      "grad_norm": 0.12978293942613067,
+      "learning_rate": 1.8876565354606402e-05,
+      "loss": 1.3418,
+      "step": 11198
+    },
+    {
+      "epoch": 0.8072514957110934,
+      "grad_norm": 0.12595766953702983,
+      "learning_rate": 1.8862915992422513e-05,
+      "loss": 1.239,
+      "step": 11199
+    },
+    {
+      "epoch": 0.8073235781734304,
+      "grad_norm": 0.12223489459102069,
+      "learning_rate": 1.8849271052935636e-05,
+      "loss": 1.2282,
+      "step": 11200
+    },
+    {
+      "epoch": 0.8073956606357673,
+      "grad_norm": 0.14290722047253068,
+      "learning_rate": 1.8835630536889437e-05,
+      "loss": 1.2976,
+      "step": 11201
+    },
+    {
+      "epoch": 0.8074677430981042,
+      "grad_norm": 0.11756006138612841,
+      "learning_rate": 1.8821994445027503e-05,
+      "loss": 1.2764,
+      "step": 11202
+    },
+    {
+      "epoch": 0.8075398255604411,
+      "grad_norm": 0.10784534832004018,
+      "learning_rate": 1.8808362778093092e-05,
+      "loss": 1.2525,
+      "step": 11203
+    },
+    {
+      "epoch": 0.807611908022778,
+      "grad_norm": 0.11471625199283583,
+      "learning_rate": 1.879473553682931e-05,
+      "loss": 1.1963,
+      "step": 11204
+    },
+    {
+      "epoch": 0.807683990485115,
+      "grad_norm": 0.12212827276613823,
+      "learning_rate": 1.87811127219789e-05,
+      "loss": 1.2886,
+      "step": 11205
+    },
+    {
+      "epoch": 0.8077560729474519,
+      "grad_norm": 0.10257887256178885,
+      "learning_rate": 1.8767494334284453e-05,
+      "loss": 1.3089,
+      "step": 11206
+    },
+    {
+      "epoch": 0.8078281554097888,
+      "grad_norm": 0.11904076289724512,
+      "learning_rate": 1.8753880374488274e-05,
+      "loss": 1.2833,
+      "step": 11207
+    },
+    {
+      "epoch": 0.8079002378721257,
+      "grad_norm": 0.1183708697679259,
+      "learning_rate": 1.874027084333251e-05,
+      "loss": 1.2631,
+      "step": 11208
+    },
+    {
+      "epoch": 0.8079723203344626,
+      "grad_norm": 0.1160640486821319,
+      "learning_rate": 1.8726665741558923e-05,
+      "loss": 1.2531,
+      "step": 11209
+    },
+    {
+      "epoch": 0.8080444027967996,
+      "grad_norm": 0.12011924539292212,
+      "learning_rate": 1.8713065069909152e-05,
+      "loss": 1.3353,
+      "step": 11210
+    },
+    {
+      "epoch": 0.8081164852591365,
+      "grad_norm": 0.12373800825839538,
+      "learning_rate": 1.8699468829124545e-05,
+      "loss": 1.1683,
+      "step": 11211
+    },
+    {
+      "epoch": 0.8081885677214734,
+      "grad_norm": 0.11409915417213333,
+      "learning_rate": 1.8685877019946217e-05,
+      "loss": 1.2374,
+      "step": 11212
+    },
+    {
+      "epoch": 0.8082606501838103,
+      "grad_norm": 0.11130304237868634,
+      "learning_rate": 1.8672289643115058e-05,
+      "loss": 1.247,
+      "step": 11213
+    },
+    {
+      "epoch": 0.8083327326461472,
+      "grad_norm": 0.11221494388334546,
+      "learning_rate": 1.8658706699371687e-05,
+      "loss": 1.2761,
+      "step": 11214
+    },
+    {
+      "epoch": 0.808404815108484,
+      "grad_norm": 0.10976268652566594,
+      "learning_rate": 1.8645128189456505e-05,
+      "loss": 1.3066,
+      "step": 11215
+    },
+    {
+      "epoch": 0.808476897570821,
+      "grad_norm": 0.11618288541734301,
+      "learning_rate": 1.863155411410965e-05,
+      "loss": 1.2759,
+      "step": 11216
+    },
+    {
+      "epoch": 0.8085489800331579,
+      "grad_norm": 0.10102150255563763,
+      "learning_rate": 1.861798447407105e-05,
+      "loss": 1.1434,
+      "step": 11217
+    },
+    {
+      "epoch": 0.8086210624954948,
+      "grad_norm": 0.11651883968802693,
+      "learning_rate": 1.8604419270080355e-05,
+      "loss": 1.3077,
+      "step": 11218
+    },
+    {
+      "epoch": 0.8086931449578317,
+      "grad_norm": 0.11335173949791802,
+      "learning_rate": 1.8590858502876983e-05,
+      "loss": 1.1905,
+      "step": 11219
+    },
+    {
+      "epoch": 0.8087652274201687,
+      "grad_norm": 0.12161277855379016,
+      "learning_rate": 1.857730217320015e-05,
+      "loss": 1.2435,
+      "step": 11220
+    },
+    {
+      "epoch": 0.8088373098825056,
+      "grad_norm": 0.1229218219268471,
+      "learning_rate": 1.8563750281788763e-05,
+      "loss": 1.2363,
+      "step": 11221
+    },
+    {
+      "epoch": 0.8089093923448425,
+      "grad_norm": 0.11986135360210125,
+      "learning_rate": 1.8550202829381546e-05,
+      "loss": 1.2322,
+      "step": 11222
+    },
+    {
+      "epoch": 0.8089814748071794,
+      "grad_norm": 0.114456423879333,
+      "learning_rate": 1.853665981671694e-05,
+      "loss": 1.2317,
+      "step": 11223
+    },
+    {
+      "epoch": 0.8090535572695163,
+      "grad_norm": 0.1247282899078032,
+      "learning_rate": 1.852312124453316e-05,
+      "loss": 1.2675,
+      "step": 11224
+    },
+    {
+      "epoch": 0.8091256397318533,
+      "grad_norm": 0.12141053704332587,
+      "learning_rate": 1.8509587113568193e-05,
+      "loss": 1.2408,
+      "step": 11225
+    },
+    {
+      "epoch": 0.8091977221941902,
+      "grad_norm": 0.13738570592422017,
+      "learning_rate": 1.8496057424559764e-05,
+      "loss": 1.2115,
+      "step": 11226
+    },
+    {
+      "epoch": 0.8092698046565271,
+      "grad_norm": 0.11431051241060464,
+      "learning_rate": 1.848253217824536e-05,
+      "loss": 1.2942,
+      "step": 11227
+    },
+    {
+      "epoch": 0.809341887118864,
+      "grad_norm": 0.11947233640447306,
+      "learning_rate": 1.8469011375362223e-05,
+      "loss": 1.1651,
+      "step": 11228
+    },
+    {
+      "epoch": 0.8094139695812009,
+      "grad_norm": 0.12565885251826386,
+      "learning_rate": 1.845549501664736e-05,
+      "loss": 1.1992,
+      "step": 11229
+    },
+    {
+      "epoch": 0.8094860520435379,
+      "grad_norm": 0.11591623605668414,
+      "learning_rate": 1.8441983102837546e-05,
+      "loss": 1.2904,
+      "step": 11230
+    },
+    {
+      "epoch": 0.8095581345058748,
+      "grad_norm": 0.12044473839089992,
+      "learning_rate": 1.8428475634669285e-05,
+      "loss": 1.2169,
+      "step": 11231
+    },
+    {
+      "epoch": 0.8096302169682116,
+      "grad_norm": 0.12816090217833717,
+      "learning_rate": 1.8414972612878867e-05,
+      "loss": 1.2896,
+      "step": 11232
+    },
+    {
+      "epoch": 0.8097022994305485,
+      "grad_norm": 0.13434509680567666,
+      "learning_rate": 1.8401474038202337e-05,
+      "loss": 1.2692,
+      "step": 11233
+    },
+    {
+      "epoch": 0.8097743818928854,
+      "grad_norm": 0.12258714564149935,
+      "learning_rate": 1.838797991137543e-05,
+      "loss": 1.2312,
+      "step": 11234
+    },
+    {
+      "epoch": 0.8098464643552223,
+      "grad_norm": 0.1251969933018942,
+      "learning_rate": 1.8374490233133768e-05,
+      "loss": 1.2408,
+      "step": 11235
+    },
+    {
+      "epoch": 0.8099185468175593,
+      "grad_norm": 0.12785737028537458,
+      "learning_rate": 1.8361005004212626e-05,
+      "loss": 1.348,
+      "step": 11236
+    },
+    {
+      "epoch": 0.8099906292798962,
+      "grad_norm": 0.12325485877559883,
+      "learning_rate": 1.834752422534711e-05,
+      "loss": 1.349,
+      "step": 11237
+    },
+    {
+      "epoch": 0.8100627117422331,
+      "grad_norm": 0.11601554130362851,
+      "learning_rate": 1.8334047897271945e-05,
+      "loss": 1.3893,
+      "step": 11238
+    },
+    {
+      "epoch": 0.81013479420457,
+      "grad_norm": 0.12230333176690682,
+      "learning_rate": 1.8320576020721804e-05,
+      "loss": 1.1395,
+      "step": 11239
+    },
+    {
+      "epoch": 0.8102068766669069,
+      "grad_norm": 0.11969611719240605,
+      "learning_rate": 1.8307108596431e-05,
+      "loss": 1.2852,
+      "step": 11240
+    },
+    {
+      "epoch": 0.8102789591292439,
+      "grad_norm": 0.12864361350121878,
+      "learning_rate": 1.8293645625133636e-05,
+      "loss": 1.2661,
+      "step": 11241
+    },
+    {
+      "epoch": 0.8103510415915808,
+      "grad_norm": 0.12600384123359964,
+      "learning_rate": 1.828018710756354e-05,
+      "loss": 1.2081,
+      "step": 11242
+    },
+    {
+      "epoch": 0.8104231240539177,
+      "grad_norm": 0.11968925508653062,
+      "learning_rate": 1.8266733044454298e-05,
+      "loss": 1.3174,
+      "step": 11243
+    },
+    {
+      "epoch": 0.8104952065162546,
+      "grad_norm": 0.10997278384086899,
+      "learning_rate": 1.8253283436539338e-05,
+      "loss": 1.1465,
+      "step": 11244
+    },
+    {
+      "epoch": 0.8105672889785915,
+      "grad_norm": 0.10973264061558373,
+      "learning_rate": 1.823983828455178e-05,
+      "loss": 1.1703,
+      "step": 11245
+    },
+    {
+      "epoch": 0.8106393714409285,
+      "grad_norm": 0.1124938784622115,
+      "learning_rate": 1.8226397589224464e-05,
+      "loss": 1.2451,
+      "step": 11246
+    },
+    {
+      "epoch": 0.8107114539032654,
+      "grad_norm": 0.12534938822564992,
+      "learning_rate": 1.8212961351290025e-05,
+      "loss": 1.1902,
+      "step": 11247
+    },
+    {
+      "epoch": 0.8107835363656023,
+      "grad_norm": 0.12627917755283072,
+      "learning_rate": 1.81995295714809e-05,
+      "loss": 1.2103,
+      "step": 11248
+    },
+    {
+      "epoch": 0.8108556188279392,
+      "grad_norm": 0.10952451148716517,
+      "learning_rate": 1.8186102250529246e-05,
+      "loss": 1.2683,
+      "step": 11249
+    },
+    {
+      "epoch": 0.810927701290276,
+      "grad_norm": 0.12386171032821752,
+      "learning_rate": 1.8172679389166925e-05,
+      "loss": 1.2107,
+      "step": 11250
+    },
+    {
+      "epoch": 0.810999783752613,
+      "grad_norm": 0.12865192030930925,
+      "learning_rate": 1.815926098812559e-05,
+      "loss": 1.2627,
+      "step": 11251
+    },
+    {
+      "epoch": 0.8110718662149499,
+      "grad_norm": 0.10934679281626486,
+      "learning_rate": 1.8145847048136734e-05,
+      "loss": 1.2643,
+      "step": 11252
+    },
+    {
+      "epoch": 0.8111439486772868,
+      "grad_norm": 0.13092727136978774,
+      "learning_rate": 1.8132437569931527e-05,
+      "loss": 1.285,
+      "step": 11253
+    },
+    {
+      "epoch": 0.8112160311396237,
+      "grad_norm": 0.12283377212460557,
+      "learning_rate": 1.8119032554240855e-05,
+      "loss": 1.3142,
+      "step": 11254
+    },
+    {
+      "epoch": 0.8112881136019606,
+      "grad_norm": 0.12324128100753069,
+      "learning_rate": 1.8105632001795435e-05,
+      "loss": 1.2463,
+      "step": 11255
+    },
+    {
+      "epoch": 0.8113601960642975,
+      "grad_norm": 0.12877587822819206,
+      "learning_rate": 1.8092235913325695e-05,
+      "loss": 1.3554,
+      "step": 11256
+    },
+    {
+      "epoch": 0.8114322785266345,
+      "grad_norm": 0.12198184844616643,
+      "learning_rate": 1.8078844289561914e-05,
+      "loss": 1.243,
+      "step": 11257
+    },
+    {
+      "epoch": 0.8115043609889714,
+      "grad_norm": 0.12341977356919265,
+      "learning_rate": 1.8065457131233987e-05,
+      "loss": 1.2023,
+      "step": 11258
+    },
+    {
+      "epoch": 0.8115764434513083,
+      "grad_norm": 0.11109629659419097,
+      "learning_rate": 1.8052074439071653e-05,
+      "loss": 1.2932,
+      "step": 11259
+    },
+    {
+      "epoch": 0.8116485259136452,
+      "grad_norm": 0.12367972020360532,
+      "learning_rate": 1.8038696213804363e-05,
+      "loss": 1.267,
+      "step": 11260
+    },
+    {
+      "epoch": 0.8117206083759821,
+      "grad_norm": 0.11368881019593381,
+      "learning_rate": 1.802532245616142e-05,
+      "loss": 1.2023,
+      "step": 11261
+    },
+    {
+      "epoch": 0.8117926908383191,
+      "grad_norm": 0.11924263466357025,
+      "learning_rate": 1.8011953166871753e-05,
+      "loss": 1.2789,
+      "step": 11262
+    },
+    {
+      "epoch": 0.811864773300656,
+      "grad_norm": 0.11323090659290148,
+      "learning_rate": 1.7998588346664115e-05,
+      "loss": 1.2388,
+      "step": 11263
+    },
+    {
+      "epoch": 0.8119368557629929,
+      "grad_norm": 0.11021319342224187,
+      "learning_rate": 1.798522799626702e-05,
+      "loss": 1.1629,
+      "step": 11264
+    },
+    {
+      "epoch": 0.8120089382253298,
+      "grad_norm": 0.12611120763509148,
+      "learning_rate": 1.7971872116408727e-05,
+      "loss": 1.2429,
+      "step": 11265
+    },
+    {
+      "epoch": 0.8120810206876667,
+      "grad_norm": 0.18437913288024094,
+      "learning_rate": 1.795852070781725e-05,
+      "loss": 1.2518,
+      "step": 11266
+    },
+    {
+      "epoch": 0.8121531031500036,
+      "grad_norm": 0.11203090997479058,
+      "learning_rate": 1.794517377122036e-05,
+      "loss": 1.2627,
+      "step": 11267
+    },
+    {
+      "epoch": 0.8122251856123405,
+      "grad_norm": 0.11981541616208088,
+      "learning_rate": 1.793183130734558e-05,
+      "loss": 1.2991,
+      "step": 11268
+    },
+    {
+      "epoch": 0.8122972680746774,
+      "grad_norm": 0.11362420955726188,
+      "learning_rate": 1.791849331692019e-05,
+      "loss": 1.2848,
+      "step": 11269
+    },
+    {
+      "epoch": 0.8123693505370143,
+      "grad_norm": 0.11673723484311903,
+      "learning_rate": 1.7905159800671246e-05,
+      "loss": 1.2752,
+      "step": 11270
+    },
+    {
+      "epoch": 0.8124414329993512,
+      "grad_norm": 0.12104886764471973,
+      "learning_rate": 1.7891830759325534e-05,
+      "loss": 1.2943,
+      "step": 11271
+    },
+    {
+      "epoch": 0.8125135154616882,
+      "grad_norm": 0.11860789158910125,
+      "learning_rate": 1.7878506193609602e-05,
+      "loss": 1.2704,
+      "step": 11272
+    },
+    {
+      "epoch": 0.8125855979240251,
+      "grad_norm": 0.12115020580373453,
+      "learning_rate": 1.7865186104249764e-05,
+      "loss": 1.2388,
+      "step": 11273
+    },
+    {
+      "epoch": 0.812657680386362,
+      "grad_norm": 0.11667036873886777,
+      "learning_rate": 1.785187049197208e-05,
+      "loss": 1.2647,
+      "step": 11274
+    },
+    {
+      "epoch": 0.8127297628486989,
+      "grad_norm": 0.10497904106495135,
+      "learning_rate": 1.783855935750238e-05,
+      "loss": 1.1479,
+      "step": 11275
+    },
+    {
+      "epoch": 0.8128018453110358,
+      "grad_norm": 0.11329066222482019,
+      "learning_rate": 1.782525270156622e-05,
+      "loss": 1.2892,
+      "step": 11276
+    },
+    {
+      "epoch": 0.8128739277733728,
+      "grad_norm": 0.16073039033074357,
+      "learning_rate": 1.7811950524888955e-05,
+      "loss": 1.2375,
+      "step": 11277
+    },
+    {
+      "epoch": 0.8129460102357097,
+      "grad_norm": 0.11588199549082959,
+      "learning_rate": 1.779865282819566e-05,
+      "loss": 1.3537,
+      "step": 11278
+    },
+    {
+      "epoch": 0.8130180926980466,
+      "grad_norm": 0.13743462305218612,
+      "learning_rate": 1.778535961221117e-05,
+      "loss": 1.2751,
+      "step": 11279
+    },
+    {
+      "epoch": 0.8130901751603835,
+      "grad_norm": 0.1074560518403456,
+      "learning_rate": 1.7772070877660107e-05,
+      "loss": 1.2586,
+      "step": 11280
+    },
+    {
+      "epoch": 0.8131622576227204,
+      "grad_norm": 0.17203889535654018,
+      "learning_rate": 1.775878662526681e-05,
+      "loss": 1.2998,
+      "step": 11281
+    },
+    {
+      "epoch": 0.8132343400850574,
+      "grad_norm": 0.13045368545031824,
+      "learning_rate": 1.774550685575539e-05,
+      "loss": 1.2683,
+      "step": 11282
+    },
+    {
+      "epoch": 0.8133064225473943,
+      "grad_norm": 0.12258305791234557,
+      "learning_rate": 1.773223156984971e-05,
+      "loss": 1.2777,
+      "step": 11283
+    },
+    {
+      "epoch": 0.8133785050097311,
+      "grad_norm": 0.18009291060931137,
+      "learning_rate": 1.7718960768273406e-05,
+      "loss": 1.2536,
+      "step": 11284
+    },
+    {
+      "epoch": 0.813450587472068,
+      "grad_norm": 0.11911719272813057,
+      "learning_rate": 1.7705694451749856e-05,
+      "loss": 1.2561,
+      "step": 11285
+    },
+    {
+      "epoch": 0.8135226699344049,
+      "grad_norm": 0.11201847194654393,
+      "learning_rate": 1.7692432621002196e-05,
+      "loss": 1.2343,
+      "step": 11286
+    },
+    {
+      "epoch": 0.8135947523967418,
+      "grad_norm": 0.10758689493895478,
+      "learning_rate": 1.7679175276753257e-05,
+      "loss": 1.3413,
+      "step": 11287
+    },
+    {
+      "epoch": 0.8136668348590788,
+      "grad_norm": 0.1147366516611864,
+      "learning_rate": 1.7665922419725757e-05,
+      "loss": 1.2632,
+      "step": 11288
+    },
+    {
+      "epoch": 0.8137389173214157,
+      "grad_norm": 0.12249848928119653,
+      "learning_rate": 1.765267405064206e-05,
+      "loss": 1.2321,
+      "step": 11289
+    },
+    {
+      "epoch": 0.8138109997837526,
+      "grad_norm": 0.1183011026530876,
+      "learning_rate": 1.7639430170224357e-05,
+      "loss": 1.2296,
+      "step": 11290
+    },
+    {
+      "epoch": 0.8138830822460895,
+      "grad_norm": 0.11377639261271037,
+      "learning_rate": 1.7626190779194475e-05,
+      "loss": 1.2435,
+      "step": 11291
+    },
+    {
+      "epoch": 0.8139551647084264,
+      "grad_norm": 0.12262292916136278,
+      "learning_rate": 1.761295587827416e-05,
+      "loss": 1.2907,
+      "step": 11292
+    },
+    {
+      "epoch": 0.8140272471707634,
+      "grad_norm": 0.12073513860952233,
+      "learning_rate": 1.7599725468184802e-05,
+      "loss": 1.3068,
+      "step": 11293
+    },
+    {
+      "epoch": 0.8140993296331003,
+      "grad_norm": 0.11158971813531406,
+      "learning_rate": 1.7586499549647607e-05,
+      "loss": 1.3915,
+      "step": 11294
+    },
+    {
+      "epoch": 0.8141714120954372,
+      "grad_norm": 0.1245481229402474,
+      "learning_rate": 1.757327812338343e-05,
+      "loss": 1.3038,
+      "step": 11295
+    },
+    {
+      "epoch": 0.8142434945577741,
+      "grad_norm": 0.11540740307079175,
+      "learning_rate": 1.756006119011303e-05,
+      "loss": 1.2616,
+      "step": 11296
+    },
+    {
+      "epoch": 0.814315577020111,
+      "grad_norm": 0.12502737471466654,
+      "learning_rate": 1.754684875055682e-05,
+      "loss": 1.2902,
+      "step": 11297
+    },
+    {
+      "epoch": 0.814387659482448,
+      "grad_norm": 0.12293313509335924,
+      "learning_rate": 1.7533640805435038e-05,
+      "loss": 1.2711,
+      "step": 11298
+    },
+    {
+      "epoch": 0.8144597419447849,
+      "grad_norm": 0.11738470105632966,
+      "learning_rate": 1.7520437355467555e-05,
+      "loss": 1.2798,
+      "step": 11299
+    },
+    {
+      "epoch": 0.8145318244071218,
+      "grad_norm": 0.12320741709860512,
+      "learning_rate": 1.7507238401374105e-05,
+      "loss": 1.2801,
+      "step": 11300
+    },
+    {
+      "epoch": 0.8146039068694587,
+      "grad_norm": 0.12278750177140405,
+      "learning_rate": 1.7494043943874183e-05,
+      "loss": 1.1813,
+      "step": 11301
+    },
+    {
+      "epoch": 0.8146759893317955,
+      "grad_norm": 0.11358076003281152,
+      "learning_rate": 1.7480853983687018e-05,
+      "loss": 1.1848,
+      "step": 11302
+    },
+    {
+      "epoch": 0.8147480717941324,
+      "grad_norm": 0.13191033811018352,
+      "learning_rate": 1.7467668521531523e-05,
+      "loss": 1.3793,
+      "step": 11303
+    },
+    {
+      "epoch": 0.8148201542564694,
+      "grad_norm": 0.11633926984721672,
+      "learning_rate": 1.745448755812641e-05,
+      "loss": 1.2725,
+      "step": 11304
+    },
+    {
+      "epoch": 0.8148922367188063,
+      "grad_norm": 0.11246095857204352,
+      "learning_rate": 1.744131109419024e-05,
+      "loss": 1.264,
+      "step": 11305
+    },
+    {
+      "epoch": 0.8149643191811432,
+      "grad_norm": 0.11248599960520585,
+      "learning_rate": 1.7428139130441224e-05,
+      "loss": 1.2901,
+      "step": 11306
+    },
+    {
+      "epoch": 0.8150364016434801,
+      "grad_norm": 0.10845969390319173,
+      "learning_rate": 1.7414971667597314e-05,
+      "loss": 1.1417,
+      "step": 11307
+    },
+    {
+      "epoch": 0.815108484105817,
+      "grad_norm": 0.1073288438396058,
+      "learning_rate": 1.7401808706376267e-05,
+      "loss": 1.2663,
+      "step": 11308
+    },
+    {
+      "epoch": 0.815180566568154,
+      "grad_norm": 0.1169313622101733,
+      "learning_rate": 1.7388650247495565e-05,
+      "loss": 1.2825,
+      "step": 11309
+    },
+    {
+      "epoch": 0.8152526490304909,
+      "grad_norm": 0.11699492779885763,
+      "learning_rate": 1.7375496291672545e-05,
+      "loss": 1.2989,
+      "step": 11310
+    },
+    {
+      "epoch": 0.8153247314928278,
+      "grad_norm": 0.11126065338967567,
+      "learning_rate": 1.7362346839624123e-05,
+      "loss": 1.1923,
+      "step": 11311
+    },
+    {
+      "epoch": 0.8153968139551647,
+      "grad_norm": 0.12062755032176538,
+      "learning_rate": 1.7349201892067103e-05,
+      "loss": 1.2632,
+      "step": 11312
+    },
+    {
+      "epoch": 0.8154688964175016,
+      "grad_norm": 0.12370223434763312,
+      "learning_rate": 1.7336061449717966e-05,
+      "loss": 1.3159,
+      "step": 11313
+    },
+    {
+      "epoch": 0.8155409788798386,
+      "grad_norm": 0.1341772400273287,
+      "learning_rate": 1.7322925513293064e-05,
+      "loss": 1.2825,
+      "step": 11314
+    },
+    {
+      "epoch": 0.8156130613421755,
+      "grad_norm": 0.12349401609108104,
+      "learning_rate": 1.7309794083508346e-05,
+      "loss": 1.2037,
+      "step": 11315
+    },
+    {
+      "epoch": 0.8156851438045124,
+      "grad_norm": 0.13116713337928337,
+      "learning_rate": 1.7296667161079615e-05,
+      "loss": 1.3633,
+      "step": 11316
+    },
+    {
+      "epoch": 0.8157572262668493,
+      "grad_norm": 0.13435487317976333,
+      "learning_rate": 1.728354474672239e-05,
+      "loss": 1.2943,
+      "step": 11317
+    },
+    {
+      "epoch": 0.8158293087291862,
+      "grad_norm": 0.12486667419732628,
+      "learning_rate": 1.7270426841152032e-05,
+      "loss": 1.3023,
+      "step": 11318
+    },
+    {
+      "epoch": 0.815901391191523,
+      "grad_norm": 0.11637132024459972,
+      "learning_rate": 1.72573134450835e-05,
+      "loss": 1.2193,
+      "step": 11319
+    },
+    {
+      "epoch": 0.81597347365386,
+      "grad_norm": 0.12443392388708123,
+      "learning_rate": 1.7244204559231624e-05,
+      "loss": 1.364,
+      "step": 11320
+    },
+    {
+      "epoch": 0.8160455561161969,
+      "grad_norm": 0.11681423223220028,
+      "learning_rate": 1.7231100184310956e-05,
+      "loss": 1.2582,
+      "step": 11321
+    },
+    {
+      "epoch": 0.8161176385785338,
+      "grad_norm": 0.11295397949622327,
+      "learning_rate": 1.7218000321035797e-05,
+      "loss": 1.194,
+      "step": 11322
+    },
+    {
+      "epoch": 0.8161897210408707,
+      "grad_norm": 0.12587927013968728,
+      "learning_rate": 1.7204904970120207e-05,
+      "loss": 1.22,
+      "step": 11323
+    },
+    {
+      "epoch": 0.8162618035032077,
+      "grad_norm": 0.11937226284177313,
+      "learning_rate": 1.7191814132278007e-05,
+      "loss": 1.2337,
+      "step": 11324
+    },
+    {
+      "epoch": 0.8163338859655446,
+      "grad_norm": 0.1138000123132776,
+      "learning_rate": 1.717872780822275e-05,
+      "loss": 1.1699,
+      "step": 11325
+    },
+    {
+      "epoch": 0.8164059684278815,
+      "grad_norm": 0.1215070269587993,
+      "learning_rate": 1.716564599866778e-05,
+      "loss": 1.3018,
+      "step": 11326
+    },
+    {
+      "epoch": 0.8164780508902184,
+      "grad_norm": 0.12272773096180618,
+      "learning_rate": 1.7152568704326156e-05,
+      "loss": 1.2157,
+      "step": 11327
+    },
+    {
+      "epoch": 0.8165501333525553,
+      "grad_norm": 0.11571281952720193,
+      "learning_rate": 1.713949592591071e-05,
+      "loss": 1.2588,
+      "step": 11328
+    },
+    {
+      "epoch": 0.8166222158148922,
+      "grad_norm": 0.11602850210983201,
+      "learning_rate": 1.7126427664134027e-05,
+      "loss": 1.2228,
+      "step": 11329
+    },
+    {
+      "epoch": 0.8166942982772292,
+      "grad_norm": 0.1184645891613467,
+      "learning_rate": 1.711336391970845e-05,
+      "loss": 1.3191,
+      "step": 11330
+    },
+    {
+      "epoch": 0.8167663807395661,
+      "grad_norm": 0.1265732565748061,
+      "learning_rate": 1.710030469334606e-05,
+      "loss": 1.3555,
+      "step": 11331
+    },
+    {
+      "epoch": 0.816838463201903,
+      "grad_norm": 0.11662144769622271,
+      "learning_rate": 1.7087249985758715e-05,
+      "loss": 1.1909,
+      "step": 11332
+    },
+    {
+      "epoch": 0.8169105456642399,
+      "grad_norm": 0.10645664999060964,
+      "learning_rate": 1.7074199797657984e-05,
+      "loss": 1.2597,
+      "step": 11333
+    },
+    {
+      "epoch": 0.8169826281265768,
+      "grad_norm": 0.12084230692230238,
+      "learning_rate": 1.7061154129755252e-05,
+      "loss": 1.3068,
+      "step": 11334
+    },
+    {
+      "epoch": 0.8170547105889138,
+      "grad_norm": 0.12718761944488122,
+      "learning_rate": 1.7048112982761622e-05,
+      "loss": 1.3138,
+      "step": 11335
+    },
+    {
+      "epoch": 0.8171267930512506,
+      "grad_norm": 0.11545785189281002,
+      "learning_rate": 1.7035076357387924e-05,
+      "loss": 1.2249,
+      "step": 11336
+    },
+    {
+      "epoch": 0.8171988755135875,
+      "grad_norm": 0.10151998350654591,
+      "learning_rate": 1.7022044254344805e-05,
+      "loss": 1.2651,
+      "step": 11337
+    },
+    {
+      "epoch": 0.8172709579759244,
+      "grad_norm": 0.11729485734839078,
+      "learning_rate": 1.7009016674342614e-05,
+      "loss": 1.3084,
+      "step": 11338
+    },
+    {
+      "epoch": 0.8173430404382613,
+      "grad_norm": 0.1166137406916192,
+      "learning_rate": 1.6995993618091467e-05,
+      "loss": 1.2701,
+      "step": 11339
+    },
+    {
+      "epoch": 0.8174151229005983,
+      "grad_norm": 0.13703235934100366,
+      "learning_rate": 1.6982975086301244e-05,
+      "loss": 1.2056,
+      "step": 11340
+    },
+    {
+      "epoch": 0.8174872053629352,
+      "grad_norm": 0.12092291553926733,
+      "learning_rate": 1.6969961079681574e-05,
+      "loss": 1.3566,
+      "step": 11341
+    },
+    {
+      "epoch": 0.8175592878252721,
+      "grad_norm": 0.12387116454979931,
+      "learning_rate": 1.6956951598941828e-05,
+      "loss": 1.2604,
+      "step": 11342
+    },
+    {
+      "epoch": 0.817631370287609,
+      "grad_norm": 0.12084111190875822,
+      "learning_rate": 1.6943946644791164e-05,
+      "loss": 1.2261,
+      "step": 11343
+    },
+    {
+      "epoch": 0.8177034527499459,
+      "grad_norm": 0.11499233802192686,
+      "learning_rate": 1.6930946217938416e-05,
+      "loss": 1.2944,
+      "step": 11344
+    },
+    {
+      "epoch": 0.8177755352122829,
+      "grad_norm": 0.10942986352138018,
+      "learning_rate": 1.691795031909227e-05,
+      "loss": 1.2184,
+      "step": 11345
+    },
+    {
+      "epoch": 0.8178476176746198,
+      "grad_norm": 0.1214086488016664,
+      "learning_rate": 1.6904958948961104e-05,
+      "loss": 1.2627,
+      "step": 11346
+    },
+    {
+      "epoch": 0.8179197001369567,
+      "grad_norm": 0.14114001089408534,
+      "learning_rate": 1.6891972108253095e-05,
+      "loss": 1.2906,
+      "step": 11347
+    },
+    {
+      "epoch": 0.8179917825992936,
+      "grad_norm": 0.13764271525631103,
+      "learning_rate": 1.687898979767606e-05,
+      "loss": 1.1861,
+      "step": 11348
+    },
+    {
+      "epoch": 0.8180638650616305,
+      "grad_norm": 0.11908310727611475,
+      "learning_rate": 1.6866012017937715e-05,
+      "loss": 1.2882,
+      "step": 11349
+    },
+    {
+      "epoch": 0.8181359475239675,
+      "grad_norm": 0.12449629905524243,
+      "learning_rate": 1.6853038769745467e-05,
+      "loss": 1.2911,
+      "step": 11350
+    },
+    {
+      "epoch": 0.8182080299863044,
+      "grad_norm": 0.12181478407836735,
+      "learning_rate": 1.684007005380649e-05,
+      "loss": 1.2186,
+      "step": 11351
+    },
+    {
+      "epoch": 0.8182801124486413,
+      "grad_norm": 0.11699153900810377,
+      "learning_rate": 1.682710587082763e-05,
+      "loss": 1.3151,
+      "step": 11352
+    },
+    {
+      "epoch": 0.8183521949109781,
+      "grad_norm": 0.1264134292293532,
+      "learning_rate": 1.681414622151556e-05,
+      "loss": 1.2105,
+      "step": 11353
+    },
+    {
+      "epoch": 0.818424277373315,
+      "grad_norm": 0.11946096326963507,
+      "learning_rate": 1.6801191106576775e-05,
+      "loss": 1.3461,
+      "step": 11354
+    },
+    {
+      "epoch": 0.8184963598356519,
+      "grad_norm": 0.11630386269471797,
+      "learning_rate": 1.6788240526717368e-05,
+      "loss": 1.2297,
+      "step": 11355
+    },
+    {
+      "epoch": 0.8185684422979889,
+      "grad_norm": 0.10744361234101214,
+      "learning_rate": 1.677529448264329e-05,
+      "loss": 1.321,
+      "step": 11356
+    },
+    {
+      "epoch": 0.8186405247603258,
+      "grad_norm": 0.11291863565535644,
+      "learning_rate": 1.676235297506018e-05,
+      "loss": 1.2924,
+      "step": 11357
+    },
+    {
+      "epoch": 0.8187126072226627,
+      "grad_norm": 0.14565883747094893,
+      "learning_rate": 1.6749416004673558e-05,
+      "loss": 1.2954,
+      "step": 11358
+    },
+    {
+      "epoch": 0.8187846896849996,
+      "grad_norm": 0.12674279089253715,
+      "learning_rate": 1.673648357218851e-05,
+      "loss": 1.3305,
+      "step": 11359
+    },
+    {
+      "epoch": 0.8188567721473365,
+      "grad_norm": 0.14196083681383653,
+      "learning_rate": 1.672355567831001e-05,
+      "loss": 1.334,
+      "step": 11360
+    },
+    {
+      "epoch": 0.8189288546096735,
+      "grad_norm": 0.12019940090468564,
+      "learning_rate": 1.6710632323742738e-05,
+      "loss": 1.3262,
+      "step": 11361
+    },
+    {
+      "epoch": 0.8190009370720104,
+      "grad_norm": 0.1208840231276566,
+      "learning_rate": 1.6697713509191125e-05,
+      "loss": 1.2157,
+      "step": 11362
+    },
+    {
+      "epoch": 0.8190730195343473,
+      "grad_norm": 0.12951756294274377,
+      "learning_rate": 1.668479923535937e-05,
+      "loss": 1.314,
+      "step": 11363
+    },
+    {
+      "epoch": 0.8191451019966842,
+      "grad_norm": 0.1167684535219832,
+      "learning_rate": 1.6671889502951432e-05,
+      "loss": 1.292,
+      "step": 11364
+    },
+    {
+      "epoch": 0.8192171844590211,
+      "grad_norm": 0.12043027132888785,
+      "learning_rate": 1.6658984312670978e-05,
+      "loss": 1.3244,
+      "step": 11365
+    },
+    {
+      "epoch": 0.8192892669213581,
+      "grad_norm": 0.12172268883814122,
+      "learning_rate": 1.664608366522147e-05,
+      "loss": 1.2324,
+      "step": 11366
+    },
+    {
+      "epoch": 0.819361349383695,
+      "grad_norm": 0.10998546410251456,
+      "learning_rate": 1.6633187561306108e-05,
+      "loss": 1.2405,
+      "step": 11367
+    },
+    {
+      "epoch": 0.8194334318460319,
+      "grad_norm": 0.11904506824571055,
+      "learning_rate": 1.6620296001627845e-05,
+      "loss": 1.2164,
+      "step": 11368
+    },
+    {
+      "epoch": 0.8195055143083688,
+      "grad_norm": 0.11101391222846792,
+      "learning_rate": 1.6607408986889384e-05,
+      "loss": 1.1454,
+      "step": 11369
+    },
+    {
+      "epoch": 0.8195775967707057,
+      "grad_norm": 0.12333156865795382,
+      "learning_rate": 1.6594526517793185e-05,
+      "loss": 1.2192,
+      "step": 11370
+    },
+    {
+      "epoch": 0.8196496792330425,
+      "grad_norm": 0.1251248950478889,
+      "learning_rate": 1.6581648595041454e-05,
+      "loss": 1.3047,
+      "step": 11371
+    },
+    {
+      "epoch": 0.8197217616953795,
+      "grad_norm": 0.11464462818723198,
+      "learning_rate": 1.6568775219336164e-05,
+      "loss": 1.2088,
+      "step": 11372
+    },
+    {
+      "epoch": 0.8197938441577164,
+      "grad_norm": 0.12001353667388534,
+      "learning_rate": 1.6555906391379015e-05,
+      "loss": 1.2881,
+      "step": 11373
+    },
+    {
+      "epoch": 0.8198659266200533,
+      "grad_norm": 0.12318522724948504,
+      "learning_rate": 1.6543042111871486e-05,
+      "loss": 1.3784,
+      "step": 11374
+    },
+    {
+      "epoch": 0.8199380090823902,
+      "grad_norm": 0.11892194749941468,
+      "learning_rate": 1.653018238151478e-05,
+      "loss": 1.149,
+      "step": 11375
+    },
+    {
+      "epoch": 0.8200100915447271,
+      "grad_norm": 0.1239432459866346,
+      "learning_rate": 1.6517327201009893e-05,
+      "loss": 1.2527,
+      "step": 11376
+    },
+    {
+      "epoch": 0.8200821740070641,
+      "grad_norm": 0.16673630715060644,
+      "learning_rate": 1.6504476571057516e-05,
+      "loss": 1.276,
+      "step": 11377
+    },
+    {
+      "epoch": 0.820154256469401,
+      "grad_norm": 0.11450057493513766,
+      "learning_rate": 1.6491630492358147e-05,
+      "loss": 1.2848,
+      "step": 11378
+    },
+    {
+      "epoch": 0.8202263389317379,
+      "grad_norm": 0.11925305050672005,
+      "learning_rate": 1.6478788965611993e-05,
+      "loss": 1.3022,
+      "step": 11379
+    },
+    {
+      "epoch": 0.8202984213940748,
+      "grad_norm": 0.11958123534549527,
+      "learning_rate": 1.6465951991519045e-05,
+      "loss": 1.2703,
+      "step": 11380
+    },
+    {
+      "epoch": 0.8203705038564117,
+      "grad_norm": 0.11740010310388614,
+      "learning_rate": 1.6453119570779042e-05,
+      "loss": 1.231,
+      "step": 11381
+    },
+    {
+      "epoch": 0.8204425863187487,
+      "grad_norm": 0.12207861336978632,
+      "learning_rate": 1.644029170409145e-05,
+      "loss": 1.2155,
+      "step": 11382
+    },
+    {
+      "epoch": 0.8205146687810856,
+      "grad_norm": 0.11838043452069563,
+      "learning_rate": 1.642746839215552e-05,
+      "loss": 1.1832,
+      "step": 11383
+    },
+    {
+      "epoch": 0.8205867512434225,
+      "grad_norm": 0.1267233748992853,
+      "learning_rate": 1.6414649635670188e-05,
+      "loss": 1.2421,
+      "step": 11384
+    },
+    {
+      "epoch": 0.8206588337057594,
+      "grad_norm": 0.11718667178113458,
+      "learning_rate": 1.6401835435334245e-05,
+      "loss": 1.2519,
+      "step": 11385
+    },
+    {
+      "epoch": 0.8207309161680963,
+      "grad_norm": 0.11830133006504769,
+      "learning_rate": 1.638902579184616e-05,
+      "loss": 1.3069,
+      "step": 11386
+    },
+    {
+      "epoch": 0.8208029986304333,
+      "grad_norm": 0.2172038119227481,
+      "learning_rate": 1.63762207059042e-05,
+      "loss": 1.2915,
+      "step": 11387
+    },
+    {
+      "epoch": 0.8208750810927701,
+      "grad_norm": 0.11526877508322696,
+      "learning_rate": 1.636342017820629e-05,
+      "loss": 1.2312,
+      "step": 11388
+    },
+    {
+      "epoch": 0.820947163555107,
+      "grad_norm": 0.1208634220877653,
+      "learning_rate": 1.6350624209450226e-05,
+      "loss": 1.3008,
+      "step": 11389
+    },
+    {
+      "epoch": 0.8210192460174439,
+      "grad_norm": 0.11700141936924519,
+      "learning_rate": 1.6337832800333496e-05,
+      "loss": 1.2295,
+      "step": 11390
+    },
+    {
+      "epoch": 0.8210913284797808,
+      "grad_norm": 0.12354086676203561,
+      "learning_rate": 1.6325045951553374e-05,
+      "loss": 1.2687,
+      "step": 11391
+    },
+    {
+      "epoch": 0.8211634109421178,
+      "grad_norm": 0.12587089199593698,
+      "learning_rate": 1.631226366380677e-05,
+      "loss": 1.2216,
+      "step": 11392
+    },
+    {
+      "epoch": 0.8212354934044547,
+      "grad_norm": 0.1203096013947627,
+      "learning_rate": 1.6299485937790502e-05,
+      "loss": 1.3407,
+      "step": 11393
+    },
+    {
+      "epoch": 0.8213075758667916,
+      "grad_norm": 0.11360701815881619,
+      "learning_rate": 1.6286712774201074e-05,
+      "loss": 1.2344,
+      "step": 11394
+    },
+    {
+      "epoch": 0.8213796583291285,
+      "grad_norm": 0.11790477688169669,
+      "learning_rate": 1.6273944173734735e-05,
+      "loss": 1.2352,
+      "step": 11395
+    },
+    {
+      "epoch": 0.8214517407914654,
+      "grad_norm": 0.11112040182619605,
+      "learning_rate": 1.626118013708745e-05,
+      "loss": 1.2538,
+      "step": 11396
+    },
+    {
+      "epoch": 0.8215238232538024,
+      "grad_norm": 0.1173379700623943,
+      "learning_rate": 1.624842066495498e-05,
+      "loss": 1.2539,
+      "step": 11397
+    },
+    {
+      "epoch": 0.8215959057161393,
+      "grad_norm": 0.12417720556552163,
+      "learning_rate": 1.623566575803286e-05,
+      "loss": 1.2299,
+      "step": 11398
+    },
+    {
+      "epoch": 0.8216679881784762,
+      "grad_norm": 0.11470915465048237,
+      "learning_rate": 1.622291541701637e-05,
+      "loss": 1.3015,
+      "step": 11399
+    },
+    {
+      "epoch": 0.8217400706408131,
+      "grad_norm": 0.12142415237773244,
+      "learning_rate": 1.621016964260045e-05,
+      "loss": 1.291,
+      "step": 11400
+    },
+    {
+      "epoch": 0.82181215310315,
+      "grad_norm": 0.12768119364047825,
+      "learning_rate": 1.6197428435479857e-05,
+      "loss": 1.348,
+      "step": 11401
+    },
+    {
+      "epoch": 0.821884235565487,
+      "grad_norm": 0.11824007594027383,
+      "learning_rate": 1.6184691796349172e-05,
+      "loss": 1.2091,
+      "step": 11402
+    },
+    {
+      "epoch": 0.8219563180278239,
+      "grad_norm": 0.12200110869319018,
+      "learning_rate": 1.617195972590263e-05,
+      "loss": 1.2216,
+      "step": 11403
+    },
+    {
+      "epoch": 0.8220284004901608,
+      "grad_norm": 0.128863150203846,
+      "learning_rate": 1.615923222483421e-05,
+      "loss": 1.2893,
+      "step": 11404
+    },
+    {
+      "epoch": 0.8221004829524976,
+      "grad_norm": 0.12468966323974265,
+      "learning_rate": 1.6146509293837698e-05,
+      "loss": 1.2825,
+      "step": 11405
+    },
+    {
+      "epoch": 0.8221725654148345,
+      "grad_norm": 0.11375433240041236,
+      "learning_rate": 1.6133790933606573e-05,
+      "loss": 1.189,
+      "step": 11406
+    },
+    {
+      "epoch": 0.8222446478771714,
+      "grad_norm": 0.11257637541363966,
+      "learning_rate": 1.6121077144834185e-05,
+      "loss": 1.3175,
+      "step": 11407
+    },
+    {
+      "epoch": 0.8223167303395084,
+      "grad_norm": 0.1258253208476569,
+      "learning_rate": 1.6108367928213476e-05,
+      "loss": 1.2119,
+      "step": 11408
+    },
+    {
+      "epoch": 0.8223888128018453,
+      "grad_norm": 0.11930435322979506,
+      "learning_rate": 1.6095663284437222e-05,
+      "loss": 1.1727,
+      "step": 11409
+    },
+    {
+      "epoch": 0.8224608952641822,
+      "grad_norm": 0.11912345268685574,
+      "learning_rate": 1.6082963214197923e-05,
+      "loss": 1.2732,
+      "step": 11410
+    },
+    {
+      "epoch": 0.8225329777265191,
+      "grad_norm": 0.1208267875092763,
+      "learning_rate": 1.6070267718187927e-05,
+      "loss": 1.2922,
+      "step": 11411
+    },
+    {
+      "epoch": 0.822605060188856,
+      "grad_norm": 0.2575512852846183,
+      "learning_rate": 1.605757679709917e-05,
+      "loss": 1.2391,
+      "step": 11412
+    },
+    {
+      "epoch": 0.822677142651193,
+      "grad_norm": 0.13627676253077278,
+      "learning_rate": 1.604489045162344e-05,
+      "loss": 1.2724,
+      "step": 11413
+    },
+    {
+      "epoch": 0.8227492251135299,
+      "grad_norm": 0.12304150993016226,
+      "learning_rate": 1.6032208682452254e-05,
+      "loss": 1.2169,
+      "step": 11414
+    },
+    {
+      "epoch": 0.8228213075758668,
+      "grad_norm": 0.10699256201228843,
+      "learning_rate": 1.601953149027693e-05,
+      "loss": 1.2403,
+      "step": 11415
+    },
+    {
+      "epoch": 0.8228933900382037,
+      "grad_norm": 0.12479236608544085,
+      "learning_rate": 1.6006858875788434e-05,
+      "loss": 1.316,
+      "step": 11416
+    },
+    {
+      "epoch": 0.8229654725005406,
+      "grad_norm": 0.13553720335997296,
+      "learning_rate": 1.5994190839677546e-05,
+      "loss": 1.2877,
+      "step": 11417
+    },
+    {
+      "epoch": 0.8230375549628776,
+      "grad_norm": 0.12593030780048226,
+      "learning_rate": 1.5981527382634808e-05,
+      "loss": 1.2984,
+      "step": 11418
+    },
+    {
+      "epoch": 0.8231096374252145,
+      "grad_norm": 0.10749224096943512,
+      "learning_rate": 1.5968868505350466e-05,
+      "loss": 1.2728,
+      "step": 11419
+    },
+    {
+      "epoch": 0.8231817198875514,
+      "grad_norm": 0.12194534255106938,
+      "learning_rate": 1.5956214208514554e-05,
+      "loss": 1.2998,
+      "step": 11420
+    },
+    {
+      "epoch": 0.8232538023498883,
+      "grad_norm": 0.11339121847524332,
+      "learning_rate": 1.5943564492816853e-05,
+      "loss": 1.3446,
+      "step": 11421
+    },
+    {
+      "epoch": 0.8233258848122252,
+      "grad_norm": 0.1274601357131661,
+      "learning_rate": 1.5930919358946873e-05,
+      "loss": 1.2864,
+      "step": 11422
+    },
+    {
+      "epoch": 0.823397967274562,
+      "grad_norm": 0.12138278107678042,
+      "learning_rate": 1.591827880759389e-05,
+      "loss": 1.3125,
+      "step": 11423
+    },
+    {
+      "epoch": 0.823470049736899,
+      "grad_norm": 0.12318905152607794,
+      "learning_rate": 1.590564283944693e-05,
+      "loss": 1.3044,
+      "step": 11424
+    },
+    {
+      "epoch": 0.8235421321992359,
+      "grad_norm": 0.11896689897479167,
+      "learning_rate": 1.5893011455194762e-05,
+      "loss": 1.1516,
+      "step": 11425
+    },
+    {
+      "epoch": 0.8236142146615728,
+      "grad_norm": 0.10770771213904713,
+      "learning_rate": 1.5880384655525914e-05,
+      "loss": 1.2003,
+      "step": 11426
+    },
+    {
+      "epoch": 0.8236862971239097,
+      "grad_norm": 0.17580096325123185,
+      "learning_rate": 1.5867762441128664e-05,
+      "loss": 1.1976,
+      "step": 11427
+    },
+    {
+      "epoch": 0.8237583795862466,
+      "grad_norm": 0.16437229925999539,
+      "learning_rate": 1.5855144812691027e-05,
+      "loss": 1.202,
+      "step": 11428
+    },
+    {
+      "epoch": 0.8238304620485836,
+      "grad_norm": 0.19280818799942498,
+      "learning_rate": 1.584253177090077e-05,
+      "loss": 1.3843,
+      "step": 11429
+    },
+    {
+      "epoch": 0.8239025445109205,
+      "grad_norm": 0.12185449618731808,
+      "learning_rate": 1.582992331644544e-05,
+      "loss": 1.3302,
+      "step": 11430
+    },
+    {
+      "epoch": 0.8239746269732574,
+      "grad_norm": 0.1054240259464193,
+      "learning_rate": 1.581731945001228e-05,
+      "loss": 1.3021,
+      "step": 11431
+    },
+    {
+      "epoch": 0.8240467094355943,
+      "grad_norm": 0.11529507606314014,
+      "learning_rate": 1.5804720172288344e-05,
+      "loss": 1.2245,
+      "step": 11432
+    },
+    {
+      "epoch": 0.8241187918979312,
+      "grad_norm": 0.12190578026371165,
+      "learning_rate": 1.5792125483960385e-05,
+      "loss": 1.2274,
+      "step": 11433
+    },
+    {
+      "epoch": 0.8241908743602682,
+      "grad_norm": 0.12475584141464194,
+      "learning_rate": 1.5779535385714927e-05,
+      "loss": 1.3332,
+      "step": 11434
+    },
+    {
+      "epoch": 0.8242629568226051,
+      "grad_norm": 0.11674610307358617,
+      "learning_rate": 1.5766949878238258e-05,
+      "loss": 1.2343,
+      "step": 11435
+    },
+    {
+      "epoch": 0.824335039284942,
+      "grad_norm": 0.10945738692807747,
+      "learning_rate": 1.5754368962216393e-05,
+      "loss": 1.2567,
+      "step": 11436
+    },
+    {
+      "epoch": 0.8244071217472789,
+      "grad_norm": 0.11684657973107787,
+      "learning_rate": 1.5741792638335095e-05,
+      "loss": 1.2951,
+      "step": 11437
+    },
+    {
+      "epoch": 0.8244792042096158,
+      "grad_norm": 0.1322237250639176,
+      "learning_rate": 1.5729220907279897e-05,
+      "loss": 1.2539,
+      "step": 11438
+    },
+    {
+      "epoch": 0.8245512866719528,
+      "grad_norm": 0.12294113983068604,
+      "learning_rate": 1.5716653769736057e-05,
+      "loss": 1.2504,
+      "step": 11439
+    },
+    {
+      "epoch": 0.8246233691342896,
+      "grad_norm": 0.11343459947728364,
+      "learning_rate": 1.570409122638864e-05,
+      "loss": 1.3106,
+      "step": 11440
+    },
+    {
+      "epoch": 0.8246954515966265,
+      "grad_norm": 0.13067199030495008,
+      "learning_rate": 1.5691533277922343e-05,
+      "loss": 1.2089,
+      "step": 11441
+    },
+    {
+      "epoch": 0.8247675340589634,
+      "grad_norm": 0.1326613465478787,
+      "learning_rate": 1.5678979925021743e-05,
+      "loss": 1.2696,
+      "step": 11442
+    },
+    {
+      "epoch": 0.8248396165213003,
+      "grad_norm": 0.13002487583472627,
+      "learning_rate": 1.5666431168371086e-05,
+      "loss": 1.2515,
+      "step": 11443
+    },
+    {
+      "epoch": 0.8249116989836373,
+      "grad_norm": 0.11758678056329246,
+      "learning_rate": 1.5653887008654444e-05,
+      "loss": 1.2361,
+      "step": 11444
+    },
+    {
+      "epoch": 0.8249837814459742,
+      "grad_norm": 0.11787084001398854,
+      "learning_rate": 1.5641347446555476e-05,
+      "loss": 1.2501,
+      "step": 11445
+    },
+    {
+      "epoch": 0.8250558639083111,
+      "grad_norm": 0.11677908511056892,
+      "learning_rate": 1.5628812482757803e-05,
+      "loss": 1.2806,
+      "step": 11446
+    },
+    {
+      "epoch": 0.825127946370648,
+      "grad_norm": 0.11565509873010157,
+      "learning_rate": 1.5616282117944648e-05,
+      "loss": 1.29,
+      "step": 11447
+    },
+    {
+      "epoch": 0.8252000288329849,
+      "grad_norm": 0.1106569877829446,
+      "learning_rate": 1.560375635279906e-05,
+      "loss": 1.2847,
+      "step": 11448
+    },
+    {
+      "epoch": 0.8252721112953219,
+      "grad_norm": 0.10772706604968978,
+      "learning_rate": 1.559123518800376e-05,
+      "loss": 1.224,
+      "step": 11449
+    },
+    {
+      "epoch": 0.8253441937576588,
+      "grad_norm": 0.11987626940395053,
+      "learning_rate": 1.5578718624241262e-05,
+      "loss": 1.2814,
+      "step": 11450
+    },
+    {
+      "epoch": 0.8254162762199957,
+      "grad_norm": 0.14555191943516715,
+      "learning_rate": 1.556620666219387e-05,
+      "loss": 1.2907,
+      "step": 11451
+    },
+    {
+      "epoch": 0.8254883586823326,
+      "grad_norm": 0.11927950432106028,
+      "learning_rate": 1.555369930254361e-05,
+      "loss": 1.1671,
+      "step": 11452
+    },
+    {
+      "epoch": 0.8255604411446695,
+      "grad_norm": 0.13077315975222648,
+      "learning_rate": 1.55411965459722e-05,
+      "loss": 1.2531,
+      "step": 11453
+    },
+    {
+      "epoch": 0.8256325236070065,
+      "grad_norm": 0.11433048248424717,
+      "learning_rate": 1.5528698393161134e-05,
+      "loss": 1.2569,
+      "step": 11454
+    },
+    {
+      "epoch": 0.8257046060693434,
+      "grad_norm": 0.12258218274385375,
+      "learning_rate": 1.551620484479174e-05,
+      "loss": 1.2877,
+      "step": 11455
+    },
+    {
+      "epoch": 0.8257766885316803,
+      "grad_norm": 0.12061061881532446,
+      "learning_rate": 1.5503715901545013e-05,
+      "loss": 1.2209,
+      "step": 11456
+    },
+    {
+      "epoch": 0.8258487709940171,
+      "grad_norm": 0.12253355416258822,
+      "learning_rate": 1.5491231564101672e-05,
+      "loss": 1.3272,
+      "step": 11457
+    },
+    {
+      "epoch": 0.825920853456354,
+      "grad_norm": 0.14055583744600578,
+      "learning_rate": 1.5478751833142224e-05,
+      "loss": 1.249,
+      "step": 11458
+    },
+    {
+      "epoch": 0.8259929359186909,
+      "grad_norm": 0.1230501238061785,
+      "learning_rate": 1.546627670934697e-05,
+      "loss": 1.3176,
+      "step": 11459
+    },
+    {
+      "epoch": 0.8260650183810279,
+      "grad_norm": 0.10854637235862989,
+      "learning_rate": 1.5453806193395925e-05,
+      "loss": 1.2246,
+      "step": 11460
+    },
+    {
+      "epoch": 0.8261371008433648,
+      "grad_norm": 0.1132724209183056,
+      "learning_rate": 1.5441340285968786e-05,
+      "loss": 1.2543,
+      "step": 11461
+    },
+    {
+      "epoch": 0.8262091833057017,
+      "grad_norm": 0.12350978971974386,
+      "learning_rate": 1.5428878987745086e-05,
+      "loss": 1.2994,
+      "step": 11462
+    },
+    {
+      "epoch": 0.8262812657680386,
+      "grad_norm": 0.11315330259249738,
+      "learning_rate": 1.5416422299404053e-05,
+      "loss": 1.2048,
+      "step": 11463
+    },
+    {
+      "epoch": 0.8263533482303755,
+      "grad_norm": 0.10996518132107409,
+      "learning_rate": 1.5403970221624763e-05,
+      "loss": 1.2573,
+      "step": 11464
+    },
+    {
+      "epoch": 0.8264254306927125,
+      "grad_norm": 0.12773819213278828,
+      "learning_rate": 1.5391522755085884e-05,
+      "loss": 1.3217,
+      "step": 11465
+    },
+    {
+      "epoch": 0.8264975131550494,
+      "grad_norm": 0.23454613221734447,
+      "learning_rate": 1.5379079900465953e-05,
+      "loss": 1.3063,
+      "step": 11466
+    },
+    {
+      "epoch": 0.8265695956173863,
+      "grad_norm": 0.1302448116496279,
+      "learning_rate": 1.536664165844317e-05,
+      "loss": 1.2296,
+      "step": 11467
+    },
+    {
+      "epoch": 0.8266416780797232,
+      "grad_norm": 0.12754866644438084,
+      "learning_rate": 1.5354208029695628e-05,
+      "loss": 1.2228,
+      "step": 11468
+    },
+    {
+      "epoch": 0.8267137605420601,
+      "grad_norm": 0.1260250426314835,
+      "learning_rate": 1.5341779014900982e-05,
+      "loss": 1.3462,
+      "step": 11469
+    },
+    {
+      "epoch": 0.8267858430043971,
+      "grad_norm": 0.11702538558044062,
+      "learning_rate": 1.5329354614736757e-05,
+      "loss": 1.3034,
+      "step": 11470
+    },
+    {
+      "epoch": 0.826857925466734,
+      "grad_norm": 0.11755253168837396,
+      "learning_rate": 1.531693482988018e-05,
+      "loss": 1.1892,
+      "step": 11471
+    },
+    {
+      "epoch": 0.8269300079290709,
+      "grad_norm": 0.1282370127270553,
+      "learning_rate": 1.5304519661008267e-05,
+      "loss": 1.332,
+      "step": 11472
+    },
+    {
+      "epoch": 0.8270020903914078,
+      "grad_norm": 0.11569700789923477,
+      "learning_rate": 1.5292109108797726e-05,
+      "loss": 1.2507,
+      "step": 11473
+    },
+    {
+      "epoch": 0.8270741728537447,
+      "grad_norm": 0.12626851455134555,
+      "learning_rate": 1.5279703173925063e-05,
+      "loss": 1.2717,
+      "step": 11474
+    },
+    {
+      "epoch": 0.8271462553160815,
+      "grad_norm": 0.10710882895605532,
+      "learning_rate": 1.5267301857066508e-05,
+      "loss": 1.2161,
+      "step": 11475
+    },
+    {
+      "epoch": 0.8272183377784185,
+      "grad_norm": 0.15609730775721017,
+      "learning_rate": 1.5254905158898037e-05,
+      "loss": 1.2734,
+      "step": 11476
+    },
+    {
+      "epoch": 0.8272904202407554,
+      "grad_norm": 0.11420178379853772,
+      "learning_rate": 1.5242513080095389e-05,
+      "loss": 1.2641,
+      "step": 11477
+    },
+    {
+      "epoch": 0.8273625027030923,
+      "grad_norm": 0.12371490690056597,
+      "learning_rate": 1.5230125621334046e-05,
+      "loss": 1.3014,
+      "step": 11478
+    },
+    {
+      "epoch": 0.8274345851654292,
+      "grad_norm": 0.11547083897030518,
+      "learning_rate": 1.5217742783289224e-05,
+      "loss": 1.3256,
+      "step": 11479
+    },
+    {
+      "epoch": 0.8275066676277661,
+      "grad_norm": 0.12063672910008277,
+      "learning_rate": 1.5205364566635904e-05,
+      "loss": 1.2015,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8275787500901031,
+      "grad_norm": 0.22602860590617868,
+      "learning_rate": 1.5192990972048804e-05,
+      "loss": 1.2266,
+      "step": 11481
+    },
+    {
+      "epoch": 0.82765083255244,
+      "grad_norm": 0.12210971159576957,
+      "learning_rate": 1.5180622000202416e-05,
+      "loss": 1.2528,
+      "step": 11482
+    },
+    {
+      "epoch": 0.8277229150147769,
+      "grad_norm": 0.2197950707270212,
+      "learning_rate": 1.5168257651770934e-05,
+      "loss": 1.2231,
+      "step": 11483
+    },
+    {
+      "epoch": 0.8277949974771138,
+      "grad_norm": 0.11727988784216406,
+      "learning_rate": 1.515589792742833e-05,
+      "loss": 1.2389,
+      "step": 11484
+    },
+    {
+      "epoch": 0.8278670799394507,
+      "grad_norm": 0.11541767255791023,
+      "learning_rate": 1.5143542827848345e-05,
+      "loss": 1.3093,
+      "step": 11485
+    },
+    {
+      "epoch": 0.8279391624017877,
+      "grad_norm": 0.1122279409074809,
+      "learning_rate": 1.5131192353704404e-05,
+      "loss": 1.123,
+      "step": 11486
+    },
+    {
+      "epoch": 0.8280112448641246,
+      "grad_norm": 0.14766547121962958,
+      "learning_rate": 1.5118846505669748e-05,
+      "loss": 1.2568,
+      "step": 11487
+    },
+    {
+      "epoch": 0.8280833273264615,
+      "grad_norm": 0.12793035703350877,
+      "learning_rate": 1.5106505284417327e-05,
+      "loss": 1.2181,
+      "step": 11488
+    },
+    {
+      "epoch": 0.8281554097887984,
+      "grad_norm": 0.1417762781847635,
+      "learning_rate": 1.5094168690619848e-05,
+      "loss": 1.2336,
+      "step": 11489
+    },
+    {
+      "epoch": 0.8282274922511353,
+      "grad_norm": 0.10840116566595118,
+      "learning_rate": 1.508183672494976e-05,
+      "loss": 1.3721,
+      "step": 11490
+    },
+    {
+      "epoch": 0.8282995747134723,
+      "grad_norm": 0.12394149553898419,
+      "learning_rate": 1.5069509388079273e-05,
+      "loss": 1.3127,
+      "step": 11491
+    },
+    {
+      "epoch": 0.8283716571758091,
+      "grad_norm": 0.11801118623583938,
+      "learning_rate": 1.5057186680680336e-05,
+      "loss": 1.23,
+      "step": 11492
+    },
+    {
+      "epoch": 0.828443739638146,
+      "grad_norm": 0.10893716003118342,
+      "learning_rate": 1.5044868603424655e-05,
+      "loss": 1.2534,
+      "step": 11493
+    },
+    {
+      "epoch": 0.8285158221004829,
+      "grad_norm": 0.13012979243015327,
+      "learning_rate": 1.5032555156983629e-05,
+      "loss": 1.3223,
+      "step": 11494
+    },
+    {
+      "epoch": 0.8285879045628198,
+      "grad_norm": 0.12915594766544866,
+      "learning_rate": 1.502024634202851e-05,
+      "loss": 1.2407,
+      "step": 11495
+    },
+    {
+      "epoch": 0.8286599870251568,
+      "grad_norm": 0.12753548073595153,
+      "learning_rate": 1.5007942159230226e-05,
+      "loss": 1.2782,
+      "step": 11496
+    },
+    {
+      "epoch": 0.8287320694874937,
+      "grad_norm": 0.12714763936165152,
+      "learning_rate": 1.4995642609259463e-05,
+      "loss": 1.2991,
+      "step": 11497
+    },
+    {
+      "epoch": 0.8288041519498306,
+      "grad_norm": 0.12886866607324848,
+      "learning_rate": 1.4983347692786609e-05,
+      "loss": 1.2894,
+      "step": 11498
+    },
+    {
+      "epoch": 0.8288762344121675,
+      "grad_norm": 0.11508598988258595,
+      "learning_rate": 1.4971057410481904e-05,
+      "loss": 1.2975,
+      "step": 11499
+    },
+    {
+      "epoch": 0.8289483168745044,
+      "grad_norm": 0.11544621648493127,
+      "learning_rate": 1.4958771763015267e-05,
+      "loss": 1.2412,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8290203993368414,
+      "grad_norm": 0.1193422815737603,
+      "learning_rate": 1.4946490751056397e-05,
+      "loss": 1.2413,
+      "step": 11501
+    },
+    {
+      "epoch": 0.8290924817991783,
+      "grad_norm": 0.11572715770732929,
+      "learning_rate": 1.4934214375274635e-05,
+      "loss": 1.2275,
+      "step": 11502
+    },
+    {
+      "epoch": 0.8291645642615152,
+      "grad_norm": 0.11542220225420972,
+      "learning_rate": 1.4921942636339236e-05,
+      "loss": 1.4203,
+      "step": 11503
+    },
+    {
+      "epoch": 0.8292366467238521,
+      "grad_norm": 0.12999859667161348,
+      "learning_rate": 1.4909675534919122e-05,
+      "loss": 1.244,
+      "step": 11504
+    },
+    {
+      "epoch": 0.829308729186189,
+      "grad_norm": 0.11377064230303294,
+      "learning_rate": 1.4897413071682897e-05,
+      "loss": 1.3064,
+      "step": 11505
+    },
+    {
+      "epoch": 0.829380811648526,
+      "grad_norm": 0.11958994879744704,
+      "learning_rate": 1.488515524729902e-05,
+      "loss": 1.3523,
+      "step": 11506
+    },
+    {
+      "epoch": 0.8294528941108629,
+      "grad_norm": 0.11726610420631783,
+      "learning_rate": 1.4872902062435612e-05,
+      "loss": 1.306,
+      "step": 11507
+    },
+    {
+      "epoch": 0.8295249765731998,
+      "grad_norm": 0.11929053793183231,
+      "learning_rate": 1.4860653517760659e-05,
+      "loss": 1.3101,
+      "step": 11508
+    },
+    {
+      "epoch": 0.8295970590355366,
+      "grad_norm": 0.10870915099917663,
+      "learning_rate": 1.4848409613941738e-05,
+      "loss": 1.2107,
+      "step": 11509
+    },
+    {
+      "epoch": 0.8296691414978735,
+      "grad_norm": 0.122431035052791,
+      "learning_rate": 1.4836170351646294e-05,
+      "loss": 1.1528,
+      "step": 11510
+    },
+    {
+      "epoch": 0.8297412239602104,
+      "grad_norm": 0.12397533423618906,
+      "learning_rate": 1.4823935731541427e-05,
+      "loss": 1.3398,
+      "step": 11511
+    },
+    {
+      "epoch": 0.8298133064225474,
+      "grad_norm": 0.12535379697984383,
+      "learning_rate": 1.4811705754294125e-05,
+      "loss": 1.3098,
+      "step": 11512
+    },
+    {
+      "epoch": 0.8298853888848843,
+      "grad_norm": 0.12444410303631548,
+      "learning_rate": 1.4799480420570943e-05,
+      "loss": 1.3342,
+      "step": 11513
+    },
+    {
+      "epoch": 0.8299574713472212,
+      "grad_norm": 0.12609482916419118,
+      "learning_rate": 1.4787259731038317e-05,
+      "loss": 1.21,
+      "step": 11514
+    },
+    {
+      "epoch": 0.8300295538095581,
+      "grad_norm": 0.111663160417308,
+      "learning_rate": 1.4775043686362367e-05,
+      "loss": 1.2284,
+      "step": 11515
+    },
+    {
+      "epoch": 0.830101636271895,
+      "grad_norm": 0.11954056609150499,
+      "learning_rate": 1.4762832287208972e-05,
+      "loss": 1.3322,
+      "step": 11516
+    },
+    {
+      "epoch": 0.830173718734232,
+      "grad_norm": 0.11994298567290436,
+      "learning_rate": 1.4750625534243779e-05,
+      "loss": 1.2648,
+      "step": 11517
+    },
+    {
+      "epoch": 0.8302458011965689,
+      "grad_norm": 0.11344791287901007,
+      "learning_rate": 1.4738423428132153e-05,
+      "loss": 1.3166,
+      "step": 11518
+    },
+    {
+      "epoch": 0.8303178836589058,
+      "grad_norm": 0.11486616813898962,
+      "learning_rate": 1.472622596953922e-05,
+      "loss": 1.3227,
+      "step": 11519
+    },
+    {
+      "epoch": 0.8303899661212427,
+      "grad_norm": 0.1294637941668031,
+      "learning_rate": 1.471403315912986e-05,
+      "loss": 1.3009,
+      "step": 11520
+    },
+    {
+      "epoch": 0.8304620485835796,
+      "grad_norm": 0.11431752916504334,
+      "learning_rate": 1.470184499756867e-05,
+      "loss": 1.2686,
+      "step": 11521
+    },
+    {
+      "epoch": 0.8305341310459166,
+      "grad_norm": 0.10836228409180118,
+      "learning_rate": 1.4689661485520035e-05,
+      "loss": 1.3353,
+      "step": 11522
+    },
+    {
+      "epoch": 0.8306062135082535,
+      "grad_norm": 0.11613323688782173,
+      "learning_rate": 1.4677482623648064e-05,
+      "loss": 1.2675,
+      "step": 11523
+    },
+    {
+      "epoch": 0.8306782959705904,
+      "grad_norm": 0.12071389062191425,
+      "learning_rate": 1.4665308412616596e-05,
+      "loss": 1.2869,
+      "step": 11524
+    },
+    {
+      "epoch": 0.8307503784329273,
+      "grad_norm": 0.11656319549522508,
+      "learning_rate": 1.4653138853089254e-05,
+      "loss": 1.3034,
+      "step": 11525
+    },
+    {
+      "epoch": 0.8308224608952642,
+      "grad_norm": 0.12239352022852285,
+      "learning_rate": 1.4640973945729374e-05,
+      "loss": 1.3244,
+      "step": 11526
+    },
+    {
+      "epoch": 0.830894543357601,
+      "grad_norm": 0.12322013848426397,
+      "learning_rate": 1.4628813691200049e-05,
+      "loss": 1.2859,
+      "step": 11527
+    },
+    {
+      "epoch": 0.830966625819938,
+      "grad_norm": 0.12530070498136756,
+      "learning_rate": 1.4616658090164149e-05,
+      "loss": 1.2675,
+      "step": 11528
+    },
+    {
+      "epoch": 0.8310387082822749,
+      "grad_norm": 0.12107116002446355,
+      "learning_rate": 1.460450714328423e-05,
+      "loss": 1.2747,
+      "step": 11529
+    },
+    {
+      "epoch": 0.8311107907446118,
+      "grad_norm": 0.10999231975515116,
+      "learning_rate": 1.4592360851222653e-05,
+      "loss": 1.2465,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8311828732069487,
+      "grad_norm": 0.11549736321397808,
+      "learning_rate": 1.4580219214641478e-05,
+      "loss": 1.2578,
+      "step": 11531
+    },
+    {
+      "epoch": 0.8312549556692856,
+      "grad_norm": 0.1422745210533065,
+      "learning_rate": 1.4568082234202551e-05,
+      "loss": 1.3951,
+      "step": 11532
+    },
+    {
+      "epoch": 0.8313270381316226,
+      "grad_norm": 0.1453257702147481,
+      "learning_rate": 1.4555949910567435e-05,
+      "loss": 1.1666,
+      "step": 11533
+    },
+    {
+      "epoch": 0.8313991205939595,
+      "grad_norm": 0.1209202715455999,
+      "learning_rate": 1.4543822244397443e-05,
+      "loss": 1.3525,
+      "step": 11534
+    },
+    {
+      "epoch": 0.8314712030562964,
+      "grad_norm": 0.11525215862979715,
+      "learning_rate": 1.4531699236353669e-05,
+      "loss": 1.3494,
+      "step": 11535
+    },
+    {
+      "epoch": 0.8315432855186333,
+      "grad_norm": 0.11777896634874653,
+      "learning_rate": 1.4519580887096895e-05,
+      "loss": 1.1766,
+      "step": 11536
+    },
+    {
+      "epoch": 0.8316153679809702,
+      "grad_norm": 0.11744012258476583,
+      "learning_rate": 1.4507467197287717e-05,
+      "loss": 1.279,
+      "step": 11537
+    },
+    {
+      "epoch": 0.8316874504433072,
+      "grad_norm": 0.1127108946681268,
+      "learning_rate": 1.449535816758637e-05,
+      "loss": 1.1997,
+      "step": 11538
+    },
+    {
+      "epoch": 0.8317595329056441,
+      "grad_norm": 0.10569133390453078,
+      "learning_rate": 1.4483253798652974e-05,
+      "loss": 1.2072,
+      "step": 11539
+    },
+    {
+      "epoch": 0.831831615367981,
+      "grad_norm": 0.12365177790129873,
+      "learning_rate": 1.447115409114731e-05,
+      "loss": 1.2959,
+      "step": 11540
+    },
+    {
+      "epoch": 0.8319036978303179,
+      "grad_norm": 0.11892044978466638,
+      "learning_rate": 1.4459059045728918e-05,
+      "loss": 1.3143,
+      "step": 11541
+    },
+    {
+      "epoch": 0.8319757802926548,
+      "grad_norm": 0.1147327369746932,
+      "learning_rate": 1.4446968663057037e-05,
+      "loss": 1.2879,
+      "step": 11542
+    },
+    {
+      "epoch": 0.8320478627549918,
+      "grad_norm": 0.11596241760296772,
+      "learning_rate": 1.4434882943790773e-05,
+      "loss": 1.339,
+      "step": 11543
+    },
+    {
+      "epoch": 0.8321199452173286,
+      "grad_norm": 0.11693927761657452,
+      "learning_rate": 1.4422801888588866e-05,
+      "loss": 1.1957,
+      "step": 11544
+    },
+    {
+      "epoch": 0.8321920276796655,
+      "grad_norm": 0.1160301076684354,
+      "learning_rate": 1.4410725498109889e-05,
+      "loss": 1.277,
+      "step": 11545
+    },
+    {
+      "epoch": 0.8322641101420024,
+      "grad_norm": 0.12495789626110736,
+      "learning_rate": 1.4398653773012017e-05,
+      "loss": 1.2848,
+      "step": 11546
+    },
+    {
+      "epoch": 0.8323361926043393,
+      "grad_norm": 0.12436199639837348,
+      "learning_rate": 1.4386586713953353e-05,
+      "loss": 1.295,
+      "step": 11547
+    },
+    {
+      "epoch": 0.8324082750666763,
+      "grad_norm": 0.11915253521452429,
+      "learning_rate": 1.4374524321591632e-05,
+      "loss": 1.2447,
+      "step": 11548
+    },
+    {
+      "epoch": 0.8324803575290132,
+      "grad_norm": 0.14221625724511155,
+      "learning_rate": 1.4362466596584389e-05,
+      "loss": 1.2823,
+      "step": 11549
+    },
+    {
+      "epoch": 0.8325524399913501,
+      "grad_norm": 0.11447907655291639,
+      "learning_rate": 1.4350413539588825e-05,
+      "loss": 1.2606,
+      "step": 11550
+    },
+    {
+      "epoch": 0.832624522453687,
+      "grad_norm": 0.11744462422683886,
+      "learning_rate": 1.4338365151261945e-05,
+      "loss": 1.3195,
+      "step": 11551
+    },
+    {
+      "epoch": 0.8326966049160239,
+      "grad_norm": 0.14094328620422605,
+      "learning_rate": 1.4326321432260537e-05,
+      "loss": 1.2713,
+      "step": 11552
+    },
+    {
+      "epoch": 0.8327686873783609,
+      "grad_norm": 0.12343948403710675,
+      "learning_rate": 1.4314282383241096e-05,
+      "loss": 1.2851,
+      "step": 11553
+    },
+    {
+      "epoch": 0.8328407698406978,
+      "grad_norm": 0.11469404671197304,
+      "learning_rate": 1.4302248004859809e-05,
+      "loss": 1.3292,
+      "step": 11554
+    },
+    {
+      "epoch": 0.8329128523030347,
+      "grad_norm": 0.12225681069466837,
+      "learning_rate": 1.429021829777265e-05,
+      "loss": 1.2721,
+      "step": 11555
+    },
+    {
+      "epoch": 0.8329849347653716,
+      "grad_norm": 0.11771334679598143,
+      "learning_rate": 1.4278193262635397e-05,
+      "loss": 1.3032,
+      "step": 11556
+    },
+    {
+      "epoch": 0.8330570172277085,
+      "grad_norm": 0.12181514423896869,
+      "learning_rate": 1.4266172900103525e-05,
+      "loss": 1.3088,
+      "step": 11557
+    },
+    {
+      "epoch": 0.8331290996900454,
+      "grad_norm": 0.12547140430838263,
+      "learning_rate": 1.4254157210832208e-05,
+      "loss": 1.1826,
+      "step": 11558
+    },
+    {
+      "epoch": 0.8332011821523824,
+      "grad_norm": 0.11012064069073516,
+      "learning_rate": 1.4242146195476425e-05,
+      "loss": 1.2411,
+      "step": 11559
+    },
+    {
+      "epoch": 0.8332732646147193,
+      "grad_norm": 0.10819753549914402,
+      "learning_rate": 1.4230139854690872e-05,
+      "loss": 1.2488,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8333453470770561,
+      "grad_norm": 0.13330243112729145,
+      "learning_rate": 1.4218138189130047e-05,
+      "loss": 1.307,
+      "step": 11561
+    },
+    {
+      "epoch": 0.833417429539393,
+      "grad_norm": 0.12176671365471914,
+      "learning_rate": 1.4206141199448109e-05,
+      "loss": 1.2221,
+      "step": 11562
+    },
+    {
+      "epoch": 0.8334895120017299,
+      "grad_norm": 0.1175520246955457,
+      "learning_rate": 1.4194148886298997e-05,
+      "loss": 1.3144,
+      "step": 11563
+    },
+    {
+      "epoch": 0.8335615944640669,
+      "grad_norm": 0.1311196579697973,
+      "learning_rate": 1.4182161250336389e-05,
+      "loss": 1.1825,
+      "step": 11564
+    },
+    {
+      "epoch": 0.8336336769264038,
+      "grad_norm": 0.11937456968560445,
+      "learning_rate": 1.4170178292213799e-05,
+      "loss": 1.2747,
+      "step": 11565
+    },
+    {
+      "epoch": 0.8337057593887407,
+      "grad_norm": 0.12050064664074071,
+      "learning_rate": 1.415820001258431e-05,
+      "loss": 1.2221,
+      "step": 11566
+    },
+    {
+      "epoch": 0.8337778418510776,
+      "grad_norm": 0.12569823276204703,
+      "learning_rate": 1.4146226412100893e-05,
+      "loss": 1.3301,
+      "step": 11567
+    },
+    {
+      "epoch": 0.8338499243134145,
+      "grad_norm": 0.11714514991257124,
+      "learning_rate": 1.413425749141618e-05,
+      "loss": 1.2772,
+      "step": 11568
+    },
+    {
+      "epoch": 0.8339220067757515,
+      "grad_norm": 0.1109732003776942,
+      "learning_rate": 1.4122293251182661e-05,
+      "loss": 1.3147,
+      "step": 11569
+    },
+    {
+      "epoch": 0.8339940892380884,
+      "grad_norm": 0.12105752089032834,
+      "learning_rate": 1.4110333692052413e-05,
+      "loss": 1.4175,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8340661717004253,
+      "grad_norm": 0.12929124212134038,
+      "learning_rate": 1.4098378814677382e-05,
+      "loss": 1.2313,
+      "step": 11571
+    },
+    {
+      "epoch": 0.8341382541627622,
+      "grad_norm": 0.27294121450856995,
+      "learning_rate": 1.408642861970919e-05,
+      "loss": 1.2924,
+      "step": 11572
+    },
+    {
+      "epoch": 0.8342103366250991,
+      "grad_norm": 0.1267667932014632,
+      "learning_rate": 1.4074483107799253e-05,
+      "loss": 1.2519,
+      "step": 11573
+    },
+    {
+      "epoch": 0.8342824190874361,
+      "grad_norm": 0.11545451400330398,
+      "learning_rate": 1.4062542279598689e-05,
+      "loss": 1.1716,
+      "step": 11574
+    },
+    {
+      "epoch": 0.834354501549773,
+      "grad_norm": 0.11431663205032279,
+      "learning_rate": 1.4050606135758392e-05,
+      "loss": 1.216,
+      "step": 11575
+    },
+    {
+      "epoch": 0.8344265840121099,
+      "grad_norm": 0.13003532841404578,
+      "learning_rate": 1.4038674676928986e-05,
+      "loss": 1.3568,
+      "step": 11576
+    },
+    {
+      "epoch": 0.8344986664744468,
+      "grad_norm": 0.12493257230434879,
+      "learning_rate": 1.402674790376085e-05,
+      "loss": 1.2214,
+      "step": 11577
+    },
+    {
+      "epoch": 0.8345707489367837,
+      "grad_norm": 0.12372210478739076,
+      "learning_rate": 1.4014825816904075e-05,
+      "loss": 1.3361,
+      "step": 11578
+    },
+    {
+      "epoch": 0.8346428313991205,
+      "grad_norm": 0.12146427617106124,
+      "learning_rate": 1.4002908417008553e-05,
+      "loss": 1.258,
+      "step": 11579
+    },
+    {
+      "epoch": 0.8347149138614575,
+      "grad_norm": 0.13494799798997084,
+      "learning_rate": 1.3990995704723875e-05,
+      "loss": 1.3048,
+      "step": 11580
+    },
+    {
+      "epoch": 0.8347869963237944,
+      "grad_norm": 0.12336258841400986,
+      "learning_rate": 1.3979087680699377e-05,
+      "loss": 1.2818,
+      "step": 11581
+    },
+    {
+      "epoch": 0.8348590787861313,
+      "grad_norm": 0.11878501908054642,
+      "learning_rate": 1.3967184345584173e-05,
+      "loss": 1.2463,
+      "step": 11582
+    },
+    {
+      "epoch": 0.8349311612484682,
+      "grad_norm": 0.11900731803478151,
+      "learning_rate": 1.3955285700027099e-05,
+      "loss": 1.378,
+      "step": 11583
+    },
+    {
+      "epoch": 0.8350032437108051,
+      "grad_norm": 0.11950696831475882,
+      "learning_rate": 1.3943391744676727e-05,
+      "loss": 1.2814,
+      "step": 11584
+    },
+    {
+      "epoch": 0.8350753261731421,
+      "grad_norm": 0.12946726387264018,
+      "learning_rate": 1.3931502480181391e-05,
+      "loss": 1.2563,
+      "step": 11585
+    },
+    {
+      "epoch": 0.835147408635479,
+      "grad_norm": 0.12340295008209115,
+      "learning_rate": 1.3919617907189154e-05,
+      "loss": 1.2812,
+      "step": 11586
+    },
+    {
+      "epoch": 0.8352194910978159,
+      "grad_norm": 0.12258812878737393,
+      "learning_rate": 1.3907738026347849e-05,
+      "loss": 1.1867,
+      "step": 11587
+    },
+    {
+      "epoch": 0.8352915735601528,
+      "grad_norm": 0.13437155138072138,
+      "learning_rate": 1.3895862838305018e-05,
+      "loss": 1.3187,
+      "step": 11588
+    },
+    {
+      "epoch": 0.8353636560224897,
+      "grad_norm": 0.11617827052636516,
+      "learning_rate": 1.3883992343707975e-05,
+      "loss": 1.3031,
+      "step": 11589
+    },
+    {
+      "epoch": 0.8354357384848267,
+      "grad_norm": 0.11966496413887093,
+      "learning_rate": 1.3872126543203755e-05,
+      "loss": 1.2328,
+      "step": 11590
+    },
+    {
+      "epoch": 0.8355078209471636,
+      "grad_norm": 0.1136571929690267,
+      "learning_rate": 1.3860265437439169e-05,
+      "loss": 1.2502,
+      "step": 11591
+    },
+    {
+      "epoch": 0.8355799034095005,
+      "grad_norm": 0.1277384552817239,
+      "learning_rate": 1.3848409027060748e-05,
+      "loss": 1.317,
+      "step": 11592
+    },
+    {
+      "epoch": 0.8356519858718374,
+      "grad_norm": 0.11167577031365052,
+      "learning_rate": 1.383655731271477e-05,
+      "loss": 1.2682,
+      "step": 11593
+    },
+    {
+      "epoch": 0.8357240683341743,
+      "grad_norm": 0.12522845357615142,
+      "learning_rate": 1.3824710295047272e-05,
+      "loss": 1.2323,
+      "step": 11594
+    },
+    {
+      "epoch": 0.8357961507965113,
+      "grad_norm": 0.13889361761481855,
+      "learning_rate": 1.381286797470397e-05,
+      "loss": 1.3383,
+      "step": 11595
+    },
+    {
+      "epoch": 0.8358682332588481,
+      "grad_norm": 0.16542383270941738,
+      "learning_rate": 1.3801030352330435e-05,
+      "loss": 1.3081,
+      "step": 11596
+    },
+    {
+      "epoch": 0.835940315721185,
+      "grad_norm": 0.1285137792110732,
+      "learning_rate": 1.3789197428571909e-05,
+      "loss": 1.2105,
+      "step": 11597
+    },
+    {
+      "epoch": 0.8360123981835219,
+      "grad_norm": 0.12733741034820406,
+      "learning_rate": 1.377736920407341e-05,
+      "loss": 1.2862,
+      "step": 11598
+    },
+    {
+      "epoch": 0.8360844806458588,
+      "grad_norm": 0.1262673776781347,
+      "learning_rate": 1.3765545679479609e-05,
+      "loss": 1.2603,
+      "step": 11599
+    },
+    {
+      "epoch": 0.8361565631081957,
+      "grad_norm": 0.12846517874038554,
+      "learning_rate": 1.3753726855435068e-05,
+      "loss": 1.1776,
+      "step": 11600
+    },
+    {
+      "epoch": 0.8362286455705327,
+      "grad_norm": 0.10985211350114323,
+      "learning_rate": 1.3741912732583994e-05,
+      "loss": 1.2624,
+      "step": 11601
+    },
+    {
+      "epoch": 0.8363007280328696,
+      "grad_norm": 0.22256875918407876,
+      "learning_rate": 1.3730103311570386e-05,
+      "loss": 1.2903,
+      "step": 11602
+    },
+    {
+      "epoch": 0.8363728104952065,
+      "grad_norm": 0.1218450205417662,
+      "learning_rate": 1.3718298593037916e-05,
+      "loss": 1.2993,
+      "step": 11603
+    },
+    {
+      "epoch": 0.8364448929575434,
+      "grad_norm": 0.11731579867065955,
+      "learning_rate": 1.3706498577630045e-05,
+      "loss": 1.2184,
+      "step": 11604
+    },
+    {
+      "epoch": 0.8365169754198803,
+      "grad_norm": 0.12781165415465687,
+      "learning_rate": 1.3694703265990039e-05,
+      "loss": 1.2418,
+      "step": 11605
+    },
+    {
+      "epoch": 0.8365890578822173,
+      "grad_norm": 0.19437390338059224,
+      "learning_rate": 1.3682912658760827e-05,
+      "loss": 1.2131,
+      "step": 11606
+    },
+    {
+      "epoch": 0.8366611403445542,
+      "grad_norm": 0.122468720387618,
+      "learning_rate": 1.3671126756585073e-05,
+      "loss": 1.2299,
+      "step": 11607
+    },
+    {
+      "epoch": 0.8367332228068911,
+      "grad_norm": 0.11604720236538837,
+      "learning_rate": 1.3659345560105196e-05,
+      "loss": 1.3051,
+      "step": 11608
+    },
+    {
+      "epoch": 0.836805305269228,
+      "grad_norm": 0.1157319850507661,
+      "learning_rate": 1.364756906996345e-05,
+      "loss": 1.237,
+      "step": 11609
+    },
+    {
+      "epoch": 0.836877387731565,
+      "grad_norm": 0.11767868759078312,
+      "learning_rate": 1.3635797286801733e-05,
+      "loss": 1.2794,
+      "step": 11610
+    },
+    {
+      "epoch": 0.8369494701939019,
+      "grad_norm": 0.11662388068363369,
+      "learning_rate": 1.3624030211261685e-05,
+      "loss": 1.2514,
+      "step": 11611
+    },
+    {
+      "epoch": 0.8370215526562388,
+      "grad_norm": 0.12485741001692993,
+      "learning_rate": 1.3612267843984706e-05,
+      "loss": 1.2814,
+      "step": 11612
+    },
+    {
+      "epoch": 0.8370936351185756,
+      "grad_norm": 0.11875623791032497,
+      "learning_rate": 1.3600510185612004e-05,
+      "loss": 1.2102,
+      "step": 11613
+    },
+    {
+      "epoch": 0.8371657175809125,
+      "grad_norm": 0.1130334939824981,
+      "learning_rate": 1.3588757236784456e-05,
+      "loss": 1.2527,
+      "step": 11614
+    },
+    {
+      "epoch": 0.8372378000432494,
+      "grad_norm": 0.11509539819754644,
+      "learning_rate": 1.3577008998142692e-05,
+      "loss": 1.3406,
+      "step": 11615
+    },
+    {
+      "epoch": 0.8373098825055864,
+      "grad_norm": 0.13171363579474082,
+      "learning_rate": 1.3565265470327082e-05,
+      "loss": 1.2933,
+      "step": 11616
+    },
+    {
+      "epoch": 0.8373819649679233,
+      "grad_norm": 0.12252232142296278,
+      "learning_rate": 1.3553526653977766e-05,
+      "loss": 1.2281,
+      "step": 11617
+    },
+    {
+      "epoch": 0.8374540474302602,
+      "grad_norm": 0.121142402516292,
+      "learning_rate": 1.3541792549734655e-05,
+      "loss": 1.2305,
+      "step": 11618
+    },
+    {
+      "epoch": 0.8375261298925971,
+      "grad_norm": 0.11256688180802751,
+      "learning_rate": 1.3530063158237305e-05,
+      "loss": 1.2668,
+      "step": 11619
+    },
+    {
+      "epoch": 0.837598212354934,
+      "grad_norm": 0.10993628166935202,
+      "learning_rate": 1.3518338480125093e-05,
+      "loss": 1.2457,
+      "step": 11620
+    },
+    {
+      "epoch": 0.837670294817271,
+      "grad_norm": 0.12358010458298606,
+      "learning_rate": 1.3506618516037106e-05,
+      "loss": 1.3635,
+      "step": 11621
+    },
+    {
+      "epoch": 0.8377423772796079,
+      "grad_norm": 0.1267580195768701,
+      "learning_rate": 1.3494903266612236e-05,
+      "loss": 1.2032,
+      "step": 11622
+    },
+    {
+      "epoch": 0.8378144597419448,
+      "grad_norm": 0.11682148659508691,
+      "learning_rate": 1.3483192732489025e-05,
+      "loss": 1.2492,
+      "step": 11623
+    },
+    {
+      "epoch": 0.8378865422042817,
+      "grad_norm": 0.11914639316691715,
+      "learning_rate": 1.3471486914305798e-05,
+      "loss": 1.3037,
+      "step": 11624
+    },
+    {
+      "epoch": 0.8379586246666186,
+      "grad_norm": 0.11845863928471831,
+      "learning_rate": 1.3459785812700654e-05,
+      "loss": 1.2899,
+      "step": 11625
+    },
+    {
+      "epoch": 0.8380307071289556,
+      "grad_norm": 0.12380649858729276,
+      "learning_rate": 1.3448089428311384e-05,
+      "loss": 1.2858,
+      "step": 11626
+    },
+    {
+      "epoch": 0.8381027895912925,
+      "grad_norm": 0.12566649197520424,
+      "learning_rate": 1.3436397761775565e-05,
+      "loss": 1.3146,
+      "step": 11627
+    },
+    {
+      "epoch": 0.8381748720536294,
+      "grad_norm": 0.1312011972456959,
+      "learning_rate": 1.342471081373048e-05,
+      "loss": 1.218,
+      "step": 11628
+    },
+    {
+      "epoch": 0.8382469545159663,
+      "grad_norm": 0.12948648862339182,
+      "learning_rate": 1.341302858481318e-05,
+      "loss": 1.3768,
+      "step": 11629
+    },
+    {
+      "epoch": 0.8383190369783032,
+      "grad_norm": 0.12162604831311202,
+      "learning_rate": 1.3401351075660452e-05,
+      "loss": 1.2717,
+      "step": 11630
+    },
+    {
+      "epoch": 0.83839111944064,
+      "grad_norm": 0.11794718276952354,
+      "learning_rate": 1.3389678286908824e-05,
+      "loss": 1.2564,
+      "step": 11631
+    },
+    {
+      "epoch": 0.838463201902977,
+      "grad_norm": 0.12922047392209052,
+      "learning_rate": 1.337801021919457e-05,
+      "loss": 1.3171,
+      "step": 11632
+    },
+    {
+      "epoch": 0.8385352843653139,
+      "grad_norm": 0.11825706233979158,
+      "learning_rate": 1.33663468731537e-05,
+      "loss": 1.2791,
+      "step": 11633
+    },
+    {
+      "epoch": 0.8386073668276508,
+      "grad_norm": 0.12225552116906398,
+      "learning_rate": 1.3354688249421977e-05,
+      "loss": 1.2434,
+      "step": 11634
+    },
+    {
+      "epoch": 0.8386794492899877,
+      "grad_norm": 0.11916339461112262,
+      "learning_rate": 1.3343034348634886e-05,
+      "loss": 1.3263,
+      "step": 11635
+    },
+    {
+      "epoch": 0.8387515317523246,
+      "grad_norm": 0.12328481332428723,
+      "learning_rate": 1.3331385171427691e-05,
+      "loss": 1.2487,
+      "step": 11636
+    },
+    {
+      "epoch": 0.8388236142146616,
+      "grad_norm": 0.11673394116819773,
+      "learning_rate": 1.3319740718435358e-05,
+      "loss": 1.303,
+      "step": 11637
+    },
+    {
+      "epoch": 0.8388956966769985,
+      "grad_norm": 0.1091948256593919,
+      "learning_rate": 1.3308100990292616e-05,
+      "loss": 1.2384,
+      "step": 11638
+    },
+    {
+      "epoch": 0.8389677791393354,
+      "grad_norm": 0.11940904414187581,
+      "learning_rate": 1.3296465987633943e-05,
+      "loss": 1.18,
+      "step": 11639
+    },
+    {
+      "epoch": 0.8390398616016723,
+      "grad_norm": 0.12273931093691626,
+      "learning_rate": 1.3284835711093535e-05,
+      "loss": 1.2797,
+      "step": 11640
+    },
+    {
+      "epoch": 0.8391119440640092,
+      "grad_norm": 0.12168652813582932,
+      "learning_rate": 1.3273210161305372e-05,
+      "loss": 1.3372,
+      "step": 11641
+    },
+    {
+      "epoch": 0.8391840265263462,
+      "grad_norm": 0.11104283849275885,
+      "learning_rate": 1.3261589338903124e-05,
+      "loss": 1.2605,
+      "step": 11642
+    },
+    {
+      "epoch": 0.8392561089886831,
+      "grad_norm": 0.11526500661286773,
+      "learning_rate": 1.3249973244520241e-05,
+      "loss": 1.3123,
+      "step": 11643
+    },
+    {
+      "epoch": 0.83932819145102,
+      "grad_norm": 0.11429306016475721,
+      "learning_rate": 1.3238361878789919e-05,
+      "loss": 1.2831,
+      "step": 11644
+    },
+    {
+      "epoch": 0.8394002739133569,
+      "grad_norm": 0.11992123653177458,
+      "learning_rate": 1.3226755242345046e-05,
+      "loss": 1.2902,
+      "step": 11645
+    },
+    {
+      "epoch": 0.8394723563756938,
+      "grad_norm": 0.11095229578233026,
+      "learning_rate": 1.321515333581832e-05,
+      "loss": 1.2216,
+      "step": 11646
+    },
+    {
+      "epoch": 0.8395444388380308,
+      "grad_norm": 0.11621266881308341,
+      "learning_rate": 1.3203556159842156e-05,
+      "loss": 1.278,
+      "step": 11647
+    },
+    {
+      "epoch": 0.8396165213003676,
+      "grad_norm": 0.12049034004263008,
+      "learning_rate": 1.3191963715048638e-05,
+      "loss": 1.3191,
+      "step": 11648
+    },
+    {
+      "epoch": 0.8396886037627045,
+      "grad_norm": 0.12315535359772099,
+      "learning_rate": 1.3180376002069728e-05,
+      "loss": 1.2386,
+      "step": 11649
+    },
+    {
+      "epoch": 0.8397606862250414,
+      "grad_norm": 0.12342214839964259,
+      "learning_rate": 1.316879302153704e-05,
+      "loss": 1.2893,
+      "step": 11650
+    },
+    {
+      "epoch": 0.8398327686873783,
+      "grad_norm": 0.11303480592135795,
+      "learning_rate": 1.3157214774081972e-05,
+      "loss": 1.2875,
+      "step": 11651
+    },
+    {
+      "epoch": 0.8399048511497152,
+      "grad_norm": 0.12141472622304661,
+      "learning_rate": 1.3145641260335562e-05,
+      "loss": 1.2499,
+      "step": 11652
+    },
+    {
+      "epoch": 0.8399769336120522,
+      "grad_norm": 0.12175240124511721,
+      "learning_rate": 1.313407248092875e-05,
+      "loss": 1.1981,
+      "step": 11653
+    },
+    {
+      "epoch": 0.8400490160743891,
+      "grad_norm": 0.11393419494238624,
+      "learning_rate": 1.3122508436492109e-05,
+      "loss": 1.3205,
+      "step": 11654
+    },
+    {
+      "epoch": 0.840121098536726,
+      "grad_norm": 0.13310638426100208,
+      "learning_rate": 1.3110949127656014e-05,
+      "loss": 1.2741,
+      "step": 11655
+    },
+    {
+      "epoch": 0.8401931809990629,
+      "grad_norm": 0.1281864080063397,
+      "learning_rate": 1.3099394555050504e-05,
+      "loss": 1.2442,
+      "step": 11656
+    },
+    {
+      "epoch": 0.8402652634613998,
+      "grad_norm": 0.12586796620006827,
+      "learning_rate": 1.3087844719305397e-05,
+      "loss": 1.2446,
+      "step": 11657
+    },
+    {
+      "epoch": 0.8403373459237368,
+      "grad_norm": 0.12092069357094158,
+      "learning_rate": 1.3076299621050336e-05,
+      "loss": 1.206,
+      "step": 11658
+    },
+    {
+      "epoch": 0.8404094283860737,
+      "grad_norm": 0.11683026860373315,
+      "learning_rate": 1.3064759260914573e-05,
+      "loss": 1.2491,
+      "step": 11659
+    },
+    {
+      "epoch": 0.8404815108484106,
+      "grad_norm": 0.12419666674058849,
+      "learning_rate": 1.3053223639527168e-05,
+      "loss": 1.3478,
+      "step": 11660
+    },
+    {
+      "epoch": 0.8405535933107475,
+      "grad_norm": 0.12364131955304668,
+      "learning_rate": 1.3041692757516898e-05,
+      "loss": 1.3452,
+      "step": 11661
+    },
+    {
+      "epoch": 0.8406256757730844,
+      "grad_norm": 0.11512487227806209,
+      "learning_rate": 1.3030166615512374e-05,
+      "loss": 1.2376,
+      "step": 11662
+    },
+    {
+      "epoch": 0.8406977582354214,
+      "grad_norm": 0.12384524660929751,
+      "learning_rate": 1.30186452141418e-05,
+      "loss": 1.3165,
+      "step": 11663
+    },
+    {
+      "epoch": 0.8407698406977583,
+      "grad_norm": 0.11942669408610539,
+      "learning_rate": 1.3007128554033222e-05,
+      "loss": 1.199,
+      "step": 11664
+    },
+    {
+      "epoch": 0.8408419231600951,
+      "grad_norm": 0.11907271184760605,
+      "learning_rate": 1.2995616635814379e-05,
+      "loss": 1.2208,
+      "step": 11665
+    },
+    {
+      "epoch": 0.840914005622432,
+      "grad_norm": 0.142294132930552,
+      "learning_rate": 1.2984109460112825e-05,
+      "loss": 1.3054,
+      "step": 11666
+    },
+    {
+      "epoch": 0.8409860880847689,
+      "grad_norm": 0.1211231441973865,
+      "learning_rate": 1.2972607027555761e-05,
+      "loss": 1.2953,
+      "step": 11667
+    },
+    {
+      "epoch": 0.8410581705471059,
+      "grad_norm": 0.11986887537703696,
+      "learning_rate": 1.2961109338770172e-05,
+      "loss": 1.3366,
+      "step": 11668
+    },
+    {
+      "epoch": 0.8411302530094428,
+      "grad_norm": 0.12354800849124781,
+      "learning_rate": 1.2949616394382802e-05,
+      "loss": 1.3178,
+      "step": 11669
+    },
+    {
+      "epoch": 0.8412023354717797,
+      "grad_norm": 0.1221339863261037,
+      "learning_rate": 1.2938128195020116e-05,
+      "loss": 1.2965,
+      "step": 11670
+    },
+    {
+      "epoch": 0.8412744179341166,
+      "grad_norm": 0.11067076594287023,
+      "learning_rate": 1.2926644741308326e-05,
+      "loss": 1.169,
+      "step": 11671
+    },
+    {
+      "epoch": 0.8413465003964535,
+      "grad_norm": 0.10519509392127725,
+      "learning_rate": 1.2915166033873371e-05,
+      "loss": 1.2727,
+      "step": 11672
+    },
+    {
+      "epoch": 0.8414185828587905,
+      "grad_norm": 0.11815027964124926,
+      "learning_rate": 1.2903692073340945e-05,
+      "loss": 1.243,
+      "step": 11673
+    },
+    {
+      "epoch": 0.8414906653211274,
+      "grad_norm": 0.11837989630117952,
+      "learning_rate": 1.289222286033649e-05,
+      "loss": 1.2893,
+      "step": 11674
+    },
+    {
+      "epoch": 0.8415627477834643,
+      "grad_norm": 0.12270392615397185,
+      "learning_rate": 1.2880758395485183e-05,
+      "loss": 1.2717,
+      "step": 11675
+    },
+    {
+      "epoch": 0.8416348302458012,
+      "grad_norm": 0.11863833552106652,
+      "learning_rate": 1.2869298679411934e-05,
+      "loss": 1.1668,
+      "step": 11676
+    },
+    {
+      "epoch": 0.8417069127081381,
+      "grad_norm": 0.17061201020529027,
+      "learning_rate": 1.2857843712741401e-05,
+      "loss": 1.3108,
+      "step": 11677
+    },
+    {
+      "epoch": 0.841778995170475,
+      "grad_norm": 0.1270135063734845,
+      "learning_rate": 1.2846393496097975e-05,
+      "loss": 1.2237,
+      "step": 11678
+    },
+    {
+      "epoch": 0.841851077632812,
+      "grad_norm": 0.11370466938085226,
+      "learning_rate": 1.2834948030105808e-05,
+      "loss": 1.2141,
+      "step": 11679
+    },
+    {
+      "epoch": 0.8419231600951489,
+      "grad_norm": 0.11742497057284354,
+      "learning_rate": 1.2823507315388772e-05,
+      "loss": 1.2845,
+      "step": 11680
+    },
+    {
+      "epoch": 0.8419952425574858,
+      "grad_norm": 0.13068139490129743,
+      "learning_rate": 1.2812071352570488e-05,
+      "loss": 1.2342,
+      "step": 11681
+    },
+    {
+      "epoch": 0.8420673250198227,
+      "grad_norm": 0.12015394721516708,
+      "learning_rate": 1.2800640142274324e-05,
+      "loss": 1.2299,
+      "step": 11682
+    },
+    {
+      "epoch": 0.8421394074821595,
+      "grad_norm": 0.12603586080380264,
+      "learning_rate": 1.278921368512338e-05,
+      "loss": 1.2346,
+      "step": 11683
+    },
+    {
+      "epoch": 0.8422114899444965,
+      "grad_norm": 0.11934041477892127,
+      "learning_rate": 1.2777791981740495e-05,
+      "loss": 1.2646,
+      "step": 11684
+    },
+    {
+      "epoch": 0.8422835724068334,
+      "grad_norm": 0.1162485751486297,
+      "learning_rate": 1.2766375032748268e-05,
+      "loss": 1.3373,
+      "step": 11685
+    },
+    {
+      "epoch": 0.8423556548691703,
+      "grad_norm": 0.12297094985190064,
+      "learning_rate": 1.2754962838769013e-05,
+      "loss": 1.2707,
+      "step": 11686
+    },
+    {
+      "epoch": 0.8424277373315072,
+      "grad_norm": 0.1125725507015629,
+      "learning_rate": 1.27435554004248e-05,
+      "loss": 1.2602,
+      "step": 11687
+    },
+    {
+      "epoch": 0.8424998197938441,
+      "grad_norm": 0.12424786236695314,
+      "learning_rate": 1.2732152718337431e-05,
+      "loss": 1.1823,
+      "step": 11688
+    },
+    {
+      "epoch": 0.8425719022561811,
+      "grad_norm": 0.12399919741674492,
+      "learning_rate": 1.2720754793128464e-05,
+      "loss": 1.2766,
+      "step": 11689
+    },
+    {
+      "epoch": 0.842643984718518,
+      "grad_norm": 0.11566677787956821,
+      "learning_rate": 1.2709361625419181e-05,
+      "loss": 1.2408,
+      "step": 11690
+    },
+    {
+      "epoch": 0.8427160671808549,
+      "grad_norm": 0.12339332010195359,
+      "learning_rate": 1.2697973215830638e-05,
+      "loss": 1.2811,
+      "step": 11691
+    },
+    {
+      "epoch": 0.8427881496431918,
+      "grad_norm": 0.11256970076190476,
+      "learning_rate": 1.268658956498353e-05,
+      "loss": 1.2174,
+      "step": 11692
+    },
+    {
+      "epoch": 0.8428602321055287,
+      "grad_norm": 0.11401604931307209,
+      "learning_rate": 1.267521067349845e-05,
+      "loss": 1.2381,
+      "step": 11693
+    },
+    {
+      "epoch": 0.8429323145678657,
+      "grad_norm": 0.117934977412739,
+      "learning_rate": 1.2663836541995611e-05,
+      "loss": 1.274,
+      "step": 11694
+    },
+    {
+      "epoch": 0.8430043970302026,
+      "grad_norm": 0.10871236747706466,
+      "learning_rate": 1.2652467171095039e-05,
+      "loss": 1.2316,
+      "step": 11695
+    },
+    {
+      "epoch": 0.8430764794925395,
+      "grad_norm": 0.13201809360684535,
+      "learning_rate": 1.2641102561416396e-05,
+      "loss": 1.2204,
+      "step": 11696
+    },
+    {
+      "epoch": 0.8431485619548764,
+      "grad_norm": 0.12494145447255675,
+      "learning_rate": 1.2629742713579219e-05,
+      "loss": 1.2675,
+      "step": 11697
+    },
+    {
+      "epoch": 0.8432206444172133,
+      "grad_norm": 0.11680997954420623,
+      "learning_rate": 1.2618387628202699e-05,
+      "loss": 1.2482,
+      "step": 11698
+    },
+    {
+      "epoch": 0.8432927268795503,
+      "grad_norm": 0.12023971147673818,
+      "learning_rate": 1.2607037305905811e-05,
+      "loss": 1.2769,
+      "step": 11699
+    },
+    {
+      "epoch": 0.8433648093418871,
+      "grad_norm": 0.10961177324777174,
+      "learning_rate": 1.2595691747307214e-05,
+      "loss": 1.3027,
+      "step": 11700
+    },
+    {
+      "epoch": 0.843436891804224,
+      "grad_norm": 0.11039364753919284,
+      "learning_rate": 1.2584350953025337e-05,
+      "loss": 1.3007,
+      "step": 11701
+    },
+    {
+      "epoch": 0.8435089742665609,
+      "grad_norm": 0.12527917321110923,
+      "learning_rate": 1.2573014923678395e-05,
+      "loss": 1.2051,
+      "step": 11702
+    },
+    {
+      "epoch": 0.8435810567288978,
+      "grad_norm": 0.11623557224319292,
+      "learning_rate": 1.2561683659884305e-05,
+      "loss": 1.1807,
+      "step": 11703
+    },
+    {
+      "epoch": 0.8436531391912347,
+      "grad_norm": 0.1220712279640005,
+      "learning_rate": 1.2550357162260672e-05,
+      "loss": 1.2154,
+      "step": 11704
+    },
+    {
+      "epoch": 0.8437252216535717,
+      "grad_norm": 0.12221780284319843,
+      "learning_rate": 1.2539035431424917e-05,
+      "loss": 1.292,
+      "step": 11705
+    },
+    {
+      "epoch": 0.8437973041159086,
+      "grad_norm": 0.11859207084799178,
+      "learning_rate": 1.2527718467994187e-05,
+      "loss": 1.2788,
+      "step": 11706
+    },
+    {
+      "epoch": 0.8438693865782455,
+      "grad_norm": 0.12037263937144266,
+      "learning_rate": 1.2516406272585368e-05,
+      "loss": 1.3579,
+      "step": 11707
+    },
+    {
+      "epoch": 0.8439414690405824,
+      "grad_norm": 0.11827787384154882,
+      "learning_rate": 1.2505098845815044e-05,
+      "loss": 1.3382,
+      "step": 11708
+    },
+    {
+      "epoch": 0.8440135515029193,
+      "grad_norm": 0.12248571804355703,
+      "learning_rate": 1.2493796188299567e-05,
+      "loss": 1.2825,
+      "step": 11709
+    },
+    {
+      "epoch": 0.8440856339652563,
+      "grad_norm": 0.1597890034726168,
+      "learning_rate": 1.2482498300655065e-05,
+      "loss": 1.3051,
+      "step": 11710
+    },
+    {
+      "epoch": 0.8441577164275932,
+      "grad_norm": 0.12389773046456987,
+      "learning_rate": 1.2471205183497381e-05,
+      "loss": 1.36,
+      "step": 11711
+    },
+    {
+      "epoch": 0.8442297988899301,
+      "grad_norm": 0.11565251667110901,
+      "learning_rate": 1.2459916837442054e-05,
+      "loss": 1.258,
+      "step": 11712
+    },
+    {
+      "epoch": 0.844301881352267,
+      "grad_norm": 0.12403046808886567,
+      "learning_rate": 1.2448633263104415e-05,
+      "loss": 1.3698,
+      "step": 11713
+    },
+    {
+      "epoch": 0.844373963814604,
+      "grad_norm": 0.13241949069218698,
+      "learning_rate": 1.2437354461099493e-05,
+      "loss": 1.2346,
+      "step": 11714
+    },
+    {
+      "epoch": 0.8444460462769409,
+      "grad_norm": 0.11034475542910782,
+      "learning_rate": 1.2426080432042153e-05,
+      "loss": 1.2034,
+      "step": 11715
+    },
+    {
+      "epoch": 0.8445181287392778,
+      "grad_norm": 0.11701132710423393,
+      "learning_rate": 1.241481117654687e-05,
+      "loss": 1.2654,
+      "step": 11716
+    },
+    {
+      "epoch": 0.8445902112016146,
+      "grad_norm": 0.11557880219461171,
+      "learning_rate": 1.240354669522793e-05,
+      "loss": 1.2841,
+      "step": 11717
+    },
+    {
+      "epoch": 0.8446622936639515,
+      "grad_norm": 0.11694895207908809,
+      "learning_rate": 1.2392286988699341e-05,
+      "loss": 1.1917,
+      "step": 11718
+    },
+    {
+      "epoch": 0.8447343761262884,
+      "grad_norm": 0.11787364409390381,
+      "learning_rate": 1.2381032057574915e-05,
+      "loss": 1.2631,
+      "step": 11719
+    },
+    {
+      "epoch": 0.8448064585886254,
+      "grad_norm": 0.216159859722899,
+      "learning_rate": 1.236978190246808e-05,
+      "loss": 1.2897,
+      "step": 11720
+    },
+    {
+      "epoch": 0.8448785410509623,
+      "grad_norm": 0.12624788154910077,
+      "learning_rate": 1.2358536523992103e-05,
+      "loss": 1.2606,
+      "step": 11721
+    },
+    {
+      "epoch": 0.8449506235132992,
+      "grad_norm": 0.1173202525028851,
+      "learning_rate": 1.2347295922759938e-05,
+      "loss": 1.2464,
+      "step": 11722
+    },
+    {
+      "epoch": 0.8450227059756361,
+      "grad_norm": 0.12239252424654459,
+      "learning_rate": 1.2336060099384306e-05,
+      "loss": 1.28,
+      "step": 11723
+    },
+    {
+      "epoch": 0.845094788437973,
+      "grad_norm": 0.12294452609754201,
+      "learning_rate": 1.2324829054477671e-05,
+      "loss": 1.2833,
+      "step": 11724
+    },
+    {
+      "epoch": 0.84516687090031,
+      "grad_norm": 0.11343863887194881,
+      "learning_rate": 1.2313602788652212e-05,
+      "loss": 1.2005,
+      "step": 11725
+    },
+    {
+      "epoch": 0.8452389533626469,
+      "grad_norm": 0.11715654703978011,
+      "learning_rate": 1.2302381302519873e-05,
+      "loss": 1.2542,
+      "step": 11726
+    },
+    {
+      "epoch": 0.8453110358249838,
+      "grad_norm": 0.12549731090478766,
+      "learning_rate": 1.2291164596692305e-05,
+      "loss": 1.2181,
+      "step": 11727
+    },
+    {
+      "epoch": 0.8453831182873207,
+      "grad_norm": 0.13161907125866912,
+      "learning_rate": 1.2279952671780948e-05,
+      "loss": 1.3181,
+      "step": 11728
+    },
+    {
+      "epoch": 0.8454552007496576,
+      "grad_norm": 0.12287148903435843,
+      "learning_rate": 1.226874552839693e-05,
+      "loss": 1.292,
+      "step": 11729
+    },
+    {
+      "epoch": 0.8455272832119946,
+      "grad_norm": 0.11866093535737354,
+      "learning_rate": 1.2257543167151142e-05,
+      "loss": 1.2773,
+      "step": 11730
+    },
+    {
+      "epoch": 0.8455993656743315,
+      "grad_norm": 0.12803049746250728,
+      "learning_rate": 1.2246345588654218e-05,
+      "loss": 1.12,
+      "step": 11731
+    },
+    {
+      "epoch": 0.8456714481366684,
+      "grad_norm": 0.12911510672387566,
+      "learning_rate": 1.2235152793516536e-05,
+      "loss": 1.3152,
+      "step": 11732
+    },
+    {
+      "epoch": 0.8457435305990053,
+      "grad_norm": 0.13218291196194143,
+      "learning_rate": 1.2223964782348185e-05,
+      "loss": 1.2614,
+      "step": 11733
+    },
+    {
+      "epoch": 0.8458156130613422,
+      "grad_norm": 0.12458299412049528,
+      "learning_rate": 1.2212781555759022e-05,
+      "loss": 1.2545,
+      "step": 11734
+    },
+    {
+      "epoch": 0.845887695523679,
+      "grad_norm": 0.29517703190117944,
+      "learning_rate": 1.2201603114358628e-05,
+      "loss": 1.3067,
+      "step": 11735
+    },
+    {
+      "epoch": 0.845959777986016,
+      "grad_norm": 0.15636900791160427,
+      "learning_rate": 1.2190429458756336e-05,
+      "loss": 1.2377,
+      "step": 11736
+    },
+    {
+      "epoch": 0.8460318604483529,
+      "grad_norm": 0.1177830815484915,
+      "learning_rate": 1.2179260589561204e-05,
+      "loss": 1.187,
+      "step": 11737
+    },
+    {
+      "epoch": 0.8461039429106898,
+      "grad_norm": 0.10822401291511112,
+      "learning_rate": 1.2168096507382043e-05,
+      "loss": 1.1634,
+      "step": 11738
+    },
+    {
+      "epoch": 0.8461760253730267,
+      "grad_norm": 0.11956136144645008,
+      "learning_rate": 1.2156937212827379e-05,
+      "loss": 1.3145,
+      "step": 11739
+    },
+    {
+      "epoch": 0.8462481078353636,
+      "grad_norm": 0.12994969186607927,
+      "learning_rate": 1.2145782706505505e-05,
+      "loss": 1.169,
+      "step": 11740
+    },
+    {
+      "epoch": 0.8463201902977006,
+      "grad_norm": 0.12768132054103015,
+      "learning_rate": 1.2134632989024441e-05,
+      "loss": 1.2685,
+      "step": 11741
+    },
+    {
+      "epoch": 0.8463922727600375,
+      "grad_norm": 0.21356124851325892,
+      "learning_rate": 1.212348806099195e-05,
+      "loss": 1.2924,
+      "step": 11742
+    },
+    {
+      "epoch": 0.8464643552223744,
+      "grad_norm": 0.11946223003002632,
+      "learning_rate": 1.211234792301552e-05,
+      "loss": 1.2485,
+      "step": 11743
+    },
+    {
+      "epoch": 0.8465364376847113,
+      "grad_norm": 0.11236290265978781,
+      "learning_rate": 1.2101212575702414e-05,
+      "loss": 1.3058,
+      "step": 11744
+    },
+    {
+      "epoch": 0.8466085201470482,
+      "grad_norm": 0.11798278338912103,
+      "learning_rate": 1.2090082019659544e-05,
+      "loss": 1.2826,
+      "step": 11745
+    },
+    {
+      "epoch": 0.8466806026093852,
+      "grad_norm": 0.12530888842305157,
+      "learning_rate": 1.2078956255493701e-05,
+      "loss": 1.2838,
+      "step": 11746
+    },
+    {
+      "epoch": 0.8467526850717221,
+      "grad_norm": 0.10706970089947107,
+      "learning_rate": 1.206783528381129e-05,
+      "loss": 1.2735,
+      "step": 11747
+    },
+    {
+      "epoch": 0.846824767534059,
+      "grad_norm": 0.12295280078226951,
+      "learning_rate": 1.2056719105218551e-05,
+      "loss": 1.2936,
+      "step": 11748
+    },
+    {
+      "epoch": 0.8468968499963959,
+      "grad_norm": 0.11690838801435566,
+      "learning_rate": 1.204560772032134e-05,
+      "loss": 1.2339,
+      "step": 11749
+    },
+    {
+      "epoch": 0.8469689324587328,
+      "grad_norm": 0.121822134484339,
+      "learning_rate": 1.2034501129725396e-05,
+      "loss": 1.2758,
+      "step": 11750
+    },
+    {
+      "epoch": 0.8470410149210698,
+      "grad_norm": 0.12954687414696947,
+      "learning_rate": 1.2023399334036089e-05,
+      "loss": 1.3247,
+      "step": 11751
+    },
+    {
+      "epoch": 0.8471130973834066,
+      "grad_norm": 0.11920716212915113,
+      "learning_rate": 1.2012302333858616e-05,
+      "loss": 1.2889,
+      "step": 11752
+    },
+    {
+      "epoch": 0.8471851798457435,
+      "grad_norm": 0.12531458240160107,
+      "learning_rate": 1.2001210129797779e-05,
+      "loss": 1.288,
+      "step": 11753
+    },
+    {
+      "epoch": 0.8472572623080804,
+      "grad_norm": 0.12018195952652445,
+      "learning_rate": 1.1990122722458264e-05,
+      "loss": 1.1902,
+      "step": 11754
+    },
+    {
+      "epoch": 0.8473293447704173,
+      "grad_norm": 0.11689721671982442,
+      "learning_rate": 1.1979040112444428e-05,
+      "loss": 1.2674,
+      "step": 11755
+    },
+    {
+      "epoch": 0.8474014272327542,
+      "grad_norm": 0.12329127773576923,
+      "learning_rate": 1.196796230036038e-05,
+      "loss": 1.231,
+      "step": 11756
+    },
+    {
+      "epoch": 0.8474735096950912,
+      "grad_norm": 0.11410515758819952,
+      "learning_rate": 1.1956889286809925e-05,
+      "loss": 1.2842,
+      "step": 11757
+    },
+    {
+      "epoch": 0.8475455921574281,
+      "grad_norm": 0.12010696252226424,
+      "learning_rate": 1.1945821072396635e-05,
+      "loss": 1.2711,
+      "step": 11758
+    },
+    {
+      "epoch": 0.847617674619765,
+      "grad_norm": 0.11816889668432089,
+      "learning_rate": 1.1934757657723882e-05,
+      "loss": 1.2339,
+      "step": 11759
+    },
+    {
+      "epoch": 0.8476897570821019,
+      "grad_norm": 0.11202187634432989,
+      "learning_rate": 1.1923699043394699e-05,
+      "loss": 1.2535,
+      "step": 11760
+    },
+    {
+      "epoch": 0.8477618395444388,
+      "grad_norm": 0.11531926200456366,
+      "learning_rate": 1.1912645230011854e-05,
+      "loss": 1.2983,
+      "step": 11761
+    },
+    {
+      "epoch": 0.8478339220067758,
+      "grad_norm": 0.12439322815049529,
+      "learning_rate": 1.190159621817788e-05,
+      "loss": 1.2527,
+      "step": 11762
+    },
+    {
+      "epoch": 0.8479060044691127,
+      "grad_norm": 0.12505106466047655,
+      "learning_rate": 1.1890552008495081e-05,
+      "loss": 1.2504,
+      "step": 11763
+    },
+    {
+      "epoch": 0.8479780869314496,
+      "grad_norm": 0.1284347901082484,
+      "learning_rate": 1.1879512601565467e-05,
+      "loss": 1.2854,
+      "step": 11764
+    },
+    {
+      "epoch": 0.8480501693937865,
+      "grad_norm": 0.13593709690176053,
+      "learning_rate": 1.1868477997990756e-05,
+      "loss": 1.2092,
+      "step": 11765
+    },
+    {
+      "epoch": 0.8481222518561234,
+      "grad_norm": 0.11933041747109853,
+      "learning_rate": 1.1857448198372423e-05,
+      "loss": 1.2967,
+      "step": 11766
+    },
+    {
+      "epoch": 0.8481943343184604,
+      "grad_norm": 0.12321936838052805,
+      "learning_rate": 1.1846423203311696e-05,
+      "loss": 1.2369,
+      "step": 11767
+    },
+    {
+      "epoch": 0.8482664167807973,
+      "grad_norm": 0.12015277008483806,
+      "learning_rate": 1.1835403013409596e-05,
+      "loss": 1.1633,
+      "step": 11768
+    },
+    {
+      "epoch": 0.8483384992431341,
+      "grad_norm": 0.11674150282655106,
+      "learning_rate": 1.1824387629266742e-05,
+      "loss": 1.2958,
+      "step": 11769
+    },
+    {
+      "epoch": 0.848410581705471,
+      "grad_norm": 0.11380991418791804,
+      "learning_rate": 1.1813377051483621e-05,
+      "loss": 1.2527,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8484826641678079,
+      "grad_norm": 0.12300798554164741,
+      "learning_rate": 1.1802371280660352e-05,
+      "loss": 1.191,
+      "step": 11771
+    },
+    {
+      "epoch": 0.8485547466301449,
+      "grad_norm": 0.11260964313041313,
+      "learning_rate": 1.1791370317396933e-05,
+      "loss": 1.2773,
+      "step": 11772
+    },
+    {
+      "epoch": 0.8486268290924818,
+      "grad_norm": 0.11739975535824891,
+      "learning_rate": 1.1780374162292951e-05,
+      "loss": 1.2041,
+      "step": 11773
+    },
+    {
+      "epoch": 0.8486989115548187,
+      "grad_norm": 0.1176981564058792,
+      "learning_rate": 1.1769382815947816e-05,
+      "loss": 1.1955,
+      "step": 11774
+    },
+    {
+      "epoch": 0.8487709940171556,
+      "grad_norm": 0.12063146971548973,
+      "learning_rate": 1.1758396278960626e-05,
+      "loss": 1.3025,
+      "step": 11775
+    },
+    {
+      "epoch": 0.8488430764794925,
+      "grad_norm": 0.11907339886174824,
+      "learning_rate": 1.1747414551930314e-05,
+      "loss": 1.2269,
+      "step": 11776
+    },
+    {
+      "epoch": 0.8489151589418295,
+      "grad_norm": 0.12189084040768988,
+      "learning_rate": 1.1736437635455411e-05,
+      "loss": 1.3103,
+      "step": 11777
+    },
+    {
+      "epoch": 0.8489872414041664,
+      "grad_norm": 0.12870278972521265,
+      "learning_rate": 1.1725465530134294e-05,
+      "loss": 1.2811,
+      "step": 11778
+    },
+    {
+      "epoch": 0.8490593238665033,
+      "grad_norm": 0.1244051552390999,
+      "learning_rate": 1.171449823656503e-05,
+      "loss": 1.2011,
+      "step": 11779
+    },
+    {
+      "epoch": 0.8491314063288402,
+      "grad_norm": 0.12487964628699205,
+      "learning_rate": 1.170353575534544e-05,
+      "loss": 1.2779,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8492034887911771,
+      "grad_norm": 0.11247499974401909,
+      "learning_rate": 1.169257808707307e-05,
+      "loss": 1.2137,
+      "step": 11781
+    },
+    {
+      "epoch": 0.849275571253514,
+      "grad_norm": 0.11803106537334963,
+      "learning_rate": 1.1681625232345217e-05,
+      "loss": 1.114,
+      "step": 11782
+    },
+    {
+      "epoch": 0.849347653715851,
+      "grad_norm": 0.11073641786643713,
+      "learning_rate": 1.1670677191758905e-05,
+      "loss": 1.3258,
+      "step": 11783
+    },
+    {
+      "epoch": 0.8494197361781879,
+      "grad_norm": 0.11133091228843695,
+      "learning_rate": 1.1659733965910902e-05,
+      "loss": 1.2597,
+      "step": 11784
+    },
+    {
+      "epoch": 0.8494918186405248,
+      "grad_norm": 0.1254694982220098,
+      "learning_rate": 1.1648795555397719e-05,
+      "loss": 1.1986,
+      "step": 11785
+    },
+    {
+      "epoch": 0.8495639011028616,
+      "grad_norm": 0.12109502322994259,
+      "learning_rate": 1.1637861960815577e-05,
+      "loss": 1.291,
+      "step": 11786
+    },
+    {
+      "epoch": 0.8496359835651985,
+      "grad_norm": 0.1318013515746823,
+      "learning_rate": 1.162693318276048e-05,
+      "loss": 1.1674,
+      "step": 11787
+    },
+    {
+      "epoch": 0.8497080660275355,
+      "grad_norm": 0.12070805980887507,
+      "learning_rate": 1.1616009221828129e-05,
+      "loss": 1.2316,
+      "step": 11788
+    },
+    {
+      "epoch": 0.8497801484898724,
+      "grad_norm": 0.1158550267956344,
+      "learning_rate": 1.1605090078613978e-05,
+      "loss": 1.226,
+      "step": 11789
+    },
+    {
+      "epoch": 0.8498522309522093,
+      "grad_norm": 0.1140811636764033,
+      "learning_rate": 1.1594175753713232e-05,
+      "loss": 1.3005,
+      "step": 11790
+    },
+    {
+      "epoch": 0.8499243134145462,
+      "grad_norm": 0.12068480124314146,
+      "learning_rate": 1.1583266247720803e-05,
+      "loss": 1.3193,
+      "step": 11791
+    },
+    {
+      "epoch": 0.8499963958768831,
+      "grad_norm": 0.12696438856859726,
+      "learning_rate": 1.157236156123136e-05,
+      "loss": 1.3452,
+      "step": 11792
+    },
+    {
+      "epoch": 0.8500684783392201,
+      "grad_norm": 0.12498716196525629,
+      "learning_rate": 1.1561461694839303e-05,
+      "loss": 1.218,
+      "step": 11793
+    },
+    {
+      "epoch": 0.850140560801557,
+      "grad_norm": 0.11914866913815608,
+      "learning_rate": 1.1550566649138783e-05,
+      "loss": 1.2969,
+      "step": 11794
+    },
+    {
+      "epoch": 0.8502126432638939,
+      "grad_norm": 0.12214690603394544,
+      "learning_rate": 1.1539676424723679e-05,
+      "loss": 1.2919,
+      "step": 11795
+    },
+    {
+      "epoch": 0.8502847257262308,
+      "grad_norm": 0.12678712941643303,
+      "learning_rate": 1.152879102218759e-05,
+      "loss": 1.2593,
+      "step": 11796
+    },
+    {
+      "epoch": 0.8503568081885677,
+      "grad_norm": 0.12135123852828612,
+      "learning_rate": 1.1517910442123881e-05,
+      "loss": 1.3662,
+      "step": 11797
+    },
+    {
+      "epoch": 0.8504288906509047,
+      "grad_norm": 0.11044134415161219,
+      "learning_rate": 1.1507034685125628e-05,
+      "loss": 1.2945,
+      "step": 11798
+    },
+    {
+      "epoch": 0.8505009731132416,
+      "grad_norm": 0.11188094731373424,
+      "learning_rate": 1.1496163751785672e-05,
+      "loss": 1.3124,
+      "step": 11799
+    },
+    {
+      "epoch": 0.8505730555755785,
+      "grad_norm": 0.11704317157685891,
+      "learning_rate": 1.1485297642696569e-05,
+      "loss": 1.1967,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8506451380379154,
+      "grad_norm": 0.13637546271811318,
+      "learning_rate": 1.1474436358450635e-05,
+      "loss": 1.2204,
+      "step": 11801
+    },
+    {
+      "epoch": 0.8507172205002523,
+      "grad_norm": 0.1550131178983262,
+      "learning_rate": 1.1463579899639854e-05,
+      "loss": 1.2979,
+      "step": 11802
+    },
+    {
+      "epoch": 0.8507893029625893,
+      "grad_norm": 0.11518631534639313,
+      "learning_rate": 1.1452728266856061e-05,
+      "loss": 1.257,
+      "step": 11803
+    },
+    {
+      "epoch": 0.8508613854249261,
+      "grad_norm": 0.1142962737751662,
+      "learning_rate": 1.1441881460690752e-05,
+      "loss": 1.3103,
+      "step": 11804
+    },
+    {
+      "epoch": 0.850933467887263,
+      "grad_norm": 0.11959911864253166,
+      "learning_rate": 1.1431039481735174e-05,
+      "loss": 1.2691,
+      "step": 11805
+    },
+    {
+      "epoch": 0.8510055503495999,
+      "grad_norm": 0.12806826376415062,
+      "learning_rate": 1.1420202330580266e-05,
+      "loss": 1.2559,
+      "step": 11806
+    },
+    {
+      "epoch": 0.8510776328119368,
+      "grad_norm": 0.12000832881199326,
+      "learning_rate": 1.140937000781681e-05,
+      "loss": 1.2423,
+      "step": 11807
+    },
+    {
+      "epoch": 0.8511497152742737,
+      "grad_norm": 0.1187145679915118,
+      "learning_rate": 1.1398542514035248e-05,
+      "loss": 1.3564,
+      "step": 11808
+    },
+    {
+      "epoch": 0.8512217977366107,
+      "grad_norm": 0.13122151446705652,
+      "learning_rate": 1.1387719849825784e-05,
+      "loss": 1.2974,
+      "step": 11809
+    },
+    {
+      "epoch": 0.8512938801989476,
+      "grad_norm": 0.1181264124202009,
+      "learning_rate": 1.1376902015778313e-05,
+      "loss": 1.2967,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8513659626612845,
+      "grad_norm": 0.11089775876624065,
+      "learning_rate": 1.1366089012482517e-05,
+      "loss": 1.184,
+      "step": 11811
+    },
+    {
+      "epoch": 0.8514380451236214,
+      "grad_norm": 0.11902969787302975,
+      "learning_rate": 1.135528084052785e-05,
+      "loss": 1.2414,
+      "step": 11812
+    },
+    {
+      "epoch": 0.8515101275859583,
+      "grad_norm": 0.14937766287757495,
+      "learning_rate": 1.1344477500503391e-05,
+      "loss": 1.2661,
+      "step": 11813
+    },
+    {
+      "epoch": 0.8515822100482953,
+      "grad_norm": 0.11549754347796971,
+      "learning_rate": 1.1333678992998043e-05,
+      "loss": 1.3519,
+      "step": 11814
+    },
+    {
+      "epoch": 0.8516542925106322,
+      "grad_norm": 0.1108501215798427,
+      "learning_rate": 1.1322885318600407e-05,
+      "loss": 1.2477,
+      "step": 11815
+    },
+    {
+      "epoch": 0.8517263749729691,
+      "grad_norm": 0.12344563105813175,
+      "learning_rate": 1.1312096477898881e-05,
+      "loss": 1.2821,
+      "step": 11816
+    },
+    {
+      "epoch": 0.851798457435306,
+      "grad_norm": 0.12057566793499434,
+      "learning_rate": 1.1301312471481507e-05,
+      "loss": 1.3153,
+      "step": 11817
+    },
+    {
+      "epoch": 0.8518705398976429,
+      "grad_norm": 0.11177753071235359,
+      "learning_rate": 1.1290533299936123e-05,
+      "loss": 1.2255,
+      "step": 11818
+    },
+    {
+      "epoch": 0.8519426223599799,
+      "grad_norm": 0.11331060104226713,
+      "learning_rate": 1.1279758963850273e-05,
+      "loss": 1.2279,
+      "step": 11819
+    },
+    {
+      "epoch": 0.8520147048223168,
+      "grad_norm": 0.12365517480881705,
+      "learning_rate": 1.1268989463811308e-05,
+      "loss": 1.2954,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8520867872846536,
+      "grad_norm": 0.11213032893172106,
+      "learning_rate": 1.1258224800406214e-05,
+      "loss": 1.1988,
+      "step": 11821
+    },
+    {
+      "epoch": 0.8521588697469905,
+      "grad_norm": 0.12134224935296874,
+      "learning_rate": 1.1247464974221778e-05,
+      "loss": 1.2388,
+      "step": 11822
+    },
+    {
+      "epoch": 0.8522309522093274,
+      "grad_norm": 0.12258262288922629,
+      "learning_rate": 1.1236709985844497e-05,
+      "loss": 1.2527,
+      "step": 11823
+    },
+    {
+      "epoch": 0.8523030346716644,
+      "grad_norm": 0.11289233050707857,
+      "learning_rate": 1.1225959835860622e-05,
+      "loss": 1.2098,
+      "step": 11824
+    },
+    {
+      "epoch": 0.8523751171340013,
+      "grad_norm": 0.11637463390280232,
+      "learning_rate": 1.1215214524856133e-05,
+      "loss": 1.2823,
+      "step": 11825
+    },
+    {
+      "epoch": 0.8524471995963382,
+      "grad_norm": 0.1158879301413899,
+      "learning_rate": 1.1204474053416737e-05,
+      "loss": 1.2865,
+      "step": 11826
+    },
+    {
+      "epoch": 0.8525192820586751,
+      "grad_norm": 0.1161905453061602,
+      "learning_rate": 1.1193738422127897e-05,
+      "loss": 1.1996,
+      "step": 11827
+    },
+    {
+      "epoch": 0.852591364521012,
+      "grad_norm": 0.115826433519389,
+      "learning_rate": 1.1183007631574805e-05,
+      "loss": 1.3037,
+      "step": 11828
+    },
+    {
+      "epoch": 0.852663446983349,
+      "grad_norm": 0.12313688102866022,
+      "learning_rate": 1.1172281682342367e-05,
+      "loss": 1.2598,
+      "step": 11829
+    },
+    {
+      "epoch": 0.8527355294456859,
+      "grad_norm": 0.11669588991867273,
+      "learning_rate": 1.116156057501525e-05,
+      "loss": 1.2365,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8528076119080228,
+      "grad_norm": 0.11881455544807545,
+      "learning_rate": 1.1150844310177865e-05,
+      "loss": 1.2186,
+      "step": 11831
+    },
+    {
+      "epoch": 0.8528796943703597,
+      "grad_norm": 0.12296412728752908,
+      "learning_rate": 1.1140132888414323e-05,
+      "loss": 1.3685,
+      "step": 11832
+    },
+    {
+      "epoch": 0.8529517768326966,
+      "grad_norm": 0.11759804755608173,
+      "learning_rate": 1.11294263103085e-05,
+      "loss": 1.2805,
+      "step": 11833
+    },
+    {
+      "epoch": 0.8530238592950335,
+      "grad_norm": 0.21262204321544667,
+      "learning_rate": 1.1118724576444006e-05,
+      "loss": 1.1467,
+      "step": 11834
+    },
+    {
+      "epoch": 0.8530959417573705,
+      "grad_norm": 0.11925659235619367,
+      "learning_rate": 1.1108027687404188e-05,
+      "loss": 1.3086,
+      "step": 11835
+    },
+    {
+      "epoch": 0.8531680242197074,
+      "grad_norm": 0.12572191279391884,
+      "learning_rate": 1.10973356437721e-05,
+      "loss": 1.2258,
+      "step": 11836
+    },
+    {
+      "epoch": 0.8532401066820443,
+      "grad_norm": 0.12040955282398923,
+      "learning_rate": 1.1086648446130566e-05,
+      "loss": 1.2072,
+      "step": 11837
+    },
+    {
+      "epoch": 0.8533121891443811,
+      "grad_norm": 0.11402907055716328,
+      "learning_rate": 1.1075966095062129e-05,
+      "loss": 1.2395,
+      "step": 11838
+    },
+    {
+      "epoch": 0.853384271606718,
+      "grad_norm": 0.14535941075226722,
+      "learning_rate": 1.106528859114908e-05,
+      "loss": 1.2183,
+      "step": 11839
+    },
+    {
+      "epoch": 0.853456354069055,
+      "grad_norm": 0.13598909949692814,
+      "learning_rate": 1.1054615934973433e-05,
+      "loss": 1.2159,
+      "step": 11840
+    },
+    {
+      "epoch": 0.8535284365313919,
+      "grad_norm": 0.12724296379996683,
+      "learning_rate": 1.1043948127116954e-05,
+      "loss": 1.3136,
+      "step": 11841
+    },
+    {
+      "epoch": 0.8536005189937288,
+      "grad_norm": 0.12156372504434813,
+      "learning_rate": 1.1033285168161123e-05,
+      "loss": 1.3368,
+      "step": 11842
+    },
+    {
+      "epoch": 0.8536726014560657,
+      "grad_norm": 0.11928943202489915,
+      "learning_rate": 1.1022627058687163e-05,
+      "loss": 1.2768,
+      "step": 11843
+    },
+    {
+      "epoch": 0.8537446839184026,
+      "grad_norm": 0.10985632690773714,
+      "learning_rate": 1.1011973799276054e-05,
+      "loss": 1.2274,
+      "step": 11844
+    },
+    {
+      "epoch": 0.8538167663807396,
+      "grad_norm": 0.11941219261890144,
+      "learning_rate": 1.1001325390508488e-05,
+      "loss": 1.356,
+      "step": 11845
+    },
+    {
+      "epoch": 0.8538888488430765,
+      "grad_norm": 0.10954441378304802,
+      "learning_rate": 1.0990681832964866e-05,
+      "loss": 1.2619,
+      "step": 11846
+    },
+    {
+      "epoch": 0.8539609313054134,
+      "grad_norm": 0.1253063809036113,
+      "learning_rate": 1.0980043127225393e-05,
+      "loss": 1.3308,
+      "step": 11847
+    },
+    {
+      "epoch": 0.8540330137677503,
+      "grad_norm": 0.11995669288500875,
+      "learning_rate": 1.0969409273869979e-05,
+      "loss": 1.2671,
+      "step": 11848
+    },
+    {
+      "epoch": 0.8541050962300872,
+      "grad_norm": 0.12293031143384492,
+      "learning_rate": 1.0958780273478253e-05,
+      "loss": 1.2322,
+      "step": 11849
+    },
+    {
+      "epoch": 0.8541771786924242,
+      "grad_norm": 0.13061017273538753,
+      "learning_rate": 1.0948156126629561e-05,
+      "loss": 1.2865,
+      "step": 11850
+    },
+    {
+      "epoch": 0.8542492611547611,
+      "grad_norm": 0.11234138145000433,
+      "learning_rate": 1.093753683390305e-05,
+      "loss": 1.2319,
+      "step": 11851
+    },
+    {
+      "epoch": 0.854321343617098,
+      "grad_norm": 0.1271479326619616,
+      "learning_rate": 1.092692239587757e-05,
+      "loss": 1.2472,
+      "step": 11852
+    },
+    {
+      "epoch": 0.8543934260794349,
+      "grad_norm": 0.11577467276916714,
+      "learning_rate": 1.0916312813131702e-05,
+      "loss": 1.22,
+      "step": 11853
+    },
+    {
+      "epoch": 0.8544655085417718,
+      "grad_norm": 0.1213199747925956,
+      "learning_rate": 1.0905708086243727e-05,
+      "loss": 1.2509,
+      "step": 11854
+    },
+    {
+      "epoch": 0.8545375910041088,
+      "grad_norm": 0.12571513581160657,
+      "learning_rate": 1.0895108215791717e-05,
+      "loss": 1.2874,
+      "step": 11855
+    },
+    {
+      "epoch": 0.8546096734664456,
+      "grad_norm": 0.12397413133020767,
+      "learning_rate": 1.0884513202353463e-05,
+      "loss": 1.3507,
+      "step": 11856
+    },
+    {
+      "epoch": 0.8546817559287825,
+      "grad_norm": 0.11324005705852873,
+      "learning_rate": 1.087392304650653e-05,
+      "loss": 1.3213,
+      "step": 11857
+    },
+    {
+      "epoch": 0.8547538383911194,
+      "grad_norm": 0.11545899987353012,
+      "learning_rate": 1.0863337748828106e-05,
+      "loss": 1.2629,
+      "step": 11858
+    },
+    {
+      "epoch": 0.8548259208534563,
+      "grad_norm": 0.1218452348865927,
+      "learning_rate": 1.0852757309895189e-05,
+      "loss": 1.2847,
+      "step": 11859
+    },
+    {
+      "epoch": 0.8548980033157932,
+      "grad_norm": 0.12026848731475302,
+      "learning_rate": 1.0842181730284562e-05,
+      "loss": 1.1205,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8549700857781302,
+      "grad_norm": 0.13188987310864808,
+      "learning_rate": 1.0831611010572674e-05,
+      "loss": 1.3069,
+      "step": 11861
+    },
+    {
+      "epoch": 0.8550421682404671,
+      "grad_norm": 0.13574711426353006,
+      "learning_rate": 1.0821045151335685e-05,
+      "loss": 1.3416,
+      "step": 11862
+    },
+    {
+      "epoch": 0.855114250702804,
+      "grad_norm": 0.11231318656020955,
+      "learning_rate": 1.0810484153149548e-05,
+      "loss": 1.3317,
+      "step": 11863
+    },
+    {
+      "epoch": 0.8551863331651409,
+      "grad_norm": 0.14423819103878857,
+      "learning_rate": 1.0799928016589945e-05,
+      "loss": 1.2124,
+      "step": 11864
+    },
+    {
+      "epoch": 0.8552584156274778,
+      "grad_norm": 0.12694530310988475,
+      "learning_rate": 1.0789376742232293e-05,
+      "loss": 1.3037,
+      "step": 11865
+    },
+    {
+      "epoch": 0.8553304980898148,
+      "grad_norm": 0.11962850312929033,
+      "learning_rate": 1.077883033065169e-05,
+      "loss": 1.3422,
+      "step": 11866
+    },
+    {
+      "epoch": 0.8554025805521517,
+      "grad_norm": 0.12324462259419919,
+      "learning_rate": 1.0768288782423031e-05,
+      "loss": 1.3061,
+      "step": 11867
+    },
+    {
+      "epoch": 0.8554746630144886,
+      "grad_norm": 0.1257155225235635,
+      "learning_rate": 1.0757752098120899e-05,
+      "loss": 1.1919,
+      "step": 11868
+    },
+    {
+      "epoch": 0.8555467454768255,
+      "grad_norm": 0.11349525353012575,
+      "learning_rate": 1.0747220278319703e-05,
+      "loss": 1.2264,
+      "step": 11869
+    },
+    {
+      "epoch": 0.8556188279391624,
+      "grad_norm": 0.12762389154039638,
+      "learning_rate": 1.0736693323593461e-05,
+      "loss": 1.3099,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8556909104014994,
+      "grad_norm": 0.11901771658047558,
+      "learning_rate": 1.0726171234516015e-05,
+      "loss": 1.3003,
+      "step": 11871
+    },
+    {
+      "epoch": 0.8557629928638363,
+      "grad_norm": 0.1314523531524553,
+      "learning_rate": 1.071565401166087e-05,
+      "loss": 1.2188,
+      "step": 11872
+    },
+    {
+      "epoch": 0.8558350753261731,
+      "grad_norm": 0.11234016284206957,
+      "learning_rate": 1.0705141655601381e-05,
+      "loss": 1.1944,
+      "step": 11873
+    },
+    {
+      "epoch": 0.85590715778851,
+      "grad_norm": 0.10960071280568821,
+      "learning_rate": 1.0694634166910511e-05,
+      "loss": 1.2693,
+      "step": 11874
+    },
+    {
+      "epoch": 0.8559792402508469,
+      "grad_norm": 0.12667397880248227,
+      "learning_rate": 1.0684131546161035e-05,
+      "loss": 1.2982,
+      "step": 11875
+    },
+    {
+      "epoch": 0.8560513227131838,
+      "grad_norm": 0.11250220244899553,
+      "learning_rate": 1.0673633793925419e-05,
+      "loss": 1.2125,
+      "step": 11876
+    },
+    {
+      "epoch": 0.8561234051755208,
+      "grad_norm": 0.1202472237890846,
+      "learning_rate": 1.0663140910775904e-05,
+      "loss": 1.2458,
+      "step": 11877
+    },
+    {
+      "epoch": 0.8561954876378577,
+      "grad_norm": 0.10863367021194871,
+      "learning_rate": 1.0652652897284443e-05,
+      "loss": 1.1466,
+      "step": 11878
+    },
+    {
+      "epoch": 0.8562675701001946,
+      "grad_norm": 0.12121830806003589,
+      "learning_rate": 1.0642169754022724e-05,
+      "loss": 1.1369,
+      "step": 11879
+    },
+    {
+      "epoch": 0.8563396525625315,
+      "grad_norm": 0.12262913870520485,
+      "learning_rate": 1.0631691481562167e-05,
+      "loss": 1.2748,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8564117350248684,
+      "grad_norm": 0.1168335770085519,
+      "learning_rate": 1.0621218080473927e-05,
+      "loss": 1.2508,
+      "step": 11881
+    },
+    {
+      "epoch": 0.8564838174872054,
+      "grad_norm": 0.12661329752938585,
+      "learning_rate": 1.0610749551328925e-05,
+      "loss": 1.2632,
+      "step": 11882
+    },
+    {
+      "epoch": 0.8565558999495423,
+      "grad_norm": 0.12203402640075454,
+      "learning_rate": 1.060028589469776e-05,
+      "loss": 1.203,
+      "step": 11883
+    },
+    {
+      "epoch": 0.8566279824118792,
+      "grad_norm": 0.1272647882763144,
+      "learning_rate": 1.0589827111150808e-05,
+      "loss": 1.1662,
+      "step": 11884
+    },
+    {
+      "epoch": 0.8567000648742161,
+      "grad_norm": 0.1258937145963107,
+      "learning_rate": 1.057937320125817e-05,
+      "loss": 1.2944,
+      "step": 11885
+    },
+    {
+      "epoch": 0.856772147336553,
+      "grad_norm": 0.12366814947856809,
+      "learning_rate": 1.0568924165589666e-05,
+      "loss": 1.2751,
+      "step": 11886
+    },
+    {
+      "epoch": 0.85684422979889,
+      "grad_norm": 0.13272750847994072,
+      "learning_rate": 1.0558480004714866e-05,
+      "loss": 1.3451,
+      "step": 11887
+    },
+    {
+      "epoch": 0.8569163122612269,
+      "grad_norm": 0.10763953253614188,
+      "learning_rate": 1.054804071920308e-05,
+      "loss": 1.2685,
+      "step": 11888
+    },
+    {
+      "epoch": 0.8569883947235638,
+      "grad_norm": 0.1245303282928214,
+      "learning_rate": 1.0537606309623338e-05,
+      "loss": 1.1981,
+      "step": 11889
+    },
+    {
+      "epoch": 0.8570604771859006,
+      "grad_norm": 0.12255544391443092,
+      "learning_rate": 1.0527176776544389e-05,
+      "loss": 1.144,
+      "step": 11890
+    },
+    {
+      "epoch": 0.8571325596482375,
+      "grad_norm": 0.12026847533637898,
+      "learning_rate": 1.0516752120534768e-05,
+      "loss": 1.2165,
+      "step": 11891
+    },
+    {
+      "epoch": 0.8572046421105745,
+      "grad_norm": 0.11621825097774405,
+      "learning_rate": 1.0506332342162695e-05,
+      "loss": 1.324,
+      "step": 11892
+    },
+    {
+      "epoch": 0.8572767245729114,
+      "grad_norm": 0.11716026532519874,
+      "learning_rate": 1.049591744199614e-05,
+      "loss": 1.2565,
+      "step": 11893
+    },
+    {
+      "epoch": 0.8573488070352483,
+      "grad_norm": 0.12755250509005298,
+      "learning_rate": 1.0485507420602803e-05,
+      "loss": 1.2917,
+      "step": 11894
+    },
+    {
+      "epoch": 0.8574208894975852,
+      "grad_norm": 0.12329319601723274,
+      "learning_rate": 1.0475102278550142e-05,
+      "loss": 1.2417,
+      "step": 11895
+    },
+    {
+      "epoch": 0.8574929719599221,
+      "grad_norm": 0.12195388806948994,
+      "learning_rate": 1.0464702016405314e-05,
+      "loss": 1.2527,
+      "step": 11896
+    },
+    {
+      "epoch": 0.857565054422259,
+      "grad_norm": 0.11523527303980609,
+      "learning_rate": 1.0454306634735234e-05,
+      "loss": 1.2471,
+      "step": 11897
+    },
+    {
+      "epoch": 0.857637136884596,
+      "grad_norm": 0.11293020976051414,
+      "learning_rate": 1.0443916134106558e-05,
+      "loss": 1.2868,
+      "step": 11898
+    },
+    {
+      "epoch": 0.8577092193469329,
+      "grad_norm": 0.12176542136580669,
+      "learning_rate": 1.043353051508561e-05,
+      "loss": 1.2193,
+      "step": 11899
+    },
+    {
+      "epoch": 0.8577813018092698,
+      "grad_norm": 0.1252332481513393,
+      "learning_rate": 1.042314977823855e-05,
+      "loss": 1.2543,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8578533842716067,
+      "grad_norm": 0.13264396583346516,
+      "learning_rate": 1.0412773924131203e-05,
+      "loss": 1.2462,
+      "step": 11901
+    },
+    {
+      "epoch": 0.8579254667339437,
+      "grad_norm": 0.11598327101294503,
+      "learning_rate": 1.0402402953329171e-05,
+      "loss": 1.3145,
+      "step": 11902
+    },
+    {
+      "epoch": 0.8579975491962806,
+      "grad_norm": 0.12121460702082713,
+      "learning_rate": 1.0392036866397703e-05,
+      "loss": 1.2804,
+      "step": 11903
+    },
+    {
+      "epoch": 0.8580696316586175,
+      "grad_norm": 0.12093896527022466,
+      "learning_rate": 1.038167566390189e-05,
+      "loss": 1.2525,
+      "step": 11904
+    },
+    {
+      "epoch": 0.8581417141209544,
+      "grad_norm": 0.11767166865228691,
+      "learning_rate": 1.0371319346406516e-05,
+      "loss": 1.3163,
+      "step": 11905
+    },
+    {
+      "epoch": 0.8582137965832913,
+      "grad_norm": 0.11445691333691674,
+      "learning_rate": 1.0360967914476083e-05,
+      "loss": 1.2391,
+      "step": 11906
+    },
+    {
+      "epoch": 0.8582858790456283,
+      "grad_norm": 0.11776975952906628,
+      "learning_rate": 1.0350621368674807e-05,
+      "loss": 1.2667,
+      "step": 11907
+    },
+    {
+      "epoch": 0.8583579615079651,
+      "grad_norm": 0.13197818051185725,
+      "learning_rate": 1.0340279709566702e-05,
+      "loss": 1.1907,
+      "step": 11908
+    },
+    {
+      "epoch": 0.858430043970302,
+      "grad_norm": 0.13267580768209156,
+      "learning_rate": 1.0329942937715475e-05,
+      "loss": 1.2021,
+      "step": 11909
+    },
+    {
+      "epoch": 0.8585021264326389,
+      "grad_norm": 0.12327605220917029,
+      "learning_rate": 1.0319611053684596e-05,
+      "loss": 1.2756,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8585742088949758,
+      "grad_norm": 0.1534143109853025,
+      "learning_rate": 1.0309284058037194e-05,
+      "loss": 1.2124,
+      "step": 11911
+    },
+    {
+      "epoch": 0.8586462913573127,
+      "grad_norm": 0.12189464893305856,
+      "learning_rate": 1.0298961951336184e-05,
+      "loss": 1.2824,
+      "step": 11912
+    },
+    {
+      "epoch": 0.8587183738196497,
+      "grad_norm": 0.12266316146050724,
+      "learning_rate": 1.0288644734144259e-05,
+      "loss": 1.2599,
+      "step": 11913
+    },
+    {
+      "epoch": 0.8587904562819866,
+      "grad_norm": 0.13714166429389796,
+      "learning_rate": 1.0278332407023794e-05,
+      "loss": 1.2625,
+      "step": 11914
+    },
+    {
+      "epoch": 0.8588625387443235,
+      "grad_norm": 0.10948888843010159,
+      "learning_rate": 1.0268024970536871e-05,
+      "loss": 1.2478,
+      "step": 11915
+    },
+    {
+      "epoch": 0.8589346212066604,
+      "grad_norm": 0.1379478349866121,
+      "learning_rate": 1.025772242524533e-05,
+      "loss": 1.2999,
+      "step": 11916
+    },
+    {
+      "epoch": 0.8590067036689973,
+      "grad_norm": 0.1124121380753039,
+      "learning_rate": 1.024742477171079e-05,
+      "loss": 1.3191,
+      "step": 11917
+    },
+    {
+      "epoch": 0.8590787861313343,
+      "grad_norm": 0.12881578840704602,
+      "learning_rate": 1.0237132010494577e-05,
+      "loss": 1.3049,
+      "step": 11918
+    },
+    {
+      "epoch": 0.8591508685936712,
+      "grad_norm": 0.11439035091736746,
+      "learning_rate": 1.0226844142157677e-05,
+      "loss": 1.2675,
+      "step": 11919
+    },
+    {
+      "epoch": 0.8592229510560081,
+      "grad_norm": 0.11438025407891303,
+      "learning_rate": 1.021656116726092e-05,
+      "loss": 1.3454,
+      "step": 11920
+    },
+    {
+      "epoch": 0.859295033518345,
+      "grad_norm": 0.11421022153283314,
+      "learning_rate": 1.0206283086364776e-05,
+      "loss": 1.2509,
+      "step": 11921
+    },
+    {
+      "epoch": 0.8593671159806819,
+      "grad_norm": 0.12634442757321415,
+      "learning_rate": 1.0196009900029568e-05,
+      "loss": 1.346,
+      "step": 11922
+    },
+    {
+      "epoch": 0.8594391984430189,
+      "grad_norm": 0.12214349108985216,
+      "learning_rate": 1.0185741608815202e-05,
+      "loss": 1.2735,
+      "step": 11923
+    },
+    {
+      "epoch": 0.8595112809053558,
+      "grad_norm": 0.12383629235137371,
+      "learning_rate": 1.017547821328143e-05,
+      "loss": 1.2376,
+      "step": 11924
+    },
+    {
+      "epoch": 0.8595833633676926,
+      "grad_norm": 0.12729249704569726,
+      "learning_rate": 1.0165219713987662e-05,
+      "loss": 1.2762,
+      "step": 11925
+    },
+    {
+      "epoch": 0.8596554458300295,
+      "grad_norm": 0.1157925220877218,
+      "learning_rate": 1.0154966111493147e-05,
+      "loss": 1.3655,
+      "step": 11926
+    },
+    {
+      "epoch": 0.8597275282923664,
+      "grad_norm": 0.12273499998249653,
+      "learning_rate": 1.0144717406356729e-05,
+      "loss": 1.3108,
+      "step": 11927
+    },
+    {
+      "epoch": 0.8597996107547033,
+      "grad_norm": 0.11679467474603057,
+      "learning_rate": 1.0134473599137084e-05,
+      "loss": 1.2508,
+      "step": 11928
+    },
+    {
+      "epoch": 0.8598716932170403,
+      "grad_norm": 0.1304593225234668,
+      "learning_rate": 1.0124234690392586e-05,
+      "loss": 1.3488,
+      "step": 11929
+    },
+    {
+      "epoch": 0.8599437756793772,
+      "grad_norm": 0.1170010189395729,
+      "learning_rate": 1.0114000680681357e-05,
+      "loss": 1.1467,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8600158581417141,
+      "grad_norm": 0.13489910598198063,
+      "learning_rate": 1.0103771570561226e-05,
+      "loss": 1.1763,
+      "step": 11931
+    },
+    {
+      "epoch": 0.860087940604051,
+      "grad_norm": 0.12985632101541592,
+      "learning_rate": 1.0093547360589784e-05,
+      "loss": 1.3383,
+      "step": 11932
+    },
+    {
+      "epoch": 0.860160023066388,
+      "grad_norm": 0.1225482636710346,
+      "learning_rate": 1.0083328051324337e-05,
+      "loss": 1.3032,
+      "step": 11933
+    },
+    {
+      "epoch": 0.8602321055287249,
+      "grad_norm": 0.10857109618147076,
+      "learning_rate": 1.0073113643321929e-05,
+      "loss": 1.3398,
+      "step": 11934
+    },
+    {
+      "epoch": 0.8603041879910618,
+      "grad_norm": 0.11648187147707469,
+      "learning_rate": 1.006290413713933e-05,
+      "loss": 1.27,
+      "step": 11935
+    },
+    {
+      "epoch": 0.8603762704533987,
+      "grad_norm": 0.13000030124900855,
+      "learning_rate": 1.0052699533333065e-05,
+      "loss": 1.2887,
+      "step": 11936
+    },
+    {
+      "epoch": 0.8604483529157356,
+      "grad_norm": 0.12290691924349538,
+      "learning_rate": 1.0042499832459362e-05,
+      "loss": 1.2965,
+      "step": 11937
+    },
+    {
+      "epoch": 0.8605204353780725,
+      "grad_norm": 0.1262228749747375,
+      "learning_rate": 1.0032305035074197e-05,
+      "loss": 1.2951,
+      "step": 11938
+    },
+    {
+      "epoch": 0.8605925178404095,
+      "grad_norm": 0.12835216562536209,
+      "learning_rate": 1.0022115141733279e-05,
+      "loss": 1.2848,
+      "step": 11939
+    },
+    {
+      "epoch": 0.8606646003027464,
+      "grad_norm": 0.10830510804570412,
+      "learning_rate": 1.0011930152992067e-05,
+      "loss": 1.2503,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8607366827650833,
+      "grad_norm": 0.11950898114439348,
+      "learning_rate": 1.0001750069405701e-05,
+      "loss": 1.2277,
+      "step": 11941
+    },
+    {
+      "epoch": 0.8608087652274201,
+      "grad_norm": 0.10895914567876408,
+      "learning_rate": 9.991574891529109e-06,
+      "loss": 1.2606,
+      "step": 11942
+    },
+    {
+      "epoch": 0.860880847689757,
+      "grad_norm": 0.11811579238812005,
+      "learning_rate": 9.981404619916934e-06,
+      "loss": 1.3489,
+      "step": 11943
+    },
+    {
+      "epoch": 0.860952930152094,
+      "grad_norm": 0.11522823532137305,
+      "learning_rate": 9.97123925512352e-06,
+      "loss": 1.2668,
+      "step": 11944
+    },
+    {
+      "epoch": 0.8610250126144309,
+      "grad_norm": 0.11637573271074916,
+      "learning_rate": 9.961078797703005e-06,
+      "loss": 1.2141,
+      "step": 11945
+    },
+    {
+      "epoch": 0.8610970950767678,
+      "grad_norm": 0.1157361572170896,
+      "learning_rate": 9.950923248209188e-06,
+      "loss": 1.3235,
+      "step": 11946
+    },
+    {
+      "epoch": 0.8611691775391047,
+      "grad_norm": 0.11233428942021292,
+      "learning_rate": 9.94077260719567e-06,
+      "loss": 1.3478,
+      "step": 11947
+    },
+    {
+      "epoch": 0.8612412600014416,
+      "grad_norm": 0.13890337431386746,
+      "learning_rate": 9.930626875215731e-06,
+      "loss": 1.3318,
+      "step": 11948
+    },
+    {
+      "epoch": 0.8613133424637786,
+      "grad_norm": 0.1389562811242946,
+      "learning_rate": 9.920486052822409e-06,
+      "loss": 1.1745,
+      "step": 11949
+    },
+    {
+      "epoch": 0.8613854249261155,
+      "grad_norm": 0.12005937467864511,
+      "learning_rate": 9.910350140568481e-06,
+      "loss": 1.1182,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8614575073884524,
+      "grad_norm": 0.1269149166063751,
+      "learning_rate": 9.90021913900645e-06,
+      "loss": 1.3179,
+      "step": 11951
+    },
+    {
+      "epoch": 0.8615295898507893,
+      "grad_norm": 0.12521712592778828,
+      "learning_rate": 9.890093048688487e-06,
+      "loss": 1.2688,
+      "step": 11952
+    },
+    {
+      "epoch": 0.8616016723131262,
+      "grad_norm": 0.13331591579746313,
+      "learning_rate": 9.879971870166626e-06,
+      "loss": 1.2791,
+      "step": 11953
+    },
+    {
+      "epoch": 0.8616737547754632,
+      "grad_norm": 0.12479518074691158,
+      "learning_rate": 9.869855603992529e-06,
+      "loss": 1.1997,
+      "step": 11954
+    },
+    {
+      "epoch": 0.8617458372378001,
+      "grad_norm": 0.12736091069969813,
+      "learning_rate": 9.859744250717651e-06,
+      "loss": 1.3755,
+      "step": 11955
+    },
+    {
+      "epoch": 0.861817919700137,
+      "grad_norm": 0.13483746244202516,
+      "learning_rate": 9.849637810893098e-06,
+      "loss": 1.2225,
+      "step": 11956
+    },
+    {
+      "epoch": 0.8618900021624739,
+      "grad_norm": 0.12586486651391404,
+      "learning_rate": 9.839536285069805e-06,
+      "loss": 1.2948,
+      "step": 11957
+    },
+    {
+      "epoch": 0.8619620846248108,
+      "grad_norm": 0.13272433272228257,
+      "learning_rate": 9.82943967379838e-06,
+      "loss": 1.2338,
+      "step": 11958
+    },
+    {
+      "epoch": 0.8620341670871478,
+      "grad_norm": 0.11953727797646115,
+      "learning_rate": 9.819347977629202e-06,
+      "loss": 1.2714,
+      "step": 11959
+    },
+    {
+      "epoch": 0.8621062495494846,
+      "grad_norm": 0.10832882600236991,
+      "learning_rate": 9.809261197112296e-06,
+      "loss": 1.2833,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8621783320118215,
+      "grad_norm": 0.12498081592241392,
+      "learning_rate": 9.79917933279756e-06,
+      "loss": 1.2344,
+      "step": 11961
+    },
+    {
+      "epoch": 0.8622504144741584,
+      "grad_norm": 0.11856486115838374,
+      "learning_rate": 9.789102385234516e-06,
+      "loss": 1.2426,
+      "step": 11962
+    },
+    {
+      "epoch": 0.8623224969364953,
+      "grad_norm": 0.12445146281943889,
+      "learning_rate": 9.779030354972419e-06,
+      "loss": 1.2154,
+      "step": 11963
+    },
+    {
+      "epoch": 0.8623945793988322,
+      "grad_norm": 0.12100470475274808,
+      "learning_rate": 9.768963242560313e-06,
+      "loss": 1.3487,
+      "step": 11964
+    },
+    {
+      "epoch": 0.8624666618611692,
+      "grad_norm": 0.12270266947178232,
+      "learning_rate": 9.75890104854692e-06,
+      "loss": 1.269,
+      "step": 11965
+    },
+    {
+      "epoch": 0.8625387443235061,
+      "grad_norm": 0.11364866972717233,
+      "learning_rate": 9.748843773480787e-06,
+      "loss": 1.2256,
+      "step": 11966
+    },
+    {
+      "epoch": 0.862610826785843,
+      "grad_norm": 0.12780629590710943,
+      "learning_rate": 9.738791417910043e-06,
+      "loss": 1.375,
+      "step": 11967
+    },
+    {
+      "epoch": 0.8626829092481799,
+      "grad_norm": 0.11052492893102824,
+      "learning_rate": 9.728743982382681e-06,
+      "loss": 1.2786,
+      "step": 11968
+    },
+    {
+      "epoch": 0.8627549917105168,
+      "grad_norm": 0.11067179199363787,
+      "learning_rate": 9.718701467446333e-06,
+      "loss": 1.2706,
+      "step": 11969
+    },
+    {
+      "epoch": 0.8628270741728538,
+      "grad_norm": 0.12077635472035328,
+      "learning_rate": 9.708663873648482e-06,
+      "loss": 1.3058,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8628991566351907,
+      "grad_norm": 0.12197637334293628,
+      "learning_rate": 9.69863120153619e-06,
+      "loss": 1.3579,
+      "step": 11971
+    },
+    {
+      "epoch": 0.8629712390975276,
+      "grad_norm": 0.1323352317413488,
+      "learning_rate": 9.688603451656364e-06,
+      "loss": 1.3403,
+      "step": 11972
+    },
+    {
+      "epoch": 0.8630433215598645,
+      "grad_norm": 0.11773342647954801,
+      "learning_rate": 9.67858062455559e-06,
+      "loss": 1.2828,
+      "step": 11973
+    },
+    {
+      "epoch": 0.8631154040222014,
+      "grad_norm": 0.12143102420365631,
+      "learning_rate": 9.668562720780217e-06,
+      "loss": 1.2991,
+      "step": 11974
+    },
+    {
+      "epoch": 0.8631874864845384,
+      "grad_norm": 0.11377459549426126,
+      "learning_rate": 9.6585497408763e-06,
+      "loss": 1.3489,
+      "step": 11975
+    },
+    {
+      "epoch": 0.8632595689468753,
+      "grad_norm": 0.12323655701986784,
+      "learning_rate": 9.648541685389645e-06,
+      "loss": 1.3225,
+      "step": 11976
+    },
+    {
+      "epoch": 0.8633316514092121,
+      "grad_norm": 0.12640185613798996,
+      "learning_rate": 9.638538554865773e-06,
+      "loss": 1.2465,
+      "step": 11977
+    },
+    {
+      "epoch": 0.863403733871549,
+      "grad_norm": 0.1263919997694618,
+      "learning_rate": 9.628540349849946e-06,
+      "loss": 1.2271,
+      "step": 11978
+    },
+    {
+      "epoch": 0.8634758163338859,
+      "grad_norm": 0.107486658495754,
+      "learning_rate": 9.618547070887162e-06,
+      "loss": 1.2441,
+      "step": 11979
+    },
+    {
+      "epoch": 0.8635478987962228,
+      "grad_norm": 0.1229072420466097,
+      "learning_rate": 9.608558718522132e-06,
+      "loss": 1.2735,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8636199812585598,
+      "grad_norm": 0.13217258526609232,
+      "learning_rate": 9.598575293299317e-06,
+      "loss": 1.2622,
+      "step": 11981
+    },
+    {
+      "epoch": 0.8636920637208967,
+      "grad_norm": 0.11393542069456275,
+      "learning_rate": 9.588596795762906e-06,
+      "loss": 1.2546,
+      "step": 11982
+    },
+    {
+      "epoch": 0.8637641461832336,
+      "grad_norm": 0.11910253195164064,
+      "learning_rate": 9.578623226456806e-06,
+      "loss": 1.1302,
+      "step": 11983
+    },
+    {
+      "epoch": 0.8638362286455705,
+      "grad_norm": 0.11790086198700572,
+      "learning_rate": 9.568654585924685e-06,
+      "loss": 1.1914,
+      "step": 11984
+    },
+    {
+      "epoch": 0.8639083111079074,
+      "grad_norm": 0.11764429951803691,
+      "learning_rate": 9.558690874709898e-06,
+      "loss": 1.2407,
+      "step": 11985
+    },
+    {
+      "epoch": 0.8639803935702444,
+      "grad_norm": 0.11719592744087157,
+      "learning_rate": 9.548732093355572e-06,
+      "loss": 1.1774,
+      "step": 11986
+    },
+    {
+      "epoch": 0.8640524760325813,
+      "grad_norm": 0.11209092957867886,
+      "learning_rate": 9.538778242404555e-06,
+      "loss": 1.2194,
+      "step": 11987
+    },
+    {
+      "epoch": 0.8641245584949182,
+      "grad_norm": 0.12483477858418865,
+      "learning_rate": 9.528829322399414e-06,
+      "loss": 1.2467,
+      "step": 11988
+    },
+    {
+      "epoch": 0.8641966409572551,
+      "grad_norm": 0.11737506873602876,
+      "learning_rate": 9.518885333882443e-06,
+      "loss": 1.1678,
+      "step": 11989
+    },
+    {
+      "epoch": 0.864268723419592,
+      "grad_norm": 0.12415128171095602,
+      "learning_rate": 9.508946277395691e-06,
+      "loss": 1.2111,
+      "step": 11990
+    },
+    {
+      "epoch": 0.864340805881929,
+      "grad_norm": 0.12141007840890959,
+      "learning_rate": 9.499012153480925e-06,
+      "loss": 1.3002,
+      "step": 11991
+    },
+    {
+      "epoch": 0.8644128883442659,
+      "grad_norm": 0.1289213086943648,
+      "learning_rate": 9.48908296267964e-06,
+      "loss": 1.3013,
+      "step": 11992
+    },
+    {
+      "epoch": 0.8644849708066028,
+      "grad_norm": 0.12230419127110476,
+      "learning_rate": 9.479158705533075e-06,
+      "loss": 1.2031,
+      "step": 11993
+    },
+    {
+      "epoch": 0.8645570532689396,
+      "grad_norm": 0.12242829295385876,
+      "learning_rate": 9.469239382582174e-06,
+      "loss": 1.2453,
+      "step": 11994
+    },
+    {
+      "epoch": 0.8646291357312765,
+      "grad_norm": 0.12015586700883399,
+      "learning_rate": 9.459324994367657e-06,
+      "loss": 1.2389,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8647012181936135,
+      "grad_norm": 0.1191389447208521,
+      "learning_rate": 9.449415541429896e-06,
+      "loss": 1.2646,
+      "step": 11996
+    },
+    {
+      "epoch": 0.8647733006559504,
+      "grad_norm": 0.1194295468304098,
+      "learning_rate": 9.439511024309088e-06,
+      "loss": 1.3362,
+      "step": 11997
+    },
+    {
+      "epoch": 0.8648453831182873,
+      "grad_norm": 0.12131500109032145,
+      "learning_rate": 9.429611443545117e-06,
+      "loss": 1.3076,
+      "step": 11998
+    },
+    {
+      "epoch": 0.8649174655806242,
+      "grad_norm": 0.11445910769887319,
+      "learning_rate": 9.419716799677603e-06,
+      "loss": 1.2244,
+      "step": 11999
+    },
+    {
+      "epoch": 0.8649895480429611,
+      "grad_norm": 0.14187043899348803,
+      "learning_rate": 9.409827093245837e-06,
+      "loss": 1.2659,
+      "step": 12000
+    },
+    {
+      "epoch": 0.865061630505298,
+      "grad_norm": 0.17221006725947652,
+      "learning_rate": 9.399942324788957e-06,
+      "loss": 1.205,
+      "step": 12001
+    },
+    {
+      "epoch": 0.865133712967635,
+      "grad_norm": 0.12138656325904112,
+      "learning_rate": 9.390062494845753e-06,
+      "loss": 1.267,
+      "step": 12002
+    },
+    {
+      "epoch": 0.8652057954299719,
+      "grad_norm": 0.11351391111883442,
+      "learning_rate": 9.380187603954771e-06,
+      "loss": 1.2397,
+      "step": 12003
+    },
+    {
+      "epoch": 0.8652778778923088,
+      "grad_norm": 0.14736425627466282,
+      "learning_rate": 9.370317652654248e-06,
+      "loss": 1.3603,
+      "step": 12004
+    },
+    {
+      "epoch": 0.8653499603546457,
+      "grad_norm": 0.11756516382107839,
+      "learning_rate": 9.36045264148222e-06,
+      "loss": 1.3202,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8654220428169827,
+      "grad_norm": 0.11614201992187513,
+      "learning_rate": 9.350592570976414e-06,
+      "loss": 1.2572,
+      "step": 12006
+    },
+    {
+      "epoch": 0.8654941252793196,
+      "grad_norm": 0.11788574082260274,
+      "learning_rate": 9.340737441674307e-06,
+      "loss": 1.2808,
+      "step": 12007
+    },
+    {
+      "epoch": 0.8655662077416565,
+      "grad_norm": 0.11709156696007804,
+      "learning_rate": 9.33088725411304e-06,
+      "loss": 1.2521,
+      "step": 12008
+    },
+    {
+      "epoch": 0.8656382902039934,
+      "grad_norm": 0.11922757440472961,
+      "learning_rate": 9.321042008829562e-06,
+      "loss": 1.2309,
+      "step": 12009
+    },
+    {
+      "epoch": 0.8657103726663303,
+      "grad_norm": 0.12685395462969776,
+      "learning_rate": 9.311201706360539e-06,
+      "loss": 1.263,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8657824551286672,
+      "grad_norm": 0.12489820216200409,
+      "learning_rate": 9.301366347242369e-06,
+      "loss": 1.3547,
+      "step": 12011
+    },
+    {
+      "epoch": 0.8658545375910041,
+      "grad_norm": 0.11999479298624546,
+      "learning_rate": 9.291535932011131e-06,
+      "loss": 1.2643,
+      "step": 12012
+    },
+    {
+      "epoch": 0.865926620053341,
+      "grad_norm": 0.12189884361879166,
+      "learning_rate": 9.281710461202664e-06,
+      "loss": 1.2128,
+      "step": 12013
+    },
+    {
+      "epoch": 0.8659987025156779,
+      "grad_norm": 0.12127562630834514,
+      "learning_rate": 9.271889935352595e-06,
+      "loss": 1.2666,
+      "step": 12014
+    },
+    {
+      "epoch": 0.8660707849780148,
+      "grad_norm": 0.12487496587748766,
+      "learning_rate": 9.262074354996209e-06,
+      "loss": 1.2341,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8661428674403517,
+      "grad_norm": 0.12311382225129028,
+      "learning_rate": 9.25226372066852e-06,
+      "loss": 1.3166,
+      "step": 12016
+    },
+    {
+      "epoch": 0.8662149499026887,
+      "grad_norm": 0.12101210901504192,
+      "learning_rate": 9.242458032904311e-06,
+      "loss": 1.2341,
+      "step": 12017
+    },
+    {
+      "epoch": 0.8662870323650256,
+      "grad_norm": 0.11914550956284352,
+      "learning_rate": 9.232657292238055e-06,
+      "loss": 1.1629,
+      "step": 12018
+    },
+    {
+      "epoch": 0.8663591148273625,
+      "grad_norm": 0.11959553469732664,
+      "learning_rate": 9.222861499204049e-06,
+      "loss": 1.3085,
+      "step": 12019
+    },
+    {
+      "epoch": 0.8664311972896994,
+      "grad_norm": 0.12430029656208437,
+      "learning_rate": 9.213070654336187e-06,
+      "loss": 1.3027,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8665032797520363,
+      "grad_norm": 0.1117493503047396,
+      "learning_rate": 9.203284758168173e-06,
+      "loss": 1.2435,
+      "step": 12021
+    },
+    {
+      "epoch": 0.8665753622143733,
+      "grad_norm": 0.11146968312386699,
+      "learning_rate": 9.193503811233417e-06,
+      "loss": 1.2751,
+      "step": 12022
+    },
+    {
+      "epoch": 0.8666474446767102,
+      "grad_norm": 0.13613071799127294,
+      "learning_rate": 9.183727814065112e-06,
+      "loss": 1.2087,
+      "step": 12023
+    },
+    {
+      "epoch": 0.8667195271390471,
+      "grad_norm": 0.12445912485936547,
+      "learning_rate": 9.17395676719609e-06,
+      "loss": 1.2007,
+      "step": 12024
+    },
+    {
+      "epoch": 0.866791609601384,
+      "grad_norm": 0.11351447725999973,
+      "learning_rate": 9.164190671158979e-06,
+      "loss": 1.2515,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8668636920637209,
+      "grad_norm": 0.12295657698845196,
+      "learning_rate": 9.154429526486098e-06,
+      "loss": 1.2656,
+      "step": 12026
+    },
+    {
+      "epoch": 0.8669357745260579,
+      "grad_norm": 0.1294227622775341,
+      "learning_rate": 9.144673333709575e-06,
+      "loss": 1.3437,
+      "step": 12027
+    },
+    {
+      "epoch": 0.8670078569883948,
+      "grad_norm": 0.1284329900084042,
+      "learning_rate": 9.134922093361154e-06,
+      "loss": 1.2389,
+      "step": 12028
+    },
+    {
+      "epoch": 0.8670799394507316,
+      "grad_norm": 0.12709209297331633,
+      "learning_rate": 9.125175805972375e-06,
+      "loss": 1.2778,
+      "step": 12029
+    },
+    {
+      "epoch": 0.8671520219130685,
+      "grad_norm": 0.11922650816930838,
+      "learning_rate": 9.115434472074503e-06,
+      "loss": 1.264,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8672241043754054,
+      "grad_norm": 0.12343248315213289,
+      "learning_rate": 9.105698092198533e-06,
+      "loss": 1.268,
+      "step": 12031
+    },
+    {
+      "epoch": 0.8672961868377423,
+      "grad_norm": 0.12285222188359331,
+      "learning_rate": 9.095966666875178e-06,
+      "loss": 1.3171,
+      "step": 12032
+    },
+    {
+      "epoch": 0.8673682693000793,
+      "grad_norm": 0.12850791226520994,
+      "learning_rate": 9.086240196634898e-06,
+      "loss": 1.2602,
+      "step": 12033
+    },
+    {
+      "epoch": 0.8674403517624162,
+      "grad_norm": 0.11645275608657671,
+      "learning_rate": 9.07651868200785e-06,
+      "loss": 1.3132,
+      "step": 12034
+    },
+    {
+      "epoch": 0.8675124342247531,
+      "grad_norm": 0.11025351221591927,
+      "learning_rate": 9.066802123523977e-06,
+      "loss": 1.2457,
+      "step": 12035
+    },
+    {
+      "epoch": 0.86758451668709,
+      "grad_norm": 0.11243785774807911,
+      "learning_rate": 9.057090521712886e-06,
+      "loss": 1.1481,
+      "step": 12036
+    },
+    {
+      "epoch": 0.8676565991494269,
+      "grad_norm": 0.1268107790266353,
+      "learning_rate": 9.047383877103965e-06,
+      "loss": 1.2417,
+      "step": 12037
+    },
+    {
+      "epoch": 0.8677286816117639,
+      "grad_norm": 0.12089755412905798,
+      "learning_rate": 9.037682190226304e-06,
+      "loss": 1.0861,
+      "step": 12038
+    },
+    {
+      "epoch": 0.8678007640741008,
+      "grad_norm": 0.11998786100622401,
+      "learning_rate": 9.027985461608746e-06,
+      "loss": 1.2297,
+      "step": 12039
+    },
+    {
+      "epoch": 0.8678728465364377,
+      "grad_norm": 0.12361410719688022,
+      "learning_rate": 9.018293691779833e-06,
+      "loss": 1.2755,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8679449289987746,
+      "grad_norm": 0.1211781664495598,
+      "learning_rate": 9.008606881267856e-06,
+      "loss": 1.2549,
+      "step": 12041
+    },
+    {
+      "epoch": 0.8680170114611115,
+      "grad_norm": 0.13155449377622205,
+      "learning_rate": 8.998925030600835e-06,
+      "loss": 1.3014,
+      "step": 12042
+    },
+    {
+      "epoch": 0.8680890939234485,
+      "grad_norm": 0.12209619564085462,
+      "learning_rate": 8.989248140306527e-06,
+      "loss": 1.3437,
+      "step": 12043
+    },
+    {
+      "epoch": 0.8681611763857854,
+      "grad_norm": 0.12219438186183497,
+      "learning_rate": 8.979576210912399e-06,
+      "loss": 1.2536,
+      "step": 12044
+    },
+    {
+      "epoch": 0.8682332588481223,
+      "grad_norm": 0.11856688428512106,
+      "learning_rate": 8.969909242945651e-06,
+      "loss": 1.2119,
+      "step": 12045
+    },
+    {
+      "epoch": 0.8683053413104591,
+      "grad_norm": 0.11213148408724645,
+      "learning_rate": 8.96024723693324e-06,
+      "loss": 1.2824,
+      "step": 12046
+    },
+    {
+      "epoch": 0.868377423772796,
+      "grad_norm": 0.12017436433871415,
+      "learning_rate": 8.95059019340181e-06,
+      "loss": 1.2893,
+      "step": 12047
+    },
+    {
+      "epoch": 0.868449506235133,
+      "grad_norm": 0.12389294465461918,
+      "learning_rate": 8.940938112877772e-06,
+      "loss": 1.3284,
+      "step": 12048
+    },
+    {
+      "epoch": 0.8685215886974699,
+      "grad_norm": 0.12242663262881691,
+      "learning_rate": 8.931290995887242e-06,
+      "loss": 1.2186,
+      "step": 12049
+    },
+    {
+      "epoch": 0.8685936711598068,
+      "grad_norm": 0.12970235572761452,
+      "learning_rate": 8.921648842956077e-06,
+      "loss": 1.2855,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8686657536221437,
+      "grad_norm": 0.12101178164727798,
+      "learning_rate": 8.912011654609852e-06,
+      "loss": 1.215,
+      "step": 12051
+    },
+    {
+      "epoch": 0.8687378360844806,
+      "grad_norm": 0.12293240296138322,
+      "learning_rate": 8.90237943137392e-06,
+      "loss": 1.2361,
+      "step": 12052
+    },
+    {
+      "epoch": 0.8688099185468176,
+      "grad_norm": 0.12178575340789771,
+      "learning_rate": 8.892752173773245e-06,
+      "loss": 1.3128,
+      "step": 12053
+    },
+    {
+      "epoch": 0.8688820010091545,
+      "grad_norm": 0.12029610759087209,
+      "learning_rate": 8.883129882332675e-06,
+      "loss": 1.2305,
+      "step": 12054
+    },
+    {
+      "epoch": 0.8689540834714914,
+      "grad_norm": 0.1119925098179401,
+      "learning_rate": 8.873512557576668e-06,
+      "loss": 1.2916,
+      "step": 12055
+    },
+    {
+      "epoch": 0.8690261659338283,
+      "grad_norm": 0.12710840904369733,
+      "learning_rate": 8.863900200029495e-06,
+      "loss": 1.2794,
+      "step": 12056
+    },
+    {
+      "epoch": 0.8690982483961652,
+      "grad_norm": 0.12079934191642279,
+      "learning_rate": 8.854292810215048e-06,
+      "loss": 1.2236,
+      "step": 12057
+    },
+    {
+      "epoch": 0.8691703308585021,
+      "grad_norm": 0.11861876005958018,
+      "learning_rate": 8.844690388657073e-06,
+      "loss": 1.1663,
+      "step": 12058
+    },
+    {
+      "epoch": 0.8692424133208391,
+      "grad_norm": 0.12411117858433857,
+      "learning_rate": 8.835092935878974e-06,
+      "loss": 1.2346,
+      "step": 12059
+    },
+    {
+      "epoch": 0.869314495783176,
+      "grad_norm": 0.11555484918034613,
+      "learning_rate": 8.825500452403901e-06,
+      "loss": 1.3046,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8693865782455129,
+      "grad_norm": 0.13417517005441287,
+      "learning_rate": 8.815912938754722e-06,
+      "loss": 1.2944,
+      "step": 12061
+    },
+    {
+      "epoch": 0.8694586607078498,
+      "grad_norm": 0.13097678545290556,
+      "learning_rate": 8.80633039545402e-06,
+      "loss": 1.3225,
+      "step": 12062
+    },
+    {
+      "epoch": 0.8695307431701867,
+      "grad_norm": 0.13242752334324667,
+      "learning_rate": 8.796752823024168e-06,
+      "loss": 1.3613,
+      "step": 12063
+    },
+    {
+      "epoch": 0.8696028256325236,
+      "grad_norm": 0.11314662983487009,
+      "learning_rate": 8.787180221987235e-06,
+      "loss": 1.3066,
+      "step": 12064
+    },
+    {
+      "epoch": 0.8696749080948605,
+      "grad_norm": 0.12414570142661277,
+      "learning_rate": 8.77761259286497e-06,
+      "loss": 1.3654,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8697469905571974,
+      "grad_norm": 0.1289538875941749,
+      "learning_rate": 8.768049936178913e-06,
+      "loss": 1.286,
+      "step": 12066
+    },
+    {
+      "epoch": 0.8698190730195343,
+      "grad_norm": 0.1328786248607706,
+      "learning_rate": 8.758492252450324e-06,
+      "loss": 1.2493,
+      "step": 12067
+    },
+    {
+      "epoch": 0.8698911554818712,
+      "grad_norm": 0.110064413615266,
+      "learning_rate": 8.748939542200196e-06,
+      "loss": 1.3136,
+      "step": 12068
+    },
+    {
+      "epoch": 0.8699632379442082,
+      "grad_norm": 0.12071902441954116,
+      "learning_rate": 8.739391805949205e-06,
+      "loss": 1.3308,
+      "step": 12069
+    },
+    {
+      "epoch": 0.8700353204065451,
+      "grad_norm": 0.12799423584680011,
+      "learning_rate": 8.729849044217786e-06,
+      "loss": 1.2793,
+      "step": 12070
+    },
+    {
+      "epoch": 0.870107402868882,
+      "grad_norm": 0.11924238952536272,
+      "learning_rate": 8.720311257526126e-06,
+      "loss": 1.2901,
+      "step": 12071
+    },
+    {
+      "epoch": 0.8701794853312189,
+      "grad_norm": 0.12174214381332082,
+      "learning_rate": 8.710778446394142e-06,
+      "loss": 1.4075,
+      "step": 12072
+    },
+    {
+      "epoch": 0.8702515677935558,
+      "grad_norm": 0.11590326189072202,
+      "learning_rate": 8.701250611341416e-06,
+      "loss": 1.2089,
+      "step": 12073
+    },
+    {
+      "epoch": 0.8703236502558928,
+      "grad_norm": 0.13555596303419357,
+      "learning_rate": 8.691727752887303e-06,
+      "loss": 1.2672,
+      "step": 12074
+    },
+    {
+      "epoch": 0.8703957327182297,
+      "grad_norm": 0.12260366032486614,
+      "learning_rate": 8.682209871550884e-06,
+      "loss": 1.326,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8704678151805666,
+      "grad_norm": 0.12889921209163033,
+      "learning_rate": 8.672696967851013e-06,
+      "loss": 1.2636,
+      "step": 12076
+    },
+    {
+      "epoch": 0.8705398976429035,
+      "grad_norm": 0.1233259953966581,
+      "learning_rate": 8.663189042306174e-06,
+      "loss": 1.237,
+      "step": 12077
+    },
+    {
+      "epoch": 0.8706119801052404,
+      "grad_norm": 0.11945989508749827,
+      "learning_rate": 8.653686095434655e-06,
+      "loss": 1.2902,
+      "step": 12078
+    },
+    {
+      "epoch": 0.8706840625675774,
+      "grad_norm": 0.12296672829728074,
+      "learning_rate": 8.644188127754427e-06,
+      "loss": 1.3763,
+      "step": 12079
+    },
+    {
+      "epoch": 0.8707561450299143,
+      "grad_norm": 0.1219337487634209,
+      "learning_rate": 8.63469513978329e-06,
+      "loss": 1.2664,
+      "step": 12080
+    },
+    {
+      "epoch": 0.8708282274922511,
+      "grad_norm": 0.12164761585428148,
+      "learning_rate": 8.625207132038615e-06,
+      "loss": 1.3055,
+      "step": 12081
+    },
+    {
+      "epoch": 0.870900309954588,
+      "grad_norm": 0.11705910795388806,
+      "learning_rate": 8.615724105037615e-06,
+      "loss": 1.3101,
+      "step": 12082
+    },
+    {
+      "epoch": 0.8709723924169249,
+      "grad_norm": 0.12107895605927753,
+      "learning_rate": 8.606246059297207e-06,
+      "loss": 1.3095,
+      "step": 12083
+    },
+    {
+      "epoch": 0.8710444748792618,
+      "grad_norm": 0.11824592557210385,
+      "learning_rate": 8.596772995334012e-06,
+      "loss": 1.2746,
+      "step": 12084
+    },
+    {
+      "epoch": 0.8711165573415988,
+      "grad_norm": 0.11519932181265263,
+      "learning_rate": 8.587304913664407e-06,
+      "loss": 1.3308,
+      "step": 12085
+    },
+    {
+      "epoch": 0.8711886398039357,
+      "grad_norm": 0.1176316189657724,
+      "learning_rate": 8.577841814804489e-06,
+      "loss": 1.2526,
+      "step": 12086
+    },
+    {
+      "epoch": 0.8712607222662726,
+      "grad_norm": 0.12905767341783528,
+      "learning_rate": 8.568383699270077e-06,
+      "loss": 1.3667,
+      "step": 12087
+    },
+    {
+      "epoch": 0.8713328047286095,
+      "grad_norm": 0.13932828271753517,
+      "learning_rate": 8.558930567576718e-06,
+      "loss": 1.2707,
+      "step": 12088
+    },
+    {
+      "epoch": 0.8714048871909464,
+      "grad_norm": 0.12090799723115254,
+      "learning_rate": 8.549482420239707e-06,
+      "loss": 1.2642,
+      "step": 12089
+    },
+    {
+      "epoch": 0.8714769696532834,
+      "grad_norm": 0.12449323083867969,
+      "learning_rate": 8.540039257774035e-06,
+      "loss": 1.3136,
+      "step": 12090
+    },
+    {
+      "epoch": 0.8715490521156203,
+      "grad_norm": 0.11283903003243556,
+      "learning_rate": 8.530601080694456e-06,
+      "loss": 1.32,
+      "step": 12091
+    },
+    {
+      "epoch": 0.8716211345779572,
+      "grad_norm": 0.11726053206115872,
+      "learning_rate": 8.521167889515424e-06,
+      "loss": 1.2749,
+      "step": 12092
+    },
+    {
+      "epoch": 0.8716932170402941,
+      "grad_norm": 0.10843048961945143,
+      "learning_rate": 8.51173968475113e-06,
+      "loss": 1.2556,
+      "step": 12093
+    },
+    {
+      "epoch": 0.871765299502631,
+      "grad_norm": 0.13281849675055885,
+      "learning_rate": 8.502316466915505e-06,
+      "loss": 1.3475,
+      "step": 12094
+    },
+    {
+      "epoch": 0.871837381964968,
+      "grad_norm": 0.12552005690565132,
+      "learning_rate": 8.492898236522196e-06,
+      "loss": 1.1678,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8719094644273049,
+      "grad_norm": 0.11873389480766569,
+      "learning_rate": 8.483484994084578e-06,
+      "loss": 1.156,
+      "step": 12096
+    },
+    {
+      "epoch": 0.8719815468896418,
+      "grad_norm": 0.11805201074325354,
+      "learning_rate": 8.474076740115754e-06,
+      "loss": 1.2065,
+      "step": 12097
+    },
+    {
+      "epoch": 0.8720536293519786,
+      "grad_norm": 0.1190002460613009,
+      "learning_rate": 8.464673475128571e-06,
+      "loss": 1.224,
+      "step": 12098
+    },
+    {
+      "epoch": 0.8721257118143155,
+      "grad_norm": 0.12803374802416598,
+      "learning_rate": 8.455275199635571e-06,
+      "loss": 1.2488,
+      "step": 12099
+    },
+    {
+      "epoch": 0.8721977942766524,
+      "grad_norm": 0.11357771526361368,
+      "learning_rate": 8.445881914149068e-06,
+      "loss": 1.2865,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8722698767389894,
+      "grad_norm": 0.12940431496827323,
+      "learning_rate": 8.436493619181051e-06,
+      "loss": 1.3559,
+      "step": 12101
+    },
+    {
+      "epoch": 0.8723419592013263,
+      "grad_norm": 0.11397273165143729,
+      "learning_rate": 8.427110315243291e-06,
+      "loss": 1.1542,
+      "step": 12102
+    },
+    {
+      "epoch": 0.8724140416636632,
+      "grad_norm": 0.1238688983090887,
+      "learning_rate": 8.417732002847255e-06,
+      "loss": 1.3874,
+      "step": 12103
+    },
+    {
+      "epoch": 0.8724861241260001,
+      "grad_norm": 0.13396332063083444,
+      "learning_rate": 8.408358682504147e-06,
+      "loss": 1.2979,
+      "step": 12104
+    },
+    {
+      "epoch": 0.872558206588337,
+      "grad_norm": 0.13310138457731457,
+      "learning_rate": 8.398990354724901e-06,
+      "loss": 1.181,
+      "step": 12105
+    },
+    {
+      "epoch": 0.872630289050674,
+      "grad_norm": 0.12309043323466148,
+      "learning_rate": 8.38962702002013e-06,
+      "loss": 1.3316,
+      "step": 12106
+    },
+    {
+      "epoch": 0.8727023715130109,
+      "grad_norm": 0.12035196511782484,
+      "learning_rate": 8.380268678900283e-06,
+      "loss": 1.3302,
+      "step": 12107
+    },
+    {
+      "epoch": 0.8727744539753478,
+      "grad_norm": 0.11006639699128291,
+      "learning_rate": 8.37091533187544e-06,
+      "loss": 1.2128,
+      "step": 12108
+    },
+    {
+      "epoch": 0.8728465364376847,
+      "grad_norm": 0.12263409741463975,
+      "learning_rate": 8.361566979455471e-06,
+      "loss": 1.2424,
+      "step": 12109
+    },
+    {
+      "epoch": 0.8729186189000216,
+      "grad_norm": 0.12332397298932539,
+      "learning_rate": 8.35222362214989e-06,
+      "loss": 1.2387,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8729907013623586,
+      "grad_norm": 0.141540445295307,
+      "learning_rate": 8.342885260468037e-06,
+      "loss": 1.1883,
+      "step": 12111
+    },
+    {
+      "epoch": 0.8730627838246955,
+      "grad_norm": 0.11718953372820953,
+      "learning_rate": 8.333551894918924e-06,
+      "loss": 1.3254,
+      "step": 12112
+    },
+    {
+      "epoch": 0.8731348662870324,
+      "grad_norm": 0.11383522671997527,
+      "learning_rate": 8.324223526011321e-06,
+      "loss": 1.3188,
+      "step": 12113
+    },
+    {
+      "epoch": 0.8732069487493693,
+      "grad_norm": 0.11236307966745607,
+      "learning_rate": 8.314900154253669e-06,
+      "loss": 1.3369,
+      "step": 12114
+    },
+    {
+      "epoch": 0.8732790312117062,
+      "grad_norm": 0.117823050045762,
+      "learning_rate": 8.305581780154204e-06,
+      "loss": 1.2061,
+      "step": 12115
+    },
+    {
+      "epoch": 0.8733511136740431,
+      "grad_norm": 0.12385868783869527,
+      "learning_rate": 8.296268404220865e-06,
+      "loss": 1.2875,
+      "step": 12116
+    },
+    {
+      "epoch": 0.87342319613638,
+      "grad_norm": 0.12889334136422204,
+      "learning_rate": 8.286960026961299e-06,
+      "loss": 1.2758,
+      "step": 12117
+    },
+    {
+      "epoch": 0.8734952785987169,
+      "grad_norm": 0.1163290219248217,
+      "learning_rate": 8.277656648882893e-06,
+      "loss": 1.219,
+      "step": 12118
+    },
+    {
+      "epoch": 0.8735673610610538,
+      "grad_norm": 0.10938682556475506,
+      "learning_rate": 8.268358270492749e-06,
+      "loss": 1.1722,
+      "step": 12119
+    },
+    {
+      "epoch": 0.8736394435233907,
+      "grad_norm": 0.1260509366464524,
+      "learning_rate": 8.259064892297775e-06,
+      "loss": 1.2046,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8737115259857277,
+      "grad_norm": 0.11313520142130946,
+      "learning_rate": 8.249776514804485e-06,
+      "loss": 1.2856,
+      "step": 12121
+    },
+    {
+      "epoch": 0.8737836084480646,
+      "grad_norm": 0.1335968354018312,
+      "learning_rate": 8.2404931385192e-06,
+      "loss": 1.2749,
+      "step": 12122
+    },
+    {
+      "epoch": 0.8738556909104015,
+      "grad_norm": 0.1370160844744164,
+      "learning_rate": 8.231214763947914e-06,
+      "loss": 1.2619,
+      "step": 12123
+    },
+    {
+      "epoch": 0.8739277733727384,
+      "grad_norm": 0.12192770256511123,
+      "learning_rate": 8.221941391596454e-06,
+      "loss": 1.3212,
+      "step": 12124
+    },
+    {
+      "epoch": 0.8739998558350753,
+      "grad_norm": 0.11390223954505649,
+      "learning_rate": 8.21267302197024e-06,
+      "loss": 1.2343,
+      "step": 12125
+    },
+    {
+      "epoch": 0.8740719382974123,
+      "grad_norm": 0.1178753140655755,
+      "learning_rate": 8.20340965557449e-06,
+      "loss": 1.2441,
+      "step": 12126
+    },
+    {
+      "epoch": 0.8741440207597492,
+      "grad_norm": 0.12251809157401783,
+      "learning_rate": 8.194151292914166e-06,
+      "loss": 1.2688,
+      "step": 12127
+    },
+    {
+      "epoch": 0.8742161032220861,
+      "grad_norm": 0.12282575099902812,
+      "learning_rate": 8.184897934493907e-06,
+      "loss": 1.2472,
+      "step": 12128
+    },
+    {
+      "epoch": 0.874288185684423,
+      "grad_norm": 0.12646200730127935,
+      "learning_rate": 8.1756495808181e-06,
+      "loss": 1.2086,
+      "step": 12129
+    },
+    {
+      "epoch": 0.8743602681467599,
+      "grad_norm": 0.18465731747393424,
+      "learning_rate": 8.166406232390889e-06,
+      "loss": 1.237,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8744323506090969,
+      "grad_norm": 0.11836791962648757,
+      "learning_rate": 8.157167889716112e-06,
+      "loss": 1.2887,
+      "step": 12131
+    },
+    {
+      "epoch": 0.8745044330714338,
+      "grad_norm": 0.11816006581879646,
+      "learning_rate": 8.14793455329732e-06,
+      "loss": 1.3007,
+      "step": 12132
+    },
+    {
+      "epoch": 0.8745765155337706,
+      "grad_norm": 0.1204405805396172,
+      "learning_rate": 8.138706223637827e-06,
+      "loss": 1.259,
+      "step": 12133
+    },
+    {
+      "epoch": 0.8746485979961075,
+      "grad_norm": 0.11102795763726805,
+      "learning_rate": 8.12948290124067e-06,
+      "loss": 1.1991,
+      "step": 12134
+    },
+    {
+      "epoch": 0.8747206804584444,
+      "grad_norm": 0.1208623277416036,
+      "learning_rate": 8.120264586608584e-06,
+      "loss": 1.2326,
+      "step": 12135
+    },
+    {
+      "epoch": 0.8747927629207813,
+      "grad_norm": 0.1090367858216786,
+      "learning_rate": 8.111051280244053e-06,
+      "loss": 1.2571,
+      "step": 12136
+    },
+    {
+      "epoch": 0.8748648453831183,
+      "grad_norm": 0.1135333322866845,
+      "learning_rate": 8.1018429826493e-06,
+      "loss": 1.2064,
+      "step": 12137
+    },
+    {
+      "epoch": 0.8749369278454552,
+      "grad_norm": 0.11901861659833171,
+      "learning_rate": 8.092639694326242e-06,
+      "loss": 1.3898,
+      "step": 12138
+    },
+    {
+      "epoch": 0.8750090103077921,
+      "grad_norm": 0.12525039867996754,
+      "learning_rate": 8.08344141577655e-06,
+      "loss": 1.3337,
+      "step": 12139
+    },
+    {
+      "epoch": 0.875081092770129,
+      "grad_norm": 0.12217253032898495,
+      "learning_rate": 8.074248147501617e-06,
+      "loss": 1.3036,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8751531752324659,
+      "grad_norm": 0.11379923144409933,
+      "learning_rate": 8.065059890002547e-06,
+      "loss": 1.2061,
+      "step": 12141
+    },
+    {
+      "epoch": 0.8752252576948029,
+      "grad_norm": 0.1207266135172142,
+      "learning_rate": 8.05587664378018e-06,
+      "loss": 1.2523,
+      "step": 12142
+    },
+    {
+      "epoch": 0.8752973401571398,
+      "grad_norm": 0.12534508442788792,
+      "learning_rate": 8.046698409335096e-06,
+      "loss": 1.2871,
+      "step": 12143
+    },
+    {
+      "epoch": 0.8753694226194767,
+      "grad_norm": 0.1389971589267299,
+      "learning_rate": 8.037525187167583e-06,
+      "loss": 1.3074,
+      "step": 12144
+    },
+    {
+      "epoch": 0.8754415050818136,
+      "grad_norm": 0.11650043707549237,
+      "learning_rate": 8.028356977777674e-06,
+      "loss": 1.2903,
+      "step": 12145
+    },
+    {
+      "epoch": 0.8755135875441505,
+      "grad_norm": 0.13430388318915606,
+      "learning_rate": 8.019193781665103e-06,
+      "loss": 1.3278,
+      "step": 12146
+    },
+    {
+      "epoch": 0.8755856700064875,
+      "grad_norm": 0.13222567573827024,
+      "learning_rate": 8.010035599329358e-06,
+      "loss": 1.2084,
+      "step": 12147
+    },
+    {
+      "epoch": 0.8756577524688244,
+      "grad_norm": 0.1221653388150714,
+      "learning_rate": 8.000882431269641e-06,
+      "loss": 1.2358,
+      "step": 12148
+    },
+    {
+      "epoch": 0.8757298349311613,
+      "grad_norm": 0.1303384341748396,
+      "learning_rate": 7.991734277984897e-06,
+      "loss": 1.3866,
+      "step": 12149
+    },
+    {
+      "epoch": 0.8758019173934981,
+      "grad_norm": 0.11757907325061916,
+      "learning_rate": 7.982591139973727e-06,
+      "loss": 1.2609,
+      "step": 12150
+    },
+    {
+      "epoch": 0.875873999855835,
+      "grad_norm": 0.1267825346777156,
+      "learning_rate": 7.973453017734567e-06,
+      "loss": 1.3142,
+      "step": 12151
+    },
+    {
+      "epoch": 0.875946082318172,
+      "grad_norm": 0.11409451701870676,
+      "learning_rate": 7.964319911765517e-06,
+      "loss": 1.3078,
+      "step": 12152
+    },
+    {
+      "epoch": 0.8760181647805089,
+      "grad_norm": 0.11585715100111707,
+      "learning_rate": 7.955191822564412e-06,
+      "loss": 1.18,
+      "step": 12153
+    },
+    {
+      "epoch": 0.8760902472428458,
+      "grad_norm": 0.10817909697982574,
+      "learning_rate": 7.946068750628788e-06,
+      "loss": 1.2435,
+      "step": 12154
+    },
+    {
+      "epoch": 0.8761623297051827,
+      "grad_norm": 0.2672370026771889,
+      "learning_rate": 7.93695069645597e-06,
+      "loss": 1.2866,
+      "step": 12155
+    },
+    {
+      "epoch": 0.8762344121675196,
+      "grad_norm": 0.1291161851964142,
+      "learning_rate": 7.927837660542947e-06,
+      "loss": 1.3312,
+      "step": 12156
+    },
+    {
+      "epoch": 0.8763064946298565,
+      "grad_norm": 0.12864370225161664,
+      "learning_rate": 7.918729643386503e-06,
+      "loss": 1.2816,
+      "step": 12157
+    },
+    {
+      "epoch": 0.8763785770921935,
+      "grad_norm": 0.12028072947844051,
+      "learning_rate": 7.909626645483037e-06,
+      "loss": 1.3208,
+      "step": 12158
+    },
+    {
+      "epoch": 0.8764506595545304,
+      "grad_norm": 0.13327848270383388,
+      "learning_rate": 7.900528667328799e-06,
+      "loss": 1.2512,
+      "step": 12159
+    },
+    {
+      "epoch": 0.8765227420168673,
+      "grad_norm": 0.13633781454177227,
+      "learning_rate": 7.891435709419691e-06,
+      "loss": 1.3038,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8765948244792042,
+      "grad_norm": 0.13740963823992405,
+      "learning_rate": 7.882347772251386e-06,
+      "loss": 1.268,
+      "step": 12161
+    },
+    {
+      "epoch": 0.8766669069415411,
+      "grad_norm": 0.12851448213144415,
+      "learning_rate": 7.873264856319218e-06,
+      "loss": 1.2655,
+      "step": 12162
+    },
+    {
+      "epoch": 0.8767389894038781,
+      "grad_norm": 0.12392672827076437,
+      "learning_rate": 7.86418696211828e-06,
+      "loss": 1.2554,
+      "step": 12163
+    },
+    {
+      "epoch": 0.876811071866215,
+      "grad_norm": 0.11503472192341271,
+      "learning_rate": 7.855114090143446e-06,
+      "loss": 1.3534,
+      "step": 12164
+    },
+    {
+      "epoch": 0.8768831543285519,
+      "grad_norm": 0.12419009428510863,
+      "learning_rate": 7.84604624088927e-06,
+      "loss": 1.2001,
+      "step": 12165
+    },
+    {
+      "epoch": 0.8769552367908888,
+      "grad_norm": 0.11334989765761144,
+      "learning_rate": 7.836983414849975e-06,
+      "loss": 1.1924,
+      "step": 12166
+    },
+    {
+      "epoch": 0.8770273192532257,
+      "grad_norm": 0.1153097492493673,
+      "learning_rate": 7.827925612519594e-06,
+      "loss": 1.2276,
+      "step": 12167
+    },
+    {
+      "epoch": 0.8770994017155626,
+      "grad_norm": 0.113425986135614,
+      "learning_rate": 7.818872834391877e-06,
+      "loss": 1.2108,
+      "step": 12168
+    },
+    {
+      "epoch": 0.8771714841778995,
+      "grad_norm": 0.1162557029580304,
+      "learning_rate": 7.809825080960276e-06,
+      "loss": 1.3614,
+      "step": 12169
+    },
+    {
+      "epoch": 0.8772435666402364,
+      "grad_norm": 0.11754212743953928,
+      "learning_rate": 7.80078235271795e-06,
+      "loss": 1.2636,
+      "step": 12170
+    },
+    {
+      "epoch": 0.8773156491025733,
+      "grad_norm": 0.11096261867328074,
+      "learning_rate": 7.791744650157829e-06,
+      "loss": 1.2851,
+      "step": 12171
+    },
+    {
+      "epoch": 0.8773877315649102,
+      "grad_norm": 0.11861650360871219,
+      "learning_rate": 7.782711973772516e-06,
+      "loss": 1.2268,
+      "step": 12172
+    },
+    {
+      "epoch": 0.8774598140272472,
+      "grad_norm": 0.12357103534513035,
+      "learning_rate": 7.773684324054442e-06,
+      "loss": 1.3441,
+      "step": 12173
+    },
+    {
+      "epoch": 0.8775318964895841,
+      "grad_norm": 0.13066311913807377,
+      "learning_rate": 7.764661701495635e-06,
+      "loss": 1.2053,
+      "step": 12174
+    },
+    {
+      "epoch": 0.877603978951921,
+      "grad_norm": 0.1212077891916301,
+      "learning_rate": 7.755644106587934e-06,
+      "loss": 1.3088,
+      "step": 12175
+    },
+    {
+      "epoch": 0.8776760614142579,
+      "grad_norm": 0.107615606956674,
+      "learning_rate": 7.746631539822835e-06,
+      "loss": 1.2381,
+      "step": 12176
+    },
+    {
+      "epoch": 0.8777481438765948,
+      "grad_norm": 0.13023241727940502,
+      "learning_rate": 7.73762400169169e-06,
+      "loss": 1.2911,
+      "step": 12177
+    },
+    {
+      "epoch": 0.8778202263389318,
+      "grad_norm": 0.10912087020633549,
+      "learning_rate": 7.728621492685418e-06,
+      "loss": 1.272,
+      "step": 12178
+    },
+    {
+      "epoch": 0.8778923088012687,
+      "grad_norm": 0.12266121190592584,
+      "learning_rate": 7.719624013294757e-06,
+      "loss": 1.2605,
+      "step": 12179
+    },
+    {
+      "epoch": 0.8779643912636056,
+      "grad_norm": 0.11241335976187637,
+      "learning_rate": 7.710631564010129e-06,
+      "loss": 1.3172,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8780364737259425,
+      "grad_norm": 0.11444256352049759,
+      "learning_rate": 7.701644145321763e-06,
+      "loss": 1.2943,
+      "step": 12181
+    },
+    {
+      "epoch": 0.8781085561882794,
+      "grad_norm": 0.12723716864735637,
+      "learning_rate": 7.6926617577195e-06,
+      "loss": 1.2456,
+      "step": 12182
+    },
+    {
+      "epoch": 0.8781806386506164,
+      "grad_norm": 0.11859028516877311,
+      "learning_rate": 7.68368440169297e-06,
+      "loss": 1.3037,
+      "step": 12183
+    },
+    {
+      "epoch": 0.8782527211129533,
+      "grad_norm": 0.11773943375294785,
+      "learning_rate": 7.674712077731538e-06,
+      "loss": 1.2498,
+      "step": 12184
+    },
+    {
+      "epoch": 0.8783248035752901,
+      "grad_norm": 0.12798179369905485,
+      "learning_rate": 7.665744786324258e-06,
+      "loss": 1.3121,
+      "step": 12185
+    },
+    {
+      "epoch": 0.878396886037627,
+      "grad_norm": 0.12724374663503973,
+      "learning_rate": 7.656782527959927e-06,
+      "loss": 1.3694,
+      "step": 12186
+    },
+    {
+      "epoch": 0.8784689684999639,
+      "grad_norm": 0.12630357220730554,
+      "learning_rate": 7.647825303127076e-06,
+      "loss": 1.2715,
+      "step": 12187
+    },
+    {
+      "epoch": 0.8785410509623008,
+      "grad_norm": 0.11700703957888875,
+      "learning_rate": 7.63887311231396e-06,
+      "loss": 1.2276,
+      "step": 12188
+    },
+    {
+      "epoch": 0.8786131334246378,
+      "grad_norm": 0.13010303782161708,
+      "learning_rate": 7.629925956008533e-06,
+      "loss": 1.2978,
+      "step": 12189
+    },
+    {
+      "epoch": 0.8786852158869747,
+      "grad_norm": 0.12572037717074863,
+      "learning_rate": 7.620983834698514e-06,
+      "loss": 1.2674,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8787572983493116,
+      "grad_norm": 0.11603773594528495,
+      "learning_rate": 7.612046748871327e-06,
+      "loss": 1.2856,
+      "step": 12191
+    },
+    {
+      "epoch": 0.8788293808116485,
+      "grad_norm": 0.1120288173108197,
+      "learning_rate": 7.6031146990141134e-06,
+      "loss": 1.2654,
+      "step": 12192
+    },
+    {
+      "epoch": 0.8789014632739854,
+      "grad_norm": 0.11236451379763947,
+      "learning_rate": 7.594187685613763e-06,
+      "loss": 1.2803,
+      "step": 12193
+    },
+    {
+      "epoch": 0.8789735457363224,
+      "grad_norm": 0.1198521844317497,
+      "learning_rate": 7.585265709156864e-06,
+      "loss": 1.336,
+      "step": 12194
+    },
+    {
+      "epoch": 0.8790456281986593,
+      "grad_norm": 0.12297758204960638,
+      "learning_rate": 7.576348770129749e-06,
+      "loss": 1.2342,
+      "step": 12195
+    },
+    {
+      "epoch": 0.8791177106609962,
+      "grad_norm": 0.12177411916026792,
+      "learning_rate": 7.567436869018462e-06,
+      "loss": 1.2653,
+      "step": 12196
+    },
+    {
+      "epoch": 0.8791897931233331,
+      "grad_norm": 0.12252287707867979,
+      "learning_rate": 7.558530006308806e-06,
+      "loss": 1.2077,
+      "step": 12197
+    },
+    {
+      "epoch": 0.87926187558567,
+      "grad_norm": 0.12544595415199766,
+      "learning_rate": 7.549628182486257e-06,
+      "loss": 1.3003,
+      "step": 12198
+    },
+    {
+      "epoch": 0.879333958048007,
+      "grad_norm": 0.1262976824979084,
+      "learning_rate": 7.540731398036072e-06,
+      "loss": 1.2984,
+      "step": 12199
+    },
+    {
+      "epoch": 0.8794060405103439,
+      "grad_norm": 0.12609886009536378,
+      "learning_rate": 7.531839653443174e-06,
+      "loss": 1.2891,
+      "step": 12200
+    },
+    {
+      "epoch": 0.8794781229726808,
+      "grad_norm": 0.11980864250191117,
+      "learning_rate": 7.522952949192275e-06,
+      "loss": 1.2111,
+      "step": 12201
+    },
+    {
+      "epoch": 0.8795502054350176,
+      "grad_norm": 0.11347295231949707,
+      "learning_rate": 7.514071285767766e-06,
+      "loss": 1.2011,
+      "step": 12202
+    },
+    {
+      "epoch": 0.8796222878973545,
+      "grad_norm": 0.10419437974629982,
+      "learning_rate": 7.505194663653747e-06,
+      "loss": 1.2038,
+      "step": 12203
+    },
+    {
+      "epoch": 0.8796943703596914,
+      "grad_norm": 0.1205461421474147,
+      "learning_rate": 7.496323083334123e-06,
+      "loss": 1.2675,
+      "step": 12204
+    },
+    {
+      "epoch": 0.8797664528220284,
+      "grad_norm": 0.1217369355521473,
+      "learning_rate": 7.487456545292437e-06,
+      "loss": 1.2826,
+      "step": 12205
+    },
+    {
+      "epoch": 0.8798385352843653,
+      "grad_norm": 0.1166357689752345,
+      "learning_rate": 7.478595050012038e-06,
+      "loss": 1.1696,
+      "step": 12206
+    },
+    {
+      "epoch": 0.8799106177467022,
+      "grad_norm": 0.11731955826349778,
+      "learning_rate": 7.4697385979758946e-06,
+      "loss": 1.336,
+      "step": 12207
+    },
+    {
+      "epoch": 0.8799827002090391,
+      "grad_norm": 0.10838536678602928,
+      "learning_rate": 7.460887189666821e-06,
+      "loss": 1.2451,
+      "step": 12208
+    },
+    {
+      "epoch": 0.880054782671376,
+      "grad_norm": 0.11817370260592656,
+      "learning_rate": 7.452040825567263e-06,
+      "loss": 1.2277,
+      "step": 12209
+    },
+    {
+      "epoch": 0.880126865133713,
+      "grad_norm": 0.12296389378023449,
+      "learning_rate": 7.44319950615946e-06,
+      "loss": 1.3796,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8801989475960499,
+      "grad_norm": 0.12793849548926609,
+      "learning_rate": 7.434363231925279e-06,
+      "loss": 1.2669,
+      "step": 12211
+    },
+    {
+      "epoch": 0.8802710300583868,
+      "grad_norm": 0.11821637034168521,
+      "learning_rate": 7.425532003346447e-06,
+      "loss": 1.3289,
+      "step": 12212
+    },
+    {
+      "epoch": 0.8803431125207237,
+      "grad_norm": 0.12203053417485994,
+      "learning_rate": 7.416705820904313e-06,
+      "loss": 1.2775,
+      "step": 12213
+    },
+    {
+      "epoch": 0.8804151949830606,
+      "grad_norm": 0.12204275408925655,
+      "learning_rate": 7.407884685079991e-06,
+      "loss": 1.3057,
+      "step": 12214
+    },
+    {
+      "epoch": 0.8804872774453976,
+      "grad_norm": 0.11392174586921114,
+      "learning_rate": 7.399068596354297e-06,
+      "loss": 1.3007,
+      "step": 12215
+    },
+    {
+      "epoch": 0.8805593599077345,
+      "grad_norm": 0.1293191391513906,
+      "learning_rate": 7.39025755520778e-06,
+      "loss": 1.2556,
+      "step": 12216
+    },
+    {
+      "epoch": 0.8806314423700714,
+      "grad_norm": 0.1156869812423095,
+      "learning_rate": 7.381451562120756e-06,
+      "loss": 1.1895,
+      "step": 12217
+    },
+    {
+      "epoch": 0.8807035248324083,
+      "grad_norm": 0.11888870595852508,
+      "learning_rate": 7.372650617573229e-06,
+      "loss": 1.29,
+      "step": 12218
+    },
+    {
+      "epoch": 0.8807756072947451,
+      "grad_norm": 0.12019367305204698,
+      "learning_rate": 7.363854722044883e-06,
+      "loss": 1.287,
+      "step": 12219
+    },
+    {
+      "epoch": 0.880847689757082,
+      "grad_norm": 0.1112169350542919,
+      "learning_rate": 7.3550638760152e-06,
+      "loss": 1.1882,
+      "step": 12220
+    },
+    {
+      "epoch": 0.880919772219419,
+      "grad_norm": 0.11511233048867142,
+      "learning_rate": 7.3462780799633645e-06,
+      "loss": 1.2598,
+      "step": 12221
+    },
+    {
+      "epoch": 0.8809918546817559,
+      "grad_norm": 0.13867580261457,
+      "learning_rate": 7.337497334368304e-06,
+      "loss": 1.3239,
+      "step": 12222
+    },
+    {
+      "epoch": 0.8810639371440928,
+      "grad_norm": 0.11648454642975865,
+      "learning_rate": 7.328721639708602e-06,
+      "loss": 1.2756,
+      "step": 12223
+    },
+    {
+      "epoch": 0.8811360196064297,
+      "grad_norm": 0.12369146317381595,
+      "learning_rate": 7.319950996462621e-06,
+      "loss": 1.1403,
+      "step": 12224
+    },
+    {
+      "epoch": 0.8812081020687667,
+      "grad_norm": 0.10942419234640495,
+      "learning_rate": 7.311185405108445e-06,
+      "loss": 1.2687,
+      "step": 12225
+    },
+    {
+      "epoch": 0.8812801845311036,
+      "grad_norm": 0.12742816268856064,
+      "learning_rate": 7.302424866123913e-06,
+      "loss": 1.1812,
+      "step": 12226
+    },
+    {
+      "epoch": 0.8813522669934405,
+      "grad_norm": 0.1178284641603396,
+      "learning_rate": 7.2936693799865004e-06,
+      "loss": 1.216,
+      "step": 12227
+    },
+    {
+      "epoch": 0.8814243494557774,
+      "grad_norm": 0.12026735781784549,
+      "learning_rate": 7.2849189471734915e-06,
+      "loss": 1.3341,
+      "step": 12228
+    },
+    {
+      "epoch": 0.8814964319181143,
+      "grad_norm": 0.12295026335695107,
+      "learning_rate": 7.276173568161826e-06,
+      "loss": 1.2854,
+      "step": 12229
+    },
+    {
+      "epoch": 0.8815685143804513,
+      "grad_norm": 0.12398285276059827,
+      "learning_rate": 7.267433243428279e-06,
+      "loss": 1.3716,
+      "step": 12230
+    },
+    {
+      "epoch": 0.8816405968427882,
+      "grad_norm": 0.12633697182663498,
+      "learning_rate": 7.2586979734492025e-06,
+      "loss": 1.2535,
+      "step": 12231
+    },
+    {
+      "epoch": 0.8817126793051251,
+      "grad_norm": 0.13978430986791665,
+      "learning_rate": 7.249967758700771e-06,
+      "loss": 1.3034,
+      "step": 12232
+    },
+    {
+      "epoch": 0.881784761767462,
+      "grad_norm": 0.12624643649389566,
+      "learning_rate": 7.241242599658849e-06,
+      "loss": 1.1972,
+      "step": 12233
+    },
+    {
+      "epoch": 0.8818568442297989,
+      "grad_norm": 0.13184010017542858,
+      "learning_rate": 7.2325224967990876e-06,
+      "loss": 1.2713,
+      "step": 12234
+    },
+    {
+      "epoch": 0.8819289266921359,
+      "grad_norm": 0.1307893740250687,
+      "learning_rate": 7.223807450596742e-06,
+      "loss": 1.2181,
+      "step": 12235
+    },
+    {
+      "epoch": 0.8820010091544728,
+      "grad_norm": 0.12923202018460436,
+      "learning_rate": 7.215097461526898e-06,
+      "loss": 1.2997,
+      "step": 12236
+    },
+    {
+      "epoch": 0.8820730916168096,
+      "grad_norm": 0.11309836847908615,
+      "learning_rate": 7.206392530064321e-06,
+      "loss": 1.2061,
+      "step": 12237
+    },
+    {
+      "epoch": 0.8821451740791465,
+      "grad_norm": 0.12424384251383062,
+      "learning_rate": 7.1976926566835075e-06,
+      "loss": 1.281,
+      "step": 12238
+    },
+    {
+      "epoch": 0.8822172565414834,
+      "grad_norm": 0.12489123859258164,
+      "learning_rate": 7.18899784185868e-06,
+      "loss": 1.3113,
+      "step": 12239
+    },
+    {
+      "epoch": 0.8822893390038203,
+      "grad_norm": 0.13066425287191255,
+      "learning_rate": 7.180308086063769e-06,
+      "loss": 1.3052,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8823614214661573,
+      "grad_norm": 0.123264260063967,
+      "learning_rate": 7.171623389772464e-06,
+      "loss": 1.2867,
+      "step": 12241
+    },
+    {
+      "epoch": 0.8824335039284942,
+      "grad_norm": 0.12030646225602114,
+      "learning_rate": 7.162943753458151e-06,
+      "loss": 1.3465,
+      "step": 12242
+    },
+    {
+      "epoch": 0.8825055863908311,
+      "grad_norm": 0.2083654048227853,
+      "learning_rate": 7.154269177593953e-06,
+      "loss": 1.1965,
+      "step": 12243
+    },
+    {
+      "epoch": 0.882577668853168,
+      "grad_norm": 0.1390768204610105,
+      "learning_rate": 7.145599662652691e-06,
+      "loss": 1.232,
+      "step": 12244
+    },
+    {
+      "epoch": 0.8826497513155049,
+      "grad_norm": 0.11600046091998084,
+      "learning_rate": 7.136935209106954e-06,
+      "loss": 1.2508,
+      "step": 12245
+    },
+    {
+      "epoch": 0.8827218337778419,
+      "grad_norm": 0.11901042555978734,
+      "learning_rate": 7.128275817429031e-06,
+      "loss": 1.2491,
+      "step": 12246
+    },
+    {
+      "epoch": 0.8827939162401788,
+      "grad_norm": 0.11668063288121484,
+      "learning_rate": 7.119621488090922e-06,
+      "loss": 1.3047,
+      "step": 12247
+    },
+    {
+      "epoch": 0.8828659987025157,
+      "grad_norm": 0.11365101839309777,
+      "learning_rate": 7.110972221564382e-06,
+      "loss": 1.2313,
+      "step": 12248
+    },
+    {
+      "epoch": 0.8829380811648526,
+      "grad_norm": 0.15337713079985238,
+      "learning_rate": 7.102328018320858e-06,
+      "loss": 1.3364,
+      "step": 12249
+    },
+    {
+      "epoch": 0.8830101636271895,
+      "grad_norm": 0.12064534323507208,
+      "learning_rate": 7.09368887883155e-06,
+      "loss": 1.3178,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8830822460895265,
+      "grad_norm": 0.12566142169757785,
+      "learning_rate": 7.085054803567348e-06,
+      "loss": 1.322,
+      "step": 12251
+    },
+    {
+      "epoch": 0.8831543285518634,
+      "grad_norm": 0.12148043330110518,
+      "learning_rate": 7.076425792998909e-06,
+      "loss": 1.2639,
+      "step": 12252
+    },
+    {
+      "epoch": 0.8832264110142003,
+      "grad_norm": 0.11363216097520765,
+      "learning_rate": 7.067801847596589e-06,
+      "loss": 1.3286,
+      "step": 12253
+    },
+    {
+      "epoch": 0.8832984934765371,
+      "grad_norm": 0.12268084501598674,
+      "learning_rate": 7.059182967830458e-06,
+      "loss": 1.2183,
+      "step": 12254
+    },
+    {
+      "epoch": 0.883370575938874,
+      "grad_norm": 0.31279062428960397,
+      "learning_rate": 7.050569154170317e-06,
+      "loss": 1.227,
+      "step": 12255
+    },
+    {
+      "epoch": 0.883442658401211,
+      "grad_norm": 0.11690564275431901,
+      "learning_rate": 7.041960407085724e-06,
+      "loss": 1.3142,
+      "step": 12256
+    },
+    {
+      "epoch": 0.8835147408635479,
+      "grad_norm": 0.11018067922147107,
+      "learning_rate": 7.033356727045903e-06,
+      "loss": 1.3154,
+      "step": 12257
+    },
+    {
+      "epoch": 0.8835868233258848,
+      "grad_norm": 0.12284120021773708,
+      "learning_rate": 7.024758114519847e-06,
+      "loss": 1.311,
+      "step": 12258
+    },
+    {
+      "epoch": 0.8836589057882217,
+      "grad_norm": 0.11684119534583474,
+      "learning_rate": 7.016164569976269e-06,
+      "loss": 1.2239,
+      "step": 12259
+    },
+    {
+      "epoch": 0.8837309882505586,
+      "grad_norm": 0.12088448625814735,
+      "learning_rate": 7.00757609388355e-06,
+      "loss": 1.2602,
+      "step": 12260
+    },
+    {
+      "epoch": 0.8838030707128955,
+      "grad_norm": 0.1209784415616847,
+      "learning_rate": 6.998992686709882e-06,
+      "loss": 1.2078,
+      "step": 12261
+    },
+    {
+      "epoch": 0.8838751531752325,
+      "grad_norm": 0.11897370158876049,
+      "learning_rate": 6.9904143489231244e-06,
+      "loss": 1.2217,
+      "step": 12262
+    },
+    {
+      "epoch": 0.8839472356375694,
+      "grad_norm": 0.12213355208315954,
+      "learning_rate": 6.981841080990892e-06,
+      "loss": 1.2369,
+      "step": 12263
+    },
+    {
+      "epoch": 0.8840193180999063,
+      "grad_norm": 0.12975354851558835,
+      "learning_rate": 6.973272883380444e-06,
+      "loss": 1.1889,
+      "step": 12264
+    },
+    {
+      "epoch": 0.8840914005622432,
+      "grad_norm": 0.11951473603436746,
+      "learning_rate": 6.964709756558885e-06,
+      "loss": 1.1837,
+      "step": 12265
+    },
+    {
+      "epoch": 0.8841634830245801,
+      "grad_norm": 0.12818727389726509,
+      "learning_rate": 6.956151700992963e-06,
+      "loss": 1.3874,
+      "step": 12266
+    },
+    {
+      "epoch": 0.8842355654869171,
+      "grad_norm": 0.12613685641377162,
+      "learning_rate": 6.947598717149184e-06,
+      "loss": 1.264,
+      "step": 12267
+    },
+    {
+      "epoch": 0.884307647949254,
+      "grad_norm": 0.11698504365150103,
+      "learning_rate": 6.93905080549373e-06,
+      "loss": 1.2885,
+      "step": 12268
+    },
+    {
+      "epoch": 0.8843797304115909,
+      "grad_norm": 0.12141664527867829,
+      "learning_rate": 6.930507966492539e-06,
+      "loss": 1.2799,
+      "step": 12269
+    },
+    {
+      "epoch": 0.8844518128739278,
+      "grad_norm": 0.12874814355254116,
+      "learning_rate": 6.921970200611317e-06,
+      "loss": 1.3084,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8845238953362646,
+      "grad_norm": 0.1291037756341175,
+      "learning_rate": 6.913437508315401e-06,
+      "loss": 1.2267,
+      "step": 12271
+    },
+    {
+      "epoch": 0.8845959777986016,
+      "grad_norm": 0.11827678680440058,
+      "learning_rate": 6.9049098900699215e-06,
+      "loss": 1.3065,
+      "step": 12272
+    },
+    {
+      "epoch": 0.8846680602609385,
+      "grad_norm": 0.11347267650659697,
+      "learning_rate": 6.896387346339683e-06,
+      "loss": 1.2133,
+      "step": 12273
+    },
+    {
+      "epoch": 0.8847401427232754,
+      "grad_norm": 0.13165659585451683,
+      "learning_rate": 6.887869877589292e-06,
+      "loss": 1.43,
+      "step": 12274
+    },
+    {
+      "epoch": 0.8848122251856123,
+      "grad_norm": 0.13499799933510256,
+      "learning_rate": 6.879357484282989e-06,
+      "loss": 1.2446,
+      "step": 12275
+    },
+    {
+      "epoch": 0.8848843076479492,
+      "grad_norm": 0.11510586852996277,
+      "learning_rate": 6.87085016688479e-06,
+      "loss": 1.1919,
+      "step": 12276
+    },
+    {
+      "epoch": 0.8849563901102862,
+      "grad_norm": 0.1286295141973087,
+      "learning_rate": 6.862347925858392e-06,
+      "loss": 1.2133,
+      "step": 12277
+    },
+    {
+      "epoch": 0.8850284725726231,
+      "grad_norm": 0.11900479586061054,
+      "learning_rate": 6.853850761667291e-06,
+      "loss": 1.2899,
+      "step": 12278
+    },
+    {
+      "epoch": 0.88510055503496,
+      "grad_norm": 0.12348660739942816,
+      "learning_rate": 6.845358674774626e-06,
+      "loss": 1.3286,
+      "step": 12279
+    },
+    {
+      "epoch": 0.8851726374972969,
+      "grad_norm": 0.13934341872439446,
+      "learning_rate": 6.836871665643297e-06,
+      "loss": 1.2818,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8852447199596338,
+      "grad_norm": 0.12373511732775452,
+      "learning_rate": 6.828389734735918e-06,
+      "loss": 1.1917,
+      "step": 12281
+    },
+    {
+      "epoch": 0.8853168024219707,
+      "grad_norm": 0.13430686855460838,
+      "learning_rate": 6.819912882514834e-06,
+      "loss": 1.3067,
+      "step": 12282
+    },
+    {
+      "epoch": 0.8853888848843077,
+      "grad_norm": 0.12371377409041348,
+      "learning_rate": 6.811441109442118e-06,
+      "loss": 1.2871,
+      "step": 12283
+    },
+    {
+      "epoch": 0.8854609673466446,
+      "grad_norm": 0.11148035307751299,
+      "learning_rate": 6.802974415979546e-06,
+      "loss": 1.1755,
+      "step": 12284
+    },
+    {
+      "epoch": 0.8855330498089815,
+      "grad_norm": 0.11397752911734543,
+      "learning_rate": 6.7945128025886375e-06,
+      "loss": 1.2476,
+      "step": 12285
+    },
+    {
+      "epoch": 0.8856051322713184,
+      "grad_norm": 0.11891311677240765,
+      "learning_rate": 6.7860562697306114e-06,
+      "loss": 1.2285,
+      "step": 12286
+    },
+    {
+      "epoch": 0.8856772147336553,
+      "grad_norm": 0.11775577246714927,
+      "learning_rate": 6.777604817866445e-06,
+      "loss": 1.1416,
+      "step": 12287
+    },
+    {
+      "epoch": 0.8857492971959923,
+      "grad_norm": 0.13350640353168539,
+      "learning_rate": 6.769158447456802e-06,
+      "loss": 1.2274,
+      "step": 12288
+    },
+    {
+      "epoch": 0.8858213796583291,
+      "grad_norm": 0.12930831539245516,
+      "learning_rate": 6.760717158962093e-06,
+      "loss": 1.239,
+      "step": 12289
+    },
+    {
+      "epoch": 0.885893462120666,
+      "grad_norm": 0.1362075961002773,
+      "learning_rate": 6.752280952842438e-06,
+      "loss": 1.2152,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8859655445830029,
+      "grad_norm": 0.12211996788814462,
+      "learning_rate": 6.7438498295576915e-06,
+      "loss": 1.3529,
+      "step": 12291
+    },
+    {
+      "epoch": 0.8860376270453398,
+      "grad_norm": 0.12719582355324013,
+      "learning_rate": 6.73542378956743e-06,
+      "loss": 1.3004,
+      "step": 12292
+    },
+    {
+      "epoch": 0.8861097095076768,
+      "grad_norm": 0.12209945446978956,
+      "learning_rate": 6.727002833330931e-06,
+      "loss": 1.2778,
+      "step": 12293
+    },
+    {
+      "epoch": 0.8861817919700137,
+      "grad_norm": 0.11878101976323117,
+      "learning_rate": 6.718586961307227e-06,
+      "loss": 1.2677,
+      "step": 12294
+    },
+    {
+      "epoch": 0.8862538744323506,
+      "grad_norm": 0.11972557918165012,
+      "learning_rate": 6.71017617395504e-06,
+      "loss": 1.2921,
+      "step": 12295
+    },
+    {
+      "epoch": 0.8863259568946875,
+      "grad_norm": 0.11951462677051931,
+      "learning_rate": 6.70177047173286e-06,
+      "loss": 1.3435,
+      "step": 12296
+    },
+    {
+      "epoch": 0.8863980393570244,
+      "grad_norm": 0.12014825242983591,
+      "learning_rate": 6.693369855098852e-06,
+      "loss": 1.3236,
+      "step": 12297
+    },
+    {
+      "epoch": 0.8864701218193614,
+      "grad_norm": 0.11417034886427697,
+      "learning_rate": 6.6849743245109396e-06,
+      "loss": 1.187,
+      "step": 12298
+    },
+    {
+      "epoch": 0.8865422042816983,
+      "grad_norm": 0.12210855633120797,
+      "learning_rate": 6.676583880426746e-06,
+      "loss": 1.3341,
+      "step": 12299
+    },
+    {
+      "epoch": 0.8866142867440352,
+      "grad_norm": 0.13910236077515967,
+      "learning_rate": 6.6681985233036145e-06,
+      "loss": 1.3096,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8866863692063721,
+      "grad_norm": 0.11905049824598567,
+      "learning_rate": 6.659818253598648e-06,
+      "loss": 1.2596,
+      "step": 12301
+    },
+    {
+      "epoch": 0.886758451668709,
+      "grad_norm": 0.12340313655824445,
+      "learning_rate": 6.651443071768626e-06,
+      "loss": 1.319,
+      "step": 12302
+    },
+    {
+      "epoch": 0.886830534131046,
+      "grad_norm": 0.14092902560623538,
+      "learning_rate": 6.6430729782700925e-06,
+      "loss": 1.2665,
+      "step": 12303
+    },
+    {
+      "epoch": 0.8869026165933829,
+      "grad_norm": 0.11970464311532593,
+      "learning_rate": 6.63470797355924e-06,
+      "loss": 1.1958,
+      "step": 12304
+    },
+    {
+      "epoch": 0.8869746990557198,
+      "grad_norm": 0.11671991876035227,
+      "learning_rate": 6.626348058092091e-06,
+      "loss": 1.2834,
+      "step": 12305
+    },
+    {
+      "epoch": 0.8870467815180566,
+      "grad_norm": 0.12180666952738178,
+      "learning_rate": 6.617993232324315e-06,
+      "loss": 1.2112,
+      "step": 12306
+    },
+    {
+      "epoch": 0.8871188639803935,
+      "grad_norm": 0.12031894435211743,
+      "learning_rate": 6.609643496711349e-06,
+      "loss": 1.3287,
+      "step": 12307
+    },
+    {
+      "epoch": 0.8871909464427304,
+      "grad_norm": 0.1211106382200576,
+      "learning_rate": 6.601298851708271e-06,
+      "loss": 1.2645,
+      "step": 12308
+    },
+    {
+      "epoch": 0.8872630289050674,
+      "grad_norm": 0.12036107126634416,
+      "learning_rate": 6.5929592977699964e-06,
+      "loss": 1.2213,
+      "step": 12309
+    },
+    {
+      "epoch": 0.8873351113674043,
+      "grad_norm": 0.11408783238533353,
+      "learning_rate": 6.584624835351072e-06,
+      "loss": 1.2881,
+      "step": 12310
+    },
+    {
+      "epoch": 0.8874071938297412,
+      "grad_norm": 0.12685693260879094,
+      "learning_rate": 6.576295464905835e-06,
+      "loss": 1.221,
+      "step": 12311
+    },
+    {
+      "epoch": 0.8874792762920781,
+      "grad_norm": 0.12416539653943359,
+      "learning_rate": 6.5679711868882774e-06,
+      "loss": 1.2186,
+      "step": 12312
+    },
+    {
+      "epoch": 0.887551358754415,
+      "grad_norm": 0.11297991435804675,
+      "learning_rate": 6.559652001752137e-06,
+      "loss": 1.3595,
+      "step": 12313
+    },
+    {
+      "epoch": 0.887623441216752,
+      "grad_norm": 0.10936745754065044,
+      "learning_rate": 6.551337909950916e-06,
+      "loss": 1.2224,
+      "step": 12314
+    },
+    {
+      "epoch": 0.8876955236790889,
+      "grad_norm": 0.10788490008225662,
+      "learning_rate": 6.54302891193781e-06,
+      "loss": 1.3273,
+      "step": 12315
+    },
+    {
+      "epoch": 0.8877676061414258,
+      "grad_norm": 0.12208713637128688,
+      "learning_rate": 6.534725008165699e-06,
+      "loss": 1.2683,
+      "step": 12316
+    },
+    {
+      "epoch": 0.8878396886037627,
+      "grad_norm": 0.13167894109390693,
+      "learning_rate": 6.526426199087232e-06,
+      "loss": 1.2189,
+      "step": 12317
+    },
+    {
+      "epoch": 0.8879117710660996,
+      "grad_norm": 0.1301037635182785,
+      "learning_rate": 6.518132485154793e-06,
+      "loss": 1.2759,
+      "step": 12318
+    },
+    {
+      "epoch": 0.8879838535284366,
+      "grad_norm": 0.141154905494827,
+      "learning_rate": 6.5098438668204535e-06,
+      "loss": 1.3234,
+      "step": 12319
+    },
+    {
+      "epoch": 0.8880559359907735,
+      "grad_norm": 0.12698314962098034,
+      "learning_rate": 6.5015603445359955e-06,
+      "loss": 1.24,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8881280184531104,
+      "grad_norm": 0.11492101118719804,
+      "learning_rate": 6.493281918752947e-06,
+      "loss": 1.23,
+      "step": 12321
+    },
+    {
+      "epoch": 0.8882001009154473,
+      "grad_norm": 0.11775932012495949,
+      "learning_rate": 6.485008589922581e-06,
+      "loss": 1.249,
+      "step": 12322
+    },
+    {
+      "epoch": 0.8882721833777841,
+      "grad_norm": 0.1292139702616653,
+      "learning_rate": 6.4767403584958804e-06,
+      "loss": 1.2459,
+      "step": 12323
+    },
+    {
+      "epoch": 0.888344265840121,
+      "grad_norm": 0.1291529291636658,
+      "learning_rate": 6.468477224923486e-06,
+      "loss": 1.2976,
+      "step": 12324
+    },
+    {
+      "epoch": 0.888416348302458,
+      "grad_norm": 0.11646069840258136,
+      "learning_rate": 6.460219189655847e-06,
+      "loss": 1.3421,
+      "step": 12325
+    },
+    {
+      "epoch": 0.8884884307647949,
+      "grad_norm": 0.12673059102112544,
+      "learning_rate": 6.451966253143082e-06,
+      "loss": 1.2687,
+      "step": 12326
+    },
+    {
+      "epoch": 0.8885605132271318,
+      "grad_norm": 0.11723602955067092,
+      "learning_rate": 6.4437184158350765e-06,
+      "loss": 1.3655,
+      "step": 12327
+    },
+    {
+      "epoch": 0.8886325956894687,
+      "grad_norm": 0.13023856800717615,
+      "learning_rate": 6.435475678181391e-06,
+      "loss": 1.2555,
+      "step": 12328
+    },
+    {
+      "epoch": 0.8887046781518056,
+      "grad_norm": 0.13457620593664002,
+      "learning_rate": 6.4272380406313334e-06,
+      "loss": 1.2238,
+      "step": 12329
+    },
+    {
+      "epoch": 0.8887767606141426,
+      "grad_norm": 0.11945305277864879,
+      "learning_rate": 6.419005503633913e-06,
+      "loss": 1.2328,
+      "step": 12330
+    },
+    {
+      "epoch": 0.8888488430764795,
+      "grad_norm": 0.1294844508707183,
+      "learning_rate": 6.410778067637924e-06,
+      "loss": 1.2928,
+      "step": 12331
+    },
+    {
+      "epoch": 0.8889209255388164,
+      "grad_norm": 0.12874379328134125,
+      "learning_rate": 6.402555733091786e-06,
+      "loss": 1.2904,
+      "step": 12332
+    },
+    {
+      "epoch": 0.8889930080011533,
+      "grad_norm": 0.12419767862048567,
+      "learning_rate": 6.394338500443697e-06,
+      "loss": 1.2893,
+      "step": 12333
+    },
+    {
+      "epoch": 0.8890650904634902,
+      "grad_norm": 0.12499399969425126,
+      "learning_rate": 6.386126370141599e-06,
+      "loss": 1.2584,
+      "step": 12334
+    },
+    {
+      "epoch": 0.8891371729258272,
+      "grad_norm": 0.11516119770193793,
+      "learning_rate": 6.377919342633098e-06,
+      "loss": 1.3009,
+      "step": 12335
+    },
+    {
+      "epoch": 0.8892092553881641,
+      "grad_norm": 0.11580193866156428,
+      "learning_rate": 6.36971741836555e-06,
+      "loss": 1.2826,
+      "step": 12336
+    },
+    {
+      "epoch": 0.889281337850501,
+      "grad_norm": 0.11977801412928292,
+      "learning_rate": 6.3615205977860526e-06,
+      "loss": 1.2361,
+      "step": 12337
+    },
+    {
+      "epoch": 0.8893534203128379,
+      "grad_norm": 0.1139543919767955,
+      "learning_rate": 6.353328881341402e-06,
+      "loss": 1.2961,
+      "step": 12338
+    },
+    {
+      "epoch": 0.8894255027751748,
+      "grad_norm": 0.12529567793099103,
+      "learning_rate": 6.345142269478111e-06,
+      "loss": 1.2985,
+      "step": 12339
+    },
+    {
+      "epoch": 0.8894975852375118,
+      "grad_norm": 0.11759916069911605,
+      "learning_rate": 6.336960762642419e-06,
+      "loss": 1.2416,
+      "step": 12340
+    },
+    {
+      "epoch": 0.8895696676998486,
+      "grad_norm": 0.12890656035012965,
+      "learning_rate": 6.328784361280304e-06,
+      "loss": 1.2271,
+      "step": 12341
+    },
+    {
+      "epoch": 0.8896417501621855,
+      "grad_norm": 0.12482868984276806,
+      "learning_rate": 6.320613065837455e-06,
+      "loss": 1.2382,
+      "step": 12342
+    },
+    {
+      "epoch": 0.8897138326245224,
+      "grad_norm": 0.12615265761510888,
+      "learning_rate": 6.312446876759259e-06,
+      "loss": 1.2543,
+      "step": 12343
+    },
+    {
+      "epoch": 0.8897859150868593,
+      "grad_norm": 0.11278239092170277,
+      "learning_rate": 6.3042857944908715e-06,
+      "loss": 1.2942,
+      "step": 12344
+    },
+    {
+      "epoch": 0.8898579975491963,
+      "grad_norm": 0.12377677975643782,
+      "learning_rate": 6.296129819477137e-06,
+      "loss": 1.2982,
+      "step": 12345
+    },
+    {
+      "epoch": 0.8899300800115332,
+      "grad_norm": 0.14846700011695874,
+      "learning_rate": 6.287978952162621e-06,
+      "loss": 1.2744,
+      "step": 12346
+    },
+    {
+      "epoch": 0.8900021624738701,
+      "grad_norm": 0.1393820592177439,
+      "learning_rate": 6.2798331929916246e-06,
+      "loss": 1.1179,
+      "step": 12347
+    },
+    {
+      "epoch": 0.890074244936207,
+      "grad_norm": 0.11414338457110812,
+      "learning_rate": 6.271692542408159e-06,
+      "loss": 1.1913,
+      "step": 12348
+    },
+    {
+      "epoch": 0.8901463273985439,
+      "grad_norm": 0.12739781883222334,
+      "learning_rate": 6.263557000855969e-06,
+      "loss": 1.2542,
+      "step": 12349
+    },
+    {
+      "epoch": 0.8902184098608809,
+      "grad_norm": 0.1296463442067619,
+      "learning_rate": 6.255426568778522e-06,
+      "loss": 1.2393,
+      "step": 12350
+    },
+    {
+      "epoch": 0.8902904923232178,
+      "grad_norm": 0.12509445127375077,
+      "learning_rate": 6.247301246618975e-06,
+      "loss": 1.2155,
+      "step": 12351
+    },
+    {
+      "epoch": 0.8903625747855547,
+      "grad_norm": 0.12192884764072334,
+      "learning_rate": 6.239181034820251e-06,
+      "loss": 1.2366,
+      "step": 12352
+    },
+    {
+      "epoch": 0.8904346572478916,
+      "grad_norm": 0.1212683921557252,
+      "learning_rate": 6.231065933824976e-06,
+      "loss": 1.2254,
+      "step": 12353
+    },
+    {
+      "epoch": 0.8905067397102285,
+      "grad_norm": 0.11158940679234623,
+      "learning_rate": 6.2229559440754815e-06,
+      "loss": 1.2517,
+      "step": 12354
+    },
+    {
+      "epoch": 0.8905788221725655,
+      "grad_norm": 0.11242550189351913,
+      "learning_rate": 6.2148510660138496e-06,
+      "loss": 1.3278,
+      "step": 12355
+    },
+    {
+      "epoch": 0.8906509046349024,
+      "grad_norm": 0.13907887509616385,
+      "learning_rate": 6.20675130008187e-06,
+      "loss": 1.2393,
+      "step": 12356
+    },
+    {
+      "epoch": 0.8907229870972393,
+      "grad_norm": 0.11552757552339593,
+      "learning_rate": 6.198656646721012e-06,
+      "loss": 1.2313,
+      "step": 12357
+    },
+    {
+      "epoch": 0.8907950695595761,
+      "grad_norm": 0.11581591647348985,
+      "learning_rate": 6.190567106372547e-06,
+      "loss": 1.2726,
+      "step": 12358
+    },
+    {
+      "epoch": 0.890867152021913,
+      "grad_norm": 0.11758891883923385,
+      "learning_rate": 6.18248267947743e-06,
+      "loss": 1.2551,
+      "step": 12359
+    },
+    {
+      "epoch": 0.8909392344842499,
+      "grad_norm": 0.12998582442003384,
+      "learning_rate": 6.1744033664763225e-06,
+      "loss": 1.2953,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8910113169465869,
+      "grad_norm": 0.11692611702942496,
+      "learning_rate": 6.166329167809604e-06,
+      "loss": 1.2395,
+      "step": 12361
+    },
+    {
+      "epoch": 0.8910833994089238,
+      "grad_norm": 0.11712074831410924,
+      "learning_rate": 6.158260083917411e-06,
+      "loss": 1.2143,
+      "step": 12362
+    },
+    {
+      "epoch": 0.8911554818712607,
+      "grad_norm": 0.12766223304250912,
+      "learning_rate": 6.150196115239581e-06,
+      "loss": 1.2068,
+      "step": 12363
+    },
+    {
+      "epoch": 0.8912275643335976,
+      "grad_norm": 0.12826363316523046,
+      "learning_rate": 6.142137262215686e-06,
+      "loss": 1.211,
+      "step": 12364
+    },
+    {
+      "epoch": 0.8912996467959345,
+      "grad_norm": 0.12499190953624244,
+      "learning_rate": 6.13408352528495e-06,
+      "loss": 1.238,
+      "step": 12365
+    },
+    {
+      "epoch": 0.8913717292582715,
+      "grad_norm": 0.12343959269167724,
+      "learning_rate": 6.126034904886424e-06,
+      "loss": 1.2781,
+      "step": 12366
+    },
+    {
+      "epoch": 0.8914438117206084,
+      "grad_norm": 0.13198059402846735,
+      "learning_rate": 6.117991401458823e-06,
+      "loss": 1.2248,
+      "step": 12367
+    },
+    {
+      "epoch": 0.8915158941829453,
+      "grad_norm": 0.11823978291544521,
+      "learning_rate": 6.109953015440595e-06,
+      "loss": 1.2203,
+      "step": 12368
+    },
+    {
+      "epoch": 0.8915879766452822,
+      "grad_norm": 0.12338482076897646,
+      "learning_rate": 6.101919747269869e-06,
+      "loss": 1.2272,
+      "step": 12369
+    },
+    {
+      "epoch": 0.8916600591076191,
+      "grad_norm": 0.10786628797515553,
+      "learning_rate": 6.093891597384549e-06,
+      "loss": 1.2747,
+      "step": 12370
+    },
+    {
+      "epoch": 0.8917321415699561,
+      "grad_norm": 0.1281762594504289,
+      "learning_rate": 6.085868566222252e-06,
+      "loss": 1.2206,
+      "step": 12371
+    },
+    {
+      "epoch": 0.891804224032293,
+      "grad_norm": 0.12482347938846608,
+      "learning_rate": 6.077850654220318e-06,
+      "loss": 1.2107,
+      "step": 12372
+    },
+    {
+      "epoch": 0.8918763064946299,
+      "grad_norm": 0.12016462730577784,
+      "learning_rate": 6.069837861815764e-06,
+      "loss": 1.1756,
+      "step": 12373
+    },
+    {
+      "epoch": 0.8919483889569668,
+      "grad_norm": 0.1144011076701129,
+      "learning_rate": 6.061830189445339e-06,
+      "loss": 1.1712,
+      "step": 12374
+    },
+    {
+      "epoch": 0.8920204714193036,
+      "grad_norm": 0.12873246657860354,
+      "learning_rate": 6.053827637545595e-06,
+      "loss": 1.2903,
+      "step": 12375
+    },
+    {
+      "epoch": 0.8920925538816405,
+      "grad_norm": 0.12560311227612583,
+      "learning_rate": 6.045830206552716e-06,
+      "loss": 1.3123,
+      "step": 12376
+    },
+    {
+      "epoch": 0.8921646363439775,
+      "grad_norm": 0.11755253727670555,
+      "learning_rate": 6.037837896902621e-06,
+      "loss": 1.2278,
+      "step": 12377
+    },
+    {
+      "epoch": 0.8922367188063144,
+      "grad_norm": 0.12906426865945217,
+      "learning_rate": 6.029850709030971e-06,
+      "loss": 1.3813,
+      "step": 12378
+    },
+    {
+      "epoch": 0.8923088012686513,
+      "grad_norm": 0.11465759477808934,
+      "learning_rate": 6.021868643373119e-06,
+      "loss": 1.228,
+      "step": 12379
+    },
+    {
+      "epoch": 0.8923808837309882,
+      "grad_norm": 0.12013471247610029,
+      "learning_rate": 6.013891700364205e-06,
+      "loss": 1.2149,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8924529661933251,
+      "grad_norm": 0.12030348360205442,
+      "learning_rate": 6.005919880439004e-06,
+      "loss": 1.3083,
+      "step": 12381
+    },
+    {
+      "epoch": 0.8925250486556621,
+      "grad_norm": 0.11487866823021714,
+      "learning_rate": 5.997953184032068e-06,
+      "loss": 1.3817,
+      "step": 12382
+    },
+    {
+      "epoch": 0.892597131117999,
+      "grad_norm": 0.12253586680783014,
+      "learning_rate": 5.989991611577639e-06,
+      "loss": 1.2503,
+      "step": 12383
+    },
+    {
+      "epoch": 0.8926692135803359,
+      "grad_norm": 0.12435155757745534,
+      "learning_rate": 5.982035163509725e-06,
+      "loss": 1.2394,
+      "step": 12384
+    },
+    {
+      "epoch": 0.8927412960426728,
+      "grad_norm": 0.1256229169021424,
+      "learning_rate": 5.974083840261991e-06,
+      "loss": 1.3351,
+      "step": 12385
+    },
+    {
+      "epoch": 0.8928133785050097,
+      "grad_norm": 0.12853021367698284,
+      "learning_rate": 5.966137642267866e-06,
+      "loss": 1.2484,
+      "step": 12386
+    },
+    {
+      "epoch": 0.8928854609673467,
+      "grad_norm": 0.13349584531828423,
+      "learning_rate": 5.958196569960483e-06,
+      "loss": 1.3023,
+      "step": 12387
+    },
+    {
+      "epoch": 0.8929575434296836,
+      "grad_norm": 0.1400223756584618,
+      "learning_rate": 5.95026062377273e-06,
+      "loss": 1.347,
+      "step": 12388
+    },
+    {
+      "epoch": 0.8930296258920205,
+      "grad_norm": 0.21261229023779085,
+      "learning_rate": 5.942329804137159e-06,
+      "loss": 1.4036,
+      "step": 12389
+    },
+    {
+      "epoch": 0.8931017083543574,
+      "grad_norm": 0.11886377062943758,
+      "learning_rate": 5.9344041114860805e-06,
+      "loss": 1.2837,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8931737908166943,
+      "grad_norm": 0.11611321172549295,
+      "learning_rate": 5.9264835462515046e-06,
+      "loss": 1.292,
+      "step": 12391
+    },
+    {
+      "epoch": 0.8932458732790313,
+      "grad_norm": 0.1224769603230322,
+      "learning_rate": 5.918568108865174e-06,
+      "loss": 1.2347,
+      "step": 12392
+    },
+    {
+      "epoch": 0.8933179557413681,
+      "grad_norm": 0.12796818827081846,
+      "learning_rate": 5.910657799758568e-06,
+      "loss": 1.1825,
+      "step": 12393
+    },
+    {
+      "epoch": 0.893390038203705,
+      "grad_norm": 0.11649020109313135,
+      "learning_rate": 5.902752619362861e-06,
+      "loss": 1.2358,
+      "step": 12394
+    },
+    {
+      "epoch": 0.8934621206660419,
+      "grad_norm": 0.11416589715312282,
+      "learning_rate": 5.894852568108955e-06,
+      "loss": 1.2781,
+      "step": 12395
+    },
+    {
+      "epoch": 0.8935342031283788,
+      "grad_norm": 0.11617329723270918,
+      "learning_rate": 5.886957646427471e-06,
+      "loss": 1.3046,
+      "step": 12396
+    },
+    {
+      "epoch": 0.8936062855907158,
+      "grad_norm": 0.13175956807311978,
+      "learning_rate": 5.879067854748743e-06,
+      "loss": 1.1781,
+      "step": 12397
+    },
+    {
+      "epoch": 0.8936783680530527,
+      "grad_norm": 0.1171592902824682,
+      "learning_rate": 5.871183193502861e-06,
+      "loss": 1.2493,
+      "step": 12398
+    },
+    {
+      "epoch": 0.8937504505153896,
+      "grad_norm": 0.1171696443383718,
+      "learning_rate": 5.8633036631195906e-06,
+      "loss": 1.2254,
+      "step": 12399
+    },
+    {
+      "epoch": 0.8938225329777265,
+      "grad_norm": 0.11417445636300076,
+      "learning_rate": 5.855429264028444e-06,
+      "loss": 1.2688,
+      "step": 12400
+    },
+    {
+      "epoch": 0.8938946154400634,
+      "grad_norm": 0.13099427418531817,
+      "learning_rate": 5.847559996658636e-06,
+      "loss": 1.1978,
+      "step": 12401
+    },
+    {
+      "epoch": 0.8939666979024004,
+      "grad_norm": 0.12379791641470839,
+      "learning_rate": 5.839695861439132e-06,
+      "loss": 1.266,
+      "step": 12402
+    },
+    {
+      "epoch": 0.8940387803647373,
+      "grad_norm": 0.1264062797795999,
+      "learning_rate": 5.831836858798579e-06,
+      "loss": 1.2368,
+      "step": 12403
+    },
+    {
+      "epoch": 0.8941108628270742,
+      "grad_norm": 0.114689856089128,
+      "learning_rate": 5.823982989165367e-06,
+      "loss": 1.272,
+      "step": 12404
+    },
+    {
+      "epoch": 0.8941829452894111,
+      "grad_norm": 0.1159767931191237,
+      "learning_rate": 5.816134252967609e-06,
+      "loss": 1.2127,
+      "step": 12405
+    },
+    {
+      "epoch": 0.894255027751748,
+      "grad_norm": 0.11895972687336734,
+      "learning_rate": 5.8082906506331306e-06,
+      "loss": 1.2727,
+      "step": 12406
+    },
+    {
+      "epoch": 0.894327110214085,
+      "grad_norm": 0.12689455912477776,
+      "learning_rate": 5.8004521825894665e-06,
+      "loss": 1.2673,
+      "step": 12407
+    },
+    {
+      "epoch": 0.8943991926764219,
+      "grad_norm": 0.11302641711847594,
+      "learning_rate": 5.792618849263898e-06,
+      "loss": 1.2477,
+      "step": 12408
+    },
+    {
+      "epoch": 0.8944712751387588,
+      "grad_norm": 0.12374317357032454,
+      "learning_rate": 5.784790651083405e-06,
+      "loss": 1.2792,
+      "step": 12409
+    },
+    {
+      "epoch": 0.8945433576010956,
+      "grad_norm": 0.12124802891287338,
+      "learning_rate": 5.776967588474691e-06,
+      "loss": 1.2236,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8946154400634325,
+      "grad_norm": 0.12348410658937195,
+      "learning_rate": 5.769149661864193e-06,
+      "loss": 1.2751,
+      "step": 12411
+    },
+    {
+      "epoch": 0.8946875225257694,
+      "grad_norm": 0.12777859455263882,
+      "learning_rate": 5.761336871678047e-06,
+      "loss": 1.2599,
+      "step": 12412
+    },
+    {
+      "epoch": 0.8947596049881064,
+      "grad_norm": 0.11468327480125379,
+      "learning_rate": 5.753529218342135e-06,
+      "loss": 1.1698,
+      "step": 12413
+    },
+    {
+      "epoch": 0.8948316874504433,
+      "grad_norm": 0.19596259077283468,
+      "learning_rate": 5.745726702282006e-06,
+      "loss": 1.2297,
+      "step": 12414
+    },
+    {
+      "epoch": 0.8949037699127802,
+      "grad_norm": 0.12768269852944678,
+      "learning_rate": 5.737929323923019e-06,
+      "loss": 1.2723,
+      "step": 12415
+    },
+    {
+      "epoch": 0.8949758523751171,
+      "grad_norm": 0.13184394544170733,
+      "learning_rate": 5.730137083690168e-06,
+      "loss": 1.2634,
+      "step": 12416
+    },
+    {
+      "epoch": 0.895047934837454,
+      "grad_norm": 0.11794190552178115,
+      "learning_rate": 5.7223499820082235e-06,
+      "loss": 1.2523,
+      "step": 12417
+    },
+    {
+      "epoch": 0.895120017299791,
+      "grad_norm": 0.11659049893424539,
+      "learning_rate": 5.714568019301603e-06,
+      "loss": 1.3097,
+      "step": 12418
+    },
+    {
+      "epoch": 0.8951920997621279,
+      "grad_norm": 0.12326393416153263,
+      "learning_rate": 5.706791195994554e-06,
+      "loss": 1.2814,
+      "step": 12419
+    },
+    {
+      "epoch": 0.8952641822244648,
+      "grad_norm": 0.12910850892331835,
+      "learning_rate": 5.69901951251095e-06,
+      "loss": 1.3608,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8953362646868017,
+      "grad_norm": 0.12414380310078814,
+      "learning_rate": 5.69125296927443e-06,
+      "loss": 1.2426,
+      "step": 12421
+    },
+    {
+      "epoch": 0.8954083471491386,
+      "grad_norm": 0.12561222166952699,
+      "learning_rate": 5.683491566708332e-06,
+      "loss": 1.3008,
+      "step": 12422
+    },
+    {
+      "epoch": 0.8954804296114756,
+      "grad_norm": 0.1182164509956659,
+      "learning_rate": 5.6757353052356964e-06,
+      "loss": 1.3074,
+      "step": 12423
+    },
+    {
+      "epoch": 0.8955525120738125,
+      "grad_norm": 0.12275977408644202,
+      "learning_rate": 5.667984185279374e-06,
+      "loss": 1.2399,
+      "step": 12424
+    },
+    {
+      "epoch": 0.8956245945361494,
+      "grad_norm": 0.12439219356191628,
+      "learning_rate": 5.660238207261814e-06,
+      "loss": 1.1729,
+      "step": 12425
+    },
+    {
+      "epoch": 0.8956966769984863,
+      "grad_norm": 0.16955798023431887,
+      "learning_rate": 5.652497371605259e-06,
+      "loss": 1.2404,
+      "step": 12426
+    },
+    {
+      "epoch": 0.8957687594608231,
+      "grad_norm": 0.1237470520912978,
+      "learning_rate": 5.644761678731647e-06,
+      "loss": 1.2727,
+      "step": 12427
+    },
+    {
+      "epoch": 0.89584084192316,
+      "grad_norm": 0.12117813652660102,
+      "learning_rate": 5.6370311290626645e-06,
+      "loss": 1.3333,
+      "step": 12428
+    },
+    {
+      "epoch": 0.895912924385497,
+      "grad_norm": 0.12525448604875566,
+      "learning_rate": 5.629305723019673e-06,
+      "loss": 1.3979,
+      "step": 12429
+    },
+    {
+      "epoch": 0.8959850068478339,
+      "grad_norm": 0.125360517225798,
+      "learning_rate": 5.6215854610237925e-06,
+      "loss": 1.2793,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8960570893101708,
+      "grad_norm": 0.11825595148633526,
+      "learning_rate": 5.6138703434958084e-06,
+      "loss": 1.2309,
+      "step": 12431
+    },
+    {
+      "epoch": 0.8961291717725077,
+      "grad_norm": 0.12305751897631312,
+      "learning_rate": 5.606160370856328e-06,
+      "loss": 1.3104,
+      "step": 12432
+    },
+    {
+      "epoch": 0.8962012542348446,
+      "grad_norm": 0.11065396282109882,
+      "learning_rate": 5.598455543525571e-06,
+      "loss": 1.1675,
+      "step": 12433
+    },
+    {
+      "epoch": 0.8962733366971816,
+      "grad_norm": 0.13110327399815475,
+      "learning_rate": 5.590755861923513e-06,
+      "loss": 1.2789,
+      "step": 12434
+    },
+    {
+      "epoch": 0.8963454191595185,
+      "grad_norm": 0.11925082212728327,
+      "learning_rate": 5.583061326469885e-06,
+      "loss": 1.2244,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8964175016218554,
+      "grad_norm": 0.11265940929454794,
+      "learning_rate": 5.5753719375840845e-06,
+      "loss": 1.3423,
+      "step": 12436
+    },
+    {
+      "epoch": 0.8964895840841923,
+      "grad_norm": 0.12627486684527225,
+      "learning_rate": 5.567687695685264e-06,
+      "loss": 1.3134,
+      "step": 12437
+    },
+    {
+      "epoch": 0.8965616665465292,
+      "grad_norm": 0.11303818909463755,
+      "learning_rate": 5.560008601192279e-06,
+      "loss": 1.2672,
+      "step": 12438
+    },
+    {
+      "epoch": 0.8966337490088662,
+      "grad_norm": 0.12717502393556188,
+      "learning_rate": 5.552334654523705e-06,
+      "loss": 1.2583,
+      "step": 12439
+    },
+    {
+      "epoch": 0.8967058314712031,
+      "grad_norm": 0.12067386120105113,
+      "learning_rate": 5.544665856097842e-06,
+      "loss": 1.3165,
+      "step": 12440
+    },
+    {
+      "epoch": 0.89677791393354,
+      "grad_norm": 0.11669850114090392,
+      "learning_rate": 5.53700220633272e-06,
+      "loss": 1.2879,
+      "step": 12441
+    },
+    {
+      "epoch": 0.8968499963958769,
+      "grad_norm": 0.11883359984873164,
+      "learning_rate": 5.529343705646073e-06,
+      "loss": 1.288,
+      "step": 12442
+    },
+    {
+      "epoch": 0.8969220788582138,
+      "grad_norm": 0.13157467558613836,
+      "learning_rate": 5.521690354455344e-06,
+      "loss": 1.1993,
+      "step": 12443
+    },
+    {
+      "epoch": 0.8969941613205508,
+      "grad_norm": 0.1281657901422245,
+      "learning_rate": 5.514042153177712e-06,
+      "loss": 1.3022,
+      "step": 12444
+    },
+    {
+      "epoch": 0.8970662437828876,
+      "grad_norm": 0.1304604517101371,
+      "learning_rate": 5.506399102230086e-06,
+      "loss": 1.317,
+      "step": 12445
+    },
+    {
+      "epoch": 0.8971383262452245,
+      "grad_norm": 0.14917819407568367,
+      "learning_rate": 5.498761202029079e-06,
+      "loss": 1.2334,
+      "step": 12446
+    },
+    {
+      "epoch": 0.8972104087075614,
+      "grad_norm": 0.11809023371842113,
+      "learning_rate": 5.4911284529910126e-06,
+      "loss": 1.3128,
+      "step": 12447
+    },
+    {
+      "epoch": 0.8972824911698983,
+      "grad_norm": 0.1182877354730926,
+      "learning_rate": 5.483500855531942e-06,
+      "loss": 1.2467,
+      "step": 12448
+    },
+    {
+      "epoch": 0.8973545736322353,
+      "grad_norm": 0.11829186272217494,
+      "learning_rate": 5.475878410067647e-06,
+      "loss": 1.3185,
+      "step": 12449
+    },
+    {
+      "epoch": 0.8974266560945722,
+      "grad_norm": 0.12151032560315808,
+      "learning_rate": 5.468261117013618e-06,
+      "loss": 1.1965,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8974987385569091,
+      "grad_norm": 0.10868586534299955,
+      "learning_rate": 5.460648976785054e-06,
+      "loss": 1.235,
+      "step": 12451
+    },
+    {
+      "epoch": 0.897570821019246,
+      "grad_norm": 0.11547340057095863,
+      "learning_rate": 5.453041989796903e-06,
+      "loss": 1.262,
+      "step": 12452
+    },
+    {
+      "epoch": 0.8976429034815829,
+      "grad_norm": 0.1149339345450557,
+      "learning_rate": 5.445440156463799e-06,
+      "loss": 1.2751,
+      "step": 12453
+    },
+    {
+      "epoch": 0.8977149859439199,
+      "grad_norm": 0.11687563684322026,
+      "learning_rate": 5.437843477200122e-06,
+      "loss": 1.2917,
+      "step": 12454
+    },
+    {
+      "epoch": 0.8977870684062568,
+      "grad_norm": 0.1359905671091966,
+      "learning_rate": 5.430251952419951e-06,
+      "loss": 1.323,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8978591508685937,
+      "grad_norm": 0.11140385711581277,
+      "learning_rate": 5.42266558253709e-06,
+      "loss": 1.1628,
+      "step": 12456
+    },
+    {
+      "epoch": 0.8979312333309306,
+      "grad_norm": 0.4168795612067133,
+      "learning_rate": 5.4150843679650845e-06,
+      "loss": 1.2994,
+      "step": 12457
+    },
+    {
+      "epoch": 0.8980033157932675,
+      "grad_norm": 0.1247566402347119,
+      "learning_rate": 5.407508309117138e-06,
+      "loss": 1.3044,
+      "step": 12458
+    },
+    {
+      "epoch": 0.8980753982556045,
+      "grad_norm": 0.1388479876527126,
+      "learning_rate": 5.399937406406241e-06,
+      "loss": 1.3203,
+      "step": 12459
+    },
+    {
+      "epoch": 0.8981474807179414,
+      "grad_norm": 0.12532314181148035,
+      "learning_rate": 5.392371660245077e-06,
+      "loss": 1.3033,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8982195631802783,
+      "grad_norm": 0.11949626390577335,
+      "learning_rate": 5.384811071046059e-06,
+      "loss": 1.2828,
+      "step": 12461
+    },
+    {
+      "epoch": 0.8982916456426151,
+      "grad_norm": 0.130703979792932,
+      "learning_rate": 5.377255639221257e-06,
+      "loss": 1.1951,
+      "step": 12462
+    },
+    {
+      "epoch": 0.898363728104952,
+      "grad_norm": 0.11880557586990917,
+      "learning_rate": 5.369705365182554e-06,
+      "loss": 1.2764,
+      "step": 12463
+    },
+    {
+      "epoch": 0.8984358105672889,
+      "grad_norm": 0.11747163764613208,
+      "learning_rate": 5.362160249341497e-06,
+      "loss": 1.2853,
+      "step": 12464
+    },
+    {
+      "epoch": 0.8985078930296259,
+      "grad_norm": 0.12534170518452847,
+      "learning_rate": 5.354620292109381e-06,
+      "loss": 1.2819,
+      "step": 12465
+    },
+    {
+      "epoch": 0.8985799754919628,
+      "grad_norm": 0.11451135069525258,
+      "learning_rate": 5.3470854938971635e-06,
+      "loss": 1.2752,
+      "step": 12466
+    },
+    {
+      "epoch": 0.8986520579542997,
+      "grad_norm": 0.11728383273874261,
+      "learning_rate": 5.339555855115563e-06,
+      "loss": 1.2383,
+      "step": 12467
+    },
+    {
+      "epoch": 0.8987241404166366,
+      "grad_norm": 0.11434980987611464,
+      "learning_rate": 5.332031376175029e-06,
+      "loss": 1.3098,
+      "step": 12468
+    },
+    {
+      "epoch": 0.8987962228789735,
+      "grad_norm": 0.11547474072273597,
+      "learning_rate": 5.324512057485731e-06,
+      "loss": 1.2923,
+      "step": 12469
+    },
+    {
+      "epoch": 0.8988683053413105,
+      "grad_norm": 0.11587616836717747,
+      "learning_rate": 5.31699789945751e-06,
+      "loss": 1.1766,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8989403878036474,
+      "grad_norm": 0.11615491743835334,
+      "learning_rate": 5.309488902499937e-06,
+      "loss": 1.227,
+      "step": 12471
+    },
+    {
+      "epoch": 0.8990124702659843,
+      "grad_norm": 0.12675230930506262,
+      "learning_rate": 5.301985067022364e-06,
+      "loss": 1.1893,
+      "step": 12472
+    },
+    {
+      "epoch": 0.8990845527283212,
+      "grad_norm": 0.1136375148766462,
+      "learning_rate": 5.294486393433806e-06,
+      "loss": 1.1638,
+      "step": 12473
+    },
+    {
+      "epoch": 0.8991566351906581,
+      "grad_norm": 0.15336290253200077,
+      "learning_rate": 5.286992882142982e-06,
+      "loss": 1.2351,
+      "step": 12474
+    },
+    {
+      "epoch": 0.8992287176529951,
+      "grad_norm": 0.1204955853198472,
+      "learning_rate": 5.279504533558388e-06,
+      "loss": 1.1485,
+      "step": 12475
+    },
+    {
+      "epoch": 0.899300800115332,
+      "grad_norm": 0.12285215279653253,
+      "learning_rate": 5.272021348088163e-06,
+      "loss": 1.2453,
+      "step": 12476
+    },
+    {
+      "epoch": 0.8993728825776689,
+      "grad_norm": 0.12210012582308268,
+      "learning_rate": 5.2645433261402695e-06,
+      "loss": 1.2837,
+      "step": 12477
+    },
+    {
+      "epoch": 0.8994449650400058,
+      "grad_norm": 0.11975350601990169,
+      "learning_rate": 5.257070468122272e-06,
+      "loss": 1.4047,
+      "step": 12478
+    },
+    {
+      "epoch": 0.8995170475023426,
+      "grad_norm": 0.11638154622060919,
+      "learning_rate": 5.249602774441542e-06,
+      "loss": 1.2529,
+      "step": 12479
+    },
+    {
+      "epoch": 0.8995891299646795,
+      "grad_norm": 0.1257457892491502,
+      "learning_rate": 5.2421402455050895e-06,
+      "loss": 1.2801,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8996612124270165,
+      "grad_norm": 0.12540192940838443,
+      "learning_rate": 5.2346828817197655e-06,
+      "loss": 1.207,
+      "step": 12481
+    },
+    {
+      "epoch": 0.8997332948893534,
+      "grad_norm": 0.12760554628880563,
+      "learning_rate": 5.227230683492002e-06,
+      "loss": 1.3109,
+      "step": 12482
+    },
+    {
+      "epoch": 0.8998053773516903,
+      "grad_norm": 0.12357235412843974,
+      "learning_rate": 5.219783651228017e-06,
+      "loss": 1.1575,
+      "step": 12483
+    },
+    {
+      "epoch": 0.8998774598140272,
+      "grad_norm": 0.11545566479511024,
+      "learning_rate": 5.212341785333752e-06,
+      "loss": 1.1411,
+      "step": 12484
+    },
+    {
+      "epoch": 0.8999495422763641,
+      "grad_norm": 0.12482938615610617,
+      "learning_rate": 5.204905086214873e-06,
+      "loss": 1.2767,
+      "step": 12485
+    },
+    {
+      "epoch": 0.9000216247387011,
+      "grad_norm": 0.13629416552878362,
+      "learning_rate": 5.197473554276711e-06,
+      "loss": 1.2136,
+      "step": 12486
+    },
+    {
+      "epoch": 0.900093707201038,
+      "grad_norm": 0.12846622906771632,
+      "learning_rate": 5.190047189924374e-06,
+      "loss": 1.2203,
+      "step": 12487
+    },
+    {
+      "epoch": 0.9001657896633749,
+      "grad_norm": 0.13422621086620848,
+      "learning_rate": 5.182625993562662e-06,
+      "loss": 1.1835,
+      "step": 12488
+    },
+    {
+      "epoch": 0.9002378721257118,
+      "grad_norm": 0.12809075603708087,
+      "learning_rate": 5.175209965596084e-06,
+      "loss": 1.197,
+      "step": 12489
+    },
+    {
+      "epoch": 0.9003099545880487,
+      "grad_norm": 0.16438088409604815,
+      "learning_rate": 5.167799106428894e-06,
+      "loss": 1.3001,
+      "step": 12490
+    },
+    {
+      "epoch": 0.9003820370503857,
+      "grad_norm": 0.13103549621499364,
+      "learning_rate": 5.160393416465048e-06,
+      "loss": 1.2456,
+      "step": 12491
+    },
+    {
+      "epoch": 0.9004541195127226,
+      "grad_norm": 0.1274173233395303,
+      "learning_rate": 5.152992896108222e-06,
+      "loss": 1.2995,
+      "step": 12492
+    },
+    {
+      "epoch": 0.9005262019750595,
+      "grad_norm": 0.12424258676591808,
+      "learning_rate": 5.145597545761805e-06,
+      "loss": 1.268,
+      "step": 12493
+    },
+    {
+      "epoch": 0.9005982844373964,
+      "grad_norm": 0.12194975830452251,
+      "learning_rate": 5.138207365828918e-06,
+      "loss": 1.2646,
+      "step": 12494
+    },
+    {
+      "epoch": 0.9006703668997333,
+      "grad_norm": 0.11136120068665879,
+      "learning_rate": 5.130822356712384e-06,
+      "loss": 1.3413,
+      "step": 12495
+    },
+    {
+      "epoch": 0.9007424493620703,
+      "grad_norm": 0.18007035996880924,
+      "learning_rate": 5.123442518814759e-06,
+      "loss": 1.3091,
+      "step": 12496
+    },
+    {
+      "epoch": 0.9008145318244071,
+      "grad_norm": 0.12568186298860817,
+      "learning_rate": 5.116067852538298e-06,
+      "loss": 1.2053,
+      "step": 12497
+    },
+    {
+      "epoch": 0.900886614286744,
+      "grad_norm": 0.12171605926839556,
+      "learning_rate": 5.108698358285002e-06,
+      "loss": 1.2486,
+      "step": 12498
+    },
+    {
+      "epoch": 0.9009586967490809,
+      "grad_norm": 0.13681307188773292,
+      "learning_rate": 5.101334036456562e-06,
+      "loss": 1.2385,
+      "step": 12499
+    },
+    {
+      "epoch": 0.9010307792114178,
+      "grad_norm": 0.11242952582491673,
+      "learning_rate": 5.093974887454411e-06,
+      "loss": 1.3414,
+      "step": 12500
+    },
+    {
+      "epoch": 0.9011028616737548,
+      "grad_norm": 0.13970645692019285,
+      "learning_rate": 5.0866209116796845e-06,
+      "loss": 1.2913,
+      "step": 12501
+    },
+    {
+      "epoch": 0.9011749441360917,
+      "grad_norm": 0.12971182447534252,
+      "learning_rate": 5.079272109533229e-06,
+      "loss": 1.3551,
+      "step": 12502
+    },
+    {
+      "epoch": 0.9012470265984286,
+      "grad_norm": 0.13559346991551702,
+      "learning_rate": 5.071928481415633e-06,
+      "loss": 1.2917,
+      "step": 12503
+    },
+    {
+      "epoch": 0.9013191090607655,
+      "grad_norm": 0.1300038557723152,
+      "learning_rate": 5.06459002772719e-06,
+      "loss": 1.3341,
+      "step": 12504
+    },
+    {
+      "epoch": 0.9013911915231024,
+      "grad_norm": 0.11900051361631211,
+      "learning_rate": 5.0572567488679005e-06,
+      "loss": 1.316,
+      "step": 12505
+    },
+    {
+      "epoch": 0.9014632739854394,
+      "grad_norm": 0.12158128268858368,
+      "learning_rate": 5.049928645237501e-06,
+      "loss": 1.2286,
+      "step": 12506
+    },
+    {
+      "epoch": 0.9015353564477763,
+      "grad_norm": 0.13515945713305896,
+      "learning_rate": 5.042605717235438e-06,
+      "loss": 1.1956,
+      "step": 12507
+    },
+    {
+      "epoch": 0.9016074389101132,
+      "grad_norm": 0.15389084954234908,
+      "learning_rate": 5.035287965260871e-06,
+      "loss": 1.2595,
+      "step": 12508
+    },
+    {
+      "epoch": 0.9016795213724501,
+      "grad_norm": 0.13497124728289575,
+      "learning_rate": 5.027975389712702e-06,
+      "loss": 1.3331,
+      "step": 12509
+    },
+    {
+      "epoch": 0.901751603834787,
+      "grad_norm": 0.11435240692782286,
+      "learning_rate": 5.020667990989525e-06,
+      "loss": 1.2132,
+      "step": 12510
+    },
+    {
+      "epoch": 0.901823686297124,
+      "grad_norm": 0.1218071306234848,
+      "learning_rate": 5.01336576948962e-06,
+      "loss": 1.316,
+      "step": 12511
+    },
+    {
+      "epoch": 0.9018957687594609,
+      "grad_norm": 0.11676463430035318,
+      "learning_rate": 5.006068725611079e-06,
+      "loss": 1.3466,
+      "step": 12512
+    },
+    {
+      "epoch": 0.9019678512217978,
+      "grad_norm": 0.12354576351874778,
+      "learning_rate": 4.998776859751619e-06,
+      "loss": 1.1817,
+      "step": 12513
+    },
+    {
+      "epoch": 0.9020399336841346,
+      "grad_norm": 0.11370040828639157,
+      "learning_rate": 4.9914901723087544e-06,
+      "loss": 1.1939,
+      "step": 12514
+    },
+    {
+      "epoch": 0.9021120161464715,
+      "grad_norm": 0.12148783214458198,
+      "learning_rate": 4.984208663679602e-06,
+      "loss": 1.2716,
+      "step": 12515
+    },
+    {
+      "epoch": 0.9021840986088084,
+      "grad_norm": 0.1211109800858113,
+      "learning_rate": 4.9769323342611414e-06,
+      "loss": 1.309,
+      "step": 12516
+    },
+    {
+      "epoch": 0.9022561810711454,
+      "grad_norm": 0.11755251179277838,
+      "learning_rate": 4.969661184449958e-06,
+      "loss": 1.2651,
+      "step": 12517
+    },
+    {
+      "epoch": 0.9023282635334823,
+      "grad_norm": 0.13314750218944252,
+      "learning_rate": 4.962395214642424e-06,
+      "loss": 1.3233,
+      "step": 12518
+    },
+    {
+      "epoch": 0.9024003459958192,
+      "grad_norm": 0.1373382709565098,
+      "learning_rate": 4.955134425234576e-06,
+      "loss": 1.3313,
+      "step": 12519
+    },
+    {
+      "epoch": 0.9024724284581561,
+      "grad_norm": 0.11887961218224538,
+      "learning_rate": 4.947878816622175e-06,
+      "loss": 1.2052,
+      "step": 12520
+    },
+    {
+      "epoch": 0.902544510920493,
+      "grad_norm": 0.12025628576324782,
+      "learning_rate": 4.940628389200752e-06,
+      "loss": 1.2888,
+      "step": 12521
+    },
+    {
+      "epoch": 0.90261659338283,
+      "grad_norm": 0.11802720843451801,
+      "learning_rate": 4.933383143365522e-06,
+      "loss": 1.288,
+      "step": 12522
+    },
+    {
+      "epoch": 0.9026886758451669,
+      "grad_norm": 0.12306231966624132,
+      "learning_rate": 4.926143079511391e-06,
+      "loss": 1.3056,
+      "step": 12523
+    },
+    {
+      "epoch": 0.9027607583075038,
+      "grad_norm": 0.13798808599289822,
+      "learning_rate": 4.918908198033001e-06,
+      "loss": 1.3664,
+      "step": 12524
+    },
+    {
+      "epoch": 0.9028328407698407,
+      "grad_norm": 0.12474130670801219,
+      "learning_rate": 4.911678499324746e-06,
+      "loss": 1.2842,
+      "step": 12525
+    },
+    {
+      "epoch": 0.9029049232321776,
+      "grad_norm": 0.13248138039158672,
+      "learning_rate": 4.90445398378071e-06,
+      "loss": 1.397,
+      "step": 12526
+    },
+    {
+      "epoch": 0.9029770056945146,
+      "grad_norm": 0.12553241847989796,
+      "learning_rate": 4.8972346517946684e-06,
+      "loss": 1.2664,
+      "step": 12527
+    },
+    {
+      "epoch": 0.9030490881568515,
+      "grad_norm": 0.11521737357344401,
+      "learning_rate": 4.89002050376014e-06,
+      "loss": 1.182,
+      "step": 12528
+    },
+    {
+      "epoch": 0.9031211706191884,
+      "grad_norm": 0.11924853050785354,
+      "learning_rate": 4.882811540070387e-06,
+      "loss": 1.3488,
+      "step": 12529
+    },
+    {
+      "epoch": 0.9031932530815253,
+      "grad_norm": 0.1127004228958744,
+      "learning_rate": 4.875607761118362e-06,
+      "loss": 1.2444,
+      "step": 12530
+    },
+    {
+      "epoch": 0.9032653355438621,
+      "grad_norm": 0.11485139760982006,
+      "learning_rate": 4.868409167296706e-06,
+      "loss": 1.3109,
+      "step": 12531
+    },
+    {
+      "epoch": 0.903337418006199,
+      "grad_norm": 0.11500888429543862,
+      "learning_rate": 4.861215758997817e-06,
+      "loss": 1.2943,
+      "step": 12532
+    },
+    {
+      "epoch": 0.903409500468536,
+      "grad_norm": 0.12094529551830756,
+      "learning_rate": 4.854027536613792e-06,
+      "loss": 1.2991,
+      "step": 12533
+    },
+    {
+      "epoch": 0.9034815829308729,
+      "grad_norm": 0.11821650307077589,
+      "learning_rate": 4.846844500536496e-06,
+      "loss": 1.3389,
+      "step": 12534
+    },
+    {
+      "epoch": 0.9035536653932098,
+      "grad_norm": 0.14003068418623207,
+      "learning_rate": 4.839666651157426e-06,
+      "loss": 1.3515,
+      "step": 12535
+    },
+    {
+      "epoch": 0.9036257478555467,
+      "grad_norm": 0.11223074850857405,
+      "learning_rate": 4.832493988867859e-06,
+      "loss": 1.2771,
+      "step": 12536
+    },
+    {
+      "epoch": 0.9036978303178836,
+      "grad_norm": 0.14267010267679553,
+      "learning_rate": 4.825326514058737e-06,
+      "loss": 1.3409,
+      "step": 12537
+    },
+    {
+      "epoch": 0.9037699127802206,
+      "grad_norm": 0.12197357249063558,
+      "learning_rate": 4.818164227120814e-06,
+      "loss": 1.1967,
+      "step": 12538
+    },
+    {
+      "epoch": 0.9038419952425575,
+      "grad_norm": 0.11142231072301254,
+      "learning_rate": 4.811007128444445e-06,
+      "loss": 1.202,
+      "step": 12539
+    },
+    {
+      "epoch": 0.9039140777048944,
+      "grad_norm": 0.11966448277159132,
+      "learning_rate": 4.8038552184197724e-06,
+      "loss": 1.343,
+      "step": 12540
+    },
+    {
+      "epoch": 0.9039861601672313,
+      "grad_norm": 0.11878584623432066,
+      "learning_rate": 4.796708497436641e-06,
+      "loss": 1.2735,
+      "step": 12541
+    },
+    {
+      "epoch": 0.9040582426295682,
+      "grad_norm": 0.11831437086930123,
+      "learning_rate": 4.789566965884607e-06,
+      "loss": 1.265,
+      "step": 12542
+    },
+    {
+      "epoch": 0.9041303250919052,
+      "grad_norm": 0.11035769388358223,
+      "learning_rate": 4.782430624152945e-06,
+      "loss": 1.2681,
+      "step": 12543
+    },
+    {
+      "epoch": 0.9042024075542421,
+      "grad_norm": 0.12157885024697629,
+      "learning_rate": 4.775299472630656e-06,
+      "loss": 1.1929,
+      "step": 12544
+    },
+    {
+      "epoch": 0.904274490016579,
+      "grad_norm": 0.11098952271638257,
+      "learning_rate": 4.768173511706464e-06,
+      "loss": 1.2648,
+      "step": 12545
+    },
+    {
+      "epoch": 0.9043465724789159,
+      "grad_norm": 0.11865947808846831,
+      "learning_rate": 4.7610527417687675e-06,
+      "loss": 1.2812,
+      "step": 12546
+    },
+    {
+      "epoch": 0.9044186549412528,
+      "grad_norm": 0.12329077168136324,
+      "learning_rate": 4.753937163205735e-06,
+      "loss": 1.2792,
+      "step": 12547
+    },
+    {
+      "epoch": 0.9044907374035898,
+      "grad_norm": 0.1111414966022464,
+      "learning_rate": 4.746826776405222e-06,
+      "loss": 1.3161,
+      "step": 12548
+    },
+    {
+      "epoch": 0.9045628198659266,
+      "grad_norm": 0.12317378150038524,
+      "learning_rate": 4.739721581754808e-06,
+      "loss": 1.2608,
+      "step": 12549
+    },
+    {
+      "epoch": 0.9046349023282635,
+      "grad_norm": 0.14610192231635308,
+      "learning_rate": 4.732621579641794e-06,
+      "loss": 1.2418,
+      "step": 12550
+    },
+    {
+      "epoch": 0.9047069847906004,
+      "grad_norm": 0.12588261155509967,
+      "learning_rate": 4.7255267704531925e-06,
+      "loss": 1.2991,
+      "step": 12551
+    },
+    {
+      "epoch": 0.9047790672529373,
+      "grad_norm": 0.1205153758477869,
+      "learning_rate": 4.718437154575728e-06,
+      "loss": 1.3081,
+      "step": 12552
+    },
+    {
+      "epoch": 0.9048511497152742,
+      "grad_norm": 0.12845457322993784,
+      "learning_rate": 4.711352732395846e-06,
+      "loss": 1.2267,
+      "step": 12553
+    },
+    {
+      "epoch": 0.9049232321776112,
+      "grad_norm": 0.11250432923457965,
+      "learning_rate": 4.704273504299727e-06,
+      "loss": 1.3292,
+      "step": 12554
+    },
+    {
+      "epoch": 0.9049953146399481,
+      "grad_norm": 0.1140413545430185,
+      "learning_rate": 4.697199470673242e-06,
+      "loss": 1.2792,
+      "step": 12555
+    },
+    {
+      "epoch": 0.905067397102285,
+      "grad_norm": 0.12142942982140031,
+      "learning_rate": 4.690130631901979e-06,
+      "loss": 1.3049,
+      "step": 12556
+    },
+    {
+      "epoch": 0.9051394795646219,
+      "grad_norm": 0.12145401472641777,
+      "learning_rate": 4.683066988371265e-06,
+      "loss": 1.3374,
+      "step": 12557
+    },
+    {
+      "epoch": 0.9052115620269588,
+      "grad_norm": 0.11438620455788989,
+      "learning_rate": 4.6760085404661365e-06,
+      "loss": 1.111,
+      "step": 12558
+    },
+    {
+      "epoch": 0.9052836444892958,
+      "grad_norm": 0.1209221634372621,
+      "learning_rate": 4.668955288571331e-06,
+      "loss": 1.211,
+      "step": 12559
+    },
+    {
+      "epoch": 0.9053557269516327,
+      "grad_norm": 0.11154830929428859,
+      "learning_rate": 4.6619072330713164e-06,
+      "loss": 1.2652,
+      "step": 12560
+    },
+    {
+      "epoch": 0.9054278094139696,
+      "grad_norm": 0.12617922488742161,
+      "learning_rate": 4.654864374350287e-06,
+      "loss": 1.2959,
+      "step": 12561
+    },
+    {
+      "epoch": 0.9054998918763065,
+      "grad_norm": 0.12887227020366002,
+      "learning_rate": 4.647826712792125e-06,
+      "loss": 1.4518,
+      "step": 12562
+    },
+    {
+      "epoch": 0.9055719743386434,
+      "grad_norm": 0.11448872773779338,
+      "learning_rate": 4.640794248780455e-06,
+      "loss": 1.3046,
+      "step": 12563
+    },
+    {
+      "epoch": 0.9056440568009804,
+      "grad_norm": 0.1267873243353109,
+      "learning_rate": 4.633766982698595e-06,
+      "loss": 1.1858,
+      "step": 12564
+    },
+    {
+      "epoch": 0.9057161392633173,
+      "grad_norm": 0.12236916138619214,
+      "learning_rate": 4.626744914929615e-06,
+      "loss": 1.3053,
+      "step": 12565
+    },
+    {
+      "epoch": 0.9057882217256541,
+      "grad_norm": 0.11040734239774924,
+      "learning_rate": 4.619728045856264e-06,
+      "loss": 1.3566,
+      "step": 12566
+    },
+    {
+      "epoch": 0.905860304187991,
+      "grad_norm": 0.12384110289426227,
+      "learning_rate": 4.6127163758610485e-06,
+      "loss": 1.3819,
+      "step": 12567
+    },
+    {
+      "epoch": 0.9059323866503279,
+      "grad_norm": 0.15051481582674436,
+      "learning_rate": 4.605709905326117e-06,
+      "loss": 1.1847,
+      "step": 12568
+    },
+    {
+      "epoch": 0.9060044691126649,
+      "grad_norm": 0.13480627670164488,
+      "learning_rate": 4.598708634633442e-06,
+      "loss": 1.2723,
+      "step": 12569
+    },
+    {
+      "epoch": 0.9060765515750018,
+      "grad_norm": 0.12048335782250474,
+      "learning_rate": 4.591712564164619e-06,
+      "loss": 1.2001,
+      "step": 12570
+    },
+    {
+      "epoch": 0.9061486340373387,
+      "grad_norm": 0.1271763745786545,
+      "learning_rate": 4.584721694301031e-06,
+      "loss": 1.3215,
+      "step": 12571
+    },
+    {
+      "epoch": 0.9062207164996756,
+      "grad_norm": 0.12875540848013978,
+      "learning_rate": 4.577736025423684e-06,
+      "loss": 1.3278,
+      "step": 12572
+    },
+    {
+      "epoch": 0.9062927989620125,
+      "grad_norm": 0.11750548760352045,
+      "learning_rate": 4.57075555791342e-06,
+      "loss": 1.2087,
+      "step": 12573
+    },
+    {
+      "epoch": 0.9063648814243495,
+      "grad_norm": 0.11708630033778405,
+      "learning_rate": 4.563780292150721e-06,
+      "loss": 1.3778,
+      "step": 12574
+    },
+    {
+      "epoch": 0.9064369638866864,
+      "grad_norm": 0.118623829763595,
+      "learning_rate": 4.556810228515773e-06,
+      "loss": 1.2804,
+      "step": 12575
+    },
+    {
+      "epoch": 0.9065090463490233,
+      "grad_norm": 0.12062255800120675,
+      "learning_rate": 4.549845367388539e-06,
+      "loss": 1.2516,
+      "step": 12576
+    },
+    {
+      "epoch": 0.9065811288113602,
+      "grad_norm": 0.12476297791333646,
+      "learning_rate": 4.542885709148636e-06,
+      "loss": 1.2364,
+      "step": 12577
+    },
+    {
+      "epoch": 0.9066532112736971,
+      "grad_norm": 0.244360980969024,
+      "learning_rate": 4.535931254175463e-06,
+      "loss": 1.3063,
+      "step": 12578
+    },
+    {
+      "epoch": 0.906725293736034,
+      "grad_norm": 0.16269724207361402,
+      "learning_rate": 4.5289820028480815e-06,
+      "loss": 1.32,
+      "step": 12579
+    },
+    {
+      "epoch": 0.906797376198371,
+      "grad_norm": 0.11164649861061338,
+      "learning_rate": 4.522037955545277e-06,
+      "loss": 1.2272,
+      "step": 12580
+    },
+    {
+      "epoch": 0.9068694586607079,
+      "grad_norm": 0.12176379948275719,
+      "learning_rate": 4.515099112645571e-06,
+      "loss": 1.2949,
+      "step": 12581
+    },
+    {
+      "epoch": 0.9069415411230448,
+      "grad_norm": 0.12231877956926142,
+      "learning_rate": 4.5081654745272266e-06,
+      "loss": 1.2752,
+      "step": 12582
+    },
+    {
+      "epoch": 0.9070136235853816,
+      "grad_norm": 0.14327678816160894,
+      "learning_rate": 4.501237041568129e-06,
+      "loss": 1.2718,
+      "step": 12583
+    },
+    {
+      "epoch": 0.9070857060477185,
+      "grad_norm": 0.12400800932789059,
+      "learning_rate": 4.494313814145978e-06,
+      "loss": 1.1801,
+      "step": 12584
+    },
+    {
+      "epoch": 0.9071577885100555,
+      "grad_norm": 0.12560600961388527,
+      "learning_rate": 4.487395792638149e-06,
+      "loss": 1.4044,
+      "step": 12585
+    },
+    {
+      "epoch": 0.9072298709723924,
+      "grad_norm": 0.11896713204503316,
+      "learning_rate": 4.480482977421718e-06,
+      "loss": 1.2853,
+      "step": 12586
+    },
+    {
+      "epoch": 0.9073019534347293,
+      "grad_norm": 0.11713350470592176,
+      "learning_rate": 4.473575368873517e-06,
+      "loss": 1.2275,
+      "step": 12587
+    },
+    {
+      "epoch": 0.9073740358970662,
+      "grad_norm": 0.13356316160331155,
+      "learning_rate": 4.466672967370056e-06,
+      "loss": 1.3175,
+      "step": 12588
+    },
+    {
+      "epoch": 0.9074461183594031,
+      "grad_norm": 0.12618155145766133,
+      "learning_rate": 4.459775773287578e-06,
+      "loss": 1.282,
+      "step": 12589
+    },
+    {
+      "epoch": 0.9075182008217401,
+      "grad_norm": 0.1262877753785203,
+      "learning_rate": 4.45288378700206e-06,
+      "loss": 1.2394,
+      "step": 12590
+    },
+    {
+      "epoch": 0.907590283284077,
+      "grad_norm": 0.1276554891914036,
+      "learning_rate": 4.445997008889158e-06,
+      "loss": 1.3021,
+      "step": 12591
+    },
+    {
+      "epoch": 0.9076623657464139,
+      "grad_norm": 0.11584801798209778,
+      "learning_rate": 4.439115439324271e-06,
+      "loss": 1.2221,
+      "step": 12592
+    },
+    {
+      "epoch": 0.9077344482087508,
+      "grad_norm": 0.12891461109954563,
+      "learning_rate": 4.432239078682499e-06,
+      "loss": 1.215,
+      "step": 12593
+    },
+    {
+      "epoch": 0.9078065306710877,
+      "grad_norm": 0.12425277666501625,
+      "learning_rate": 4.425367927338675e-06,
+      "loss": 1.2351,
+      "step": 12594
+    },
+    {
+      "epoch": 0.9078786131334247,
+      "grad_norm": 0.1129574274893692,
+      "learning_rate": 4.418501985667334e-06,
+      "loss": 1.1106,
+      "step": 12595
+    },
+    {
+      "epoch": 0.9079506955957616,
+      "grad_norm": 0.1262212967419009,
+      "learning_rate": 4.411641254042731e-06,
+      "loss": 1.3733,
+      "step": 12596
+    },
+    {
+      "epoch": 0.9080227780580985,
+      "grad_norm": 0.11841461453785244,
+      "learning_rate": 4.404785732838846e-06,
+      "loss": 1.4162,
+      "step": 12597
+    },
+    {
+      "epoch": 0.9080948605204354,
+      "grad_norm": 0.12423996820305219,
+      "learning_rate": 4.397935422429345e-06,
+      "loss": 1.2711,
+      "step": 12598
+    },
+    {
+      "epoch": 0.9081669429827723,
+      "grad_norm": 0.12916267739619422,
+      "learning_rate": 4.3910903231876524e-06,
+      "loss": 1.2722,
+      "step": 12599
+    },
+    {
+      "epoch": 0.9082390254451093,
+      "grad_norm": 0.11868976874878548,
+      "learning_rate": 4.38425043548687e-06,
+      "loss": 1.2891,
+      "step": 12600
+    },
+    {
+      "epoch": 0.9083111079074461,
+      "grad_norm": 0.12635282637848638,
+      "learning_rate": 4.377415759699854e-06,
+      "loss": 1.2809,
+      "step": 12601
+    },
+    {
+      "epoch": 0.908383190369783,
+      "grad_norm": 0.12560094477369937,
+      "learning_rate": 4.37058629619913e-06,
+      "loss": 1.3198,
+      "step": 12602
+    },
+    {
+      "epoch": 0.9084552728321199,
+      "grad_norm": 0.12907784232458688,
+      "learning_rate": 4.3637620453569896e-06,
+      "loss": 1.3024,
+      "step": 12603
+    },
+    {
+      "epoch": 0.9085273552944568,
+      "grad_norm": 0.1139247327355935,
+      "learning_rate": 4.3569430075454e-06,
+      "loss": 1.2598,
+      "step": 12604
+    },
+    {
+      "epoch": 0.9085994377567937,
+      "grad_norm": 0.11945011242793961,
+      "learning_rate": 4.3501291831360646e-06,
+      "loss": 1.2517,
+      "step": 12605
+    },
+    {
+      "epoch": 0.9086715202191307,
+      "grad_norm": 0.11673160048546326,
+      "learning_rate": 4.3433205725003975e-06,
+      "loss": 1.2788,
+      "step": 12606
+    },
+    {
+      "epoch": 0.9087436026814676,
+      "grad_norm": 0.11429560214094013,
+      "learning_rate": 4.336517176009535e-06,
+      "loss": 1.171,
+      "step": 12607
+    },
+    {
+      "epoch": 0.9088156851438045,
+      "grad_norm": 0.11488335228092797,
+      "learning_rate": 4.329718994034293e-06,
+      "loss": 1.2621,
+      "step": 12608
+    },
+    {
+      "epoch": 0.9088877676061414,
+      "grad_norm": 0.1162487704479823,
+      "learning_rate": 4.3229260269452734e-06,
+      "loss": 1.3243,
+      "step": 12609
+    },
+    {
+      "epoch": 0.9089598500684783,
+      "grad_norm": 0.11525711526677579,
+      "learning_rate": 4.316138275112735e-06,
+      "loss": 1.185,
+      "step": 12610
+    },
+    {
+      "epoch": 0.9090319325308153,
+      "grad_norm": 0.11259384465691276,
+      "learning_rate": 4.309355738906684e-06,
+      "loss": 1.3074,
+      "step": 12611
+    },
+    {
+      "epoch": 0.9091040149931522,
+      "grad_norm": 0.12223459227114702,
+      "learning_rate": 4.30257841869679e-06,
+      "loss": 1.2266,
+      "step": 12612
+    },
+    {
+      "epoch": 0.9091760974554891,
+      "grad_norm": 0.11955279219391268,
+      "learning_rate": 4.295806314852524e-06,
+      "loss": 1.277,
+      "step": 12613
+    },
+    {
+      "epoch": 0.909248179917826,
+      "grad_norm": 0.11536585943293028,
+      "learning_rate": 4.289039427743014e-06,
+      "loss": 1.2953,
+      "step": 12614
+    },
+    {
+      "epoch": 0.909320262380163,
+      "grad_norm": 0.11534154663409553,
+      "learning_rate": 4.282277757737108e-06,
+      "loss": 1.3055,
+      "step": 12615
+    },
+    {
+      "epoch": 0.9093923448424999,
+      "grad_norm": 0.11584695809280861,
+      "learning_rate": 4.275521305203367e-06,
+      "loss": 1.2609,
+      "step": 12616
+    },
+    {
+      "epoch": 0.9094644273048368,
+      "grad_norm": 0.12104961498821921,
+      "learning_rate": 4.268770070510097e-06,
+      "loss": 1.2526,
+      "step": 12617
+    },
+    {
+      "epoch": 0.9095365097671736,
+      "grad_norm": 0.11009074359692153,
+      "learning_rate": 4.2620240540252935e-06,
+      "loss": 1.2229,
+      "step": 12618
+    },
+    {
+      "epoch": 0.9096085922295105,
+      "grad_norm": 0.12798731438036146,
+      "learning_rate": 4.255283256116682e-06,
+      "loss": 1.313,
+      "step": 12619
+    },
+    {
+      "epoch": 0.9096806746918474,
+      "grad_norm": 0.12715262005774283,
+      "learning_rate": 4.248547677151693e-06,
+      "loss": 1.2765,
+      "step": 12620
+    },
+    {
+      "epoch": 0.9097527571541844,
+      "grad_norm": 0.11518820697250215,
+      "learning_rate": 4.241817317497443e-06,
+      "loss": 1.2701,
+      "step": 12621
+    },
+    {
+      "epoch": 0.9098248396165213,
+      "grad_norm": 0.12383160814557378,
+      "learning_rate": 4.235092177520849e-06,
+      "loss": 1.2998,
+      "step": 12622
+    },
+    {
+      "epoch": 0.9098969220788582,
+      "grad_norm": 0.1252532141589141,
+      "learning_rate": 4.228372257588487e-06,
+      "loss": 1.2621,
+      "step": 12623
+    },
+    {
+      "epoch": 0.9099690045411951,
+      "grad_norm": 0.11518453359100471,
+      "learning_rate": 4.221657558066605e-06,
+      "loss": 1.2854,
+      "step": 12624
+    },
+    {
+      "epoch": 0.910041087003532,
+      "grad_norm": 0.12026642121732509,
+      "learning_rate": 4.214948079321246e-06,
+      "loss": 1.2545,
+      "step": 12625
+    },
+    {
+      "epoch": 0.910113169465869,
+      "grad_norm": 0.12369570808890891,
+      "learning_rate": 4.2082438217181385e-06,
+      "loss": 1.2899,
+      "step": 12626
+    },
+    {
+      "epoch": 0.9101852519282059,
+      "grad_norm": 0.13093880900109087,
+      "learning_rate": 4.201544785622735e-06,
+      "loss": 1.289,
+      "step": 12627
+    },
+    {
+      "epoch": 0.9102573343905428,
+      "grad_norm": 0.13932886214519338,
+      "learning_rate": 4.194850971400166e-06,
+      "loss": 1.2104,
+      "step": 12628
+    },
+    {
+      "epoch": 0.9103294168528797,
+      "grad_norm": 0.1350176069631949,
+      "learning_rate": 4.188162379415317e-06,
+      "loss": 1.2666,
+      "step": 12629
+    },
+    {
+      "epoch": 0.9104014993152166,
+      "grad_norm": 0.12821286491911743,
+      "learning_rate": 4.181479010032763e-06,
+      "loss": 1.3013,
+      "step": 12630
+    },
+    {
+      "epoch": 0.9104735817775536,
+      "grad_norm": 0.13452602025988836,
+      "learning_rate": 4.1748008636168455e-06,
+      "loss": 1.2234,
+      "step": 12631
+    },
+    {
+      "epoch": 0.9105456642398905,
+      "grad_norm": 0.12100868308091478,
+      "learning_rate": 4.168127940531541e-06,
+      "loss": 1.2692,
+      "step": 12632
+    },
+    {
+      "epoch": 0.9106177467022274,
+      "grad_norm": 0.115048049968701,
+      "learning_rate": 4.161460241140602e-06,
+      "loss": 1.2935,
+      "step": 12633
+    },
+    {
+      "epoch": 0.9106898291645643,
+      "grad_norm": 0.12256125770385815,
+      "learning_rate": 4.154797765807472e-06,
+      "loss": 1.2802,
+      "step": 12634
+    },
+    {
+      "epoch": 0.9107619116269011,
+      "grad_norm": 0.11742106068747711,
+      "learning_rate": 4.148140514895338e-06,
+      "loss": 1.275,
+      "step": 12635
+    },
+    {
+      "epoch": 0.910833994089238,
+      "grad_norm": 0.13576026845856623,
+      "learning_rate": 4.141488488767042e-06,
+      "loss": 1.2732,
+      "step": 12636
+    },
+    {
+      "epoch": 0.910906076551575,
+      "grad_norm": 0.12691323780117691,
+      "learning_rate": 4.134841687785207e-06,
+      "loss": 1.3089,
+      "step": 12637
+    },
+    {
+      "epoch": 0.9109781590139119,
+      "grad_norm": 0.1390361117718763,
+      "learning_rate": 4.128200112312119e-06,
+      "loss": 1.2377,
+      "step": 12638
+    },
+    {
+      "epoch": 0.9110502414762488,
+      "grad_norm": 0.12535297253982772,
+      "learning_rate": 4.121563762709846e-06,
+      "loss": 1.2375,
+      "step": 12639
+    },
+    {
+      "epoch": 0.9111223239385857,
+      "grad_norm": 0.11570382773592326,
+      "learning_rate": 4.114932639340074e-06,
+      "loss": 1.2133,
+      "step": 12640
+    },
+    {
+      "epoch": 0.9111944064009226,
+      "grad_norm": 0.1244525654235654,
+      "learning_rate": 4.1083067425643054e-06,
+      "loss": 1.3179,
+      "step": 12641
+    },
+    {
+      "epoch": 0.9112664888632596,
+      "grad_norm": 0.12029137527868337,
+      "learning_rate": 4.1016860727436826e-06,
+      "loss": 1.2829,
+      "step": 12642
+    },
+    {
+      "epoch": 0.9113385713255965,
+      "grad_norm": 0.10893842857818826,
+      "learning_rate": 4.095070630239095e-06,
+      "loss": 1.4201,
+      "step": 12643
+    },
+    {
+      "epoch": 0.9114106537879334,
+      "grad_norm": 0.1143015578351742,
+      "learning_rate": 4.088460415411155e-06,
+      "loss": 1.3247,
+      "step": 12644
+    },
+    {
+      "epoch": 0.9114827362502703,
+      "grad_norm": 0.11974366256582955,
+      "learning_rate": 4.081855428620174e-06,
+      "loss": 1.2775,
+      "step": 12645
+    },
+    {
+      "epoch": 0.9115548187126072,
+      "grad_norm": 0.11874022409124303,
+      "learning_rate": 4.075255670226175e-06,
+      "loss": 1.2738,
+      "step": 12646
+    },
+    {
+      "epoch": 0.9116269011749442,
+      "grad_norm": 0.12470986398856565,
+      "learning_rate": 4.068661140588903e-06,
+      "loss": 1.2461,
+      "step": 12647
+    },
+    {
+      "epoch": 0.9116989836372811,
+      "grad_norm": 0.11740440866851633,
+      "learning_rate": 4.062071840067838e-06,
+      "loss": 1.2549,
+      "step": 12648
+    },
+    {
+      "epoch": 0.911771066099618,
+      "grad_norm": 0.1298116936011028,
+      "learning_rate": 4.055487769022137e-06,
+      "loss": 1.2676,
+      "step": 12649
+    },
+    {
+      "epoch": 0.9118431485619549,
+      "grad_norm": 0.12258171264300179,
+      "learning_rate": 4.04890892781069e-06,
+      "loss": 1.1911,
+      "step": 12650
+    },
+    {
+      "epoch": 0.9119152310242918,
+      "grad_norm": 0.12361837159551048,
+      "learning_rate": 4.042335316792123e-06,
+      "loss": 1.2311,
+      "step": 12651
+    },
+    {
+      "epoch": 0.9119873134866286,
+      "grad_norm": 0.110744438521004,
+      "learning_rate": 4.035766936324736e-06,
+      "loss": 1.2127,
+      "step": 12652
+    },
+    {
+      "epoch": 0.9120593959489656,
+      "grad_norm": 0.129944849497318,
+      "learning_rate": 4.029203786766567e-06,
+      "loss": 1.2557,
+      "step": 12653
+    },
+    {
+      "epoch": 0.9121314784113025,
+      "grad_norm": 0.11159187866485551,
+      "learning_rate": 4.022645868475372e-06,
+      "loss": 1.3454,
+      "step": 12654
+    },
+    {
+      "epoch": 0.9122035608736394,
+      "grad_norm": 0.12371184897282685,
+      "learning_rate": 4.016093181808623e-06,
+      "loss": 1.1903,
+      "step": 12655
+    },
+    {
+      "epoch": 0.9122756433359763,
+      "grad_norm": 0.1400456225104207,
+      "learning_rate": 4.009545727123487e-06,
+      "loss": 1.2926,
+      "step": 12656
+    },
+    {
+      "epoch": 0.9123477257983132,
+      "grad_norm": 0.12673882049915153,
+      "learning_rate": 4.00300350477687e-06,
+      "loss": 1.3482,
+      "step": 12657
+    },
+    {
+      "epoch": 0.9124198082606502,
+      "grad_norm": 0.12187766208780185,
+      "learning_rate": 3.996466515125363e-06,
+      "loss": 1.2252,
+      "step": 12658
+    },
+    {
+      "epoch": 0.9124918907229871,
+      "grad_norm": 0.11749919284031618,
+      "learning_rate": 3.989934758525315e-06,
+      "loss": 1.2818,
+      "step": 12659
+    },
+    {
+      "epoch": 0.912563973185324,
+      "grad_norm": 0.11546257356930577,
+      "learning_rate": 3.983408235332753e-06,
+      "loss": 1.2351,
+      "step": 12660
+    },
+    {
+      "epoch": 0.9126360556476609,
+      "grad_norm": 0.12201749097188394,
+      "learning_rate": 3.9768869459034245e-06,
+      "loss": 1.2346,
+      "step": 12661
+    },
+    {
+      "epoch": 0.9127081381099978,
+      "grad_norm": 0.12434800604487113,
+      "learning_rate": 3.9703708905928005e-06,
+      "loss": 1.1998,
+      "step": 12662
+    },
+    {
+      "epoch": 0.9127802205723348,
+      "grad_norm": 0.13388068306876652,
+      "learning_rate": 3.963860069756075e-06,
+      "loss": 1.252,
+      "step": 12663
+    },
+    {
+      "epoch": 0.9128523030346717,
+      "grad_norm": 0.11779077367002215,
+      "learning_rate": 3.957354483748154e-06,
+      "loss": 1.2699,
+      "step": 12664
+    },
+    {
+      "epoch": 0.9129243854970086,
+      "grad_norm": 0.12246896536918742,
+      "learning_rate": 3.95085413292361e-06,
+      "loss": 1.3573,
+      "step": 12665
+    },
+    {
+      "epoch": 0.9129964679593455,
+      "grad_norm": 0.11903394603238339,
+      "learning_rate": 3.944359017636812e-06,
+      "loss": 1.2689,
+      "step": 12666
+    },
+    {
+      "epoch": 0.9130685504216824,
+      "grad_norm": 0.11061819578328999,
+      "learning_rate": 3.9378691382417784e-06,
+      "loss": 1.2793,
+      "step": 12667
+    },
+    {
+      "epoch": 0.9131406328840194,
+      "grad_norm": 0.1269506921527756,
+      "learning_rate": 3.931384495092283e-06,
+      "loss": 1.3431,
+      "step": 12668
+    },
+    {
+      "epoch": 0.9132127153463563,
+      "grad_norm": 0.13305421148671473,
+      "learning_rate": 3.924905088541775e-06,
+      "loss": 1.2674,
+      "step": 12669
+    },
+    {
+      "epoch": 0.9132847978086931,
+      "grad_norm": 0.12006960537117371,
+      "learning_rate": 3.91843091894345e-06,
+      "loss": 1.2788,
+      "step": 12670
+    },
+    {
+      "epoch": 0.91335688027103,
+      "grad_norm": 0.12160309362463957,
+      "learning_rate": 3.911961986650226e-06,
+      "loss": 1.3352,
+      "step": 12671
+    },
+    {
+      "epoch": 0.9134289627333669,
+      "grad_norm": 0.1399501289962117,
+      "learning_rate": 3.905498292014698e-06,
+      "loss": 1.2527,
+      "step": 12672
+    },
+    {
+      "epoch": 0.9135010451957039,
+      "grad_norm": 0.11331175015125362,
+      "learning_rate": 3.899039835389196e-06,
+      "loss": 1.3345,
+      "step": 12673
+    },
+    {
+      "epoch": 0.9135731276580408,
+      "grad_norm": 0.12038366962347845,
+      "learning_rate": 3.89258661712576e-06,
+      "loss": 1.3097,
+      "step": 12674
+    },
+    {
+      "epoch": 0.9136452101203777,
+      "grad_norm": 0.11679852419083188,
+      "learning_rate": 3.886138637576153e-06,
+      "loss": 1.3176,
+      "step": 12675
+    },
+    {
+      "epoch": 0.9137172925827146,
+      "grad_norm": 0.11621855823393566,
+      "learning_rate": 3.879695897091873e-06,
+      "loss": 1.2216,
+      "step": 12676
+    },
+    {
+      "epoch": 0.9137893750450515,
+      "grad_norm": 0.12283108517748567,
+      "learning_rate": 3.87325839602406e-06,
+      "loss": 1.2874,
+      "step": 12677
+    },
+    {
+      "epoch": 0.9138614575073885,
+      "grad_norm": 0.13142917350808295,
+      "learning_rate": 3.866826134723633e-06,
+      "loss": 1.1954,
+      "step": 12678
+    },
+    {
+      "epoch": 0.9139335399697254,
+      "grad_norm": 0.12122612416956831,
+      "learning_rate": 3.860399113541235e-06,
+      "loss": 1.223,
+      "step": 12679
+    },
+    {
+      "epoch": 0.9140056224320623,
+      "grad_norm": 0.11603732175053286,
+      "learning_rate": 3.8539773328271745e-06,
+      "loss": 1.246,
+      "step": 12680
+    },
+    {
+      "epoch": 0.9140777048943992,
+      "grad_norm": 0.11617461472506209,
+      "learning_rate": 3.847560792931482e-06,
+      "loss": 1.3648,
+      "step": 12681
+    },
+    {
+      "epoch": 0.9141497873567361,
+      "grad_norm": 0.13406638503105098,
+      "learning_rate": 3.841149494203933e-06,
+      "loss": 1.2178,
+      "step": 12682
+    },
+    {
+      "epoch": 0.914221869819073,
+      "grad_norm": 0.11041139989755805,
+      "learning_rate": 3.834743436993993e-06,
+      "loss": 1.2074,
+      "step": 12683
+    },
+    {
+      "epoch": 0.91429395228141,
+      "grad_norm": 0.14407138545863038,
+      "learning_rate": 3.828342621650882e-06,
+      "loss": 1.3077,
+      "step": 12684
+    },
+    {
+      "epoch": 0.9143660347437469,
+      "grad_norm": 0.11798153612572927,
+      "learning_rate": 3.8219470485234555e-06,
+      "loss": 1.3054,
+      "step": 12685
+    },
+    {
+      "epoch": 0.9144381172060838,
+      "grad_norm": 0.12046842435373238,
+      "learning_rate": 3.815556717960345e-06,
+      "loss": 1.2577,
+      "step": 12686
+    },
+    {
+      "epoch": 0.9145101996684206,
+      "grad_norm": 0.12627353693575447,
+      "learning_rate": 3.8091716303098714e-06,
+      "loss": 1.2218,
+      "step": 12687
+    },
+    {
+      "epoch": 0.9145822821307575,
+      "grad_norm": 0.12491289113322324,
+      "learning_rate": 3.8027917859201234e-06,
+      "loss": 1.266,
+      "step": 12688
+    },
+    {
+      "epoch": 0.9146543645930945,
+      "grad_norm": 0.12304839707378515,
+      "learning_rate": 3.796417185138812e-06,
+      "loss": 1.3006,
+      "step": 12689
+    },
+    {
+      "epoch": 0.9147264470554314,
+      "grad_norm": 0.11917835660305336,
+      "learning_rate": 3.7900478283134255e-06,
+      "loss": 1.2437,
+      "step": 12690
+    },
+    {
+      "epoch": 0.9147985295177683,
+      "grad_norm": 0.1264589918978637,
+      "learning_rate": 3.783683715791142e-06,
+      "loss": 1.3453,
+      "step": 12691
+    },
+    {
+      "epoch": 0.9148706119801052,
+      "grad_norm": 0.14409667385021674,
+      "learning_rate": 3.777324847918895e-06,
+      "loss": 1.3199,
+      "step": 12692
+    },
+    {
+      "epoch": 0.9149426944424421,
+      "grad_norm": 0.12671153695867318,
+      "learning_rate": 3.7709712250432627e-06,
+      "loss": 1.0988,
+      "step": 12693
+    },
+    {
+      "epoch": 0.9150147769047791,
+      "grad_norm": 0.10835001102173372,
+      "learning_rate": 3.7646228475105905e-06,
+      "loss": 1.293,
+      "step": 12694
+    },
+    {
+      "epoch": 0.915086859367116,
+      "grad_norm": 0.11772814487387873,
+      "learning_rate": 3.7582797156669125e-06,
+      "loss": 1.1818,
+      "step": 12695
+    },
+    {
+      "epoch": 0.9151589418294529,
+      "grad_norm": 0.11327834513595982,
+      "learning_rate": 3.7519418298579966e-06,
+      "loss": 1.2904,
+      "step": 12696
+    },
+    {
+      "epoch": 0.9152310242917898,
+      "grad_norm": 0.13027380421216184,
+      "learning_rate": 3.7456091904293113e-06,
+      "loss": 1.3675,
+      "step": 12697
+    },
+    {
+      "epoch": 0.9153031067541267,
+      "grad_norm": 0.12394881131882035,
+      "learning_rate": 3.7392817977260464e-06,
+      "loss": 1.3111,
+      "step": 12698
+    },
+    {
+      "epoch": 0.9153751892164637,
+      "grad_norm": 0.1279395831666903,
+      "learning_rate": 3.732959652093093e-06,
+      "loss": 1.2307,
+      "step": 12699
+    },
+    {
+      "epoch": 0.9154472716788006,
+      "grad_norm": 0.1185352262655508,
+      "learning_rate": 3.7266427538750647e-06,
+      "loss": 1.259,
+      "step": 12700
+    },
+    {
+      "epoch": 0.9155193541411375,
+      "grad_norm": 0.11612127808845552,
+      "learning_rate": 3.720331103416297e-06,
+      "loss": 1.2506,
+      "step": 12701
+    },
+    {
+      "epoch": 0.9155914366034744,
+      "grad_norm": 0.12956201230888909,
+      "learning_rate": 3.7140247010608255e-06,
+      "loss": 1.2588,
+      "step": 12702
+    },
+    {
+      "epoch": 0.9156635190658113,
+      "grad_norm": 0.11944981759833538,
+      "learning_rate": 3.7077235471523974e-06,
+      "loss": 1.3629,
+      "step": 12703
+    },
+    {
+      "epoch": 0.9157356015281481,
+      "grad_norm": 0.11438152126662122,
+      "learning_rate": 3.701427642034505e-06,
+      "loss": 1.4054,
+      "step": 12704
+    },
+    {
+      "epoch": 0.9158076839904851,
+      "grad_norm": 0.11345694734113804,
+      "learning_rate": 3.695136986050307e-06,
+      "loss": 1.2341,
+      "step": 12705
+    },
+    {
+      "epoch": 0.915879766452822,
+      "grad_norm": 0.1208836632659667,
+      "learning_rate": 3.688851579542718e-06,
+      "loss": 1.2616,
+      "step": 12706
+    },
+    {
+      "epoch": 0.9159518489151589,
+      "grad_norm": 0.12899314092269754,
+      "learning_rate": 3.6825714228543416e-06,
+      "loss": 1.2607,
+      "step": 12707
+    },
+    {
+      "epoch": 0.9160239313774958,
+      "grad_norm": 0.12835176509410184,
+      "learning_rate": 3.6762965163275044e-06,
+      "loss": 1.2072,
+      "step": 12708
+    },
+    {
+      "epoch": 0.9160960138398327,
+      "grad_norm": 0.1281469589173964,
+      "learning_rate": 3.6700268603042433e-06,
+      "loss": 1.2833,
+      "step": 12709
+    },
+    {
+      "epoch": 0.9161680963021697,
+      "grad_norm": 0.12215874627521088,
+      "learning_rate": 3.6637624551263073e-06,
+      "loss": 1.2722,
+      "step": 12710
+    },
+    {
+      "epoch": 0.9162401787645066,
+      "grad_norm": 0.11802475793356564,
+      "learning_rate": 3.657503301135168e-06,
+      "loss": 1.2863,
+      "step": 12711
+    },
+    {
+      "epoch": 0.9163122612268435,
+      "grad_norm": 0.11856162711185761,
+      "learning_rate": 3.6512493986719963e-06,
+      "loss": 1.2549,
+      "step": 12712
+    },
+    {
+      "epoch": 0.9163843436891804,
+      "grad_norm": 0.18260606732293505,
+      "learning_rate": 3.6450007480777093e-06,
+      "loss": 1.2966,
+      "step": 12713
+    },
+    {
+      "epoch": 0.9164564261515173,
+      "grad_norm": 0.11093851723010625,
+      "learning_rate": 3.6387573496928784e-06,
+      "loss": 1.2726,
+      "step": 12714
+    },
+    {
+      "epoch": 0.9165285086138543,
+      "grad_norm": 0.12250648430666056,
+      "learning_rate": 3.6325192038578536e-06,
+      "loss": 1.2895,
+      "step": 12715
+    },
+    {
+      "epoch": 0.9166005910761912,
+      "grad_norm": 0.12067752987582743,
+      "learning_rate": 3.626286310912663e-06,
+      "loss": 1.349,
+      "step": 12716
+    },
+    {
+      "epoch": 0.9166726735385281,
+      "grad_norm": 0.11820760538228982,
+      "learning_rate": 3.6200586711970685e-06,
+      "loss": 1.3144,
+      "step": 12717
+    },
+    {
+      "epoch": 0.916744756000865,
+      "grad_norm": 0.11718490016837609,
+      "learning_rate": 3.613836285050487e-06,
+      "loss": 1.1431,
+      "step": 12718
+    },
+    {
+      "epoch": 0.9168168384632019,
+      "grad_norm": 0.11022844664399696,
+      "learning_rate": 3.607619152812125e-06,
+      "loss": 1.2301,
+      "step": 12719
+    },
+    {
+      "epoch": 0.9168889209255389,
+      "grad_norm": 0.11506048250795452,
+      "learning_rate": 3.601407274820878e-06,
+      "loss": 1.2072,
+      "step": 12720
+    },
+    {
+      "epoch": 0.9169610033878758,
+      "grad_norm": 0.1219700920478559,
+      "learning_rate": 3.5952006514153645e-06,
+      "loss": 1.232,
+      "step": 12721
+    },
+    {
+      "epoch": 0.9170330858502126,
+      "grad_norm": 0.14165693386359962,
+      "learning_rate": 3.588999282933836e-06,
+      "loss": 1.2806,
+      "step": 12722
+    },
+    {
+      "epoch": 0.9171051683125495,
+      "grad_norm": 0.12009082832946888,
+      "learning_rate": 3.582803169714388e-06,
+      "loss": 1.323,
+      "step": 12723
+    },
+    {
+      "epoch": 0.9171772507748864,
+      "grad_norm": 0.12413368716811721,
+      "learning_rate": 3.576612312094729e-06,
+      "loss": 1.3193,
+      "step": 12724
+    },
+    {
+      "epoch": 0.9172493332372234,
+      "grad_norm": 0.10620019416868291,
+      "learning_rate": 3.5704267104123444e-06,
+      "loss": 1.2499,
+      "step": 12725
+    },
+    {
+      "epoch": 0.9173214156995603,
+      "grad_norm": 0.1330549624160414,
+      "learning_rate": 3.5642463650043644e-06,
+      "loss": 1.3047,
+      "step": 12726
+    },
+    {
+      "epoch": 0.9173934981618972,
+      "grad_norm": 0.12125491551213162,
+      "learning_rate": 3.5580712762077084e-06,
+      "loss": 1.3502,
+      "step": 12727
+    },
+    {
+      "epoch": 0.9174655806242341,
+      "grad_norm": 0.12445240382793114,
+      "learning_rate": 3.551901444358974e-06,
+      "loss": 1.2142,
+      "step": 12728
+    },
+    {
+      "epoch": 0.917537663086571,
+      "grad_norm": 0.12465924111564067,
+      "learning_rate": 3.545736869794458e-06,
+      "loss": 1.2799,
+      "step": 12729
+    },
+    {
+      "epoch": 0.917609745548908,
+      "grad_norm": 0.10865593558814976,
+      "learning_rate": 3.5395775528501706e-06,
+      "loss": 1.2676,
+      "step": 12730
+    },
+    {
+      "epoch": 0.9176818280112449,
+      "grad_norm": 0.12000435063878002,
+      "learning_rate": 3.533423493861865e-06,
+      "loss": 1.2374,
+      "step": 12731
+    },
+    {
+      "epoch": 0.9177539104735818,
+      "grad_norm": 0.1106686726239581,
+      "learning_rate": 3.527274693165017e-06,
+      "loss": 1.258,
+      "step": 12732
+    },
+    {
+      "epoch": 0.9178259929359187,
+      "grad_norm": 0.3899600406894309,
+      "learning_rate": 3.5211311510947475e-06,
+      "loss": 1.303,
+      "step": 12733
+    },
+    {
+      "epoch": 0.9178980753982556,
+      "grad_norm": 0.13311171897889593,
+      "learning_rate": 3.5149928679859555e-06,
+      "loss": 1.241,
+      "step": 12734
+    },
+    {
+      "epoch": 0.9179701578605926,
+      "grad_norm": 0.12162970180966133,
+      "learning_rate": 3.508859844173229e-06,
+      "loss": 1.2386,
+      "step": 12735
+    },
+    {
+      "epoch": 0.9180422403229295,
+      "grad_norm": 0.11493589892773699,
+      "learning_rate": 3.5027320799909e-06,
+      "loss": 1.2297,
+      "step": 12736
+    },
+    {
+      "epoch": 0.9181143227852664,
+      "grad_norm": 0.12467790643649514,
+      "learning_rate": 3.4966095757729356e-06,
+      "loss": 1.3269,
+      "step": 12737
+    },
+    {
+      "epoch": 0.9181864052476033,
+      "grad_norm": 0.12011611510679737,
+      "learning_rate": 3.490492331853101e-06,
+      "loss": 1.1917,
+      "step": 12738
+    },
+    {
+      "epoch": 0.9182584877099401,
+      "grad_norm": 0.11959875468354088,
+      "learning_rate": 3.4843803485648417e-06,
+      "loss": 1.2077,
+      "step": 12739
+    },
+    {
+      "epoch": 0.918330570172277,
+      "grad_norm": 0.12055512138428091,
+      "learning_rate": 3.478273626241302e-06,
+      "loss": 1.2347,
+      "step": 12740
+    },
+    {
+      "epoch": 0.918402652634614,
+      "grad_norm": 0.11942080141765944,
+      "learning_rate": 3.472172165215348e-06,
+      "loss": 1.3029,
+      "step": 12741
+    },
+    {
+      "epoch": 0.9184747350969509,
+      "grad_norm": 0.11456660292458586,
+      "learning_rate": 3.466075965819582e-06,
+      "loss": 1.1075,
+      "step": 12742
+    },
+    {
+      "epoch": 0.9185468175592878,
+      "grad_norm": 0.12324119160328667,
+      "learning_rate": 3.459985028386292e-06,
+      "loss": 1.2166,
+      "step": 12743
+    },
+    {
+      "epoch": 0.9186189000216247,
+      "grad_norm": 0.12503315676011686,
+      "learning_rate": 3.4538993532474918e-06,
+      "loss": 1.2603,
+      "step": 12744
+    },
+    {
+      "epoch": 0.9186909824839616,
+      "grad_norm": 0.12708129347947977,
+      "learning_rate": 3.447818940734904e-06,
+      "loss": 1.2898,
+      "step": 12745
+    },
+    {
+      "epoch": 0.9187630649462986,
+      "grad_norm": 0.11724184893709828,
+      "learning_rate": 3.4417437911799523e-06,
+      "loss": 1.2587,
+      "step": 12746
+    },
+    {
+      "epoch": 0.9188351474086355,
+      "grad_norm": 0.11503704990511177,
+      "learning_rate": 3.4356739049138166e-06,
+      "loss": 1.1964,
+      "step": 12747
+    },
+    {
+      "epoch": 0.9189072298709724,
+      "grad_norm": 0.11570192300235396,
+      "learning_rate": 3.4296092822673322e-06,
+      "loss": 1.2713,
+      "step": 12748
+    },
+    {
+      "epoch": 0.9189793123333093,
+      "grad_norm": 0.12024458974743159,
+      "learning_rate": 3.4235499235710788e-06,
+      "loss": 1.2967,
+      "step": 12749
+    },
+    {
+      "epoch": 0.9190513947956462,
+      "grad_norm": 0.11890989226418812,
+      "learning_rate": 3.4174958291553593e-06,
+      "loss": 1.1756,
+      "step": 12750
+    },
+    {
+      "epoch": 0.9191234772579832,
+      "grad_norm": 0.11791598510306942,
+      "learning_rate": 3.411446999350176e-06,
+      "loss": 1.2755,
+      "step": 12751
+    },
+    {
+      "epoch": 0.9191955597203201,
+      "grad_norm": 0.11823956601565266,
+      "learning_rate": 3.4054034344852326e-06,
+      "loss": 1.2595,
+      "step": 12752
+    },
+    {
+      "epoch": 0.919267642182657,
+      "grad_norm": 0.11977544622416043,
+      "learning_rate": 3.3993651348899538e-06,
+      "loss": 1.2985,
+      "step": 12753
+    },
+    {
+      "epoch": 0.9193397246449939,
+      "grad_norm": 0.12292515348605502,
+      "learning_rate": 3.3933321008934984e-06,
+      "loss": 1.2603,
+      "step": 12754
+    },
+    {
+      "epoch": 0.9194118071073308,
+      "grad_norm": 0.12419771605235642,
+      "learning_rate": 3.3873043328247032e-06,
+      "loss": 1.264,
+      "step": 12755
+    },
+    {
+      "epoch": 0.9194838895696676,
+      "grad_norm": 0.10627417414556013,
+      "learning_rate": 3.3812818310121507e-06,
+      "loss": 1.2658,
+      "step": 12756
+    },
+    {
+      "epoch": 0.9195559720320046,
+      "grad_norm": 0.11684774830442621,
+      "learning_rate": 3.3752645957841112e-06,
+      "loss": 1.1662,
+      "step": 12757
+    },
+    {
+      "epoch": 0.9196280544943415,
+      "grad_norm": 0.11415783304889757,
+      "learning_rate": 3.3692526274685776e-06,
+      "loss": 1.2191,
+      "step": 12758
+    },
+    {
+      "epoch": 0.9197001369566784,
+      "grad_norm": 0.10796808964388845,
+      "learning_rate": 3.3632459263932657e-06,
+      "loss": 1.2182,
+      "step": 12759
+    },
+    {
+      "epoch": 0.9197722194190153,
+      "grad_norm": 0.1210965466682326,
+      "learning_rate": 3.357244492885592e-06,
+      "loss": 1.2283,
+      "step": 12760
+    },
+    {
+      "epoch": 0.9198443018813522,
+      "grad_norm": 0.12030752734956839,
+      "learning_rate": 3.3512483272726826e-06,
+      "loss": 1.299,
+      "step": 12761
+    },
+    {
+      "epoch": 0.9199163843436892,
+      "grad_norm": 0.18670996675436677,
+      "learning_rate": 3.3452574298813653e-06,
+      "loss": 1.2595,
+      "step": 12762
+    },
+    {
+      "epoch": 0.9199884668060261,
+      "grad_norm": 0.12732525819850243,
+      "learning_rate": 3.339271801038235e-06,
+      "loss": 1.3251,
+      "step": 12763
+    },
+    {
+      "epoch": 0.920060549268363,
+      "grad_norm": 0.1667695416836441,
+      "learning_rate": 3.3332914410695415e-06,
+      "loss": 1.2708,
+      "step": 12764
+    },
+    {
+      "epoch": 0.9201326317306999,
+      "grad_norm": 0.1209115604231939,
+      "learning_rate": 3.327316350301268e-06,
+      "loss": 1.1834,
+      "step": 12765
+    },
+    {
+      "epoch": 0.9202047141930368,
+      "grad_norm": 0.11676126111605656,
+      "learning_rate": 3.3213465290590996e-06,
+      "loss": 1.233,
+      "step": 12766
+    },
+    {
+      "epoch": 0.9202767966553738,
+      "grad_norm": 0.1315598611803584,
+      "learning_rate": 3.315381977668475e-06,
+      "loss": 1.2768,
+      "step": 12767
+    },
+    {
+      "epoch": 0.9203488791177107,
+      "grad_norm": 0.1237033109933592,
+      "learning_rate": 3.3094226964544893e-06,
+      "loss": 1.2181,
+      "step": 12768
+    },
+    {
+      "epoch": 0.9204209615800476,
+      "grad_norm": 0.12473999833031862,
+      "learning_rate": 3.303468685741995e-06,
+      "loss": 1.3356,
+      "step": 12769
+    },
+    {
+      "epoch": 0.9204930440423845,
+      "grad_norm": 0.1270917362328902,
+      "learning_rate": 3.2975199458555207e-06,
+      "loss": 1.2237,
+      "step": 12770
+    },
+    {
+      "epoch": 0.9205651265047214,
+      "grad_norm": 0.11766280676609372,
+      "learning_rate": 3.2915764771193292e-06,
+      "loss": 1.2325,
+      "step": 12771
+    },
+    {
+      "epoch": 0.9206372089670584,
+      "grad_norm": 0.11757287036760997,
+      "learning_rate": 3.2856382798573947e-06,
+      "loss": 1.2285,
+      "step": 12772
+    },
+    {
+      "epoch": 0.9207092914293953,
+      "grad_norm": 0.11270783559348514,
+      "learning_rate": 3.2797053543934255e-06,
+      "loss": 1.1705,
+      "step": 12773
+    },
+    {
+      "epoch": 0.9207813738917321,
+      "grad_norm": 0.12175821831988622,
+      "learning_rate": 3.273777701050795e-06,
+      "loss": 1.2631,
+      "step": 12774
+    },
+    {
+      "epoch": 0.920853456354069,
+      "grad_norm": 0.11995173886480348,
+      "learning_rate": 3.2678553201525907e-06,
+      "loss": 1.288,
+      "step": 12775
+    },
+    {
+      "epoch": 0.9209255388164059,
+      "grad_norm": 0.11241699757141987,
+      "learning_rate": 3.2619382120216868e-06,
+      "loss": 1.1257,
+      "step": 12776
+    },
+    {
+      "epoch": 0.9209976212787429,
+      "grad_norm": 0.13030461602454335,
+      "learning_rate": 3.256026376980592e-06,
+      "loss": 1.2576,
+      "step": 12777
+    },
+    {
+      "epoch": 0.9210697037410798,
+      "grad_norm": 0.1325041414619262,
+      "learning_rate": 3.2501198153515598e-06,
+      "loss": 1.2202,
+      "step": 12778
+    },
+    {
+      "epoch": 0.9211417862034167,
+      "grad_norm": 0.11773860834994414,
+      "learning_rate": 3.2442185274565217e-06,
+      "loss": 1.2542,
+      "step": 12779
+    },
+    {
+      "epoch": 0.9212138686657536,
+      "grad_norm": 0.15113113726401128,
+      "learning_rate": 3.238322513617198e-06,
+      "loss": 1.2845,
+      "step": 12780
+    },
+    {
+      "epoch": 0.9212859511280905,
+      "grad_norm": 0.12931244552362678,
+      "learning_rate": 3.2324317741549537e-06,
+      "loss": 1.302,
+      "step": 12781
+    },
+    {
+      "epoch": 0.9213580335904274,
+      "grad_norm": 0.1218816840982623,
+      "learning_rate": 3.2265463093908766e-06,
+      "loss": 1.294,
+      "step": 12782
+    },
+    {
+      "epoch": 0.9214301160527644,
+      "grad_norm": 0.11776456652296295,
+      "learning_rate": 3.2206661196457876e-06,
+      "loss": 1.2435,
+      "step": 12783
+    },
+    {
+      "epoch": 0.9215021985151013,
+      "grad_norm": 0.1229073913661809,
+      "learning_rate": 3.2147912052401972e-06,
+      "loss": 1.2654,
+      "step": 12784
+    },
+    {
+      "epoch": 0.9215742809774382,
+      "grad_norm": 0.11432635890597997,
+      "learning_rate": 3.2089215664943716e-06,
+      "loss": 1.1866,
+      "step": 12785
+    },
+    {
+      "epoch": 0.9216463634397751,
+      "grad_norm": 0.11885053315319445,
+      "learning_rate": 3.2030572037282325e-06,
+      "loss": 1.2802,
+      "step": 12786
+    },
+    {
+      "epoch": 0.921718445902112,
+      "grad_norm": 0.13126340104798842,
+      "learning_rate": 3.1971981172614464e-06,
+      "loss": 1.2696,
+      "step": 12787
+    },
+    {
+      "epoch": 0.921790528364449,
+      "grad_norm": 0.12296975270184284,
+      "learning_rate": 3.1913443074133687e-06,
+      "loss": 1.3067,
+      "step": 12788
+    },
+    {
+      "epoch": 0.9218626108267859,
+      "grad_norm": 0.12677621068494435,
+      "learning_rate": 3.1854957745031223e-06,
+      "loss": 1.3752,
+      "step": 12789
+    },
+    {
+      "epoch": 0.9219346932891228,
+      "grad_norm": 0.18565815634980162,
+      "learning_rate": 3.1796525188494743e-06,
+      "loss": 1.3067,
+      "step": 12790
+    },
+    {
+      "epoch": 0.9220067757514596,
+      "grad_norm": 0.13362880645248895,
+      "learning_rate": 3.1738145407709476e-06,
+      "loss": 1.1962,
+      "step": 12791
+    },
+    {
+      "epoch": 0.9220788582137965,
+      "grad_norm": 0.11990020007801058,
+      "learning_rate": 3.1679818405857544e-06,
+      "loss": 1.1914,
+      "step": 12792
+    },
+    {
+      "epoch": 0.9221509406761335,
+      "grad_norm": 0.1327450163935354,
+      "learning_rate": 3.1621544186118403e-06,
+      "loss": 1.2781,
+      "step": 12793
+    },
+    {
+      "epoch": 0.9222230231384704,
+      "grad_norm": 0.11532208145633942,
+      "learning_rate": 3.15633227516684e-06,
+      "loss": 1.1772,
+      "step": 12794
+    },
+    {
+      "epoch": 0.9222951056008073,
+      "grad_norm": 0.12350617643660419,
+      "learning_rate": 3.1505154105681223e-06,
+      "loss": 1.3229,
+      "step": 12795
+    },
+    {
+      "epoch": 0.9223671880631442,
+      "grad_norm": 0.13229727576273204,
+      "learning_rate": 3.1447038251327554e-06,
+      "loss": 1.2474,
+      "step": 12796
+    },
+    {
+      "epoch": 0.9224392705254811,
+      "grad_norm": 0.1211061754199928,
+      "learning_rate": 3.1388975191775196e-06,
+      "loss": 1.2863,
+      "step": 12797
+    },
+    {
+      "epoch": 0.9225113529878181,
+      "grad_norm": 0.12928923369282552,
+      "learning_rate": 3.1330964930189168e-06,
+      "loss": 1.2431,
+      "step": 12798
+    },
+    {
+      "epoch": 0.922583435450155,
+      "grad_norm": 0.11795205161292623,
+      "learning_rate": 3.12730074697315e-06,
+      "loss": 1.2122,
+      "step": 12799
+    },
+    {
+      "epoch": 0.9226555179124919,
+      "grad_norm": 0.12333597561632446,
+      "learning_rate": 3.1215102813561435e-06,
+      "loss": 1.2899,
+      "step": 12800
+    },
+    {
+      "epoch": 0.9227276003748288,
+      "grad_norm": 0.1327859397728355,
+      "learning_rate": 3.1157250964835237e-06,
+      "loss": 1.2126,
+      "step": 12801
+    },
+    {
+      "epoch": 0.9227996828371657,
+      "grad_norm": 0.1200891241592412,
+      "learning_rate": 3.109945192670638e-06,
+      "loss": 1.227,
+      "step": 12802
+    },
+    {
+      "epoch": 0.9228717652995027,
+      "grad_norm": 0.11992180694006269,
+      "learning_rate": 3.1041705702325343e-06,
+      "loss": 1.2882,
+      "step": 12803
+    },
+    {
+      "epoch": 0.9229438477618396,
+      "grad_norm": 0.12320085616077905,
+      "learning_rate": 3.0984012294839937e-06,
+      "loss": 1.3023,
+      "step": 12804
+    },
+    {
+      "epoch": 0.9230159302241765,
+      "grad_norm": 0.13526240925137115,
+      "learning_rate": 3.0926371707394985e-06,
+      "loss": 1.3199,
+      "step": 12805
+    },
+    {
+      "epoch": 0.9230880126865134,
+      "grad_norm": 0.12566754984801906,
+      "learning_rate": 3.0868783943132305e-06,
+      "loss": 1.2691,
+      "step": 12806
+    },
+    {
+      "epoch": 0.9231600951488503,
+      "grad_norm": 0.12134638346770103,
+      "learning_rate": 3.0811249005191056e-06,
+      "loss": 1.2972,
+      "step": 12807
+    },
+    {
+      "epoch": 0.9232321776111871,
+      "grad_norm": 0.11075564032276525,
+      "learning_rate": 3.075376689670728e-06,
+      "loss": 1.3273,
+      "step": 12808
+    },
+    {
+      "epoch": 0.9233042600735241,
+      "grad_norm": 0.2208494331693833,
+      "learning_rate": 3.0696337620814364e-06,
+      "loss": 1.2783,
+      "step": 12809
+    },
+    {
+      "epoch": 0.923376342535861,
+      "grad_norm": 0.11610370354132939,
+      "learning_rate": 3.0638961180642687e-06,
+      "loss": 1.2703,
+      "step": 12810
+    },
+    {
+      "epoch": 0.9234484249981979,
+      "grad_norm": 0.12305062239367705,
+      "learning_rate": 3.0581637579319755e-06,
+      "loss": 1.2989,
+      "step": 12811
+    },
+    {
+      "epoch": 0.9235205074605348,
+      "grad_norm": 0.12060789517447905,
+      "learning_rate": 3.0524366819970287e-06,
+      "loss": 1.233,
+      "step": 12812
+    },
+    {
+      "epoch": 0.9235925899228717,
+      "grad_norm": 0.11757929216124076,
+      "learning_rate": 3.0467148905716003e-06,
+      "loss": 1.2765,
+      "step": 12813
+    },
+    {
+      "epoch": 0.9236646723852087,
+      "grad_norm": 0.1194126606054105,
+      "learning_rate": 3.040998383967586e-06,
+      "loss": 1.3365,
+      "step": 12814
+    },
+    {
+      "epoch": 0.9237367548475456,
+      "grad_norm": 0.13829866277615918,
+      "learning_rate": 3.0352871624965584e-06,
+      "loss": 1.3643,
+      "step": 12815
+    },
+    {
+      "epoch": 0.9238088373098825,
+      "grad_norm": 0.1357677391154202,
+      "learning_rate": 3.0295812264698687e-06,
+      "loss": 1.2857,
+      "step": 12816
+    },
+    {
+      "epoch": 0.9238809197722194,
+      "grad_norm": 0.12157790382326908,
+      "learning_rate": 3.023880576198512e-06,
+      "loss": 1.1864,
+      "step": 12817
+    },
+    {
+      "epoch": 0.9239530022345563,
+      "grad_norm": 0.12570515112521685,
+      "learning_rate": 3.018185211993252e-06,
+      "loss": 1.2577,
+      "step": 12818
+    },
+    {
+      "epoch": 0.9240250846968933,
+      "grad_norm": 0.11787911269263215,
+      "learning_rate": 3.012495134164506e-06,
+      "loss": 1.2039,
+      "step": 12819
+    },
+    {
+      "epoch": 0.9240971671592302,
+      "grad_norm": 0.1372202853599814,
+      "learning_rate": 3.006810343022459e-06,
+      "loss": 1.2869,
+      "step": 12820
+    },
+    {
+      "epoch": 0.9241692496215671,
+      "grad_norm": 0.12553225056009465,
+      "learning_rate": 3.0011308388769754e-06,
+      "loss": 1.2552,
+      "step": 12821
+    },
+    {
+      "epoch": 0.924241332083904,
+      "grad_norm": 0.12078323791580617,
+      "learning_rate": 2.9954566220376514e-06,
+      "loss": 1.245,
+      "step": 12822
+    },
+    {
+      "epoch": 0.9243134145462409,
+      "grad_norm": 0.1109276669817611,
+      "learning_rate": 2.989787692813739e-06,
+      "loss": 1.1993,
+      "step": 12823
+    },
+    {
+      "epoch": 0.9243854970085779,
+      "grad_norm": 0.12750117667175592,
+      "learning_rate": 2.984124051514292e-06,
+      "loss": 1.2331,
+      "step": 12824
+    },
+    {
+      "epoch": 0.9244575794709148,
+      "grad_norm": 0.11179343741332848,
+      "learning_rate": 2.9784656984480074e-06,
+      "loss": 1.2473,
+      "step": 12825
+    },
+    {
+      "epoch": 0.9245296619332516,
+      "grad_norm": 0.12147872878040614,
+      "learning_rate": 2.9728126339233385e-06,
+      "loss": 1.2672,
+      "step": 12826
+    },
+    {
+      "epoch": 0.9246017443955885,
+      "grad_norm": 0.12535979763531965,
+      "learning_rate": 2.9671648582483945e-06,
+      "loss": 1.2582,
+      "step": 12827
+    },
+    {
+      "epoch": 0.9246738268579254,
+      "grad_norm": 0.12388175461675605,
+      "learning_rate": 2.96152237173104e-06,
+      "loss": 1.3217,
+      "step": 12828
+    },
+    {
+      "epoch": 0.9247459093202623,
+      "grad_norm": 0.114552276170525,
+      "learning_rate": 2.9558851746788517e-06,
+      "loss": 1.2386,
+      "step": 12829
+    },
+    {
+      "epoch": 0.9248179917825993,
+      "grad_norm": 0.13243752769248182,
+      "learning_rate": 2.9502532673991278e-06,
+      "loss": 1.3294,
+      "step": 12830
+    },
+    {
+      "epoch": 0.9248900742449362,
+      "grad_norm": 0.11949140668789736,
+      "learning_rate": 2.944626650198812e-06,
+      "loss": 1.2344,
+      "step": 12831
+    },
+    {
+      "epoch": 0.9249621567072731,
+      "grad_norm": 0.1207024410544054,
+      "learning_rate": 2.9390053233846136e-06,
+      "loss": 1.2961,
+      "step": 12832
+    },
+    {
+      "epoch": 0.92503423916961,
+      "grad_norm": 0.11361411243702829,
+      "learning_rate": 2.9333892872629664e-06,
+      "loss": 1.354,
+      "step": 12833
+    },
+    {
+      "epoch": 0.925106321631947,
+      "grad_norm": 0.13705110589121205,
+      "learning_rate": 2.9277785421400027e-06,
+      "loss": 1.297,
+      "step": 12834
+    },
+    {
+      "epoch": 0.9251784040942839,
+      "grad_norm": 0.11511792950925115,
+      "learning_rate": 2.9221730883215337e-06,
+      "loss": 1.2077,
+      "step": 12835
+    },
+    {
+      "epoch": 0.9252504865566208,
+      "grad_norm": 0.1209612855456371,
+      "learning_rate": 2.9165729261131035e-06,
+      "loss": 1.2342,
+      "step": 12836
+    },
+    {
+      "epoch": 0.9253225690189577,
+      "grad_norm": 0.12002201163737376,
+      "learning_rate": 2.9109780558199794e-06,
+      "loss": 1.2894,
+      "step": 12837
+    },
+    {
+      "epoch": 0.9253946514812946,
+      "grad_norm": 0.11609806949385794,
+      "learning_rate": 2.9053884777471506e-06,
+      "loss": 1.211,
+      "step": 12838
+    },
+    {
+      "epoch": 0.9254667339436315,
+      "grad_norm": 0.11112261663157309,
+      "learning_rate": 2.8998041921992735e-06,
+      "loss": 1.1815,
+      "step": 12839
+    },
+    {
+      "epoch": 0.9255388164059685,
+      "grad_norm": 0.10508569142509197,
+      "learning_rate": 2.89422519948076e-06,
+      "loss": 1.1837,
+      "step": 12840
+    },
+    {
+      "epoch": 0.9256108988683054,
+      "grad_norm": 0.11490560819925522,
+      "learning_rate": 2.8886514998956894e-06,
+      "loss": 1.2547,
+      "step": 12841
+    },
+    {
+      "epoch": 0.9256829813306423,
+      "grad_norm": 0.1272328130952905,
+      "learning_rate": 2.883083093747907e-06,
+      "loss": 1.3747,
+      "step": 12842
+    },
+    {
+      "epoch": 0.9257550637929791,
+      "grad_norm": 0.12214627806072943,
+      "learning_rate": 2.877519981340926e-06,
+      "loss": 1.2809,
+      "step": 12843
+    },
+    {
+      "epoch": 0.925827146255316,
+      "grad_norm": 0.12615384376384797,
+      "learning_rate": 2.8719621629779927e-06,
+      "loss": 1.2279,
+      "step": 12844
+    },
+    {
+      "epoch": 0.925899228717653,
+      "grad_norm": 0.12504483113082357,
+      "learning_rate": 2.866409638962042e-06,
+      "loss": 1.3637,
+      "step": 12845
+    },
+    {
+      "epoch": 0.9259713111799899,
+      "grad_norm": 0.12389108971494876,
+      "learning_rate": 2.860862409595766e-06,
+      "loss": 1.2499,
+      "step": 12846
+    },
+    {
+      "epoch": 0.9260433936423268,
+      "grad_norm": 0.10964807188810281,
+      "learning_rate": 2.855320475181511e-06,
+      "loss": 1.2865,
+      "step": 12847
+    },
+    {
+      "epoch": 0.9261154761046637,
+      "grad_norm": 0.15658599070859403,
+      "learning_rate": 2.849783836021369e-06,
+      "loss": 1.2845,
+      "step": 12848
+    },
+    {
+      "epoch": 0.9261875585670006,
+      "grad_norm": 0.12545522378492344,
+      "learning_rate": 2.844252492417143e-06,
+      "loss": 1.291,
+      "step": 12849
+    },
+    {
+      "epoch": 0.9262596410293376,
+      "grad_norm": 0.12369452515495276,
+      "learning_rate": 2.8387264446703256e-06,
+      "loss": 1.286,
+      "step": 12850
+    },
+    {
+      "epoch": 0.9263317234916745,
+      "grad_norm": 0.11192949290729025,
+      "learning_rate": 2.833205693082164e-06,
+      "loss": 1.3222,
+      "step": 12851
+    },
+    {
+      "epoch": 0.9264038059540114,
+      "grad_norm": 0.1151548171822793,
+      "learning_rate": 2.827690237953562e-06,
+      "loss": 1.2399,
+      "step": 12852
+    },
+    {
+      "epoch": 0.9264758884163483,
+      "grad_norm": 0.11563142357065864,
+      "learning_rate": 2.822180079585168e-06,
+      "loss": 1.2485,
+      "step": 12853
+    },
+    {
+      "epoch": 0.9265479708786852,
+      "grad_norm": 0.12816092632447895,
+      "learning_rate": 2.8166752182773315e-06,
+      "loss": 1.2495,
+      "step": 12854
+    },
+    {
+      "epoch": 0.9266200533410222,
+      "grad_norm": 0.12454384939382686,
+      "learning_rate": 2.811175654330134e-06,
+      "loss": 1.2519,
+      "step": 12855
+    },
+    {
+      "epoch": 0.9266921358033591,
+      "grad_norm": 0.12192004169767441,
+      "learning_rate": 2.8056813880433354e-06,
+      "loss": 1.2715,
+      "step": 12856
+    },
+    {
+      "epoch": 0.926764218265696,
+      "grad_norm": 0.13177199854148441,
+      "learning_rate": 2.80019241971643e-06,
+      "loss": 1.2714,
+      "step": 12857
+    },
+    {
+      "epoch": 0.9268363007280329,
+      "grad_norm": 0.12047996810709886,
+      "learning_rate": 2.7947087496486e-06,
+      "loss": 1.2254,
+      "step": 12858
+    },
+    {
+      "epoch": 0.9269083831903698,
+      "grad_norm": 0.11064057276458304,
+      "learning_rate": 2.7892303781387853e-06,
+      "loss": 1.2699,
+      "step": 12859
+    },
+    {
+      "epoch": 0.9269804656527066,
+      "grad_norm": 0.12113480642232764,
+      "learning_rate": 2.783757305485568e-06,
+      "loss": 1.3656,
+      "step": 12860
+    },
+    {
+      "epoch": 0.9270525481150436,
+      "grad_norm": 0.1223667180137061,
+      "learning_rate": 2.7782895319873105e-06,
+      "loss": 1.3071,
+      "step": 12861
+    },
+    {
+      "epoch": 0.9271246305773805,
+      "grad_norm": 0.13253754283201902,
+      "learning_rate": 2.77282705794204e-06,
+      "loss": 1.3047,
+      "step": 12862
+    },
+    {
+      "epoch": 0.9271967130397174,
+      "grad_norm": 0.12310301653843171,
+      "learning_rate": 2.767369883647519e-06,
+      "loss": 1.3284,
+      "step": 12863
+    },
+    {
+      "epoch": 0.9272687955020543,
+      "grad_norm": 0.12051965946972117,
+      "learning_rate": 2.7619180094011986e-06,
+      "loss": 1.2724,
+      "step": 12864
+    },
+    {
+      "epoch": 0.9273408779643912,
+      "grad_norm": 0.11494957286076866,
+      "learning_rate": 2.756471435500274e-06,
+      "loss": 1.291,
+      "step": 12865
+    },
+    {
+      "epoch": 0.9274129604267282,
+      "grad_norm": 0.12178083252966145,
+      "learning_rate": 2.751030162241619e-06,
+      "loss": 1.2569,
+      "step": 12866
+    },
+    {
+      "epoch": 0.9274850428890651,
+      "grad_norm": 0.13291155365032986,
+      "learning_rate": 2.7455941899218298e-06,
+      "loss": 1.1992,
+      "step": 12867
+    },
+    {
+      "epoch": 0.927557125351402,
+      "grad_norm": 0.12085894945778619,
+      "learning_rate": 2.7401635188372244e-06,
+      "loss": 1.2128,
+      "step": 12868
+    },
+    {
+      "epoch": 0.9276292078137389,
+      "grad_norm": 0.11469424728044443,
+      "learning_rate": 2.7347381492838107e-06,
+      "loss": 1.2395,
+      "step": 12869
+    },
+    {
+      "epoch": 0.9277012902760758,
+      "grad_norm": 0.12401905520815966,
+      "learning_rate": 2.7293180815573304e-06,
+      "loss": 1.2712,
+      "step": 12870
+    },
+    {
+      "epoch": 0.9277733727384128,
+      "grad_norm": 0.124024564548043,
+      "learning_rate": 2.7239033159532358e-06,
+      "loss": 1.2725,
+      "step": 12871
+    },
+    {
+      "epoch": 0.9278454552007497,
+      "grad_norm": 0.13018072913607548,
+      "learning_rate": 2.7184938527666457e-06,
+      "loss": 1.3409,
+      "step": 12872
+    },
+    {
+      "epoch": 0.9279175376630866,
+      "grad_norm": 0.11968600170342815,
+      "learning_rate": 2.7130896922924587e-06,
+      "loss": 1.258,
+      "step": 12873
+    },
+    {
+      "epoch": 0.9279896201254235,
+      "grad_norm": 0.1136039091442331,
+      "learning_rate": 2.7076908348252383e-06,
+      "loss": 1.2891,
+      "step": 12874
+    },
+    {
+      "epoch": 0.9280617025877604,
+      "grad_norm": 0.1302455577365472,
+      "learning_rate": 2.702297280659272e-06,
+      "loss": 1.2438,
+      "step": 12875
+    },
+    {
+      "epoch": 0.9281337850500974,
+      "grad_norm": 0.120754716093362,
+      "learning_rate": 2.6969090300885458e-06,
+      "loss": 1.3195,
+      "step": 12876
+    },
+    {
+      "epoch": 0.9282058675124343,
+      "grad_norm": 0.11755966513221323,
+      "learning_rate": 2.69152608340677e-06,
+      "loss": 1.1605,
+      "step": 12877
+    },
+    {
+      "epoch": 0.9282779499747711,
+      "grad_norm": 0.11959389590573387,
+      "learning_rate": 2.686148440907377e-06,
+      "loss": 1.25,
+      "step": 12878
+    },
+    {
+      "epoch": 0.928350032437108,
+      "grad_norm": 0.1285891465456606,
+      "learning_rate": 2.680776102883509e-06,
+      "loss": 1.2043,
+      "step": 12879
+    },
+    {
+      "epoch": 0.9284221148994449,
+      "grad_norm": 0.11800083802357889,
+      "learning_rate": 2.6754090696279653e-06,
+      "loss": 1.2751,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9284941973617818,
+      "grad_norm": 0.12066749438928959,
+      "learning_rate": 2.6700473414333018e-06,
+      "loss": 1.3055,
+      "step": 12881
+    },
+    {
+      "epoch": 0.9285662798241188,
+      "grad_norm": 0.13115768492964547,
+      "learning_rate": 2.664690918591828e-06,
+      "loss": 1.2118,
+      "step": 12882
+    },
+    {
+      "epoch": 0.9286383622864557,
+      "grad_norm": 0.11654345134710228,
+      "learning_rate": 2.659339801395466e-06,
+      "loss": 1.2898,
+      "step": 12883
+    },
+    {
+      "epoch": 0.9287104447487926,
+      "grad_norm": 0.10974827845655105,
+      "learning_rate": 2.653993990135928e-06,
+      "loss": 1.3737,
+      "step": 12884
+    },
+    {
+      "epoch": 0.9287825272111295,
+      "grad_norm": 0.11965859024176763,
+      "learning_rate": 2.6486534851045907e-06,
+      "loss": 1.2335,
+      "step": 12885
+    },
+    {
+      "epoch": 0.9288546096734664,
+      "grad_norm": 0.1343810711863585,
+      "learning_rate": 2.6433182865925777e-06,
+      "loss": 1.3396,
+      "step": 12886
+    },
+    {
+      "epoch": 0.9289266921358034,
+      "grad_norm": 0.13907174093292,
+      "learning_rate": 2.6379883948907e-06,
+      "loss": 1.2694,
+      "step": 12887
+    },
+    {
+      "epoch": 0.9289987745981403,
+      "grad_norm": 0.11451092786236645,
+      "learning_rate": 2.6326638102894706e-06,
+      "loss": 1.2665,
+      "step": 12888
+    },
+    {
+      "epoch": 0.9290708570604772,
+      "grad_norm": 0.12488232551720914,
+      "learning_rate": 2.6273445330791234e-06,
+      "loss": 1.2088,
+      "step": 12889
+    },
+    {
+      "epoch": 0.9291429395228141,
+      "grad_norm": 0.11749205729311406,
+      "learning_rate": 2.622030563549649e-06,
+      "loss": 1.2662,
+      "step": 12890
+    },
+    {
+      "epoch": 0.929215021985151,
+      "grad_norm": 0.13970832755841686,
+      "learning_rate": 2.6167219019906597e-06,
+      "loss": 1.3388,
+      "step": 12891
+    },
+    {
+      "epoch": 0.929287104447488,
+      "grad_norm": 0.12762251226775204,
+      "learning_rate": 2.6114185486915465e-06,
+      "loss": 1.2172,
+      "step": 12892
+    },
+    {
+      "epoch": 0.9293591869098249,
+      "grad_norm": 0.1153990898855855,
+      "learning_rate": 2.6061205039413893e-06,
+      "loss": 1.2328,
+      "step": 12893
+    },
+    {
+      "epoch": 0.9294312693721618,
+      "grad_norm": 0.11745290834725099,
+      "learning_rate": 2.600827768028968e-06,
+      "loss": 1.2946,
+      "step": 12894
+    },
+    {
+      "epoch": 0.9295033518344986,
+      "grad_norm": 0.12425392140791396,
+      "learning_rate": 2.595540341242797e-06,
+      "loss": 1.1992,
+      "step": 12895
+    },
+    {
+      "epoch": 0.9295754342968355,
+      "grad_norm": 0.11514056035507153,
+      "learning_rate": 2.5902582238710783e-06,
+      "loss": 1.2282,
+      "step": 12896
+    },
+    {
+      "epoch": 0.9296475167591725,
+      "grad_norm": 0.11151644504633738,
+      "learning_rate": 2.584981416201737e-06,
+      "loss": 1.2309,
+      "step": 12897
+    },
+    {
+      "epoch": 0.9297195992215094,
+      "grad_norm": 0.126129897326848,
+      "learning_rate": 2.5797099185224218e-06,
+      "loss": 1.1754,
+      "step": 12898
+    },
+    {
+      "epoch": 0.9297916816838463,
+      "grad_norm": 0.11472336218568326,
+      "learning_rate": 2.574443731120446e-06,
+      "loss": 1.373,
+      "step": 12899
+    },
+    {
+      "epoch": 0.9298637641461832,
+      "grad_norm": 0.14036877081463442,
+      "learning_rate": 2.5691828542828922e-06,
+      "loss": 1.3383,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9299358466085201,
+      "grad_norm": 0.12788959437046227,
+      "learning_rate": 2.563927288296508e-06,
+      "loss": 1.2492,
+      "step": 12901
+    },
+    {
+      "epoch": 0.930007929070857,
+      "grad_norm": 0.12471117559174709,
+      "learning_rate": 2.558677033447776e-06,
+      "loss": 1.2708,
+      "step": 12902
+    },
+    {
+      "epoch": 0.930080011533194,
+      "grad_norm": 0.11399171726719154,
+      "learning_rate": 2.553432090022878e-06,
+      "loss": 1.168,
+      "step": 12903
+    },
+    {
+      "epoch": 0.9301520939955309,
+      "grad_norm": 0.10601135550560024,
+      "learning_rate": 2.548192458307719e-06,
+      "loss": 1.2058,
+      "step": 12904
+    },
+    {
+      "epoch": 0.9302241764578678,
+      "grad_norm": 0.11615500315260899,
+      "learning_rate": 2.542958138587892e-06,
+      "loss": 1.2448,
+      "step": 12905
+    },
+    {
+      "epoch": 0.9302962589202047,
+      "grad_norm": 0.11276585795526713,
+      "learning_rate": 2.537729131148725e-06,
+      "loss": 1.2138,
+      "step": 12906
+    },
+    {
+      "epoch": 0.9303683413825417,
+      "grad_norm": 0.11997710416681427,
+      "learning_rate": 2.532505436275245e-06,
+      "loss": 1.3316,
+      "step": 12907
+    },
+    {
+      "epoch": 0.9304404238448786,
+      "grad_norm": 0.12456146715680691,
+      "learning_rate": 2.527287054252192e-06,
+      "loss": 1.2193,
+      "step": 12908
+    },
+    {
+      "epoch": 0.9305125063072155,
+      "grad_norm": 0.11836252111776403,
+      "learning_rate": 2.522073985364004e-06,
+      "loss": 1.1341,
+      "step": 12909
+    },
+    {
+      "epoch": 0.9305845887695524,
+      "grad_norm": 0.11631684075990537,
+      "learning_rate": 2.516866229894843e-06,
+      "loss": 1.2083,
+      "step": 12910
+    },
+    {
+      "epoch": 0.9306566712318893,
+      "grad_norm": 0.13486718387146046,
+      "learning_rate": 2.511663788128582e-06,
+      "loss": 1.223,
+      "step": 12911
+    },
+    {
+      "epoch": 0.9307287536942261,
+      "grad_norm": 0.11420858228013507,
+      "learning_rate": 2.506466660348805e-06,
+      "loss": 1.3626,
+      "step": 12912
+    },
+    {
+      "epoch": 0.9308008361565631,
+      "grad_norm": 0.12503165393783786,
+      "learning_rate": 2.501274846838797e-06,
+      "loss": 1.1555,
+      "step": 12913
+    },
+    {
+      "epoch": 0.9308729186189,
+      "grad_norm": 0.12059095509239798,
+      "learning_rate": 2.4960883478815642e-06,
+      "loss": 1.2357,
+      "step": 12914
+    },
+    {
+      "epoch": 0.9309450010812369,
+      "grad_norm": 0.137107017188828,
+      "learning_rate": 2.490907163759815e-06,
+      "loss": 1.3767,
+      "step": 12915
+    },
+    {
+      "epoch": 0.9310170835435738,
+      "grad_norm": 0.11927886942338674,
+      "learning_rate": 2.4857312947559443e-06,
+      "loss": 1.2825,
+      "step": 12916
+    },
+    {
+      "epoch": 0.9310891660059107,
+      "grad_norm": 0.11541272374298507,
+      "learning_rate": 2.480560741152116e-06,
+      "loss": 1.2954,
+      "step": 12917
+    },
+    {
+      "epoch": 0.9311612484682477,
+      "grad_norm": 0.11686805926323258,
+      "learning_rate": 2.47539550323016e-06,
+      "loss": 1.2242,
+      "step": 12918
+    },
+    {
+      "epoch": 0.9312333309305846,
+      "grad_norm": 0.1259115995358895,
+      "learning_rate": 2.4702355812716404e-06,
+      "loss": 1.3367,
+      "step": 12919
+    },
+    {
+      "epoch": 0.9313054133929215,
+      "grad_norm": 0.13038956129676138,
+      "learning_rate": 2.4650809755577873e-06,
+      "loss": 1.2509,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9313774958552584,
+      "grad_norm": 0.11783479711269476,
+      "learning_rate": 2.45993168636961e-06,
+      "loss": 1.222,
+      "step": 12921
+    },
+    {
+      "epoch": 0.9314495783175953,
+      "grad_norm": 0.11979017546223084,
+      "learning_rate": 2.4547877139877717e-06,
+      "loss": 1.2959,
+      "step": 12922
+    },
+    {
+      "epoch": 0.9315216607799323,
+      "grad_norm": 0.12128200220142062,
+      "learning_rate": 2.4496490586926715e-06,
+      "loss": 1.2102,
+      "step": 12923
+    },
+    {
+      "epoch": 0.9315937432422692,
+      "grad_norm": 0.12753697108081638,
+      "learning_rate": 2.4445157207644065e-06,
+      "loss": 1.3086,
+      "step": 12924
+    },
+    {
+      "epoch": 0.9316658257046061,
+      "grad_norm": 0.11532855716075616,
+      "learning_rate": 2.439387700482776e-06,
+      "loss": 1.2027,
+      "step": 12925
+    },
+    {
+      "epoch": 0.931737908166943,
+      "grad_norm": 0.1157329508291151,
+      "learning_rate": 2.4342649981273226e-06,
+      "loss": 1.2182,
+      "step": 12926
+    },
+    {
+      "epoch": 0.9318099906292799,
+      "grad_norm": 0.11289024096560558,
+      "learning_rate": 2.4291476139772897e-06,
+      "loss": 1.2291,
+      "step": 12927
+    },
+    {
+      "epoch": 0.9318820730916169,
+      "grad_norm": 0.1208812720300191,
+      "learning_rate": 2.4240355483115982e-06,
+      "loss": 1.2811,
+      "step": 12928
+    },
+    {
+      "epoch": 0.9319541555539538,
+      "grad_norm": 0.12104800570912598,
+      "learning_rate": 2.418928801408904e-06,
+      "loss": 1.248,
+      "step": 12929
+    },
+    {
+      "epoch": 0.9320262380162906,
+      "grad_norm": 0.1397034207281515,
+      "learning_rate": 2.4138273735475727e-06,
+      "loss": 1.3084,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9320983204786275,
+      "grad_norm": 0.12146878892516262,
+      "learning_rate": 2.4087312650056926e-06,
+      "loss": 1.2149,
+      "step": 12931
+    },
+    {
+      "epoch": 0.9321704029409644,
+      "grad_norm": 0.12305263427684565,
+      "learning_rate": 2.403640476061031e-06,
+      "loss": 1.2817,
+      "step": 12932
+    },
+    {
+      "epoch": 0.9322424854033013,
+      "grad_norm": 0.10344038688087034,
+      "learning_rate": 2.3985550069910767e-06,
+      "loss": 1.2551,
+      "step": 12933
+    },
+    {
+      "epoch": 0.9323145678656383,
+      "grad_norm": 0.11526746257968616,
+      "learning_rate": 2.3934748580730414e-06,
+      "loss": 1.2334,
+      "step": 12934
+    },
+    {
+      "epoch": 0.9323866503279752,
+      "grad_norm": 0.13751140663162065,
+      "learning_rate": 2.3884000295838593e-06,
+      "loss": 1.208,
+      "step": 12935
+    },
+    {
+      "epoch": 0.9324587327903121,
+      "grad_norm": 0.12124453183581148,
+      "learning_rate": 2.3833305218001202e-06,
+      "loss": 1.2542,
+      "step": 12936
+    },
+    {
+      "epoch": 0.932530815252649,
+      "grad_norm": 0.11585272498774281,
+      "learning_rate": 2.3782663349981693e-06,
+      "loss": 1.2376,
+      "step": 12937
+    },
+    {
+      "epoch": 0.932602897714986,
+      "grad_norm": 0.12589634238128808,
+      "learning_rate": 2.3732074694540527e-06,
+      "loss": 1.2149,
+      "step": 12938
+    },
+    {
+      "epoch": 0.9326749801773229,
+      "grad_norm": 0.12422029493134494,
+      "learning_rate": 2.368153925443528e-06,
+      "loss": 1.3053,
+      "step": 12939
+    },
+    {
+      "epoch": 0.9327470626396598,
+      "grad_norm": 0.11658384965174164,
+      "learning_rate": 2.3631057032420523e-06,
+      "loss": 1.2597,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9328191451019967,
+      "grad_norm": 0.1133413861830397,
+      "learning_rate": 2.3580628031248054e-06,
+      "loss": 1.3429,
+      "step": 12941
+    },
+    {
+      "epoch": 0.9328912275643336,
+      "grad_norm": 0.12410074993215461,
+      "learning_rate": 2.353025225366645e-06,
+      "loss": 1.2908,
+      "step": 12942
+    },
+    {
+      "epoch": 0.9329633100266705,
+      "grad_norm": 0.11214133766332797,
+      "learning_rate": 2.3479929702422077e-06,
+      "loss": 1.242,
+      "step": 12943
+    },
+    {
+      "epoch": 0.9330353924890075,
+      "grad_norm": 0.10819641896966582,
+      "learning_rate": 2.342966038025762e-06,
+      "loss": 1.279,
+      "step": 12944
+    },
+    {
+      "epoch": 0.9331074749513444,
+      "grad_norm": 0.1173932672871508,
+      "learning_rate": 2.3379444289913342e-06,
+      "loss": 1.174,
+      "step": 12945
+    },
+    {
+      "epoch": 0.9331795574136813,
+      "grad_norm": 0.11473999153884729,
+      "learning_rate": 2.3329281434126493e-06,
+      "loss": 1.2407,
+      "step": 12946
+    },
+    {
+      "epoch": 0.9332516398760181,
+      "grad_norm": 0.11356053580572237,
+      "learning_rate": 2.327917181563133e-06,
+      "loss": 1.2538,
+      "step": 12947
+    },
+    {
+      "epoch": 0.933323722338355,
+      "grad_norm": 0.11367028371183038,
+      "learning_rate": 2.3229115437159223e-06,
+      "loss": 1.2687,
+      "step": 12948
+    },
+    {
+      "epoch": 0.933395804800692,
+      "grad_norm": 0.11973467078890204,
+      "learning_rate": 2.3179112301438878e-06,
+      "loss": 1.3227,
+      "step": 12949
+    },
+    {
+      "epoch": 0.9334678872630289,
+      "grad_norm": 0.12865495794659218,
+      "learning_rate": 2.312916241119567e-06,
+      "loss": 1.2875,
+      "step": 12950
+    },
+    {
+      "epoch": 0.9335399697253658,
+      "grad_norm": 0.10956362977324324,
+      "learning_rate": 2.3079265769152536e-06,
+      "loss": 1.3023,
+      "step": 12951
+    },
+    {
+      "epoch": 0.9336120521877027,
+      "grad_norm": 0.13851762831857076,
+      "learning_rate": 2.3029422378029184e-06,
+      "loss": 1.2203,
+      "step": 12952
+    },
+    {
+      "epoch": 0.9336841346500396,
+      "grad_norm": 0.12233443510105893,
+      "learning_rate": 2.2979632240542557e-06,
+      "loss": 1.3548,
+      "step": 12953
+    },
+    {
+      "epoch": 0.9337562171123766,
+      "grad_norm": 0.11946248042824446,
+      "learning_rate": 2.2929895359406595e-06,
+      "loss": 1.3227,
+      "step": 12954
+    },
+    {
+      "epoch": 0.9338282995747135,
+      "grad_norm": 0.11659562970290523,
+      "learning_rate": 2.288021173733246e-06,
+      "loss": 1.3063,
+      "step": 12955
+    },
+    {
+      "epoch": 0.9339003820370504,
+      "grad_norm": 0.12767669756963332,
+      "learning_rate": 2.2830581377028424e-06,
+      "loss": 1.2206,
+      "step": 12956
+    },
+    {
+      "epoch": 0.9339724644993873,
+      "grad_norm": 0.12223607259696354,
+      "learning_rate": 2.278100428119967e-06,
+      "loss": 1.3162,
+      "step": 12957
+    },
+    {
+      "epoch": 0.9340445469617242,
+      "grad_norm": 0.12635958231853484,
+      "learning_rate": 2.2731480452548582e-06,
+      "loss": 1.1796,
+      "step": 12958
+    },
+    {
+      "epoch": 0.9341166294240612,
+      "grad_norm": 0.10444805516316207,
+      "learning_rate": 2.268200989377478e-06,
+      "loss": 1.3229,
+      "step": 12959
+    },
+    {
+      "epoch": 0.9341887118863981,
+      "grad_norm": 0.11777222172097847,
+      "learning_rate": 2.2632592607574774e-06,
+      "loss": 1.2886,
+      "step": 12960
+    },
+    {
+      "epoch": 0.934260794348735,
+      "grad_norm": 0.12378226293612962,
+      "learning_rate": 2.258322859664219e-06,
+      "loss": 1.1075,
+      "step": 12961
+    },
+    {
+      "epoch": 0.9343328768110719,
+      "grad_norm": 0.11574856489829669,
+      "learning_rate": 2.2533917863667877e-06,
+      "loss": 1.2345,
+      "step": 12962
+    },
+    {
+      "epoch": 0.9344049592734088,
+      "grad_norm": 0.11562610310451828,
+      "learning_rate": 2.2484660411339787e-06,
+      "loss": 1.1694,
+      "step": 12963
+    },
+    {
+      "epoch": 0.9344770417357456,
+      "grad_norm": 0.12376649852226608,
+      "learning_rate": 2.2435456242342666e-06,
+      "loss": 1.2818,
+      "step": 12964
+    },
+    {
+      "epoch": 0.9345491241980826,
+      "grad_norm": 0.1258792580529584,
+      "learning_rate": 2.2386305359358816e-06,
+      "loss": 1.315,
+      "step": 12965
+    },
+    {
+      "epoch": 0.9346212066604195,
+      "grad_norm": 0.12364922113192656,
+      "learning_rate": 2.2337207765067314e-06,
+      "loss": 1.2846,
+      "step": 12966
+    },
+    {
+      "epoch": 0.9346932891227564,
+      "grad_norm": 0.11675040871259591,
+      "learning_rate": 2.2288163462144352e-06,
+      "loss": 1.3324,
+      "step": 12967
+    },
+    {
+      "epoch": 0.9347653715850933,
+      "grad_norm": 0.1115804972646361,
+      "learning_rate": 2.223917245326357e-06,
+      "loss": 1.2586,
+      "step": 12968
+    },
+    {
+      "epoch": 0.9348374540474302,
+      "grad_norm": 0.12127722846501282,
+      "learning_rate": 2.219023474109483e-06,
+      "loss": 1.2032,
+      "step": 12969
+    },
+    {
+      "epoch": 0.9349095365097672,
+      "grad_norm": 0.1225295972750148,
+      "learning_rate": 2.2141350328306222e-06,
+      "loss": 1.2985,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9349816189721041,
+      "grad_norm": 0.2702406486632978,
+      "learning_rate": 2.2092519217562168e-06,
+      "loss": 1.2554,
+      "step": 12971
+    },
+    {
+      "epoch": 0.935053701434441,
+      "grad_norm": 0.12065750625424676,
+      "learning_rate": 2.204374141152454e-06,
+      "loss": 1.2596,
+      "step": 12972
+    },
+    {
+      "epoch": 0.9351257838967779,
+      "grad_norm": 0.1319514693451251,
+      "learning_rate": 2.1995016912851775e-06,
+      "loss": 1.3342,
+      "step": 12973
+    },
+    {
+      "epoch": 0.9351978663591148,
+      "grad_norm": 0.12367440475783156,
+      "learning_rate": 2.194634572420029e-06,
+      "loss": 1.2781,
+      "step": 12974
+    },
+    {
+      "epoch": 0.9352699488214518,
+      "grad_norm": 0.1283516989184775,
+      "learning_rate": 2.189772784822286e-06,
+      "loss": 1.2859,
+      "step": 12975
+    },
+    {
+      "epoch": 0.9353420312837887,
+      "grad_norm": 0.12832065296686332,
+      "learning_rate": 2.1849163287569586e-06,
+      "loss": 1.2703,
+      "step": 12976
+    },
+    {
+      "epoch": 0.9354141137461256,
+      "grad_norm": 0.12033155416723923,
+      "learning_rate": 2.180065204488768e-06,
+      "loss": 1.2474,
+      "step": 12977
+    },
+    {
+      "epoch": 0.9354861962084625,
+      "grad_norm": 0.1192936930484963,
+      "learning_rate": 2.175219412282148e-06,
+      "loss": 1.1809,
+      "step": 12978
+    },
+    {
+      "epoch": 0.9355582786707994,
+      "grad_norm": 0.11571677832119016,
+      "learning_rate": 2.170378952401231e-06,
+      "loss": 1.2817,
+      "step": 12979
+    },
+    {
+      "epoch": 0.9356303611331364,
+      "grad_norm": 0.1341713894908137,
+      "learning_rate": 2.165543825109884e-06,
+      "loss": 1.2496,
+      "step": 12980
+    },
+    {
+      "epoch": 0.9357024435954733,
+      "grad_norm": 0.12948358041969463,
+      "learning_rate": 2.16071403067164e-06,
+      "loss": 1.205,
+      "step": 12981
+    },
+    {
+      "epoch": 0.9357745260578101,
+      "grad_norm": 0.117187311063467,
+      "learning_rate": 2.155889569349767e-06,
+      "loss": 1.3034,
+      "step": 12982
+    },
+    {
+      "epoch": 0.935846608520147,
+      "grad_norm": 0.11725878347222277,
+      "learning_rate": 2.1510704414072656e-06,
+      "loss": 1.3602,
+      "step": 12983
+    },
+    {
+      "epoch": 0.9359186909824839,
+      "grad_norm": 0.1252141637670251,
+      "learning_rate": 2.1462566471068037e-06,
+      "loss": 1.3123,
+      "step": 12984
+    },
+    {
+      "epoch": 0.9359907734448208,
+      "grad_norm": 0.11052935466895013,
+      "learning_rate": 2.141448186710782e-06,
+      "loss": 1.3948,
+      "step": 12985
+    },
+    {
+      "epoch": 0.9360628559071578,
+      "grad_norm": 0.1242196859078926,
+      "learning_rate": 2.1366450604812905e-06,
+      "loss": 1.2733,
+      "step": 12986
+    },
+    {
+      "epoch": 0.9361349383694947,
+      "grad_norm": 0.11501053235562238,
+      "learning_rate": 2.131847268680154e-06,
+      "loss": 1.2274,
+      "step": 12987
+    },
+    {
+      "epoch": 0.9362070208318316,
+      "grad_norm": 0.12135741244935551,
+      "learning_rate": 2.127054811568907e-06,
+      "loss": 1.2292,
+      "step": 12988
+    },
+    {
+      "epoch": 0.9362791032941685,
+      "grad_norm": 0.11962046068075449,
+      "learning_rate": 2.1222676894087633e-06,
+      "loss": 1.187,
+      "step": 12989
+    },
+    {
+      "epoch": 0.9363511857565054,
+      "grad_norm": 0.12677401512116118,
+      "learning_rate": 2.11748590246067e-06,
+      "loss": 1.2978,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9364232682188424,
+      "grad_norm": 0.12001680502984545,
+      "learning_rate": 2.112709450985273e-06,
+      "loss": 1.1855,
+      "step": 12991
+    },
+    {
+      "epoch": 0.9364953506811793,
+      "grad_norm": 0.1185020279094667,
+      "learning_rate": 2.1079383352429536e-06,
+      "loss": 1.1879,
+      "step": 12992
+    },
+    {
+      "epoch": 0.9365674331435162,
+      "grad_norm": 0.12269585441501617,
+      "learning_rate": 2.1031725554937377e-06,
+      "loss": 1.2159,
+      "step": 12993
+    },
+    {
+      "epoch": 0.9366395156058531,
+      "grad_norm": 0.10918260668421048,
+      "learning_rate": 2.0984121119974386e-06,
+      "loss": 1.2963,
+      "step": 12994
+    },
+    {
+      "epoch": 0.93671159806819,
+      "grad_norm": 0.12129890409897648,
+      "learning_rate": 2.0936570050135273e-06,
+      "loss": 1.2605,
+      "step": 12995
+    },
+    {
+      "epoch": 0.936783680530527,
+      "grad_norm": 0.14086561872041692,
+      "learning_rate": 2.0889072348012184e-06,
+      "loss": 1.2844,
+      "step": 12996
+    },
+    {
+      "epoch": 0.9368557629928639,
+      "grad_norm": 0.12318004357203514,
+      "learning_rate": 2.0841628016193936e-06,
+      "loss": 1.1934,
+      "step": 12997
+    },
+    {
+      "epoch": 0.9369278454552008,
+      "grad_norm": 0.13237739928441747,
+      "learning_rate": 2.0794237057266796e-06,
+      "loss": 1.3602,
+      "step": 12998
+    },
+    {
+      "epoch": 0.9369999279175376,
+      "grad_norm": 0.12809322797076314,
+      "learning_rate": 2.0746899473813807e-06,
+      "loss": 1.2521,
+      "step": 12999
+    },
+    {
+      "epoch": 0.9370720103798745,
+      "grad_norm": 0.11562812336563724,
+      "learning_rate": 2.0699615268415684e-06,
+      "loss": 1.2758,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9371440928422115,
+      "grad_norm": 0.1289573433547281,
+      "learning_rate": 2.0652384443649474e-06,
+      "loss": 1.3686,
+      "step": 13001
+    },
+    {
+      "epoch": 0.9372161753045484,
+      "grad_norm": 0.1710020972760547,
+      "learning_rate": 2.060520700208979e-06,
+      "loss": 1.2465,
+      "step": 13002
+    },
+    {
+      "epoch": 0.9372882577668853,
+      "grad_norm": 0.12334634041009553,
+      "learning_rate": 2.0558082946308232e-06,
+      "loss": 1.25,
+      "step": 13003
+    },
+    {
+      "epoch": 0.9373603402292222,
+      "grad_norm": 0.1139070864881562,
+      "learning_rate": 2.051101227887353e-06,
+      "loss": 1.2288,
+      "step": 13004
+    },
+    {
+      "epoch": 0.9374324226915591,
+      "grad_norm": 0.11120434595129694,
+      "learning_rate": 2.046399500235141e-06,
+      "loss": 1.2311,
+      "step": 13005
+    },
+    {
+      "epoch": 0.937504505153896,
+      "grad_norm": 0.11445168380545882,
+      "learning_rate": 2.041703111930471e-06,
+      "loss": 1.374,
+      "step": 13006
+    },
+    {
+      "epoch": 0.937576587616233,
+      "grad_norm": 0.12182227242353105,
+      "learning_rate": 2.037012063229349e-06,
+      "loss": 1.2227,
+      "step": 13007
+    },
+    {
+      "epoch": 0.9376486700785699,
+      "grad_norm": 0.3359029482603968,
+      "learning_rate": 2.0323263543874592e-06,
+      "loss": 1.2339,
+      "step": 13008
+    },
+    {
+      "epoch": 0.9377207525409068,
+      "grad_norm": 0.1244948302649898,
+      "learning_rate": 2.0276459856602424e-06,
+      "loss": 1.2957,
+      "step": 13009
+    },
+    {
+      "epoch": 0.9377928350032437,
+      "grad_norm": 0.12044080215877487,
+      "learning_rate": 2.0229709573027943e-06,
+      "loss": 1.3185,
+      "step": 13010
+    },
+    {
+      "epoch": 0.9378649174655806,
+      "grad_norm": 0.11935661957285842,
+      "learning_rate": 2.0183012695699554e-06,
+      "loss": 1.1861,
+      "step": 13011
+    },
+    {
+      "epoch": 0.9379369999279176,
+      "grad_norm": 0.14010898047733109,
+      "learning_rate": 2.0136369227162777e-06,
+      "loss": 1.2862,
+      "step": 13012
+    },
+    {
+      "epoch": 0.9380090823902545,
+      "grad_norm": 0.12403605324937468,
+      "learning_rate": 2.008977916995991e-06,
+      "loss": 1.2235,
+      "step": 13013
+    },
+    {
+      "epoch": 0.9380811648525914,
+      "grad_norm": 0.12316966288582838,
+      "learning_rate": 2.004324252663059e-06,
+      "loss": 1.2839,
+      "step": 13014
+    },
+    {
+      "epoch": 0.9381532473149283,
+      "grad_norm": 0.1388069129642839,
+      "learning_rate": 1.9996759299711567e-06,
+      "loss": 1.2818,
+      "step": 13015
+    },
+    {
+      "epoch": 0.9382253297772651,
+      "grad_norm": 0.1250348052526242,
+      "learning_rate": 1.9950329491736476e-06,
+      "loss": 1.2346,
+      "step": 13016
+    },
+    {
+      "epoch": 0.9382974122396021,
+      "grad_norm": 0.12251711034880446,
+      "learning_rate": 1.9903953105236186e-06,
+      "loss": 1.2578,
+      "step": 13017
+    },
+    {
+      "epoch": 0.938369494701939,
+      "grad_norm": 0.11721145157853498,
+      "learning_rate": 1.985763014273867e-06,
+      "loss": 1.2788,
+      "step": 13018
+    },
+    {
+      "epoch": 0.9384415771642759,
+      "grad_norm": 0.14233832466552931,
+      "learning_rate": 1.9811360606768805e-06,
+      "loss": 1.331,
+      "step": 13019
+    },
+    {
+      "epoch": 0.9385136596266128,
+      "grad_norm": 0.1135272076671253,
+      "learning_rate": 1.97651444998489e-06,
+      "loss": 1.2492,
+      "step": 13020
+    },
+    {
+      "epoch": 0.9385857420889497,
+      "grad_norm": 0.14389163729108073,
+      "learning_rate": 1.971898182449805e-06,
+      "loss": 1.1899,
+      "step": 13021
+    },
+    {
+      "epoch": 0.9386578245512867,
+      "grad_norm": 0.12958526875801582,
+      "learning_rate": 1.967287258323258e-06,
+      "loss": 1.2419,
+      "step": 13022
+    },
+    {
+      "epoch": 0.9387299070136236,
+      "grad_norm": 0.12295513022612299,
+      "learning_rate": 1.9626816778565813e-06,
+      "loss": 1.3208,
+      "step": 13023
+    },
+    {
+      "epoch": 0.9388019894759605,
+      "grad_norm": 0.11334974554510913,
+      "learning_rate": 1.9580814413008075e-06,
+      "loss": 1.3703,
+      "step": 13024
+    },
+    {
+      "epoch": 0.9388740719382974,
+      "grad_norm": 0.12106478997708013,
+      "learning_rate": 1.9534865489067246e-06,
+      "loss": 1.2852,
+      "step": 13025
+    },
+    {
+      "epoch": 0.9389461544006343,
+      "grad_norm": 0.11944133605587277,
+      "learning_rate": 1.9488970009247544e-06,
+      "loss": 1.3296,
+      "step": 13026
+    },
+    {
+      "epoch": 0.9390182368629713,
+      "grad_norm": 0.11359700362659503,
+      "learning_rate": 1.944312797605097e-06,
+      "loss": 1.3099,
+      "step": 13027
+    },
+    {
+      "epoch": 0.9390903193253082,
+      "grad_norm": 0.1267597359778977,
+      "learning_rate": 1.939733939197619e-06,
+      "loss": 1.3376,
+      "step": 13028
+    },
+    {
+      "epoch": 0.9391624017876451,
+      "grad_norm": 0.12086268293580935,
+      "learning_rate": 1.9351604259519206e-06,
+      "loss": 1.2366,
+      "step": 13029
+    },
+    {
+      "epoch": 0.939234484249982,
+      "grad_norm": 0.12113592064890608,
+      "learning_rate": 1.930592258117281e-06,
+      "loss": 1.2589,
+      "step": 13030
+    },
+    {
+      "epoch": 0.9393065667123189,
+      "grad_norm": 0.12033132734966305,
+      "learning_rate": 1.9260294359427223e-06,
+      "loss": 1.2699,
+      "step": 13031
+    },
+    {
+      "epoch": 0.9393786491746559,
+      "grad_norm": 0.12880713349429923,
+      "learning_rate": 1.921471959676957e-06,
+      "loss": 1.2565,
+      "step": 13032
+    },
+    {
+      "epoch": 0.9394507316369927,
+      "grad_norm": 0.12054798052931168,
+      "learning_rate": 1.9169198295684087e-06,
+      "loss": 1.3011,
+      "step": 13033
+    },
+    {
+      "epoch": 0.9395228140993296,
+      "grad_norm": 0.11273540653316906,
+      "learning_rate": 1.912373045865201e-06,
+      "loss": 1.298,
+      "step": 13034
+    },
+    {
+      "epoch": 0.9395948965616665,
+      "grad_norm": 0.1297287159683806,
+      "learning_rate": 1.9078316088151693e-06,
+      "loss": 1.3321,
+      "step": 13035
+    },
+    {
+      "epoch": 0.9396669790240034,
+      "grad_norm": 0.1266167885833357,
+      "learning_rate": 1.903295518665893e-06,
+      "loss": 1.2316,
+      "step": 13036
+    },
+    {
+      "epoch": 0.9397390614863403,
+      "grad_norm": 0.1235833479574025,
+      "learning_rate": 1.8987647756645966e-06,
+      "loss": 1.2553,
+      "step": 13037
+    },
+    {
+      "epoch": 0.9398111439486773,
+      "grad_norm": 0.13111747333370322,
+      "learning_rate": 1.8942393800582603e-06,
+      "loss": 1.1976,
+      "step": 13038
+    },
+    {
+      "epoch": 0.9398832264110142,
+      "grad_norm": 0.11933938374960461,
+      "learning_rate": 1.8897193320935536e-06,
+      "loss": 1.2116,
+      "step": 13039
+    },
+    {
+      "epoch": 0.9399553088733511,
+      "grad_norm": 0.11156616286557317,
+      "learning_rate": 1.885204632016868e-06,
+      "loss": 1.2724,
+      "step": 13040
+    },
+    {
+      "epoch": 0.940027391335688,
+      "grad_norm": 0.11946716852170196,
+      "learning_rate": 1.8806952800742849e-06,
+      "loss": 1.2461,
+      "step": 13041
+    },
+    {
+      "epoch": 0.9400994737980249,
+      "grad_norm": 0.12178506564677352,
+      "learning_rate": 1.8761912765116185e-06,
+      "loss": 1.1919,
+      "step": 13042
+    },
+    {
+      "epoch": 0.9401715562603619,
+      "grad_norm": 0.1354626786253028,
+      "learning_rate": 1.8716926215743612e-06,
+      "loss": 1.211,
+      "step": 13043
+    },
+    {
+      "epoch": 0.9402436387226988,
+      "grad_norm": 0.12312936348979113,
+      "learning_rate": 1.8671993155077393e-06,
+      "loss": 1.2733,
+      "step": 13044
+    },
+    {
+      "epoch": 0.9403157211850357,
+      "grad_norm": 0.1223954375619494,
+      "learning_rate": 1.862711358556679e-06,
+      "loss": 1.285,
+      "step": 13045
+    },
+    {
+      "epoch": 0.9403878036473726,
+      "grad_norm": 0.12296619317827748,
+      "learning_rate": 1.858228750965807e-06,
+      "loss": 1.2743,
+      "step": 13046
+    },
+    {
+      "epoch": 0.9404598861097095,
+      "grad_norm": 0.12737844172488116,
+      "learning_rate": 1.8537514929794608e-06,
+      "loss": 1.3378,
+      "step": 13047
+    },
+    {
+      "epoch": 0.9405319685720465,
+      "grad_norm": 0.12153927036040583,
+      "learning_rate": 1.8492795848417122e-06,
+      "loss": 1.1939,
+      "step": 13048
+    },
+    {
+      "epoch": 0.9406040510343834,
+      "grad_norm": 0.13413566001457292,
+      "learning_rate": 1.8448130267962992e-06,
+      "loss": 1.2427,
+      "step": 13049
+    },
+    {
+      "epoch": 0.9406761334967203,
+      "grad_norm": 0.12249102247742952,
+      "learning_rate": 1.8403518190867053e-06,
+      "loss": 1.2206,
+      "step": 13050
+    },
+    {
+      "epoch": 0.9407482159590571,
+      "grad_norm": 0.11577996462494887,
+      "learning_rate": 1.8358959619560911e-06,
+      "loss": 1.1485,
+      "step": 13051
+    },
+    {
+      "epoch": 0.940820298421394,
+      "grad_norm": 0.10878864938581818,
+      "learning_rate": 1.8314454556473625e-06,
+      "loss": 1.2392,
+      "step": 13052
+    },
+    {
+      "epoch": 0.940892380883731,
+      "grad_norm": 0.129542975945424,
+      "learning_rate": 1.827000300403081e-06,
+      "loss": 1.2802,
+      "step": 13053
+    },
+    {
+      "epoch": 0.9409644633460679,
+      "grad_norm": 0.12337926533376216,
+      "learning_rate": 1.8225604964655752e-06,
+      "loss": 1.3083,
+      "step": 13054
+    },
+    {
+      "epoch": 0.9410365458084048,
+      "grad_norm": 0.11551978774899022,
+      "learning_rate": 1.81812604407684e-06,
+      "loss": 1.3039,
+      "step": 13055
+    },
+    {
+      "epoch": 0.9411086282707417,
+      "grad_norm": 0.11393045082615468,
+      "learning_rate": 1.8136969434785932e-06,
+      "loss": 1.3302,
+      "step": 13056
+    },
+    {
+      "epoch": 0.9411807107330786,
+      "grad_norm": 0.1634013990247883,
+      "learning_rate": 1.8092731949122532e-06,
+      "loss": 1.2403,
+      "step": 13057
+    },
+    {
+      "epoch": 0.9412527931954155,
+      "grad_norm": 0.1293041515427013,
+      "learning_rate": 1.8048547986189822e-06,
+      "loss": 1.2789,
+      "step": 13058
+    },
+    {
+      "epoch": 0.9413248756577525,
+      "grad_norm": 0.12317332225565761,
+      "learning_rate": 1.800441754839588e-06,
+      "loss": 1.3492,
+      "step": 13059
+    },
+    {
+      "epoch": 0.9413969581200894,
+      "grad_norm": 0.1341938940664328,
+      "learning_rate": 1.7960340638146444e-06,
+      "loss": 1.2318,
+      "step": 13060
+    },
+    {
+      "epoch": 0.9414690405824263,
+      "grad_norm": 0.10959207283947572,
+      "learning_rate": 1.7916317257844039e-06,
+      "loss": 1.2527,
+      "step": 13061
+    },
+    {
+      "epoch": 0.9415411230447632,
+      "grad_norm": 0.12228882410847139,
+      "learning_rate": 1.7872347409888301e-06,
+      "loss": 1.3401,
+      "step": 13062
+    },
+    {
+      "epoch": 0.9416132055071001,
+      "grad_norm": 0.12308139228436829,
+      "learning_rate": 1.782843109667598e-06,
+      "loss": 1.2033,
+      "step": 13063
+    },
+    {
+      "epoch": 0.9416852879694371,
+      "grad_norm": 0.12104666467845064,
+      "learning_rate": 1.7784568320600825e-06,
+      "loss": 1.3608,
+      "step": 13064
+    },
+    {
+      "epoch": 0.941757370431774,
+      "grad_norm": 0.11874869612873025,
+      "learning_rate": 1.7740759084054037e-06,
+      "loss": 1.2451,
+      "step": 13065
+    },
+    {
+      "epoch": 0.9418294528941109,
+      "grad_norm": 0.1729747090212049,
+      "learning_rate": 1.7697003389423262e-06,
+      "loss": 1.2858,
+      "step": 13066
+    },
+    {
+      "epoch": 0.9419015353564478,
+      "grad_norm": 0.13476426039709966,
+      "learning_rate": 1.765330123909381e-06,
+      "loss": 1.2222,
+      "step": 13067
+    },
+    {
+      "epoch": 0.9419736178187846,
+      "grad_norm": 0.124621625155239,
+      "learning_rate": 1.7609652635447782e-06,
+      "loss": 1.1977,
+      "step": 13068
+    },
+    {
+      "epoch": 0.9420457002811216,
+      "grad_norm": 0.11539525289759742,
+      "learning_rate": 1.7566057580864382e-06,
+      "loss": 1.2058,
+      "step": 13069
+    },
+    {
+      "epoch": 0.9421177827434585,
+      "grad_norm": 0.14029787233900184,
+      "learning_rate": 1.752251607771993e-06,
+      "loss": 1.212,
+      "step": 13070
+    },
+    {
+      "epoch": 0.9421898652057954,
+      "grad_norm": 0.11164084052517223,
+      "learning_rate": 1.7479028128387865e-06,
+      "loss": 1.3693,
+      "step": 13071
+    },
+    {
+      "epoch": 0.9422619476681323,
+      "grad_norm": 0.13075558680412877,
+      "learning_rate": 1.743559373523862e-06,
+      "loss": 1.2423,
+      "step": 13072
+    },
+    {
+      "epoch": 0.9423340301304692,
+      "grad_norm": 0.1371150044797531,
+      "learning_rate": 1.7392212900639859e-06,
+      "loss": 1.2292,
+      "step": 13073
+    },
+    {
+      "epoch": 0.9424061125928062,
+      "grad_norm": 0.1232247903115217,
+      "learning_rate": 1.7348885626956024e-06,
+      "loss": 1.2438,
+      "step": 13074
+    },
+    {
+      "epoch": 0.9424781950551431,
+      "grad_norm": 0.1289606977444358,
+      "learning_rate": 1.7305611916549114e-06,
+      "loss": 1.2427,
+      "step": 13075
+    },
+    {
+      "epoch": 0.94255027751748,
+      "grad_norm": 0.11226890446891863,
+      "learning_rate": 1.7262391771777687e-06,
+      "loss": 1.1968,
+      "step": 13076
+    },
+    {
+      "epoch": 0.9426223599798169,
+      "grad_norm": 0.12234350689470327,
+      "learning_rate": 1.7219225194997857e-06,
+      "loss": 1.3609,
+      "step": 13077
+    },
+    {
+      "epoch": 0.9426944424421538,
+      "grad_norm": 0.11256161651852702,
+      "learning_rate": 1.717611218856241e-06,
+      "loss": 1.2806,
+      "step": 13078
+    },
+    {
+      "epoch": 0.9427665249044908,
+      "grad_norm": 0.12969591096924943,
+      "learning_rate": 1.7133052754821355e-06,
+      "loss": 1.2954,
+      "step": 13079
+    },
+    {
+      "epoch": 0.9428386073668277,
+      "grad_norm": 0.12693087734468333,
+      "learning_rate": 1.7090046896121925e-06,
+      "loss": 1.2828,
+      "step": 13080
+    },
+    {
+      "epoch": 0.9429106898291646,
+      "grad_norm": 0.11648336643559024,
+      "learning_rate": 1.7047094614808357e-06,
+      "loss": 1.296,
+      "step": 13081
+    },
+    {
+      "epoch": 0.9429827722915015,
+      "grad_norm": 0.12538825825078978,
+      "learning_rate": 1.7004195913221888e-06,
+      "loss": 1.3592,
+      "step": 13082
+    },
+    {
+      "epoch": 0.9430548547538384,
+      "grad_norm": 0.12231941331537471,
+      "learning_rate": 1.696135079370076e-06,
+      "loss": 1.2024,
+      "step": 13083
+    },
+    {
+      "epoch": 0.9431269372161754,
+      "grad_norm": 0.12380947656631823,
+      "learning_rate": 1.6918559258580657e-06,
+      "loss": 1.3155,
+      "step": 13084
+    },
+    {
+      "epoch": 0.9431990196785122,
+      "grad_norm": 0.12717982855213744,
+      "learning_rate": 1.687582131019394e-06,
+      "loss": 1.3042,
+      "step": 13085
+    },
+    {
+      "epoch": 0.9432711021408491,
+      "grad_norm": 0.1267144075378771,
+      "learning_rate": 1.68331369508703e-06,
+      "loss": 1.4003,
+      "step": 13086
+    },
+    {
+      "epoch": 0.943343184603186,
+      "grad_norm": 0.12845973157024254,
+      "learning_rate": 1.6790506182936316e-06,
+      "loss": 1.2112,
+      "step": 13087
+    },
+    {
+      "epoch": 0.9434152670655229,
+      "grad_norm": 0.11434139786849351,
+      "learning_rate": 1.674792900871569e-06,
+      "loss": 1.2194,
+      "step": 13088
+    },
+    {
+      "epoch": 0.9434873495278598,
+      "grad_norm": 0.12451212140403616,
+      "learning_rate": 1.6705405430529452e-06,
+      "loss": 1.2825,
+      "step": 13089
+    },
+    {
+      "epoch": 0.9435594319901968,
+      "grad_norm": 0.1190600355416014,
+      "learning_rate": 1.6662935450695416e-06,
+      "loss": 1.2547,
+      "step": 13090
+    },
+    {
+      "epoch": 0.9436315144525337,
+      "grad_norm": 0.12349549294821027,
+      "learning_rate": 1.6620519071528506e-06,
+      "loss": 1.2939,
+      "step": 13091
+    },
+    {
+      "epoch": 0.9437035969148706,
+      "grad_norm": 0.12318328525938449,
+      "learning_rate": 1.6578156295340874e-06,
+      "loss": 1.3622,
+      "step": 13092
+    },
+    {
+      "epoch": 0.9437756793772075,
+      "grad_norm": 0.12175550951274684,
+      "learning_rate": 1.6535847124441783e-06,
+      "loss": 1.2768,
+      "step": 13093
+    },
+    {
+      "epoch": 0.9438477618395444,
+      "grad_norm": 0.13018128673400253,
+      "learning_rate": 1.6493591561137388e-06,
+      "loss": 1.2843,
+      "step": 13094
+    },
+    {
+      "epoch": 0.9439198443018814,
+      "grad_norm": 0.1179962926965692,
+      "learning_rate": 1.6451389607730849e-06,
+      "loss": 1.2333,
+      "step": 13095
+    },
+    {
+      "epoch": 0.9439919267642183,
+      "grad_norm": 0.12708836174139898,
+      "learning_rate": 1.6409241266522546e-06,
+      "loss": 1.3422,
+      "step": 13096
+    },
+    {
+      "epoch": 0.9440640092265552,
+      "grad_norm": 0.11173997637313562,
+      "learning_rate": 1.6367146539810196e-06,
+      "loss": 1.2481,
+      "step": 13097
+    },
+    {
+      "epoch": 0.9441360916888921,
+      "grad_norm": 0.12716666500899426,
+      "learning_rate": 1.6325105429888188e-06,
+      "loss": 1.2869,
+      "step": 13098
+    },
+    {
+      "epoch": 0.944208174151229,
+      "grad_norm": 0.13379187408664128,
+      "learning_rate": 1.6283117939048131e-06,
+      "loss": 1.2801,
+      "step": 13099
+    },
+    {
+      "epoch": 0.944280256613566,
+      "grad_norm": 0.12170247355878312,
+      "learning_rate": 1.6241184069578641e-06,
+      "loss": 1.3448,
+      "step": 13100
+    },
+    {
+      "epoch": 0.9443523390759029,
+      "grad_norm": 0.13108109155563058,
+      "learning_rate": 1.6199303823765665e-06,
+      "loss": 1.1968,
+      "step": 13101
+    },
+    {
+      "epoch": 0.9444244215382398,
+      "grad_norm": 0.12523210791497505,
+      "learning_rate": 1.6157477203891936e-06,
+      "loss": 1.2818,
+      "step": 13102
+    },
+    {
+      "epoch": 0.9444965040005766,
+      "grad_norm": 0.10991735547226031,
+      "learning_rate": 1.6115704212237403e-06,
+      "loss": 1.2705,
+      "step": 13103
+    },
+    {
+      "epoch": 0.9445685864629135,
+      "grad_norm": 0.11128681571258303,
+      "learning_rate": 1.6073984851079139e-06,
+      "loss": 1.2859,
+      "step": 13104
+    },
+    {
+      "epoch": 0.9446406689252504,
+      "grad_norm": 0.12449771139042304,
+      "learning_rate": 1.6032319122691097e-06,
+      "loss": 1.2388,
+      "step": 13105
+    },
+    {
+      "epoch": 0.9447127513875874,
+      "grad_norm": 0.10673256292017894,
+      "learning_rate": 1.5990707029344464e-06,
+      "loss": 1.2809,
+      "step": 13106
+    },
+    {
+      "epoch": 0.9447848338499243,
+      "grad_norm": 0.14023317188396173,
+      "learning_rate": 1.5949148573307538e-06,
+      "loss": 1.3245,
+      "step": 13107
+    },
+    {
+      "epoch": 0.9448569163122612,
+      "grad_norm": 0.1207928878770872,
+      "learning_rate": 1.5907643756845503e-06,
+      "loss": 1.1709,
+      "step": 13108
+    },
+    {
+      "epoch": 0.9449289987745981,
+      "grad_norm": 0.12761303347801306,
+      "learning_rate": 1.5866192582220884e-06,
+      "loss": 1.2748,
+      "step": 13109
+    },
+    {
+      "epoch": 0.945001081236935,
+      "grad_norm": 0.11614890696140653,
+      "learning_rate": 1.5824795051693098e-06,
+      "loss": 1.2766,
+      "step": 13110
+    },
+    {
+      "epoch": 0.945073163699272,
+      "grad_norm": 0.11094606639340836,
+      "learning_rate": 1.578345116751867e-06,
+      "loss": 1.4055,
+      "step": 13111
+    },
+    {
+      "epoch": 0.9451452461616089,
+      "grad_norm": 0.13574835702126262,
+      "learning_rate": 1.5742160931951133e-06,
+      "loss": 1.2489,
+      "step": 13112
+    },
+    {
+      "epoch": 0.9452173286239458,
+      "grad_norm": 0.12903157598265133,
+      "learning_rate": 1.5700924347241353e-06,
+      "loss": 1.2435,
+      "step": 13113
+    },
+    {
+      "epoch": 0.9452894110862827,
+      "grad_norm": 0.12404897993328885,
+      "learning_rate": 1.5659741415636975e-06,
+      "loss": 1.3127,
+      "step": 13114
+    },
+    {
+      "epoch": 0.9453614935486196,
+      "grad_norm": 0.12986251362688045,
+      "learning_rate": 1.561861213938276e-06,
+      "loss": 1.2246,
+      "step": 13115
+    },
+    {
+      "epoch": 0.9454335760109566,
+      "grad_norm": 0.11575965661660095,
+      "learning_rate": 1.5577536520720803e-06,
+      "loss": 1.3509,
+      "step": 13116
+    },
+    {
+      "epoch": 0.9455056584732935,
+      "grad_norm": 0.12297670876512007,
+      "learning_rate": 1.5536514561889982e-06,
+      "loss": 1.2845,
+      "step": 13117
+    },
+    {
+      "epoch": 0.9455777409356304,
+      "grad_norm": 0.11533412806509262,
+      "learning_rate": 1.5495546265126283e-06,
+      "loss": 1.292,
+      "step": 13118
+    },
+    {
+      "epoch": 0.9456498233979673,
+      "grad_norm": 0.11675938167164081,
+      "learning_rate": 1.545463163266303e-06,
+      "loss": 1.2375,
+      "step": 13119
+    },
+    {
+      "epoch": 0.9457219058603041,
+      "grad_norm": 0.12674584461672983,
+      "learning_rate": 1.5413770666730331e-06,
+      "loss": 1.2341,
+      "step": 13120
+    },
+    {
+      "epoch": 0.945793988322641,
+      "grad_norm": 0.11380818215581136,
+      "learning_rate": 1.5372963369555516e-06,
+      "loss": 1.2822,
+      "step": 13121
+    },
+    {
+      "epoch": 0.945866070784978,
+      "grad_norm": 0.12703441863389786,
+      "learning_rate": 1.5332209743362914e-06,
+      "loss": 1.3038,
+      "step": 13122
+    },
+    {
+      "epoch": 0.9459381532473149,
+      "grad_norm": 0.12403258990601569,
+      "learning_rate": 1.529150979037397e-06,
+      "loss": 1.29,
+      "step": 13123
+    },
+    {
+      "epoch": 0.9460102357096518,
+      "grad_norm": 0.13686828123232406,
+      "learning_rate": 1.5250863512807134e-06,
+      "loss": 1.3366,
+      "step": 13124
+    },
+    {
+      "epoch": 0.9460823181719887,
+      "grad_norm": 0.13424757464620574,
+      "learning_rate": 1.5210270912878077e-06,
+      "loss": 1.1573,
+      "step": 13125
+    },
+    {
+      "epoch": 0.9461544006343257,
+      "grad_norm": 0.12357593668995885,
+      "learning_rate": 1.5169731992799585e-06,
+      "loss": 1.3466,
+      "step": 13126
+    },
+    {
+      "epoch": 0.9462264830966626,
+      "grad_norm": 0.12858651721424721,
+      "learning_rate": 1.5129246754781001e-06,
+      "loss": 1.2023,
+      "step": 13127
+    },
+    {
+      "epoch": 0.9462985655589995,
+      "grad_norm": 0.12640494319773393,
+      "learning_rate": 1.508881520102956e-06,
+      "loss": 1.1831,
+      "step": 13128
+    },
+    {
+      "epoch": 0.9463706480213364,
+      "grad_norm": 0.12122013406581023,
+      "learning_rate": 1.5048437333748833e-06,
+      "loss": 1.2994,
+      "step": 13129
+    },
+    {
+      "epoch": 0.9464427304836733,
+      "grad_norm": 0.11299238595851202,
+      "learning_rate": 1.5008113155140058e-06,
+      "loss": 1.2986,
+      "step": 13130
+    },
+    {
+      "epoch": 0.9465148129460103,
+      "grad_norm": 0.12594571563307122,
+      "learning_rate": 1.4967842667401034e-06,
+      "loss": 1.2854,
+      "step": 13131
+    },
+    {
+      "epoch": 0.9465868954083472,
+      "grad_norm": 0.11701716237436746,
+      "learning_rate": 1.492762587272678e-06,
+      "loss": 1.2238,
+      "step": 13132
+    },
+    {
+      "epoch": 0.9466589778706841,
+      "grad_norm": 0.1356011347216867,
+      "learning_rate": 1.4887462773309767e-06,
+      "loss": 1.2545,
+      "step": 13133
+    },
+    {
+      "epoch": 0.946731060333021,
+      "grad_norm": 0.14423919677413724,
+      "learning_rate": 1.4847353371339135e-06,
+      "loss": 1.2378,
+      "step": 13134
+    },
+    {
+      "epoch": 0.9468031427953579,
+      "grad_norm": 0.12004331472427973,
+      "learning_rate": 1.480729766900113e-06,
+      "loss": 1.209,
+      "step": 13135
+    },
+    {
+      "epoch": 0.9468752252576949,
+      "grad_norm": 0.1272238345454641,
+      "learning_rate": 1.476729566847912e-06,
+      "loss": 1.3164,
+      "step": 13136
+    },
+    {
+      "epoch": 0.9469473077200317,
+      "grad_norm": 0.11962770455021281,
+      "learning_rate": 1.4727347371953692e-06,
+      "loss": 1.2399,
+      "step": 13137
+    },
+    {
+      "epoch": 0.9470193901823686,
+      "grad_norm": 0.10970770546382808,
+      "learning_rate": 1.468745278160244e-06,
+      "loss": 1.2463,
+      "step": 13138
+    },
+    {
+      "epoch": 0.9470914726447055,
+      "grad_norm": 0.13071531445477794,
+      "learning_rate": 1.4647611899599734e-06,
+      "loss": 1.3018,
+      "step": 13139
+    },
+    {
+      "epoch": 0.9471635551070424,
+      "grad_norm": 0.12865469934866616,
+      "learning_rate": 1.4607824728117392e-06,
+      "loss": 1.3287,
+      "step": 13140
+    },
+    {
+      "epoch": 0.9472356375693793,
+      "grad_norm": 0.13736423283690688,
+      "learning_rate": 1.4568091269324124e-06,
+      "loss": 1.272,
+      "step": 13141
+    },
+    {
+      "epoch": 0.9473077200317163,
+      "grad_norm": 0.12187481084859432,
+      "learning_rate": 1.4528411525385976e-06,
+      "loss": 1.2973,
+      "step": 13142
+    },
+    {
+      "epoch": 0.9473798024940532,
+      "grad_norm": 0.14596506522036534,
+      "learning_rate": 1.4488785498465663e-06,
+      "loss": 1.2577,
+      "step": 13143
+    },
+    {
+      "epoch": 0.9474518849563901,
+      "grad_norm": 0.12501283239261343,
+      "learning_rate": 1.4449213190723121e-06,
+      "loss": 1.283,
+      "step": 13144
+    },
+    {
+      "epoch": 0.947523967418727,
+      "grad_norm": 0.14556448307427083,
+      "learning_rate": 1.4409694604315294e-06,
+      "loss": 1.2551,
+      "step": 13145
+    },
+    {
+      "epoch": 0.9475960498810639,
+      "grad_norm": 0.11985542512891886,
+      "learning_rate": 1.437022974139668e-06,
+      "loss": 1.275,
+      "step": 13146
+    },
+    {
+      "epoch": 0.9476681323434009,
+      "grad_norm": 0.13765949374681005,
+      "learning_rate": 1.4330818604118113e-06,
+      "loss": 1.2687,
+      "step": 13147
+    },
+    {
+      "epoch": 0.9477402148057378,
+      "grad_norm": 0.11979652221949123,
+      "learning_rate": 1.4291461194628098e-06,
+      "loss": 1.2391,
+      "step": 13148
+    },
+    {
+      "epoch": 0.9478122972680747,
+      "grad_norm": 0.13004442084537718,
+      "learning_rate": 1.4252157515071697e-06,
+      "loss": 1.2321,
+      "step": 13149
+    },
+    {
+      "epoch": 0.9478843797304116,
+      "grad_norm": 0.12690256676438033,
+      "learning_rate": 1.4212907567591638e-06,
+      "loss": 1.2368,
+      "step": 13150
+    },
+    {
+      "epoch": 0.9479564621927485,
+      "grad_norm": 0.10512681293472481,
+      "learning_rate": 1.4173711354327101e-06,
+      "loss": 1.1626,
+      "step": 13151
+    },
+    {
+      "epoch": 0.9480285446550855,
+      "grad_norm": 0.12933841125048154,
+      "learning_rate": 1.4134568877414711e-06,
+      "loss": 1.3026,
+      "step": 13152
+    },
+    {
+      "epoch": 0.9481006271174224,
+      "grad_norm": 0.13866475424683572,
+      "learning_rate": 1.4095480138988203e-06,
+      "loss": 1.2064,
+      "step": 13153
+    },
+    {
+      "epoch": 0.9481727095797593,
+      "grad_norm": 0.1221542122849753,
+      "learning_rate": 1.4056445141178208e-06,
+      "loss": 1.3281,
+      "step": 13154
+    },
+    {
+      "epoch": 0.9482447920420961,
+      "grad_norm": 0.12529034182805976,
+      "learning_rate": 1.401746388611236e-06,
+      "loss": 1.3549,
+      "step": 13155
+    },
+    {
+      "epoch": 0.948316874504433,
+      "grad_norm": 0.12576194497198484,
+      "learning_rate": 1.397853637591562e-06,
+      "loss": 1.2751,
+      "step": 13156
+    },
+    {
+      "epoch": 0.94838895696677,
+      "grad_norm": 0.11715095555527194,
+      "learning_rate": 1.393966261270996e-06,
+      "loss": 1.2559,
+      "step": 13157
+    },
+    {
+      "epoch": 0.9484610394291069,
+      "grad_norm": 0.1285191294460064,
+      "learning_rate": 1.3900842598614128e-06,
+      "loss": 1.2127,
+      "step": 13158
+    },
+    {
+      "epoch": 0.9485331218914438,
+      "grad_norm": 0.11232450388653953,
+      "learning_rate": 1.3862076335744211e-06,
+      "loss": 1.3053,
+      "step": 13159
+    },
+    {
+      "epoch": 0.9486052043537807,
+      "grad_norm": 0.1277110821752712,
+      "learning_rate": 1.382336382621341e-06,
+      "loss": 1.2432,
+      "step": 13160
+    },
+    {
+      "epoch": 0.9486772868161176,
+      "grad_norm": 0.12924334532682485,
+      "learning_rate": 1.3784705072131921e-06,
+      "loss": 1.2729,
+      "step": 13161
+    },
+    {
+      "epoch": 0.9487493692784545,
+      "grad_norm": 0.1330799437774743,
+      "learning_rate": 1.3746100075606949e-06,
+      "loss": 1.3703,
+      "step": 13162
+    },
+    {
+      "epoch": 0.9488214517407915,
+      "grad_norm": 0.12405870471471188,
+      "learning_rate": 1.3707548838742811e-06,
+      "loss": 1.2055,
+      "step": 13163
+    },
+    {
+      "epoch": 0.9488935342031284,
+      "grad_norm": 0.11547003770247387,
+      "learning_rate": 1.3669051363640827e-06,
+      "loss": 1.2727,
+      "step": 13164
+    },
+    {
+      "epoch": 0.9489656166654653,
+      "grad_norm": 0.12883187388537962,
+      "learning_rate": 1.363060765239954e-06,
+      "loss": 1.1949,
+      "step": 13165
+    },
+    {
+      "epoch": 0.9490376991278022,
+      "grad_norm": 0.13732579302470282,
+      "learning_rate": 1.3592217707114385e-06,
+      "loss": 1.2399,
+      "step": 13166
+    },
+    {
+      "epoch": 0.9491097815901391,
+      "grad_norm": 0.11607316620466322,
+      "learning_rate": 1.3553881529878132e-06,
+      "loss": 1.1115,
+      "step": 13167
+    },
+    {
+      "epoch": 0.9491818640524761,
+      "grad_norm": 0.11953973522842734,
+      "learning_rate": 1.351559912278022e-06,
+      "loss": 1.2601,
+      "step": 13168
+    },
+    {
+      "epoch": 0.949253946514813,
+      "grad_norm": 0.13125038676087586,
+      "learning_rate": 1.3477370487907537e-06,
+      "loss": 1.241,
+      "step": 13169
+    },
+    {
+      "epoch": 0.9493260289771499,
+      "grad_norm": 0.13253238476438017,
+      "learning_rate": 1.343919562734397e-06,
+      "loss": 1.2622,
+      "step": 13170
+    },
+    {
+      "epoch": 0.9493981114394868,
+      "grad_norm": 0.11129772355030061,
+      "learning_rate": 1.3401074543170078e-06,
+      "loss": 1.2674,
+      "step": 13171
+    },
+    {
+      "epoch": 0.9494701939018236,
+      "grad_norm": 0.12899773278186172,
+      "learning_rate": 1.3363007237464086e-06,
+      "loss": 1.2633,
+      "step": 13172
+    },
+    {
+      "epoch": 0.9495422763641606,
+      "grad_norm": 0.11325648818698522,
+      "learning_rate": 1.3324993712300892e-06,
+      "loss": 1.3691,
+      "step": 13173
+    },
+    {
+      "epoch": 0.9496143588264975,
+      "grad_norm": 0.12020811522592234,
+      "learning_rate": 1.3287033969752616e-06,
+      "loss": 1.2317,
+      "step": 13174
+    },
+    {
+      "epoch": 0.9496864412888344,
+      "grad_norm": 0.12643321146851874,
+      "learning_rate": 1.324912801188838e-06,
+      "loss": 1.3023,
+      "step": 13175
+    },
+    {
+      "epoch": 0.9497585237511713,
+      "grad_norm": 0.12666572008514884,
+      "learning_rate": 1.3211275840774307e-06,
+      "loss": 1.3243,
+      "step": 13176
+    },
+    {
+      "epoch": 0.9498306062135082,
+      "grad_norm": 0.1099644234545342,
+      "learning_rate": 1.317347745847386e-06,
+      "loss": 1.1968,
+      "step": 13177
+    },
+    {
+      "epoch": 0.9499026886758452,
+      "grad_norm": 0.12449543256731932,
+      "learning_rate": 1.313573286704728e-06,
+      "loss": 1.2171,
+      "step": 13178
+    },
+    {
+      "epoch": 0.9499747711381821,
+      "grad_norm": 0.10803908763650132,
+      "learning_rate": 1.309804206855203e-06,
+      "loss": 1.2744,
+      "step": 13179
+    },
+    {
+      "epoch": 0.950046853600519,
+      "grad_norm": 0.12112694478954235,
+      "learning_rate": 1.3060405065042358e-06,
+      "loss": 1.2696,
+      "step": 13180
+    },
+    {
+      "epoch": 0.9501189360628559,
+      "grad_norm": 0.12426706637816438,
+      "learning_rate": 1.3022821858570178e-06,
+      "loss": 1.3079,
+      "step": 13181
+    },
+    {
+      "epoch": 0.9501910185251928,
+      "grad_norm": 0.12253065258736538,
+      "learning_rate": 1.2985292451183962e-06,
+      "loss": 1.2863,
+      "step": 13182
+    },
+    {
+      "epoch": 0.9502631009875298,
+      "grad_norm": 0.12029212025701856,
+      "learning_rate": 1.2947816844929517e-06,
+      "loss": 1.2532,
+      "step": 13183
+    },
+    {
+      "epoch": 0.9503351834498667,
+      "grad_norm": 0.11653488458441551,
+      "learning_rate": 1.291039504184921e-06,
+      "loss": 1.1299,
+      "step": 13184
+    },
+    {
+      "epoch": 0.9504072659122036,
+      "grad_norm": 0.11927574665917591,
+      "learning_rate": 1.28730270439833e-06,
+      "loss": 1.3446,
+      "step": 13185
+    },
+    {
+      "epoch": 0.9504793483745405,
+      "grad_norm": 0.12836820656059744,
+      "learning_rate": 1.2835712853368487e-06,
+      "loss": 1.2717,
+      "step": 13186
+    },
+    {
+      "epoch": 0.9505514308368774,
+      "grad_norm": 0.1280288888986278,
+      "learning_rate": 1.2798452472038814e-06,
+      "loss": 1.2498,
+      "step": 13187
+    },
+    {
+      "epoch": 0.9506235132992144,
+      "grad_norm": 0.12634806342454116,
+      "learning_rate": 1.27612459020251e-06,
+      "loss": 1.2151,
+      "step": 13188
+    },
+    {
+      "epoch": 0.9506955957615512,
+      "grad_norm": 0.12281409604884937,
+      "learning_rate": 1.272409314535561e-06,
+      "loss": 1.2859,
+      "step": 13189
+    },
+    {
+      "epoch": 0.9507676782238881,
+      "grad_norm": 0.13342663536483962,
+      "learning_rate": 1.2686994204055503e-06,
+      "loss": 1.3022,
+      "step": 13190
+    },
+    {
+      "epoch": 0.950839760686225,
+      "grad_norm": 0.12153359980140123,
+      "learning_rate": 1.2649949080146938e-06,
+      "loss": 1.3847,
+      "step": 13191
+    },
+    {
+      "epoch": 0.9509118431485619,
+      "grad_norm": 0.11432562991233494,
+      "learning_rate": 1.2612957775649302e-06,
+      "loss": 1.2102,
+      "step": 13192
+    },
+    {
+      "epoch": 0.9509839256108988,
+      "grad_norm": 0.11385883544835144,
+      "learning_rate": 1.2576020292578649e-06,
+      "loss": 1.2834,
+      "step": 13193
+    },
+    {
+      "epoch": 0.9510560080732358,
+      "grad_norm": 0.11378009093780986,
+      "learning_rate": 1.2539136632948923e-06,
+      "loss": 1.2588,
+      "step": 13194
+    },
+    {
+      "epoch": 0.9511280905355727,
+      "grad_norm": 0.11224078590774277,
+      "learning_rate": 1.250230679877018e-06,
+      "loss": 1.302,
+      "step": 13195
+    },
+    {
+      "epoch": 0.9512001729979096,
+      "grad_norm": 0.21187020495144884,
+      "learning_rate": 1.2465530792050262e-06,
+      "loss": 1.2335,
+      "step": 13196
+    },
+    {
+      "epoch": 0.9512722554602465,
+      "grad_norm": 0.11929282181937421,
+      "learning_rate": 1.2428808614793452e-06,
+      "loss": 1.3399,
+      "step": 13197
+    },
+    {
+      "epoch": 0.9513443379225834,
+      "grad_norm": 0.12918189305189362,
+      "learning_rate": 1.2392140269001817e-06,
+      "loss": 1.3044,
+      "step": 13198
+    },
+    {
+      "epoch": 0.9514164203849204,
+      "grad_norm": 0.11379584158962186,
+      "learning_rate": 1.2355525756673869e-06,
+      "loss": 1.2643,
+      "step": 13199
+    },
+    {
+      "epoch": 0.9514885028472573,
+      "grad_norm": 0.12382081652029664,
+      "learning_rate": 1.2318965079805456e-06,
+      "loss": 1.2727,
+      "step": 13200
+    },
+    {
+      "epoch": 0.9515605853095942,
+      "grad_norm": 0.11299703389283458,
+      "learning_rate": 1.2282458240389427e-06,
+      "loss": 1.1592,
+      "step": 13201
+    },
+    {
+      "epoch": 0.9516326677719311,
+      "grad_norm": 0.1232839518968892,
+      "learning_rate": 1.2246005240415858e-06,
+      "loss": 1.1699,
+      "step": 13202
+    },
+    {
+      "epoch": 0.951704750234268,
+      "grad_norm": 0.1245331131532943,
+      "learning_rate": 1.2209606081871717e-06,
+      "loss": 1.2908,
+      "step": 13203
+    },
+    {
+      "epoch": 0.951776832696605,
+      "grad_norm": 0.12139797428087543,
+      "learning_rate": 1.217326076674108e-06,
+      "loss": 1.1954,
+      "step": 13204
+    },
+    {
+      "epoch": 0.9518489151589419,
+      "grad_norm": 0.1324547574247266,
+      "learning_rate": 1.2136969297005141e-06,
+      "loss": 1.3293,
+      "step": 13205
+    },
+    {
+      "epoch": 0.9519209976212788,
+      "grad_norm": 0.11796604423880601,
+      "learning_rate": 1.2100731674641984e-06,
+      "loss": 1.1704,
+      "step": 13206
+    },
+    {
+      "epoch": 0.9519930800836156,
+      "grad_norm": 0.11557833598734023,
+      "learning_rate": 1.206454790162692e-06,
+      "loss": 1.2789,
+      "step": 13207
+    },
+    {
+      "epoch": 0.9520651625459525,
+      "grad_norm": 0.12198625487380538,
+      "learning_rate": 1.2028417979932372e-06,
+      "loss": 1.3597,
+      "step": 13208
+    },
+    {
+      "epoch": 0.9521372450082894,
+      "grad_norm": 0.12706998705982275,
+      "learning_rate": 1.1992341911527649e-06,
+      "loss": 1.3553,
+      "step": 13209
+    },
+    {
+      "epoch": 0.9522093274706264,
+      "grad_norm": 0.12447979491139924,
+      "learning_rate": 1.1956319698379293e-06,
+      "loss": 1.2705,
+      "step": 13210
+    },
+    {
+      "epoch": 0.9522814099329633,
+      "grad_norm": 0.13294813947787276,
+      "learning_rate": 1.192035134245073e-06,
+      "loss": 1.3172,
+      "step": 13211
+    },
+    {
+      "epoch": 0.9523534923953002,
+      "grad_norm": 0.12458706095837785,
+      "learning_rate": 1.188443684570273e-06,
+      "loss": 1.2225,
+      "step": 13212
+    },
+    {
+      "epoch": 0.9524255748576371,
+      "grad_norm": 0.12005593549681114,
+      "learning_rate": 1.1848576210092833e-06,
+      "loss": 1.2077,
+      "step": 13213
+    },
+    {
+      "epoch": 0.952497657319974,
+      "grad_norm": 0.13223397596735248,
+      "learning_rate": 1.1812769437575811e-06,
+      "loss": 1.2737,
+      "step": 13214
+    },
+    {
+      "epoch": 0.952569739782311,
+      "grad_norm": 0.12469778122179548,
+      "learning_rate": 1.1777016530103325e-06,
+      "loss": 1.2355,
+      "step": 13215
+    },
+    {
+      "epoch": 0.9526418222446479,
+      "grad_norm": 0.12854355311915747,
+      "learning_rate": 1.174131748962437e-06,
+      "loss": 1.2152,
+      "step": 13216
+    },
+    {
+      "epoch": 0.9527139047069848,
+      "grad_norm": 0.12631174689361782,
+      "learning_rate": 1.1705672318084947e-06,
+      "loss": 1.2015,
+      "step": 13217
+    },
+    {
+      "epoch": 0.9527859871693217,
+      "grad_norm": 0.11321160342372617,
+      "learning_rate": 1.1670081017427836e-06,
+      "loss": 1.2195,
+      "step": 13218
+    },
+    {
+      "epoch": 0.9528580696316586,
+      "grad_norm": 0.12345913684620101,
+      "learning_rate": 1.163454358959315e-06,
+      "loss": 1.3428,
+      "step": 13219
+    },
+    {
+      "epoch": 0.9529301520939956,
+      "grad_norm": 0.23451546602488862,
+      "learning_rate": 1.1599060036518005e-06,
+      "loss": 1.2248,
+      "step": 13220
+    },
+    {
+      "epoch": 0.9530022345563325,
+      "grad_norm": 0.1318520230024285,
+      "learning_rate": 1.1563630360136523e-06,
+      "loss": 1.1513,
+      "step": 13221
+    },
+    {
+      "epoch": 0.9530743170186694,
+      "grad_norm": 0.13038066082577804,
+      "learning_rate": 1.1528254562380048e-06,
+      "loss": 1.2273,
+      "step": 13222
+    },
+    {
+      "epoch": 0.9531463994810063,
+      "grad_norm": 0.11855208456784533,
+      "learning_rate": 1.1492932645176924e-06,
+      "loss": 1.3312,
+      "step": 13223
+    },
+    {
+      "epoch": 0.9532184819433431,
+      "grad_norm": 0.12662037077156466,
+      "learning_rate": 1.1457664610452168e-06,
+      "loss": 1.2306,
+      "step": 13224
+    },
+    {
+      "epoch": 0.95329056440568,
+      "grad_norm": 0.11265521611071352,
+      "learning_rate": 1.1422450460128687e-06,
+      "loss": 1.2927,
+      "step": 13225
+    },
+    {
+      "epoch": 0.953362646868017,
+      "grad_norm": 0.11136915983190218,
+      "learning_rate": 1.138729019612561e-06,
+      "loss": 1.32,
+      "step": 13226
+    },
+    {
+      "epoch": 0.9534347293303539,
+      "grad_norm": 0.12580268309316195,
+      "learning_rate": 1.135218382035963e-06,
+      "loss": 1.3479,
+      "step": 13227
+    },
+    {
+      "epoch": 0.9535068117926908,
+      "grad_norm": 0.1167799017098658,
+      "learning_rate": 1.1317131334744212e-06,
+      "loss": 1.2917,
+      "step": 13228
+    },
+    {
+      "epoch": 0.9535788942550277,
+      "grad_norm": 0.11727473178199266,
+      "learning_rate": 1.1282132741190278e-06,
+      "loss": 1.23,
+      "step": 13229
+    },
+    {
+      "epoch": 0.9536509767173647,
+      "grad_norm": 0.13025119594127937,
+      "learning_rate": 1.1247188041605405e-06,
+      "loss": 1.2443,
+      "step": 13230
+    },
+    {
+      "epoch": 0.9537230591797016,
+      "grad_norm": 0.1261176890392211,
+      "learning_rate": 1.121229723789441e-06,
+      "loss": 1.2424,
+      "step": 13231
+    },
+    {
+      "epoch": 0.9537951416420385,
+      "grad_norm": 0.11591993269460926,
+      "learning_rate": 1.1177460331959212e-06,
+      "loss": 1.2241,
+      "step": 13232
+    },
+    {
+      "epoch": 0.9538672241043754,
+      "grad_norm": 0.12936515052132394,
+      "learning_rate": 1.1142677325698515e-06,
+      "loss": 1.2344,
+      "step": 13233
+    },
+    {
+      "epoch": 0.9539393065667123,
+      "grad_norm": 0.13995165680652763,
+      "learning_rate": 1.110794822100858e-06,
+      "loss": 1.2179,
+      "step": 13234
+    },
+    {
+      "epoch": 0.9540113890290493,
+      "grad_norm": 0.12955460296382315,
+      "learning_rate": 1.107327301978245e-06,
+      "loss": 1.293,
+      "step": 13235
+    },
+    {
+      "epoch": 0.9540834714913862,
+      "grad_norm": 0.12179729350330452,
+      "learning_rate": 1.1038651723910055e-06,
+      "loss": 1.2156,
+      "step": 13236
+    },
+    {
+      "epoch": 0.9541555539537231,
+      "grad_norm": 0.11493712663183085,
+      "learning_rate": 1.1004084335278553e-06,
+      "loss": 1.2506,
+      "step": 13237
+    },
+    {
+      "epoch": 0.95422763641606,
+      "grad_norm": 0.204857802253856,
+      "learning_rate": 1.0969570855772327e-06,
+      "loss": 1.217,
+      "step": 13238
+    },
+    {
+      "epoch": 0.9542997188783969,
+      "grad_norm": 0.12853260956995455,
+      "learning_rate": 1.093511128727276e-06,
+      "loss": 1.1923,
+      "step": 13239
+    },
+    {
+      "epoch": 0.9543718013407338,
+      "grad_norm": 0.12919513432939503,
+      "learning_rate": 1.0900705631657904e-06,
+      "loss": 1.2508,
+      "step": 13240
+    },
+    {
+      "epoch": 0.9544438838030707,
+      "grad_norm": 0.1162233900658442,
+      "learning_rate": 1.086635389080326e-06,
+      "loss": 1.2831,
+      "step": 13241
+    },
+    {
+      "epoch": 0.9545159662654076,
+      "grad_norm": 0.11791920104182664,
+      "learning_rate": 1.0832056066581442e-06,
+      "loss": 1.2775,
+      "step": 13242
+    },
+    {
+      "epoch": 0.9545880487277445,
+      "grad_norm": 0.12107875813923959,
+      "learning_rate": 1.0797812160861954e-06,
+      "loss": 1.289,
+      "step": 13243
+    },
+    {
+      "epoch": 0.9546601311900814,
+      "grad_norm": 0.11281499399859585,
+      "learning_rate": 1.0763622175511413e-06,
+      "loss": 1.2365,
+      "step": 13244
+    },
+    {
+      "epoch": 0.9547322136524183,
+      "grad_norm": 0.11783832881484402,
+      "learning_rate": 1.0729486112393327e-06,
+      "loss": 1.2776,
+      "step": 13245
+    },
+    {
+      "epoch": 0.9548042961147553,
+      "grad_norm": 0.12019129244527663,
+      "learning_rate": 1.0695403973368434e-06,
+      "loss": 1.2836,
+      "step": 13246
+    },
+    {
+      "epoch": 0.9548763785770922,
+      "grad_norm": 0.12742366029297383,
+      "learning_rate": 1.0661375760294801e-06,
+      "loss": 1.2701,
+      "step": 13247
+    },
+    {
+      "epoch": 0.9549484610394291,
+      "grad_norm": 0.11459061253695331,
+      "learning_rate": 1.0627401475026944e-06,
+      "loss": 1.3327,
+      "step": 13248
+    },
+    {
+      "epoch": 0.955020543501766,
+      "grad_norm": 0.11124892880500692,
+      "learning_rate": 1.0593481119416826e-06,
+      "loss": 1.2954,
+      "step": 13249
+    },
+    {
+      "epoch": 0.9550926259641029,
+      "grad_norm": 0.1158355211313951,
+      "learning_rate": 1.0559614695313525e-06,
+      "loss": 1.3132,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9551647084264399,
+      "grad_norm": 0.14529694194134574,
+      "learning_rate": 1.0525802204563118e-06,
+      "loss": 1.3302,
+      "step": 13251
+    },
+    {
+      "epoch": 0.9552367908887768,
+      "grad_norm": 0.12146282231604526,
+      "learning_rate": 1.0492043649008464e-06,
+      "loss": 1.2675,
+      "step": 13252
+    },
+    {
+      "epoch": 0.9553088733511137,
+      "grad_norm": 0.13390531630131786,
+      "learning_rate": 1.0458339030489872e-06,
+      "loss": 1.3245,
+      "step": 13253
+    },
+    {
+      "epoch": 0.9553809558134506,
+      "grad_norm": 0.11253288711300731,
+      "learning_rate": 1.0424688350844425e-06,
+      "loss": 1.2631,
+      "step": 13254
+    },
+    {
+      "epoch": 0.9554530382757875,
+      "grad_norm": 0.11388204274402922,
+      "learning_rate": 1.0391091611906544e-06,
+      "loss": 1.3689,
+      "step": 13255
+    },
+    {
+      "epoch": 0.9555251207381245,
+      "grad_norm": 0.11823525269818456,
+      "learning_rate": 1.0357548815507434e-06,
+      "loss": 1.2535,
+      "step": 13256
+    },
+    {
+      "epoch": 0.9555972032004614,
+      "grad_norm": 0.11063655910143945,
+      "learning_rate": 1.0324059963475518e-06,
+      "loss": 1.1258,
+      "step": 13257
+    },
+    {
+      "epoch": 0.9556692856627983,
+      "grad_norm": 0.12585074038953983,
+      "learning_rate": 1.0290625057636227e-06,
+      "loss": 1.2589,
+      "step": 13258
+    },
+    {
+      "epoch": 0.9557413681251351,
+      "grad_norm": 0.1296642559264927,
+      "learning_rate": 1.025724409981199e-06,
+      "loss": 1.1369,
+      "step": 13259
+    },
+    {
+      "epoch": 0.955813450587472,
+      "grad_norm": 0.11521535053089993,
+      "learning_rate": 1.0223917091822466e-06,
+      "loss": 1.2009,
+      "step": 13260
+    },
+    {
+      "epoch": 0.9558855330498089,
+      "grad_norm": 0.1203018225433087,
+      "learning_rate": 1.0190644035484308e-06,
+      "loss": 1.2693,
+      "step": 13261
+    },
+    {
+      "epoch": 0.9559576155121459,
+      "grad_norm": 0.11023201002407801,
+      "learning_rate": 1.015742493261107e-06,
+      "loss": 1.3405,
+      "step": 13262
+    },
+    {
+      "epoch": 0.9560296979744828,
+      "grad_norm": 0.12077412246139554,
+      "learning_rate": 1.0124259785013635e-06,
+      "loss": 1.3232,
+      "step": 13263
+    },
+    {
+      "epoch": 0.9561017804368197,
+      "grad_norm": 0.11718137894061184,
+      "learning_rate": 1.0091148594499666e-06,
+      "loss": 1.2244,
+      "step": 13264
+    },
+    {
+      "epoch": 0.9561738628991566,
+      "grad_norm": 0.12509684505008623,
+      "learning_rate": 1.0058091362874168e-06,
+      "loss": 1.2476,
+      "step": 13265
+    },
+    {
+      "epoch": 0.9562459453614935,
+      "grad_norm": 0.12290466525163611,
+      "learning_rate": 1.0025088091938806e-06,
+      "loss": 1.2375,
+      "step": 13266
+    },
+    {
+      "epoch": 0.9563180278238305,
+      "grad_norm": 0.1120026035003126,
+      "learning_rate": 9.99213878349281e-07,
+      "loss": 1.2805,
+      "step": 13267
+    },
+    {
+      "epoch": 0.9563901102861674,
+      "grad_norm": 0.12882809620642144,
+      "learning_rate": 9.95924343933219e-07,
+      "loss": 1.2339,
+      "step": 13268
+    },
+    {
+      "epoch": 0.9564621927485043,
+      "grad_norm": 0.12061230223016113,
+      "learning_rate": 9.92640206124984e-07,
+      "loss": 1.211,
+      "step": 13269
+    },
+    {
+      "epoch": 0.9565342752108412,
+      "grad_norm": 0.1259500615586966,
+      "learning_rate": 9.893614651035998e-07,
+      "loss": 1.3182,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9566063576731781,
+      "grad_norm": 0.12490862097088644,
+      "learning_rate": 9.860881210478012e-07,
+      "loss": 1.2554,
+      "step": 13271
+    },
+    {
+      "epoch": 0.9566784401355151,
+      "grad_norm": 0.12451350049743555,
+      "learning_rate": 9.82820174136001e-07,
+      "loss": 1.3489,
+      "step": 13272
+    },
+    {
+      "epoch": 0.956750522597852,
+      "grad_norm": 0.14743132868639444,
+      "learning_rate": 9.795576245463345e-07,
+      "loss": 1.2227,
+      "step": 13273
+    },
+    {
+      "epoch": 0.9568226050601889,
+      "grad_norm": 0.11138983303304953,
+      "learning_rate": 9.76300472456637e-07,
+      "loss": 1.2618,
+      "step": 13274
+    },
+    {
+      "epoch": 0.9568946875225258,
+      "grad_norm": 0.13870263753888756,
+      "learning_rate": 9.730487180444558e-07,
+      "loss": 1.184,
+      "step": 13275
+    },
+    {
+      "epoch": 0.9569667699848626,
+      "grad_norm": 0.12047113145791505,
+      "learning_rate": 9.698023614870488e-07,
+      "loss": 1.2568,
+      "step": 13276
+    },
+    {
+      "epoch": 0.9570388524471996,
+      "grad_norm": 0.11993859440197259,
+      "learning_rate": 9.665614029613413e-07,
+      "loss": 1.2801,
+      "step": 13277
+    },
+    {
+      "epoch": 0.9571109349095365,
+      "grad_norm": 0.12967686465578876,
+      "learning_rate": 9.63325842644025e-07,
+      "loss": 1.1971,
+      "step": 13278
+    },
+    {
+      "epoch": 0.9571830173718734,
+      "grad_norm": 0.118040451529965,
+      "learning_rate": 9.600956807114703e-07,
+      "loss": 1.2684,
+      "step": 13279
+    },
+    {
+      "epoch": 0.9572550998342103,
+      "grad_norm": 0.11655762708019256,
+      "learning_rate": 9.56870917339725e-07,
+      "loss": 1.0524,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9573271822965472,
+      "grad_norm": 0.12286487439848146,
+      "learning_rate": 9.53651552704582e-07,
+      "loss": 1.2806,
+      "step": 13281
+    },
+    {
+      "epoch": 0.9573992647588841,
+      "grad_norm": 0.12539930572048139,
+      "learning_rate": 9.504375869815229e-07,
+      "loss": 1.1903,
+      "step": 13282
+    },
+    {
+      "epoch": 0.9574713472212211,
+      "grad_norm": 0.11607262000788954,
+      "learning_rate": 9.472290203457412e-07,
+      "loss": 1.2489,
+      "step": 13283
+    },
+    {
+      "epoch": 0.957543429683558,
+      "grad_norm": 0.1373907278099186,
+      "learning_rate": 9.440258529721302e-07,
+      "loss": 1.244,
+      "step": 13284
+    },
+    {
+      "epoch": 0.9576155121458949,
+      "grad_norm": 0.12282129075079226,
+      "learning_rate": 9.408280850352836e-07,
+      "loss": 1.3095,
+      "step": 13285
+    },
+    {
+      "epoch": 0.9576875946082318,
+      "grad_norm": 0.11427202051429662,
+      "learning_rate": 9.376357167095174e-07,
+      "loss": 1.2371,
+      "step": 13286
+    },
+    {
+      "epoch": 0.9577596770705687,
+      "grad_norm": 0.13367061480177125,
+      "learning_rate": 9.34448748168848e-07,
+      "loss": 1.2394,
+      "step": 13287
+    },
+    {
+      "epoch": 0.9578317595329057,
+      "grad_norm": 0.12651887491718433,
+      "learning_rate": 9.31267179586981e-07,
+      "loss": 1.1535,
+      "step": 13288
+    },
+    {
+      "epoch": 0.9579038419952426,
+      "grad_norm": 0.11734035649388876,
+      "learning_rate": 9.280910111373553e-07,
+      "loss": 1.3112,
+      "step": 13289
+    },
+    {
+      "epoch": 0.9579759244575795,
+      "grad_norm": 0.12767234673265773,
+      "learning_rate": 9.24920242993077e-07,
+      "loss": 1.2146,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9580480069199164,
+      "grad_norm": 0.1271220933758285,
+      "learning_rate": 9.217548753270078e-07,
+      "loss": 1.3221,
+      "step": 13291
+    },
+    {
+      "epoch": 0.9581200893822533,
+      "grad_norm": 0.12880527968896482,
+      "learning_rate": 9.185949083116763e-07,
+      "loss": 1.4173,
+      "step": 13292
+    },
+    {
+      "epoch": 0.9581921718445902,
+      "grad_norm": 0.12275363047690094,
+      "learning_rate": 9.154403421193225e-07,
+      "loss": 1.2041,
+      "step": 13293
+    },
+    {
+      "epoch": 0.9582642543069271,
+      "grad_norm": 0.10655226769489351,
+      "learning_rate": 9.122911769219089e-07,
+      "loss": 1.2897,
+      "step": 13294
+    },
+    {
+      "epoch": 0.958336336769264,
+      "grad_norm": 0.11398551738898768,
+      "learning_rate": 9.091474128910982e-07,
+      "loss": 1.2802,
+      "step": 13295
+    },
+    {
+      "epoch": 0.9584084192316009,
+      "grad_norm": 0.12770658651764724,
+      "learning_rate": 9.060090501982421e-07,
+      "loss": 1.2796,
+      "step": 13296
+    },
+    {
+      "epoch": 0.9584805016939378,
+      "grad_norm": 0.11774792630875158,
+      "learning_rate": 9.02876089014415e-07,
+      "loss": 1.2563,
+      "step": 13297
+    },
+    {
+      "epoch": 0.9585525841562748,
+      "grad_norm": 0.11987664978025407,
+      "learning_rate": 8.997485295103802e-07,
+      "loss": 1.2733,
+      "step": 13298
+    },
+    {
+      "epoch": 0.9586246666186117,
+      "grad_norm": 0.18940155739974684,
+      "learning_rate": 8.966263718566237e-07,
+      "loss": 1.2323,
+      "step": 13299
+    },
+    {
+      "epoch": 0.9586967490809486,
+      "grad_norm": 0.11314294708997577,
+      "learning_rate": 8.935096162233536e-07,
+      "loss": 1.174,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9587688315432855,
+      "grad_norm": 0.11466770197703516,
+      "learning_rate": 8.90398262780423e-07,
+      "loss": 1.3456,
+      "step": 13301
+    },
+    {
+      "epoch": 0.9588409140056224,
+      "grad_norm": 0.1205836732599071,
+      "learning_rate": 8.872923116974518e-07,
+      "loss": 1.2832,
+      "step": 13302
+    },
+    {
+      "epoch": 0.9589129964679594,
+      "grad_norm": 0.12584484759589637,
+      "learning_rate": 8.84191763143738e-07,
+      "loss": 1.2877,
+      "step": 13303
+    },
+    {
+      "epoch": 0.9589850789302963,
+      "grad_norm": 0.1202343092288857,
+      "learning_rate": 8.810966172882907e-07,
+      "loss": 1.2379,
+      "step": 13304
+    },
+    {
+      "epoch": 0.9590571613926332,
+      "grad_norm": 0.12227197460276049,
+      "learning_rate": 8.780068742998193e-07,
+      "loss": 1.2482,
+      "step": 13305
+    },
+    {
+      "epoch": 0.9591292438549701,
+      "grad_norm": 0.12287167747126439,
+      "learning_rate": 8.749225343467449e-07,
+      "loss": 1.2845,
+      "step": 13306
+    },
+    {
+      "epoch": 0.959201326317307,
+      "grad_norm": 0.11692825736696025,
+      "learning_rate": 8.718435975971994e-07,
+      "loss": 1.2799,
+      "step": 13307
+    },
+    {
+      "epoch": 0.959273408779644,
+      "grad_norm": 0.1172927681719898,
+      "learning_rate": 8.687700642189934e-07,
+      "loss": 1.2383,
+      "step": 13308
+    },
+    {
+      "epoch": 0.9593454912419809,
+      "grad_norm": 0.12241112930709616,
+      "learning_rate": 8.657019343796702e-07,
+      "loss": 1.2676,
+      "step": 13309
+    },
+    {
+      "epoch": 0.9594175737043178,
+      "grad_norm": 0.1155584831294855,
+      "learning_rate": 8.62639208246474e-07,
+      "loss": 1.2656,
+      "step": 13310
+    },
+    {
+      "epoch": 0.9594896561666546,
+      "grad_norm": 0.1222420810683719,
+      "learning_rate": 8.595818859863491e-07,
+      "loss": 1.2313,
+      "step": 13311
+    },
+    {
+      "epoch": 0.9595617386289915,
+      "grad_norm": 0.11176801089130423,
+      "learning_rate": 8.565299677659511e-07,
+      "loss": 1.2552,
+      "step": 13312
+    },
+    {
+      "epoch": 0.9596338210913284,
+      "grad_norm": 0.21243092785522188,
+      "learning_rate": 8.534834537516245e-07,
+      "loss": 1.2435,
+      "step": 13313
+    },
+    {
+      "epoch": 0.9597059035536654,
+      "grad_norm": 0.12583195538875364,
+      "learning_rate": 8.504423441094478e-07,
+      "loss": 1.3486,
+      "step": 13314
+    },
+    {
+      "epoch": 0.9597779860160023,
+      "grad_norm": 0.1180148633077521,
+      "learning_rate": 8.474066390051771e-07,
+      "loss": 1.2647,
+      "step": 13315
+    },
+    {
+      "epoch": 0.9598500684783392,
+      "grad_norm": 0.11622932689287334,
+      "learning_rate": 8.443763386042914e-07,
+      "loss": 1.3061,
+      "step": 13316
+    },
+    {
+      "epoch": 0.9599221509406761,
+      "grad_norm": 0.12610850606989718,
+      "learning_rate": 8.413514430719694e-07,
+      "loss": 1.2163,
+      "step": 13317
+    },
+    {
+      "epoch": 0.959994233403013,
+      "grad_norm": 0.11776529056753164,
+      "learning_rate": 8.383319525730793e-07,
+      "loss": 1.2419,
+      "step": 13318
+    },
+    {
+      "epoch": 0.96006631586535,
+      "grad_norm": 0.1245448687305674,
+      "learning_rate": 8.353178672722339e-07,
+      "loss": 1.2375,
+      "step": 13319
+    },
+    {
+      "epoch": 0.9601383983276869,
+      "grad_norm": 0.11230454664816848,
+      "learning_rate": 8.323091873337241e-07,
+      "loss": 1.3946,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9602104807900238,
+      "grad_norm": 0.10822161238749088,
+      "learning_rate": 8.293059129215297e-07,
+      "loss": 1.2472,
+      "step": 13321
+    },
+    {
+      "epoch": 0.9602825632523607,
+      "grad_norm": 0.11990014159855375,
+      "learning_rate": 8.263080441993753e-07,
+      "loss": 1.1956,
+      "step": 13322
+    },
+    {
+      "epoch": 0.9603546457146976,
+      "grad_norm": 0.11810745830905277,
+      "learning_rate": 8.233155813306637e-07,
+      "loss": 1.2299,
+      "step": 13323
+    },
+    {
+      "epoch": 0.9604267281770346,
+      "grad_norm": 0.1339608754916297,
+      "learning_rate": 8.203285244785198e-07,
+      "loss": 1.2378,
+      "step": 13324
+    },
+    {
+      "epoch": 0.9604988106393715,
+      "grad_norm": 0.12266606175523101,
+      "learning_rate": 8.173468738057466e-07,
+      "loss": 1.2111,
+      "step": 13325
+    },
+    {
+      "epoch": 0.9605708931017084,
+      "grad_norm": 0.11760239884974084,
+      "learning_rate": 8.14370629474892e-07,
+      "loss": 1.2689,
+      "step": 13326
+    },
+    {
+      "epoch": 0.9606429755640453,
+      "grad_norm": 0.12556371200757666,
+      "learning_rate": 8.113997916481708e-07,
+      "loss": 1.0943,
+      "step": 13327
+    },
+    {
+      "epoch": 0.9607150580263821,
+      "grad_norm": 0.11889358345511576,
+      "learning_rate": 8.084343604875311e-07,
+      "loss": 1.224,
+      "step": 13328
+    },
+    {
+      "epoch": 0.960787140488719,
+      "grad_norm": 0.12288522244674403,
+      "learning_rate": 8.054743361546213e-07,
+      "loss": 1.1941,
+      "step": 13329
+    },
+    {
+      "epoch": 0.960859222951056,
+      "grad_norm": 0.1170964192812272,
+      "learning_rate": 8.02519718810757e-07,
+      "loss": 1.1923,
+      "step": 13330
+    },
+    {
+      "epoch": 0.9609313054133929,
+      "grad_norm": 0.12558760540555522,
+      "learning_rate": 7.995705086170313e-07,
+      "loss": 1.3023,
+      "step": 13331
+    },
+    {
+      "epoch": 0.9610033878757298,
+      "grad_norm": 0.1208271025999879,
+      "learning_rate": 7.966267057341826e-07,
+      "loss": 1.2812,
+      "step": 13332
+    },
+    {
+      "epoch": 0.9610754703380667,
+      "grad_norm": 0.12421718829154339,
+      "learning_rate": 7.936883103226822e-07,
+      "loss": 1.2032,
+      "step": 13333
+    },
+    {
+      "epoch": 0.9611475528004036,
+      "grad_norm": 0.1332723297746252,
+      "learning_rate": 7.907553225426801e-07,
+      "loss": 1.2512,
+      "step": 13334
+    },
+    {
+      "epoch": 0.9612196352627406,
+      "grad_norm": 0.11237725818941041,
+      "learning_rate": 7.878277425540703e-07,
+      "loss": 1.2223,
+      "step": 13335
+    },
+    {
+      "epoch": 0.9612917177250775,
+      "grad_norm": 0.12131453495940604,
+      "learning_rate": 7.849055705164255e-07,
+      "loss": 1.2074,
+      "step": 13336
+    },
+    {
+      "epoch": 0.9613638001874144,
+      "grad_norm": 0.10932992229365063,
+      "learning_rate": 7.819888065890402e-07,
+      "loss": 1.2514,
+      "step": 13337
+    },
+    {
+      "epoch": 0.9614358826497513,
+      "grad_norm": 0.11842786128840709,
+      "learning_rate": 7.790774509308873e-07,
+      "loss": 1.2491,
+      "step": 13338
+    },
+    {
+      "epoch": 0.9615079651120882,
+      "grad_norm": 0.13384493033859055,
+      "learning_rate": 7.761715037006623e-07,
+      "loss": 1.2548,
+      "step": 13339
+    },
+    {
+      "epoch": 0.9615800475744252,
+      "grad_norm": 0.1124070068787289,
+      "learning_rate": 7.732709650567826e-07,
+      "loss": 1.3055,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9616521300367621,
+      "grad_norm": 0.11897579066048065,
+      "learning_rate": 7.703758351573332e-07,
+      "loss": 1.3402,
+      "step": 13341
+    },
+    {
+      "epoch": 0.961724212499099,
+      "grad_norm": 0.11477184655713661,
+      "learning_rate": 7.67486114160143e-07,
+      "loss": 1.3306,
+      "step": 13342
+    },
+    {
+      "epoch": 0.9617962949614359,
+      "grad_norm": 0.12099556291076766,
+      "learning_rate": 7.646018022227086e-07,
+      "loss": 1.2659,
+      "step": 13343
+    },
+    {
+      "epoch": 0.9618683774237728,
+      "grad_norm": 0.13263081802399992,
+      "learning_rate": 7.617228995022707e-07,
+      "loss": 1.2536,
+      "step": 13344
+    },
+    {
+      "epoch": 0.9619404598861097,
+      "grad_norm": 0.11828785114939883,
+      "learning_rate": 7.588494061557483e-07,
+      "loss": 1.2858,
+      "step": 13345
+    },
+    {
+      "epoch": 0.9620125423484466,
+      "grad_norm": 0.11199676707135717,
+      "learning_rate": 7.559813223397605e-07,
+      "loss": 1.2858,
+      "step": 13346
+    },
+    {
+      "epoch": 0.9620846248107835,
+      "grad_norm": 0.11937172178860898,
+      "learning_rate": 7.5311864821066e-07,
+      "loss": 1.3349,
+      "step": 13347
+    },
+    {
+      "epoch": 0.9621567072731204,
+      "grad_norm": 0.3270698501039945,
+      "learning_rate": 7.502613839244776e-07,
+      "loss": 1.3201,
+      "step": 13348
+    },
+    {
+      "epoch": 0.9622287897354573,
+      "grad_norm": 0.11750212176836822,
+      "learning_rate": 7.474095296369665e-07,
+      "loss": 1.3068,
+      "step": 13349
+    },
+    {
+      "epoch": 0.9623008721977943,
+      "grad_norm": 0.12091425867900711,
+      "learning_rate": 7.44563085503569e-07,
+      "loss": 1.2651,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9623729546601312,
+      "grad_norm": 0.11515446009819416,
+      "learning_rate": 7.417220516794499e-07,
+      "loss": 1.1611,
+      "step": 13351
+    },
+    {
+      "epoch": 0.9624450371224681,
+      "grad_norm": 0.14147404058235963,
+      "learning_rate": 7.388864283194741e-07,
+      "loss": 1.2698,
+      "step": 13352
+    },
+    {
+      "epoch": 0.962517119584805,
+      "grad_norm": 0.12512345565429187,
+      "learning_rate": 7.360562155782069e-07,
+      "loss": 1.3684,
+      "step": 13353
+    },
+    {
+      "epoch": 0.9625892020471419,
+      "grad_norm": 0.14281148093603566,
+      "learning_rate": 7.332314136099027e-07,
+      "loss": 1.2095,
+      "step": 13354
+    },
+    {
+      "epoch": 0.9626612845094789,
+      "grad_norm": 0.14374928865899775,
+      "learning_rate": 7.304120225685607e-07,
+      "loss": 1.2965,
+      "step": 13355
+    },
+    {
+      "epoch": 0.9627333669718158,
+      "grad_norm": 0.12624335427218014,
+      "learning_rate": 7.275980426078577e-07,
+      "loss": 1.267,
+      "step": 13356
+    },
+    {
+      "epoch": 0.9628054494341527,
+      "grad_norm": 0.11426131652319103,
+      "learning_rate": 7.247894738811711e-07,
+      "loss": 1.2699,
+      "step": 13357
+    },
+    {
+      "epoch": 0.9628775318964896,
+      "grad_norm": 0.11932646839619085,
+      "learning_rate": 7.219863165416008e-07,
+      "loss": 1.2651,
+      "step": 13358
+    },
+    {
+      "epoch": 0.9629496143588265,
+      "grad_norm": 0.11728162461544522,
+      "learning_rate": 7.191885707419465e-07,
+      "loss": 1.2557,
+      "step": 13359
+    },
+    {
+      "epoch": 0.9630216968211635,
+      "grad_norm": 0.11764921202625009,
+      "learning_rate": 7.163962366346977e-07,
+      "loss": 1.3199,
+      "step": 13360
+    },
+    {
+      "epoch": 0.9630937792835004,
+      "grad_norm": 0.13107174658413706,
+      "learning_rate": 7.136093143720767e-07,
+      "loss": 1.2844,
+      "step": 13361
+    },
+    {
+      "epoch": 0.9631658617458373,
+      "grad_norm": 0.1222894181443584,
+      "learning_rate": 7.108278041059957e-07,
+      "loss": 1.3431,
+      "step": 13362
+    },
+    {
+      "epoch": 0.9632379442081741,
+      "grad_norm": 0.12769284189033836,
+      "learning_rate": 7.080517059880553e-07,
+      "loss": 1.2709,
+      "step": 13363
+    },
+    {
+      "epoch": 0.963310026670511,
+      "grad_norm": 0.1222359127126574,
+      "learning_rate": 7.052810201695903e-07,
+      "loss": 1.2835,
+      "step": 13364
+    },
+    {
+      "epoch": 0.9633821091328479,
+      "grad_norm": 0.13005788771178967,
+      "learning_rate": 7.025157468016241e-07,
+      "loss": 1.2562,
+      "step": 13365
+    },
+    {
+      "epoch": 0.9634541915951849,
+      "grad_norm": 0.12480631626129601,
+      "learning_rate": 6.99755886034903e-07,
+      "loss": 1.2739,
+      "step": 13366
+    },
+    {
+      "epoch": 0.9635262740575218,
+      "grad_norm": 0.12619848901564662,
+      "learning_rate": 6.970014380198397e-07,
+      "loss": 1.2506,
+      "step": 13367
+    },
+    {
+      "epoch": 0.9635983565198587,
+      "grad_norm": 0.13096257616147142,
+      "learning_rate": 6.942524029065922e-07,
+      "loss": 1.3531,
+      "step": 13368
+    },
+    {
+      "epoch": 0.9636704389821956,
+      "grad_norm": 0.11681470363850442,
+      "learning_rate": 6.91508780844996e-07,
+      "loss": 1.2498,
+      "step": 13369
+    },
+    {
+      "epoch": 0.9637425214445325,
+      "grad_norm": 0.14018180867473526,
+      "learning_rate": 6.887705719846205e-07,
+      "loss": 1.2711,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9638146039068695,
+      "grad_norm": 0.12040799196179769,
+      "learning_rate": 6.860377764747017e-07,
+      "loss": 1.1913,
+      "step": 13371
+    },
+    {
+      "epoch": 0.9638866863692064,
+      "grad_norm": 0.1717367275498919,
+      "learning_rate": 6.833103944642205e-07,
+      "loss": 1.2041,
+      "step": 13372
+    },
+    {
+      "epoch": 0.9639587688315433,
+      "grad_norm": 0.1349733183125019,
+      "learning_rate": 6.805884261018358e-07,
+      "loss": 1.24,
+      "step": 13373
+    },
+    {
+      "epoch": 0.9640308512938802,
+      "grad_norm": 0.12523281949665285,
+      "learning_rate": 6.778718715359178e-07,
+      "loss": 1.2324,
+      "step": 13374
+    },
+    {
+      "epoch": 0.9641029337562171,
+      "grad_norm": 0.11221214517252373,
+      "learning_rate": 6.751607309145369e-07,
+      "loss": 1.2578,
+      "step": 13375
+    },
+    {
+      "epoch": 0.9641750162185541,
+      "grad_norm": 0.12206010896720415,
+      "learning_rate": 6.72455004385486e-07,
+      "loss": 1.2634,
+      "step": 13376
+    },
+    {
+      "epoch": 0.964247098680891,
+      "grad_norm": 0.1286488505045056,
+      "learning_rate": 6.697546920962583e-07,
+      "loss": 1.2424,
+      "step": 13377
+    },
+    {
+      "epoch": 0.9643191811432279,
+      "grad_norm": 0.1252690631037353,
+      "learning_rate": 6.670597941940137e-07,
+      "loss": 1.2568,
+      "step": 13378
+    },
+    {
+      "epoch": 0.9643912636055648,
+      "grad_norm": 0.13778118678470427,
+      "learning_rate": 6.643703108256794e-07,
+      "loss": 1.2407,
+      "step": 13379
+    },
+    {
+      "epoch": 0.9644633460679016,
+      "grad_norm": 0.11394622220436865,
+      "learning_rate": 6.616862421378489e-07,
+      "loss": 1.1704,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9645354285302385,
+      "grad_norm": 0.12323758208680678,
+      "learning_rate": 6.590075882768166e-07,
+      "loss": 1.2271,
+      "step": 13381
+    },
+    {
+      "epoch": 0.9646075109925755,
+      "grad_norm": 0.12136655658148544,
+      "learning_rate": 6.563343493886099e-07,
+      "loss": 1.2245,
+      "step": 13382
+    },
+    {
+      "epoch": 0.9646795934549124,
+      "grad_norm": 0.126339273682229,
+      "learning_rate": 6.536665256189234e-07,
+      "loss": 1.3159,
+      "step": 13383
+    },
+    {
+      "epoch": 0.9647516759172493,
+      "grad_norm": 0.1212494996844135,
+      "learning_rate": 6.510041171131853e-07,
+      "loss": 1.2467,
+      "step": 13384
+    },
+    {
+      "epoch": 0.9648237583795862,
+      "grad_norm": 0.12347423180180696,
+      "learning_rate": 6.483471240165462e-07,
+      "loss": 1.2918,
+      "step": 13385
+    },
+    {
+      "epoch": 0.9648958408419231,
+      "grad_norm": 0.13051900899457694,
+      "learning_rate": 6.456955464737902e-07,
+      "loss": 1.3348,
+      "step": 13386
+    },
+    {
+      "epoch": 0.9649679233042601,
+      "grad_norm": 0.11420003872672518,
+      "learning_rate": 6.430493846294904e-07,
+      "loss": 1.2585,
+      "step": 13387
+    },
+    {
+      "epoch": 0.965040005766597,
+      "grad_norm": 0.1260694320042724,
+      "learning_rate": 6.404086386278651e-07,
+      "loss": 1.3249,
+      "step": 13388
+    },
+    {
+      "epoch": 0.9651120882289339,
+      "grad_norm": 0.1325843783708166,
+      "learning_rate": 6.377733086128767e-07,
+      "loss": 1.2157,
+      "step": 13389
+    },
+    {
+      "epoch": 0.9651841706912708,
+      "grad_norm": 0.12130417607949123,
+      "learning_rate": 6.351433947281438e-07,
+      "loss": 1.2345,
+      "step": 13390
+    },
+    {
+      "epoch": 0.9652562531536077,
+      "grad_norm": 0.11972382404380684,
+      "learning_rate": 6.325188971170515e-07,
+      "loss": 1.1915,
+      "step": 13391
+    },
+    {
+      "epoch": 0.9653283356159447,
+      "grad_norm": 0.1260708320895102,
+      "learning_rate": 6.29899815922641e-07,
+      "loss": 1.4203,
+      "step": 13392
+    },
+    {
+      "epoch": 0.9654004180782816,
+      "grad_norm": 0.12595964328622714,
+      "learning_rate": 6.272861512876871e-07,
+      "loss": 1.2975,
+      "step": 13393
+    },
+    {
+      "epoch": 0.9654725005406185,
+      "grad_norm": 0.11632094044481758,
+      "learning_rate": 6.246779033546535e-07,
+      "loss": 1.3237,
+      "step": 13394
+    },
+    {
+      "epoch": 0.9655445830029554,
+      "grad_norm": 0.11702904898425677,
+      "learning_rate": 6.220750722657043e-07,
+      "loss": 1.1931,
+      "step": 13395
+    },
+    {
+      "epoch": 0.9656166654652923,
+      "grad_norm": 0.13038287648444788,
+      "learning_rate": 6.19477658162726e-07,
+      "loss": 1.2368,
+      "step": 13396
+    },
+    {
+      "epoch": 0.9656887479276292,
+      "grad_norm": 0.12785432113428669,
+      "learning_rate": 6.168856611872942e-07,
+      "loss": 1.2454,
+      "step": 13397
+    },
+    {
+      "epoch": 0.9657608303899661,
+      "grad_norm": 0.1218538770299535,
+      "learning_rate": 6.14299081480707e-07,
+      "loss": 1.2291,
+      "step": 13398
+    },
+    {
+      "epoch": 0.965832912852303,
+      "grad_norm": 0.12374801175461686,
+      "learning_rate": 6.117179191839517e-07,
+      "loss": 1.1675,
+      "step": 13399
+    },
+    {
+      "epoch": 0.9659049953146399,
+      "grad_norm": 0.13730029528036036,
+      "learning_rate": 6.091421744377157e-07,
+      "loss": 1.2577,
+      "step": 13400
+    },
+    {
+      "epoch": 0.9659770777769768,
+      "grad_norm": 0.11718075867341822,
+      "learning_rate": 6.065718473824089e-07,
+      "loss": 1.2509,
+      "step": 13401
+    },
+    {
+      "epoch": 0.9660491602393138,
+      "grad_norm": 0.13230883926435216,
+      "learning_rate": 6.040069381581415e-07,
+      "loss": 1.1625,
+      "step": 13402
+    },
+    {
+      "epoch": 0.9661212427016507,
+      "grad_norm": 0.17579504903078283,
+      "learning_rate": 6.014474469047126e-07,
+      "loss": 1.2064,
+      "step": 13403
+    },
+    {
+      "epoch": 0.9661933251639876,
+      "grad_norm": 0.12657675374599153,
+      "learning_rate": 5.988933737616442e-07,
+      "loss": 1.2084,
+      "step": 13404
+    },
+    {
+      "epoch": 0.9662654076263245,
+      "grad_norm": 0.12374614391194984,
+      "learning_rate": 5.963447188681581e-07,
+      "loss": 1.2272,
+      "step": 13405
+    },
+    {
+      "epoch": 0.9663374900886614,
+      "grad_norm": 0.11824157549623311,
+      "learning_rate": 5.938014823631655e-07,
+      "loss": 1.2673,
+      "step": 13406
+    },
+    {
+      "epoch": 0.9664095725509984,
+      "grad_norm": 0.10907387178269536,
+      "learning_rate": 5.912636643853109e-07,
+      "loss": 1.2593,
+      "step": 13407
+    },
+    {
+      "epoch": 0.9664816550133353,
+      "grad_norm": 0.11865000819796373,
+      "learning_rate": 5.887312650729281e-07,
+      "loss": 1.3468,
+      "step": 13408
+    },
+    {
+      "epoch": 0.9665537374756722,
+      "grad_norm": 0.11353033580465426,
+      "learning_rate": 5.862042845640403e-07,
+      "loss": 1.2442,
+      "step": 13409
+    },
+    {
+      "epoch": 0.9666258199380091,
+      "grad_norm": 0.1265004724025559,
+      "learning_rate": 5.836827229964037e-07,
+      "loss": 1.295,
+      "step": 13410
+    },
+    {
+      "epoch": 0.966697902400346,
+      "grad_norm": 0.1255599171656908,
+      "learning_rate": 5.811665805074639e-07,
+      "loss": 1.3037,
+      "step": 13411
+    },
+    {
+      "epoch": 0.966769984862683,
+      "grad_norm": 0.13256408271780193,
+      "learning_rate": 5.78655857234367e-07,
+      "loss": 1.3426,
+      "step": 13412
+    },
+    {
+      "epoch": 0.9668420673250199,
+      "grad_norm": 0.11424056327216729,
+      "learning_rate": 5.76150553313981e-07,
+      "loss": 1.3084,
+      "step": 13413
+    },
+    {
+      "epoch": 0.9669141497873568,
+      "grad_norm": 0.1112634868575318,
+      "learning_rate": 5.736506688828636e-07,
+      "loss": 1.227,
+      "step": 13414
+    },
+    {
+      "epoch": 0.9669862322496936,
+      "grad_norm": 0.12857884446816698,
+      "learning_rate": 5.711562040772723e-07,
+      "loss": 1.2737,
+      "step": 13415
+    },
+    {
+      "epoch": 0.9670583147120305,
+      "grad_norm": 0.11916034976655492,
+      "learning_rate": 5.686671590331871e-07,
+      "loss": 1.2129,
+      "step": 13416
+    },
+    {
+      "epoch": 0.9671303971743674,
+      "grad_norm": 0.1481642514078646,
+      "learning_rate": 5.661835338862775e-07,
+      "loss": 1.3154,
+      "step": 13417
+    },
+    {
+      "epoch": 0.9672024796367044,
+      "grad_norm": 0.11603403143206249,
+      "learning_rate": 5.63705328771924e-07,
+      "loss": 1.3512,
+      "step": 13418
+    },
+    {
+      "epoch": 0.9672745620990413,
+      "grad_norm": 0.12489974423456367,
+      "learning_rate": 5.612325438252186e-07,
+      "loss": 1.3165,
+      "step": 13419
+    },
+    {
+      "epoch": 0.9673466445613782,
+      "grad_norm": 0.13566892014705875,
+      "learning_rate": 5.587651791809312e-07,
+      "loss": 1.2544,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9674187270237151,
+      "grad_norm": 0.11460003482273416,
+      "learning_rate": 5.563032349735875e-07,
+      "loss": 1.1821,
+      "step": 13421
+    },
+    {
+      "epoch": 0.967490809486052,
+      "grad_norm": 0.12958177375965657,
+      "learning_rate": 5.538467113373469e-07,
+      "loss": 1.3487,
+      "step": 13422
+    },
+    {
+      "epoch": 0.967562891948389,
+      "grad_norm": 0.12111384719661217,
+      "learning_rate": 5.513956084061467e-07,
+      "loss": 1.3007,
+      "step": 13423
+    },
+    {
+      "epoch": 0.9676349744107259,
+      "grad_norm": 0.13330867791756995,
+      "learning_rate": 5.48949926313569e-07,
+      "loss": 1.2814,
+      "step": 13424
+    },
+    {
+      "epoch": 0.9677070568730628,
+      "grad_norm": 0.10024319034160944,
+      "learning_rate": 5.465096651929402e-07,
+      "loss": 1.1822,
+      "step": 13425
+    },
+    {
+      "epoch": 0.9677791393353997,
+      "grad_norm": 0.13021705217094112,
+      "learning_rate": 5.440748251772764e-07,
+      "loss": 1.2634,
+      "step": 13426
+    },
+    {
+      "epoch": 0.9678512217977366,
+      "grad_norm": 0.17168848811892182,
+      "learning_rate": 5.416454063992715e-07,
+      "loss": 1.3206,
+      "step": 13427
+    },
+    {
+      "epoch": 0.9679233042600736,
+      "grad_norm": 0.11808882221126839,
+      "learning_rate": 5.392214089913861e-07,
+      "loss": 1.2023,
+      "step": 13428
+    },
+    {
+      "epoch": 0.9679953867224105,
+      "grad_norm": 0.11537135796168732,
+      "learning_rate": 5.368028330857367e-07,
+      "loss": 1.2787,
+      "step": 13429
+    },
+    {
+      "epoch": 0.9680674691847474,
+      "grad_norm": 0.12042236098002501,
+      "learning_rate": 5.343896788141622e-07,
+      "loss": 1.2599,
+      "step": 13430
+    },
+    {
+      "epoch": 0.9681395516470843,
+      "grad_norm": 0.13457762588543817,
+      "learning_rate": 5.319819463081799e-07,
+      "loss": 1.3076,
+      "step": 13431
+    },
+    {
+      "epoch": 0.9682116341094211,
+      "grad_norm": 0.1279404889068502,
+      "learning_rate": 5.295796356990512e-07,
+      "loss": 1.2561,
+      "step": 13432
+    },
+    {
+      "epoch": 0.968283716571758,
+      "grad_norm": 0.12590188823802198,
+      "learning_rate": 5.271827471177271e-07,
+      "loss": 1.1751,
+      "step": 13433
+    },
+    {
+      "epoch": 0.968355799034095,
+      "grad_norm": 0.12155842141108095,
+      "learning_rate": 5.247912806948585e-07,
+      "loss": 1.2899,
+      "step": 13434
+    },
+    {
+      "epoch": 0.9684278814964319,
+      "grad_norm": 0.11022482692137657,
+      "learning_rate": 5.224052365607746e-07,
+      "loss": 1.2008,
+      "step": 13435
+    },
+    {
+      "epoch": 0.9684999639587688,
+      "grad_norm": 0.11521400167115435,
+      "learning_rate": 5.200246148455712e-07,
+      "loss": 1.3213,
+      "step": 13436
+    },
+    {
+      "epoch": 0.9685720464211057,
+      "grad_norm": 0.1250661528431893,
+      "learning_rate": 5.176494156790002e-07,
+      "loss": 1.242,
+      "step": 13437
+    },
+    {
+      "epoch": 0.9686441288834426,
+      "grad_norm": 0.12601835477499065,
+      "learning_rate": 5.152796391905357e-07,
+      "loss": 1.2996,
+      "step": 13438
+    },
+    {
+      "epoch": 0.9687162113457796,
+      "grad_norm": 0.12850925811140623,
+      "learning_rate": 5.1291528550933e-07,
+      "loss": 1.2951,
+      "step": 13439
+    },
+    {
+      "epoch": 0.9687882938081165,
+      "grad_norm": 0.1274401382793274,
+      "learning_rate": 5.105563547642911e-07,
+      "loss": 1.4016,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9688603762704534,
+      "grad_norm": 0.11087398970567729,
+      "learning_rate": 5.082028470839828e-07,
+      "loss": 1.3266,
+      "step": 13441
+    },
+    {
+      "epoch": 0.9689324587327903,
+      "grad_norm": 0.2120380697819456,
+      "learning_rate": 5.058547625967025e-07,
+      "loss": 1.295,
+      "step": 13442
+    },
+    {
+      "epoch": 0.9690045411951272,
+      "grad_norm": 0.12136414932221565,
+      "learning_rate": 5.035121014304367e-07,
+      "loss": 1.1609,
+      "step": 13443
+    },
+    {
+      "epoch": 0.9690766236574642,
+      "grad_norm": 0.13042836556980258,
+      "learning_rate": 5.011748637128722e-07,
+      "loss": 1.2462,
+      "step": 13444
+    },
+    {
+      "epoch": 0.9691487061198011,
+      "grad_norm": 0.11676608956860571,
+      "learning_rate": 4.988430495714292e-07,
+      "loss": 1.1545,
+      "step": 13445
+    },
+    {
+      "epoch": 0.969220788582138,
+      "grad_norm": 0.11838483074171935,
+      "learning_rate": 4.96516659133206e-07,
+      "loss": 1.185,
+      "step": 13446
+    },
+    {
+      "epoch": 0.9692928710444749,
+      "grad_norm": 0.12046190318536265,
+      "learning_rate": 4.941956925250013e-07,
+      "loss": 1.2371,
+      "step": 13447
+    },
+    {
+      "epoch": 0.9693649535068118,
+      "grad_norm": 0.11565636507043166,
+      "learning_rate": 4.918801498733472e-07,
+      "loss": 1.2123,
+      "step": 13448
+    },
+    {
+      "epoch": 0.9694370359691487,
+      "grad_norm": 0.12362728876254163,
+      "learning_rate": 4.895700313044315e-07,
+      "loss": 1.2844,
+      "step": 13449
+    },
+    {
+      "epoch": 0.9695091184314856,
+      "grad_norm": 0.1287992859711041,
+      "learning_rate": 4.872653369442092e-07,
+      "loss": 1.2395,
+      "step": 13450
+    },
+    {
+      "epoch": 0.9695812008938225,
+      "grad_norm": 0.12751734806580037,
+      "learning_rate": 4.849660669182909e-07,
+      "loss": 1.2744,
+      "step": 13451
+    },
+    {
+      "epoch": 0.9696532833561594,
+      "grad_norm": 0.12298590931573894,
+      "learning_rate": 4.826722213520096e-07,
+      "loss": 1.3241,
+      "step": 13452
+    },
+    {
+      "epoch": 0.9697253658184963,
+      "grad_norm": 0.11796524688040913,
+      "learning_rate": 4.803838003703876e-07,
+      "loss": 1.2665,
+      "step": 13453
+    },
+    {
+      "epoch": 0.9697974482808333,
+      "grad_norm": 0.12546808421840067,
+      "learning_rate": 4.781008040981915e-07,
+      "loss": 1.2053,
+      "step": 13454
+    },
+    {
+      "epoch": 0.9698695307431702,
+      "grad_norm": 0.11838185216123502,
+      "learning_rate": 4.758232326598444e-07,
+      "loss": 1.2897,
+      "step": 13455
+    },
+    {
+      "epoch": 0.9699416132055071,
+      "grad_norm": 0.13134480323801007,
+      "learning_rate": 4.735510861795023e-07,
+      "loss": 1.2766,
+      "step": 13456
+    },
+    {
+      "epoch": 0.970013695667844,
+      "grad_norm": 0.12371834734762734,
+      "learning_rate": 4.7128436478099945e-07,
+      "loss": 1.2938,
+      "step": 13457
+    },
+    {
+      "epoch": 0.9700857781301809,
+      "grad_norm": 0.11806082041986318,
+      "learning_rate": 4.69023068587926e-07,
+      "loss": 1.2291,
+      "step": 13458
+    },
+    {
+      "epoch": 0.9701578605925179,
+      "grad_norm": 0.13132623540527263,
+      "learning_rate": 4.6676719772352775e-07,
+      "loss": 1.3565,
+      "step": 13459
+    },
+    {
+      "epoch": 0.9702299430548548,
+      "grad_norm": 0.11679870477838866,
+      "learning_rate": 4.6451675231075077e-07,
+      "loss": 1.3789,
+      "step": 13460
+    },
+    {
+      "epoch": 0.9703020255171917,
+      "grad_norm": 0.11632700215284268,
+      "learning_rate": 4.6227173247228583e-07,
+      "loss": 1.2969,
+      "step": 13461
+    },
+    {
+      "epoch": 0.9703741079795286,
+      "grad_norm": 0.13167947815329614,
+      "learning_rate": 4.6003213833051284e-07,
+      "loss": 1.2687,
+      "step": 13462
+    },
+    {
+      "epoch": 0.9704461904418655,
+      "grad_norm": 0.10993151230328663,
+      "learning_rate": 4.577979700074897e-07,
+      "loss": 1.2273,
+      "step": 13463
+    },
+    {
+      "epoch": 0.9705182729042025,
+      "grad_norm": 0.11692569155823959,
+      "learning_rate": 4.555692276250079e-07,
+      "loss": 1.2939,
+      "step": 13464
+    },
+    {
+      "epoch": 0.9705903553665394,
+      "grad_norm": 0.12234044664484843,
+      "learning_rate": 4.5334591130454795e-07,
+      "loss": 1.2945,
+      "step": 13465
+    },
+    {
+      "epoch": 0.9706624378288762,
+      "grad_norm": 0.10807981722626943,
+      "learning_rate": 4.51128021167313e-07,
+      "loss": 1.2076,
+      "step": 13466
+    },
+    {
+      "epoch": 0.9707345202912131,
+      "grad_norm": 0.11554687068511836,
+      "learning_rate": 4.489155573341841e-07,
+      "loss": 1.2703,
+      "step": 13467
+    },
+    {
+      "epoch": 0.97080660275355,
+      "grad_norm": 0.12176635281404954,
+      "learning_rate": 4.467085199257759e-07,
+      "loss": 1.2666,
+      "step": 13468
+    },
+    {
+      "epoch": 0.9708786852158869,
+      "grad_norm": 0.10738991820217417,
+      "learning_rate": 4.445069090623699e-07,
+      "loss": 1.2881,
+      "step": 13469
+    },
+    {
+      "epoch": 0.9709507676782239,
+      "grad_norm": 0.11640353370704294,
+      "learning_rate": 4.423107248639924e-07,
+      "loss": 1.2431,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9710228501405608,
+      "grad_norm": 0.11431269001465284,
+      "learning_rate": 4.4011996745034755e-07,
+      "loss": 1.1358,
+      "step": 13471
+    },
+    {
+      "epoch": 0.9710949326028977,
+      "grad_norm": 0.13311207830149255,
+      "learning_rate": 4.379346369408621e-07,
+      "loss": 1.2237,
+      "step": 13472
+    },
+    {
+      "epoch": 0.9711670150652346,
+      "grad_norm": 0.1277274145315061,
+      "learning_rate": 4.357547334546408e-07,
+      "loss": 1.2568,
+      "step": 13473
+    },
+    {
+      "epoch": 0.9712390975275715,
+      "grad_norm": 0.2978790738107425,
+      "learning_rate": 4.335802571105108e-07,
+      "loss": 1.2556,
+      "step": 13474
+    },
+    {
+      "epoch": 0.9713111799899085,
+      "grad_norm": 0.12843956620489175,
+      "learning_rate": 4.314112080269994e-07,
+      "loss": 1.2318,
+      "step": 13475
+    },
+    {
+      "epoch": 0.9713832624522454,
+      "grad_norm": 0.12144656771747765,
+      "learning_rate": 4.292475863223455e-07,
+      "loss": 1.166,
+      "step": 13476
+    },
+    {
+      "epoch": 0.9714553449145823,
+      "grad_norm": 0.108571170615696,
+      "learning_rate": 4.2708939211448806e-07,
+      "loss": 1.3567,
+      "step": 13477
+    },
+    {
+      "epoch": 0.9715274273769192,
+      "grad_norm": 0.12750786511063894,
+      "learning_rate": 4.2493662552106627e-07,
+      "loss": 1.3019,
+      "step": 13478
+    },
+    {
+      "epoch": 0.9715995098392561,
+      "grad_norm": 0.13947378617216735,
+      "learning_rate": 4.2278928665941964e-07,
+      "loss": 1.3417,
+      "step": 13479
+    },
+    {
+      "epoch": 0.9716715923015931,
+      "grad_norm": 0.12109240274643314,
+      "learning_rate": 4.2064737564659896e-07,
+      "loss": 1.2711,
+      "step": 13480
+    },
+    {
+      "epoch": 0.97174367476393,
+      "grad_norm": 0.1292890988420961,
+      "learning_rate": 4.185108925993553e-07,
+      "loss": 1.3471,
+      "step": 13481
+    },
+    {
+      "epoch": 0.9718157572262669,
+      "grad_norm": 0.11157216474660013,
+      "learning_rate": 4.16379837634151e-07,
+      "loss": 1.3286,
+      "step": 13482
+    },
+    {
+      "epoch": 0.9718878396886038,
+      "grad_norm": 0.12096925884579729,
+      "learning_rate": 4.142542108671599e-07,
+      "loss": 1.2255,
+      "step": 13483
+    },
+    {
+      "epoch": 0.9719599221509406,
+      "grad_norm": 0.12180572311727558,
+      "learning_rate": 4.1213401241421147e-07,
+      "loss": 1.2975,
+      "step": 13484
+    },
+    {
+      "epoch": 0.9720320046132775,
+      "grad_norm": 0.13268596013879305,
+      "learning_rate": 4.100192423909133e-07,
+      "loss": 1.2158,
+      "step": 13485
+    },
+    {
+      "epoch": 0.9721040870756145,
+      "grad_norm": 0.13801210932380453,
+      "learning_rate": 4.0790990091251755e-07,
+      "loss": 1.1868,
+      "step": 13486
+    },
+    {
+      "epoch": 0.9721761695379514,
+      "grad_norm": 0.12027938676306384,
+      "learning_rate": 4.0580598809401013e-07,
+      "loss": 1.2977,
+      "step": 13487
+    },
+    {
+      "epoch": 0.9722482520002883,
+      "grad_norm": 0.12770807229974654,
+      "learning_rate": 4.0370750405006595e-07,
+      "loss": 1.2332,
+      "step": 13488
+    },
+    {
+      "epoch": 0.9723203344626252,
+      "grad_norm": 0.12021284982538909,
+      "learning_rate": 4.0161444889508237e-07,
+      "loss": 1.2945,
+      "step": 13489
+    },
+    {
+      "epoch": 0.9723924169249621,
+      "grad_norm": 0.12818957762194502,
+      "learning_rate": 3.995268227431459e-07,
+      "loss": 1.2771,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9724644993872991,
+      "grad_norm": 0.12463082457132556,
+      "learning_rate": 3.974446257080433e-07,
+      "loss": 1.3271,
+      "step": 13491
+    },
+    {
+      "epoch": 0.972536581849636,
+      "grad_norm": 0.11253482724006313,
+      "learning_rate": 3.953678579032838e-07,
+      "loss": 1.2366,
+      "step": 13492
+    },
+    {
+      "epoch": 0.9726086643119729,
+      "grad_norm": 0.15483066975689982,
+      "learning_rate": 3.9329651944205456e-07,
+      "loss": 1.2614,
+      "step": 13493
+    },
+    {
+      "epoch": 0.9726807467743098,
+      "grad_norm": 0.1274416327525558,
+      "learning_rate": 3.9123061043728757e-07,
+      "loss": 1.2402,
+      "step": 13494
+    },
+    {
+      "epoch": 0.9727528292366467,
+      "grad_norm": 0.12035609102156683,
+      "learning_rate": 3.8917013100157054e-07,
+      "loss": 1.2554,
+      "step": 13495
+    },
+    {
+      "epoch": 0.9728249116989837,
+      "grad_norm": 0.12219799660501417,
+      "learning_rate": 3.871150812472246e-07,
+      "loss": 1.2362,
+      "step": 13496
+    },
+    {
+      "epoch": 0.9728969941613206,
+      "grad_norm": 0.12565390101390994,
+      "learning_rate": 3.8506546128626034e-07,
+      "loss": 1.1901,
+      "step": 13497
+    },
+    {
+      "epoch": 0.9729690766236575,
+      "grad_norm": 0.11947296121390705,
+      "learning_rate": 3.8302127123042155e-07,
+      "loss": 1.242,
+      "step": 13498
+    },
+    {
+      "epoch": 0.9730411590859944,
+      "grad_norm": 0.12289568989415481,
+      "learning_rate": 3.809825111911192e-07,
+      "loss": 1.2856,
+      "step": 13499
+    },
+    {
+      "epoch": 0.9731132415483313,
+      "grad_norm": 0.11650698266719614,
+      "learning_rate": 3.7894918127948655e-07,
+      "loss": 1.2005,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9731853240106682,
+      "grad_norm": 0.118672233553519,
+      "learning_rate": 3.769212816063572e-07,
+      "loss": 1.3046,
+      "step": 13501
+    },
+    {
+      "epoch": 0.9732574064730051,
+      "grad_norm": 0.1248290880195455,
+      "learning_rate": 3.748988122822761e-07,
+      "loss": 1.2175,
+      "step": 13502
+    },
+    {
+      "epoch": 0.973329488935342,
+      "grad_norm": 0.11013416135027829,
+      "learning_rate": 3.728817734174772e-07,
+      "loss": 1.2527,
+      "step": 13503
+    },
+    {
+      "epoch": 0.9734015713976789,
+      "grad_norm": 0.12000445789299669,
+      "learning_rate": 3.7087016512191705e-07,
+      "loss": 1.2721,
+      "step": 13504
+    },
+    {
+      "epoch": 0.9734736538600158,
+      "grad_norm": 0.12686590933469583,
+      "learning_rate": 3.688639875052302e-07,
+      "loss": 1.2825,
+      "step": 13505
+    },
+    {
+      "epoch": 0.9735457363223528,
+      "grad_norm": 0.12879411482227093,
+      "learning_rate": 3.668632406767958e-07,
+      "loss": 1.3497,
+      "step": 13506
+    },
+    {
+      "epoch": 0.9736178187846897,
+      "grad_norm": 0.12583291390880133,
+      "learning_rate": 3.648679247456488e-07,
+      "loss": 1.2114,
+      "step": 13507
+    },
+    {
+      "epoch": 0.9736899012470266,
+      "grad_norm": 0.11900209157300261,
+      "learning_rate": 3.628780398205578e-07,
+      "loss": 1.2857,
+      "step": 13508
+    },
+    {
+      "epoch": 0.9737619837093635,
+      "grad_norm": 0.1127593003038851,
+      "learning_rate": 3.6089358601000265e-07,
+      "loss": 1.1213,
+      "step": 13509
+    },
+    {
+      "epoch": 0.9738340661717004,
+      "grad_norm": 0.11348843733376508,
+      "learning_rate": 3.589145634221414e-07,
+      "loss": 1.2709,
+      "step": 13510
+    },
+    {
+      "epoch": 0.9739061486340373,
+      "grad_norm": 0.12427951144742716,
+      "learning_rate": 3.5694097216484314e-07,
+      "loss": 1.1827,
+      "step": 13511
+    },
+    {
+      "epoch": 0.9739782310963743,
+      "grad_norm": 0.11503963050687535,
+      "learning_rate": 3.5497281234569965e-07,
+      "loss": 1.3134,
+      "step": 13512
+    },
+    {
+      "epoch": 0.9740503135587112,
+      "grad_norm": 0.11775755027963154,
+      "learning_rate": 3.5301008407199187e-07,
+      "loss": 1.3714,
+      "step": 13513
+    },
+    {
+      "epoch": 0.9741223960210481,
+      "grad_norm": 0.11447267822968794,
+      "learning_rate": 3.510527874506897e-07,
+      "loss": 1.3142,
+      "step": 13514
+    },
+    {
+      "epoch": 0.974194478483385,
+      "grad_norm": 0.13150903877576295,
+      "learning_rate": 3.491009225885078e-07,
+      "loss": 1.254,
+      "step": 13515
+    },
+    {
+      "epoch": 0.974266560945722,
+      "grad_norm": 0.11621722481127363,
+      "learning_rate": 3.471544895918277e-07,
+      "loss": 1.3237,
+      "step": 13516
+    },
+    {
+      "epoch": 0.9743386434080589,
+      "grad_norm": 0.1118527059163569,
+      "learning_rate": 3.452134885667424e-07,
+      "loss": 1.3108,
+      "step": 13517
+    },
+    {
+      "epoch": 0.9744107258703957,
+      "grad_norm": 0.11585532283266121,
+      "learning_rate": 3.432779196190561e-07,
+      "loss": 1.3397,
+      "step": 13518
+    },
+    {
+      "epoch": 0.9744828083327326,
+      "grad_norm": 0.11816404638567458,
+      "learning_rate": 3.4134778285428437e-07,
+      "loss": 1.2613,
+      "step": 13519
+    },
+    {
+      "epoch": 0.9745548907950695,
+      "grad_norm": 0.1290464652848942,
+      "learning_rate": 3.39423078377632e-07,
+      "loss": 1.2696,
+      "step": 13520
+    },
+    {
+      "epoch": 0.9746269732574064,
+      "grad_norm": 0.10359197071975608,
+      "learning_rate": 3.375038062940039e-07,
+      "loss": 1.1756,
+      "step": 13521
+    },
+    {
+      "epoch": 0.9746990557197434,
+      "grad_norm": 0.12387173519758736,
+      "learning_rate": 3.3558996670802757e-07,
+      "loss": 1.28,
+      "step": 13522
+    },
+    {
+      "epoch": 0.9747711381820803,
+      "grad_norm": 0.14313733146479488,
+      "learning_rate": 3.3368155972401946e-07,
+      "loss": 1.3202,
+      "step": 13523
+    },
+    {
+      "epoch": 0.9748432206444172,
+      "grad_norm": 0.13030770202351186,
+      "learning_rate": 3.317785854460076e-07,
+      "loss": 1.2745,
+      "step": 13524
+    },
+    {
+      "epoch": 0.9749153031067541,
+      "grad_norm": 0.12098373405429093,
+      "learning_rate": 3.298810439777311e-07,
+      "loss": 1.2219,
+      "step": 13525
+    },
+    {
+      "epoch": 0.974987385569091,
+      "grad_norm": 0.11366419568088122,
+      "learning_rate": 3.2798893542259624e-07,
+      "loss": 1.1669,
+      "step": 13526
+    },
+    {
+      "epoch": 0.975059468031428,
+      "grad_norm": 0.12446096471222542,
+      "learning_rate": 3.261022598837649e-07,
+      "loss": 1.2634,
+      "step": 13527
+    },
+    {
+      "epoch": 0.9751315504937649,
+      "grad_norm": 0.1075603417460193,
+      "learning_rate": 3.2422101746406586e-07,
+      "loss": 1.1988,
+      "step": 13528
+    },
+    {
+      "epoch": 0.9752036329561018,
+      "grad_norm": 0.12602081110513652,
+      "learning_rate": 3.223452082660394e-07,
+      "loss": 1.3352,
+      "step": 13529
+    },
+    {
+      "epoch": 0.9752757154184387,
+      "grad_norm": 0.12736745988751363,
+      "learning_rate": 3.204748323919482e-07,
+      "loss": 1.2554,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9753477978807756,
+      "grad_norm": 0.1225519894770376,
+      "learning_rate": 3.18609889943744e-07,
+      "loss": 1.2395,
+      "step": 13531
+    },
+    {
+      "epoch": 0.9754198803431126,
+      "grad_norm": 0.11984854979383255,
+      "learning_rate": 3.1675038102306765e-07,
+      "loss": 1.1799,
+      "step": 13532
+    },
+    {
+      "epoch": 0.9754919628054495,
+      "grad_norm": 0.18501575122622962,
+      "learning_rate": 3.148963057312826e-07,
+      "loss": 1.2924,
+      "step": 13533
+    },
+    {
+      "epoch": 0.9755640452677864,
+      "grad_norm": 0.12547582561337658,
+      "learning_rate": 3.130476641694635e-07,
+      "loss": 1.3125,
+      "step": 13534
+    },
+    {
+      "epoch": 0.9756361277301233,
+      "grad_norm": 0.12219894079405205,
+      "learning_rate": 3.112044564383743e-07,
+      "loss": 1.2046,
+      "step": 13535
+    },
+    {
+      "epoch": 0.9757082101924601,
+      "grad_norm": 0.11865217845033238,
+      "learning_rate": 3.093666826384789e-07,
+      "loss": 1.257,
+      "step": 13536
+    },
+    {
+      "epoch": 0.975780292654797,
+      "grad_norm": 0.1257643511376259,
+      "learning_rate": 3.0753434286995284e-07,
+      "loss": 1.2139,
+      "step": 13537
+    },
+    {
+      "epoch": 0.975852375117134,
+      "grad_norm": 0.12295791810769507,
+      "learning_rate": 3.05707437232694e-07,
+      "loss": 1.3228,
+      "step": 13538
+    },
+    {
+      "epoch": 0.9759244575794709,
+      "grad_norm": 0.1295087228347947,
+      "learning_rate": 3.0388596582625605e-07,
+      "loss": 1.2489,
+      "step": 13539
+    },
+    {
+      "epoch": 0.9759965400418078,
+      "grad_norm": 0.11660508223772906,
+      "learning_rate": 3.020699287499484e-07,
+      "loss": 1.1586,
+      "step": 13540
+    },
+    {
+      "epoch": 0.9760686225041447,
+      "grad_norm": 0.108411505889635,
+      "learning_rate": 3.0025932610274754e-07,
+      "loss": 1.253,
+      "step": 13541
+    },
+    {
+      "epoch": 0.9761407049664816,
+      "grad_norm": 0.13199133233521254,
+      "learning_rate": 2.984541579833522e-07,
+      "loss": 1.2906,
+      "step": 13542
+    },
+    {
+      "epoch": 0.9762127874288186,
+      "grad_norm": 0.12261076174031658,
+      "learning_rate": 2.9665442449017257e-07,
+      "loss": 1.25,
+      "step": 13543
+    },
+    {
+      "epoch": 0.9762848698911555,
+      "grad_norm": 0.12714222531178446,
+      "learning_rate": 2.9486012572128573e-07,
+      "loss": 1.258,
+      "step": 13544
+    },
+    {
+      "epoch": 0.9763569523534924,
+      "grad_norm": 0.12570083708179464,
+      "learning_rate": 2.930712617745135e-07,
+      "loss": 1.2461,
+      "step": 13545
+    },
+    {
+      "epoch": 0.9764290348158293,
+      "grad_norm": 0.11376108036046924,
+      "learning_rate": 2.912878327473556e-07,
+      "loss": 1.1815,
+      "step": 13546
+    },
+    {
+      "epoch": 0.9765011172781662,
+      "grad_norm": 0.1198050468081189,
+      "learning_rate": 2.895098387370454e-07,
+      "loss": 1.2167,
+      "step": 13547
+    },
+    {
+      "epoch": 0.9765731997405032,
+      "grad_norm": 0.12125761711430441,
+      "learning_rate": 2.87737279840472e-07,
+      "loss": 1.2335,
+      "step": 13548
+    },
+    {
+      "epoch": 0.9766452822028401,
+      "grad_norm": 0.12808001810974481,
+      "learning_rate": 2.8597015615426937e-07,
+      "loss": 1.2818,
+      "step": 13549
+    },
+    {
+      "epoch": 0.976717364665177,
+      "grad_norm": 0.13087192613193332,
+      "learning_rate": 2.842084677747603e-07,
+      "loss": 1.2565,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9767894471275139,
+      "grad_norm": 0.17799803986494564,
+      "learning_rate": 2.8245221479797913e-07,
+      "loss": 1.3238,
+      "step": 13551
+    },
+    {
+      "epoch": 0.9768615295898508,
+      "grad_norm": 0.11309147833051693,
+      "learning_rate": 2.807013973196493e-07,
+      "loss": 1.324,
+      "step": 13552
+    },
+    {
+      "epoch": 0.9769336120521876,
+      "grad_norm": 0.12300146800666945,
+      "learning_rate": 2.7895601543520553e-07,
+      "loss": 1.2575,
+      "step": 13553
+    },
+    {
+      "epoch": 0.9770056945145246,
+      "grad_norm": 0.1211882495245944,
+      "learning_rate": 2.7721606923978293e-07,
+      "loss": 1.2551,
+      "step": 13554
+    },
+    {
+      "epoch": 0.9770777769768615,
+      "grad_norm": 0.24907263061381557,
+      "learning_rate": 2.754815588282389e-07,
+      "loss": 1.3451,
+      "step": 13555
+    },
+    {
+      "epoch": 0.9771498594391984,
+      "grad_norm": 0.11138774860070354,
+      "learning_rate": 2.737524842951089e-07,
+      "loss": 1.3403,
+      "step": 13556
+    },
+    {
+      "epoch": 0.9772219419015353,
+      "grad_norm": 0.13396544066733931,
+      "learning_rate": 2.720288457346398e-07,
+      "loss": 1.2585,
+      "step": 13557
+    },
+    {
+      "epoch": 0.9772940243638722,
+      "grad_norm": 0.11737577132211058,
+      "learning_rate": 2.703106432408009e-07,
+      "loss": 1.2321,
+      "step": 13558
+    },
+    {
+      "epoch": 0.9773661068262092,
+      "grad_norm": 0.11950943265340232,
+      "learning_rate": 2.685978769072284e-07,
+      "loss": 1.2291,
+      "step": 13559
+    },
+    {
+      "epoch": 0.9774381892885461,
+      "grad_norm": 0.12077114745786346,
+      "learning_rate": 2.6689054682730306e-07,
+      "loss": 1.2367,
+      "step": 13560
+    },
+    {
+      "epoch": 0.977510271750883,
+      "grad_norm": 0.11032860811314746,
+      "learning_rate": 2.651886530940728e-07,
+      "loss": 1.171,
+      "step": 13561
+    },
+    {
+      "epoch": 0.9775823542132199,
+      "grad_norm": 0.12519816831449618,
+      "learning_rate": 2.6349219580031893e-07,
+      "loss": 1.2886,
+      "step": 13562
+    },
+    {
+      "epoch": 0.9776544366755568,
+      "grad_norm": 0.12981541986574688,
+      "learning_rate": 2.6180117503850077e-07,
+      "loss": 1.2532,
+      "step": 13563
+    },
+    {
+      "epoch": 0.9777265191378938,
+      "grad_norm": 0.11843627455924659,
+      "learning_rate": 2.601155909008113e-07,
+      "loss": 1.3285,
+      "step": 13564
+    },
+    {
+      "epoch": 0.9777986016002307,
+      "grad_norm": 0.11685455526623295,
+      "learning_rate": 2.584354434791214e-07,
+      "loss": 1.1925,
+      "step": 13565
+    },
+    {
+      "epoch": 0.9778706840625676,
+      "grad_norm": 0.11448002597396244,
+      "learning_rate": 2.5676073286500237e-07,
+      "loss": 1.2669,
+      "step": 13566
+    },
+    {
+      "epoch": 0.9779427665249045,
+      "grad_norm": 0.11698068386958334,
+      "learning_rate": 2.5509145914975885e-07,
+      "loss": 1.2231,
+      "step": 13567
+    },
+    {
+      "epoch": 0.9780148489872414,
+      "grad_norm": 0.11370789742766149,
+      "learning_rate": 2.534276224243848e-07,
+      "loss": 1.2633,
+      "step": 13568
+    },
+    {
+      "epoch": 0.9780869314495784,
+      "grad_norm": 0.1495428681736477,
+      "learning_rate": 2.517692227795521e-07,
+      "loss": 1.257,
+      "step": 13569
+    },
+    {
+      "epoch": 0.9781590139119152,
+      "grad_norm": 0.10843659519865925,
+      "learning_rate": 2.5011626030566614e-07,
+      "loss": 1.2909,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9782310963742521,
+      "grad_norm": 0.11240781451514151,
+      "learning_rate": 2.4846873509284383e-07,
+      "loss": 1.2598,
+      "step": 13571
+    },
+    {
+      "epoch": 0.978303178836589,
+      "grad_norm": 0.13182124209275076,
+      "learning_rate": 2.468266472308689e-07,
+      "loss": 1.1854,
+      "step": 13572
+    },
+    {
+      "epoch": 0.9783752612989259,
+      "grad_norm": 0.12730340683555288,
+      "learning_rate": 2.451899968092586e-07,
+      "loss": 1.2221,
+      "step": 13573
+    },
+    {
+      "epoch": 0.9784473437612629,
+      "grad_norm": 0.12325784557723257,
+      "learning_rate": 2.435587839172304e-07,
+      "loss": 1.3144,
+      "step": 13574
+    },
+    {
+      "epoch": 0.9785194262235998,
+      "grad_norm": 0.13428966735348788,
+      "learning_rate": 2.4193300864369104e-07,
+      "loss": 1.2664,
+      "step": 13575
+    },
+    {
+      "epoch": 0.9785915086859367,
+      "grad_norm": 0.1300839012648295,
+      "learning_rate": 2.4031267107726964e-07,
+      "loss": 1.3526,
+      "step": 13576
+    },
+    {
+      "epoch": 0.9786635911482736,
+      "grad_norm": 0.11815623443782619,
+      "learning_rate": 2.3869777130627324e-07,
+      "loss": 1.3125,
+      "step": 13577
+    },
+    {
+      "epoch": 0.9787356736106105,
+      "grad_norm": 0.10660100938482805,
+      "learning_rate": 2.370883094187537e-07,
+      "loss": 1.2516,
+      "step": 13578
+    },
+    {
+      "epoch": 0.9788077560729475,
+      "grad_norm": 0.12536343151775095,
+      "learning_rate": 2.3548428550241865e-07,
+      "loss": 1.2793,
+      "step": 13579
+    },
+    {
+      "epoch": 0.9788798385352844,
+      "grad_norm": 0.12819590798243982,
+      "learning_rate": 2.3388569964470918e-07,
+      "loss": 1.1678,
+      "step": 13580
+    },
+    {
+      "epoch": 0.9789519209976213,
+      "grad_norm": 0.11633547600117192,
+      "learning_rate": 2.3229255193275569e-07,
+      "loss": 1.2963,
+      "step": 13581
+    },
+    {
+      "epoch": 0.9790240034599582,
+      "grad_norm": 0.14535674135641036,
+      "learning_rate": 2.3070484245341084e-07,
+      "loss": 1.3685,
+      "step": 13582
+    },
+    {
+      "epoch": 0.9790960859222951,
+      "grad_norm": 0.13289092874192074,
+      "learning_rate": 2.2912257129320547e-07,
+      "loss": 1.2803,
+      "step": 13583
+    },
+    {
+      "epoch": 0.979168168384632,
+      "grad_norm": 0.13414149407247447,
+      "learning_rate": 2.2754573853841498e-07,
+      "loss": 1.2879,
+      "step": 13584
+    },
+    {
+      "epoch": 0.979240250846969,
+      "grad_norm": 0.1236279015029758,
+      "learning_rate": 2.2597434427495956e-07,
+      "loss": 1.2913,
+      "step": 13585
+    },
+    {
+      "epoch": 0.9793123333093059,
+      "grad_norm": 0.12192577474191073,
+      "learning_rate": 2.24408388588504e-07,
+      "loss": 1.2151,
+      "step": 13586
+    },
+    {
+      "epoch": 0.9793844157716428,
+      "grad_norm": 0.13042283017597492,
+      "learning_rate": 2.2284787156441333e-07,
+      "loss": 1.271,
+      "step": 13587
+    },
+    {
+      "epoch": 0.9794564982339796,
+      "grad_norm": 0.12225888798867482,
+      "learning_rate": 2.2129279328774176e-07,
+      "loss": 1.2298,
+      "step": 13588
+    },
+    {
+      "epoch": 0.9795285806963165,
+      "grad_norm": 0.12058189794830808,
+      "learning_rate": 2.1974315384325485e-07,
+      "loss": 1.2691,
+      "step": 13589
+    },
+    {
+      "epoch": 0.9796006631586535,
+      "grad_norm": 0.12271290489887551,
+      "learning_rate": 2.181989533154405e-07,
+      "loss": 1.2996,
+      "step": 13590
+    },
+    {
+      "epoch": 0.9796727456209904,
+      "grad_norm": 0.12283918301424729,
+      "learning_rate": 2.1666019178844255e-07,
+      "loss": 1.324,
+      "step": 13591
+    },
+    {
+      "epoch": 0.9797448280833273,
+      "grad_norm": 0.12651286486872096,
+      "learning_rate": 2.1512686934616055e-07,
+      "loss": 1.2352,
+      "step": 13592
+    },
+    {
+      "epoch": 0.9798169105456642,
+      "grad_norm": 0.1289484598466807,
+      "learning_rate": 2.13598986072161e-07,
+      "loss": 1.2404,
+      "step": 13593
+    },
+    {
+      "epoch": 0.9798889930080011,
+      "grad_norm": 0.141872955396687,
+      "learning_rate": 2.120765420497328e-07,
+      "loss": 1.3073,
+      "step": 13594
+    },
+    {
+      "epoch": 0.9799610754703381,
+      "grad_norm": 0.12093007064256543,
+      "learning_rate": 2.1055953736186517e-07,
+      "loss": 1.2837,
+      "step": 13595
+    },
+    {
+      "epoch": 0.980033157932675,
+      "grad_norm": 0.12826375024828424,
+      "learning_rate": 2.090479720912364e-07,
+      "loss": 1.2157,
+      "step": 13596
+    },
+    {
+      "epoch": 0.9801052403950119,
+      "grad_norm": 0.11534548263323967,
+      "learning_rate": 2.075418463202583e-07,
+      "loss": 1.3145,
+      "step": 13597
+    },
+    {
+      "epoch": 0.9801773228573488,
+      "grad_norm": 0.1248793870114669,
+      "learning_rate": 2.0604116013100972e-07,
+      "loss": 1.1598,
+      "step": 13598
+    },
+    {
+      "epoch": 0.9802494053196857,
+      "grad_norm": 0.1327677652425344,
+      "learning_rate": 2.0454591360530297e-07,
+      "loss": 1.2767,
+      "step": 13599
+    },
+    {
+      "epoch": 0.9803214877820227,
+      "grad_norm": 0.14440813117019197,
+      "learning_rate": 2.030561068246395e-07,
+      "loss": 1.3,
+      "step": 13600
+    },
+    {
+      "epoch": 0.9803935702443596,
+      "grad_norm": 0.12923765786112884,
+      "learning_rate": 2.0157173987022104e-07,
+      "loss": 1.2057,
+      "step": 13601
+    },
+    {
+      "epoch": 0.9804656527066965,
+      "grad_norm": 0.11099280632831454,
+      "learning_rate": 2.0009281282297176e-07,
+      "loss": 1.2509,
+      "step": 13602
+    },
+    {
+      "epoch": 0.9805377351690334,
+      "grad_norm": 0.11863812776732732,
+      "learning_rate": 1.9861932576350494e-07,
+      "loss": 1.2768,
+      "step": 13603
+    },
+    {
+      "epoch": 0.9806098176313703,
+      "grad_norm": 0.13020003743117184,
+      "learning_rate": 1.9715127877212304e-07,
+      "loss": 1.2404,
+      "step": 13604
+    },
+    {
+      "epoch": 0.9806819000937071,
+      "grad_norm": 0.1181264038228879,
+      "learning_rate": 1.9568867192886197e-07,
+      "loss": 1.3717,
+      "step": 13605
+    },
+    {
+      "epoch": 0.9807539825560441,
+      "grad_norm": 0.1128734654093173,
+      "learning_rate": 1.9423150531344692e-07,
+      "loss": 1.3756,
+      "step": 13606
+    },
+    {
+      "epoch": 0.980826065018381,
+      "grad_norm": 0.11131299472718256,
+      "learning_rate": 1.927797790053032e-07,
+      "loss": 1.2532,
+      "step": 13607
+    },
+    {
+      "epoch": 0.9808981474807179,
+      "grad_norm": 0.12214703822409165,
+      "learning_rate": 1.913334930835564e-07,
+      "loss": 1.2839,
+      "step": 13608
+    },
+    {
+      "epoch": 0.9809702299430548,
+      "grad_norm": 0.1182241460494459,
+      "learning_rate": 1.898926476270546e-07,
+      "loss": 1.2479,
+      "step": 13609
+    },
+    {
+      "epoch": 0.9810423124053917,
+      "grad_norm": 0.11434944819185756,
+      "learning_rate": 1.8845724271432386e-07,
+      "loss": 1.2059,
+      "step": 13610
+    },
+    {
+      "epoch": 0.9811143948677287,
+      "grad_norm": 0.11242026778095662,
+      "learning_rate": 1.8702727842361267e-07,
+      "loss": 1.3539,
+      "step": 13611
+    },
+    {
+      "epoch": 0.9811864773300656,
+      "grad_norm": 0.13177810710464413,
+      "learning_rate": 1.8560275483286982e-07,
+      "loss": 1.2105,
+      "step": 13612
+    },
+    {
+      "epoch": 0.9812585597924025,
+      "grad_norm": 0.1217774800825044,
+      "learning_rate": 1.841836720197443e-07,
+      "loss": 1.2369,
+      "step": 13613
+    },
+    {
+      "epoch": 0.9813306422547394,
+      "grad_norm": 0.12249668361903189,
+      "learning_rate": 1.8277003006158532e-07,
+      "loss": 1.2829,
+      "step": 13614
+    },
+    {
+      "epoch": 0.9814027247170763,
+      "grad_norm": 0.12322279540992466,
+      "learning_rate": 1.813618290354535e-07,
+      "loss": 1.2931,
+      "step": 13615
+    },
+    {
+      "epoch": 0.9814748071794133,
+      "grad_norm": 0.12136142017726972,
+      "learning_rate": 1.7995906901809857e-07,
+      "loss": 1.2386,
+      "step": 13616
+    },
+    {
+      "epoch": 0.9815468896417502,
+      "grad_norm": 0.12116391755321698,
+      "learning_rate": 1.7856175008599264e-07,
+      "loss": 1.2347,
+      "step": 13617
+    },
+    {
+      "epoch": 0.9816189721040871,
+      "grad_norm": 0.11424647718648888,
+      "learning_rate": 1.7716987231529703e-07,
+      "loss": 1.2261,
+      "step": 13618
+    },
+    {
+      "epoch": 0.981691054566424,
+      "grad_norm": 0.11255886283718254,
+      "learning_rate": 1.7578343578188438e-07,
+      "loss": 1.2656,
+      "step": 13619
+    },
+    {
+      "epoch": 0.981763137028761,
+      "grad_norm": 0.12222719338766753,
+      "learning_rate": 1.744024405613276e-07,
+      "loss": 1.1303,
+      "step": 13620
+    },
+    {
+      "epoch": 0.9818352194910979,
+      "grad_norm": 0.11863252733278307,
+      "learning_rate": 1.7302688672889977e-07,
+      "loss": 1.2627,
+      "step": 13621
+    },
+    {
+      "epoch": 0.9819073019534347,
+      "grad_norm": 0.11950083241025616,
+      "learning_rate": 1.716567743595965e-07,
+      "loss": 1.1779,
+      "step": 13622
+    },
+    {
+      "epoch": 0.9819793844157716,
+      "grad_norm": 0.11785516886444253,
+      "learning_rate": 1.7029210352808023e-07,
+      "loss": 1.3109,
+      "step": 13623
+    },
+    {
+      "epoch": 0.9820514668781085,
+      "grad_norm": 0.11686122421031479,
+      "learning_rate": 1.6893287430874704e-07,
+      "loss": 1.1918,
+      "step": 13624
+    },
+    {
+      "epoch": 0.9821235493404454,
+      "grad_norm": 0.12991534829899068,
+      "learning_rate": 1.6757908677568214e-07,
+      "loss": 1.2828,
+      "step": 13625
+    },
+    {
+      "epoch": 0.9821956318027824,
+      "grad_norm": 0.12706815215001654,
+      "learning_rate": 1.662307410026931e-07,
+      "loss": 1.3382,
+      "step": 13626
+    },
+    {
+      "epoch": 0.9822677142651193,
+      "grad_norm": 0.11059091240826005,
+      "learning_rate": 1.648878370632656e-07,
+      "loss": 1.3001,
+      "step": 13627
+    },
+    {
+      "epoch": 0.9823397967274562,
+      "grad_norm": 0.12388676497834232,
+      "learning_rate": 1.6355037503059667e-07,
+      "loss": 1.3179,
+      "step": 13628
+    },
+    {
+      "epoch": 0.9824118791897931,
+      "grad_norm": 0.12503959104734688,
+      "learning_rate": 1.6221835497759462e-07,
+      "loss": 1.2824,
+      "step": 13629
+    },
+    {
+      "epoch": 0.98248396165213,
+      "grad_norm": 0.1255630264700669,
+      "learning_rate": 1.6089177697686808e-07,
+      "loss": 1.3098,
+      "step": 13630
+    },
+    {
+      "epoch": 0.982556044114467,
+      "grad_norm": 0.1224357748275856,
+      "learning_rate": 1.5957064110072584e-07,
+      "loss": 1.2102,
+      "step": 13631
+    },
+    {
+      "epoch": 0.9826281265768039,
+      "grad_norm": 0.13545341841272604,
+      "learning_rate": 1.5825494742117697e-07,
+      "loss": 1.1765,
+      "step": 13632
+    },
+    {
+      "epoch": 0.9827002090391408,
+      "grad_norm": 0.12354556909240418,
+      "learning_rate": 1.56944696009953e-07,
+      "loss": 1.2724,
+      "step": 13633
+    },
+    {
+      "epoch": 0.9827722915014777,
+      "grad_norm": 0.1250945272463245,
+      "learning_rate": 1.5563988693845233e-07,
+      "loss": 1.2135,
+      "step": 13634
+    },
+    {
+      "epoch": 0.9828443739638146,
+      "grad_norm": 0.1563937820748285,
+      "learning_rate": 1.543405202778181e-07,
+      "loss": 1.1795,
+      "step": 13635
+    },
+    {
+      "epoch": 0.9829164564261516,
+      "grad_norm": 0.12278779934011227,
+      "learning_rate": 1.530465960988714e-07,
+      "loss": 1.3532,
+      "step": 13636
+    },
+    {
+      "epoch": 0.9829885388884885,
+      "grad_norm": 0.1140209790734105,
+      "learning_rate": 1.517581144721447e-07,
+      "loss": 1.2781,
+      "step": 13637
+    },
+    {
+      "epoch": 0.9830606213508254,
+      "grad_norm": 0.11119577749463262,
+      "learning_rate": 1.5047507546785966e-07,
+      "loss": 1.2839,
+      "step": 13638
+    },
+    {
+      "epoch": 0.9831327038131623,
+      "grad_norm": 0.11925050806854791,
+      "learning_rate": 1.4919747915597138e-07,
+      "loss": 1.2854,
+      "step": 13639
+    },
+    {
+      "epoch": 0.9832047862754991,
+      "grad_norm": 0.11650200183708856,
+      "learning_rate": 1.4792532560610195e-07,
+      "loss": 1.216,
+      "step": 13640
+    },
+    {
+      "epoch": 0.983276868737836,
+      "grad_norm": 0.11269331013976258,
+      "learning_rate": 1.4665861488761813e-07,
+      "loss": 1.2656,
+      "step": 13641
+    },
+    {
+      "epoch": 0.983348951200173,
+      "grad_norm": 0.129675781041011,
+      "learning_rate": 1.4539734706954244e-07,
+      "loss": 1.2935,
+      "step": 13642
+    },
+    {
+      "epoch": 0.9834210336625099,
+      "grad_norm": 0.1070307216177849,
+      "learning_rate": 1.4414152222063105e-07,
+      "loss": 1.1709,
+      "step": 13643
+    },
+    {
+      "epoch": 0.9834931161248468,
+      "grad_norm": 0.11229796443779894,
+      "learning_rate": 1.4289114040936246e-07,
+      "loss": 1.2249,
+      "step": 13644
+    },
+    {
+      "epoch": 0.9835651985871837,
+      "grad_norm": 0.14955718786947048,
+      "learning_rate": 1.4164620170386e-07,
+      "loss": 1.3229,
+      "step": 13645
+    },
+    {
+      "epoch": 0.9836372810495206,
+      "grad_norm": 0.11784048360225508,
+      "learning_rate": 1.404067061719916e-07,
+      "loss": 1.2709,
+      "step": 13646
+    },
+    {
+      "epoch": 0.9837093635118576,
+      "grad_norm": 0.10589409189975167,
+      "learning_rate": 1.391726538813365e-07,
+      "loss": 1.2128,
+      "step": 13647
+    },
+    {
+      "epoch": 0.9837814459741945,
+      "grad_norm": 0.12741288858517366,
+      "learning_rate": 1.3794404489915203e-07,
+      "loss": 1.3793,
+      "step": 13648
+    },
+    {
+      "epoch": 0.9838535284365314,
+      "grad_norm": 0.11787369870120935,
+      "learning_rate": 1.3672087929240685e-07,
+      "loss": 1.2312,
+      "step": 13649
+    },
+    {
+      "epoch": 0.9839256108988683,
+      "grad_norm": 0.12503961136482458,
+      "learning_rate": 1.3550315712778093e-07,
+      "loss": 1.2053,
+      "step": 13650
+    },
+    {
+      "epoch": 0.9839976933612052,
+      "grad_norm": 0.10956548396194096,
+      "learning_rate": 1.3429087847163234e-07,
+      "loss": 1.2994,
+      "step": 13651
+    },
+    {
+      "epoch": 0.9840697758235422,
+      "grad_norm": 0.12663744603638838,
+      "learning_rate": 1.3308404339007484e-07,
+      "loss": 1.2978,
+      "step": 13652
+    },
+    {
+      "epoch": 0.9841418582858791,
+      "grad_norm": 0.12160037137151623,
+      "learning_rate": 1.3188265194885586e-07,
+      "loss": 1.2775,
+      "step": 13653
+    },
+    {
+      "epoch": 0.984213940748216,
+      "grad_norm": 0.12698735863822294,
+      "learning_rate": 1.3068670421348962e-07,
+      "loss": 1.3106,
+      "step": 13654
+    },
+    {
+      "epoch": 0.9842860232105529,
+      "grad_norm": 0.12925655349843135,
+      "learning_rate": 1.294962002491462e-07,
+      "loss": 1.2644,
+      "step": 13655
+    },
+    {
+      "epoch": 0.9843581056728898,
+      "grad_norm": 0.13521272231053635,
+      "learning_rate": 1.2831114012072932e-07,
+      "loss": 1.1735,
+      "step": 13656
+    },
+    {
+      "epoch": 0.9844301881352266,
+      "grad_norm": 0.11205169270918179,
+      "learning_rate": 1.2713152389283167e-07,
+      "loss": 1.3138,
+      "step": 13657
+    },
+    {
+      "epoch": 0.9845022705975636,
+      "grad_norm": 0.14655394857144752,
+      "learning_rate": 1.2595735162975743e-07,
+      "loss": 1.3833,
+      "step": 13658
+    },
+    {
+      "epoch": 0.9845743530599005,
+      "grad_norm": 0.12015993060905206,
+      "learning_rate": 1.247886233954998e-07,
+      "loss": 1.244,
+      "step": 13659
+    },
+    {
+      "epoch": 0.9846464355222374,
+      "grad_norm": 0.11759875757253654,
+      "learning_rate": 1.2362533925377452e-07,
+      "loss": 1.233,
+      "step": 13660
+    },
+    {
+      "epoch": 0.9847185179845743,
+      "grad_norm": 0.12508565659728413,
+      "learning_rate": 1.224674992679975e-07,
+      "loss": 1.319,
+      "step": 13661
+    },
+    {
+      "epoch": 0.9847906004469112,
+      "grad_norm": 0.12018619516445597,
+      "learning_rate": 1.2131510350126274e-07,
+      "loss": 1.2696,
+      "step": 13662
+    },
+    {
+      "epoch": 0.9848626829092482,
+      "grad_norm": 0.19669471820215786,
+      "learning_rate": 1.2016815201639775e-07,
+      "loss": 1.1986,
+      "step": 13663
+    },
+    {
+      "epoch": 0.9849347653715851,
+      "grad_norm": 0.1112763874514656,
+      "learning_rate": 1.1902664487590809e-07,
+      "loss": 1.3795,
+      "step": 13664
+    },
+    {
+      "epoch": 0.985006847833922,
+      "grad_norm": 0.14083694701212257,
+      "learning_rate": 1.1789058214204396e-07,
+      "loss": 1.3948,
+      "step": 13665
+    },
+    {
+      "epoch": 0.9850789302962589,
+      "grad_norm": 0.141614659520403,
+      "learning_rate": 1.167599638767003e-07,
+      "loss": 1.1659,
+      "step": 13666
+    },
+    {
+      "epoch": 0.9851510127585958,
+      "grad_norm": 0.11085460167557797,
+      "learning_rate": 1.1563479014152778e-07,
+      "loss": 1.2944,
+      "step": 13667
+    },
+    {
+      "epoch": 0.9852230952209328,
+      "grad_norm": 0.11361721096451899,
+      "learning_rate": 1.1451506099784404e-07,
+      "loss": 1.2138,
+      "step": 13668
+    },
+    {
+      "epoch": 0.9852951776832697,
+      "grad_norm": 0.11061917780424257,
+      "learning_rate": 1.1340077650670022e-07,
+      "loss": 1.2586,
+      "step": 13669
+    },
+    {
+      "epoch": 0.9853672601456066,
+      "grad_norm": 0.12040107222498024,
+      "learning_rate": 1.1229193672881444e-07,
+      "loss": 1.3205,
+      "step": 13670
+    },
+    {
+      "epoch": 0.9854393426079435,
+      "grad_norm": 0.11950881939792489,
+      "learning_rate": 1.1118854172464942e-07,
+      "loss": 1.3707,
+      "step": 13671
+    },
+    {
+      "epoch": 0.9855114250702804,
+      "grad_norm": 0.1193989933010721,
+      "learning_rate": 1.1009059155433488e-07,
+      "loss": 1.2214,
+      "step": 13672
+    },
+    {
+      "epoch": 0.9855835075326174,
+      "grad_norm": 0.11390631325187302,
+      "learning_rate": 1.089980862777229e-07,
+      "loss": 1.3126,
+      "step": 13673
+    },
+    {
+      "epoch": 0.9856555899949542,
+      "grad_norm": 0.12183173050722486,
+      "learning_rate": 1.0791102595436587e-07,
+      "loss": 1.2749,
+      "step": 13674
+    },
+    {
+      "epoch": 0.9857276724572911,
+      "grad_norm": 0.13675054369986814,
+      "learning_rate": 1.0682941064351637e-07,
+      "loss": 1.2746,
+      "step": 13675
+    },
+    {
+      "epoch": 0.985799754919628,
+      "grad_norm": 0.11698017625555468,
+      "learning_rate": 1.0575324040413836e-07,
+      "loss": 1.2401,
+      "step": 13676
+    },
+    {
+      "epoch": 0.9858718373819649,
+      "grad_norm": 0.11919433314122066,
+      "learning_rate": 1.0468251529488492e-07,
+      "loss": 1.1365,
+      "step": 13677
+    },
+    {
+      "epoch": 0.9859439198443019,
+      "grad_norm": 0.1194594478213137,
+      "learning_rate": 1.0361723537412049e-07,
+      "loss": 1.2591,
+      "step": 13678
+    },
+    {
+      "epoch": 0.9860160023066388,
+      "grad_norm": 0.12949305188747146,
+      "learning_rate": 1.0255740069992081e-07,
+      "loss": 1.2551,
+      "step": 13679
+    },
+    {
+      "epoch": 0.9860880847689757,
+      "grad_norm": 0.12857441443563528,
+      "learning_rate": 1.015030113300508e-07,
+      "loss": 1.212,
+      "step": 13680
+    },
+    {
+      "epoch": 0.9861601672313126,
+      "grad_norm": 0.10813794311961425,
+      "learning_rate": 1.0045406732197559e-07,
+      "loss": 1.2454,
+      "step": 13681
+    },
+    {
+      "epoch": 0.9862322496936495,
+      "grad_norm": 0.11474906798939989,
+      "learning_rate": 9.941056873289389e-08,
+      "loss": 1.2136,
+      "step": 13682
+    },
+    {
+      "epoch": 0.9863043321559865,
+      "grad_norm": 0.12007013857209653,
+      "learning_rate": 9.83725156196602e-08,
+      "loss": 1.2647,
+      "step": 13683
+    },
+    {
+      "epoch": 0.9863764146183234,
+      "grad_norm": 0.1172175632487065,
+      "learning_rate": 9.733990803887371e-08,
+      "loss": 1.1647,
+      "step": 13684
+    },
+    {
+      "epoch": 0.9864484970806603,
+      "grad_norm": 0.1395387906411243,
+      "learning_rate": 9.631274604682273e-08,
+      "loss": 1.2588,
+      "step": 13685
+    },
+    {
+      "epoch": 0.9865205795429972,
+      "grad_norm": 0.12981564905433643,
+      "learning_rate": 9.529102969948467e-08,
+      "loss": 1.2512,
+      "step": 13686
+    },
+    {
+      "epoch": 0.9865926620053341,
+      "grad_norm": 0.23741113072709755,
+      "learning_rate": 9.427475905255944e-08,
+      "loss": 1.3037,
+      "step": 13687
+    },
+    {
+      "epoch": 0.986664744467671,
+      "grad_norm": 0.13078512518028387,
+      "learning_rate": 9.326393416144719e-08,
+      "loss": 1.2544,
+      "step": 13688
+    },
+    {
+      "epoch": 0.986736826930008,
+      "grad_norm": 0.1128892201846755,
+      "learning_rate": 9.225855508123715e-08,
+      "loss": 1.3313,
+      "step": 13689
+    },
+    {
+      "epoch": 0.9868089093923449,
+      "grad_norm": 0.12278527522276242,
+      "learning_rate": 9.125862186674106e-08,
+      "loss": 1.2103,
+      "step": 13690
+    },
+    {
+      "epoch": 0.9868809918546818,
+      "grad_norm": 0.11794963926453664,
+      "learning_rate": 9.026413457244864e-08,
+      "loss": 1.3007,
+      "step": 13691
+    },
+    {
+      "epoch": 0.9869530743170186,
+      "grad_norm": 0.12344792427635891,
+      "learning_rate": 8.92750932525832e-08,
+      "loss": 1.237,
+      "step": 13692
+    },
+    {
+      "epoch": 0.9870251567793555,
+      "grad_norm": 0.11174273756713132,
+      "learning_rate": 8.829149796105718e-08,
+      "loss": 1.2798,
+      "step": 13693
+    },
+    {
+      "epoch": 0.9870972392416925,
+      "grad_norm": 0.1247553914676472,
+      "learning_rate": 8.731334875148322e-08,
+      "loss": 1.338,
+      "step": 13694
+    },
+    {
+      "epoch": 0.9871693217040294,
+      "grad_norm": 0.11795271792716897,
+      "learning_rate": 8.634064567716315e-08,
+      "loss": 1.2969,
+      "step": 13695
+    },
+    {
+      "epoch": 0.9872414041663663,
+      "grad_norm": 0.11660903332909059,
+      "learning_rate": 8.537338879114342e-08,
+      "loss": 1.3301,
+      "step": 13696
+    },
+    {
+      "epoch": 0.9873134866287032,
+      "grad_norm": 0.1190851024569839,
+      "learning_rate": 8.441157814612632e-08,
+      "loss": 1.2151,
+      "step": 13697
+    },
+    {
+      "epoch": 0.9873855690910401,
+      "grad_norm": 0.12389914848261076,
+      "learning_rate": 8.345521379455879e-08,
+      "loss": 1.2751,
+      "step": 13698
+    },
+    {
+      "epoch": 0.9874576515533771,
+      "grad_norm": 0.12690384782591038,
+      "learning_rate": 8.250429578855467e-08,
+      "loss": 1.4322,
+      "step": 13699
+    },
+    {
+      "epoch": 0.987529734015714,
+      "grad_norm": 0.12282895143673163,
+      "learning_rate": 8.155882417995032e-08,
+      "loss": 1.2702,
+      "step": 13700
+    },
+    {
+      "epoch": 0.9876018164780509,
+      "grad_norm": 0.10980953812587127,
+      "learning_rate": 8.061879902029335e-08,
+      "loss": 1.233,
+      "step": 13701
+    },
+    {
+      "epoch": 0.9876738989403878,
+      "grad_norm": 0.11915302471774931,
+      "learning_rate": 7.968422036080946e-08,
+      "loss": 1.2684,
+      "step": 13702
+    },
+    {
+      "epoch": 0.9877459814027247,
+      "grad_norm": 0.1192686354750224,
+      "learning_rate": 7.875508825244682e-08,
+      "loss": 1.2168,
+      "step": 13703
+    },
+    {
+      "epoch": 0.9878180638650617,
+      "grad_norm": 0.11016660884726014,
+      "learning_rate": 7.783140274585377e-08,
+      "loss": 1.1981,
+      "step": 13704
+    },
+    {
+      "epoch": 0.9878901463273986,
+      "grad_norm": 0.14164197240400137,
+      "learning_rate": 7.691316389139003e-08,
+      "loss": 1.3841,
+      "step": 13705
+    },
+    {
+      "epoch": 0.9879622287897355,
+      "grad_norm": 0.1118224164791065,
+      "learning_rate": 7.600037173909336e-08,
+      "loss": 1.226,
+      "step": 13706
+    },
+    {
+      "epoch": 0.9880343112520724,
+      "grad_norm": 0.12995793232685243,
+      "learning_rate": 7.509302633871284e-08,
+      "loss": 1.1943,
+      "step": 13707
+    },
+    {
+      "epoch": 0.9881063937144093,
+      "grad_norm": 0.11657533270651234,
+      "learning_rate": 7.419112773972003e-08,
+      "loss": 1.2191,
+      "step": 13708
+    },
+    {
+      "epoch": 0.9881784761767461,
+      "grad_norm": 0.20973565746778738,
+      "learning_rate": 7.32946759912756e-08,
+      "loss": 1.2311,
+      "step": 13709
+    },
+    {
+      "epoch": 0.9882505586390831,
+      "grad_norm": 0.1579076699850002,
+      "learning_rate": 7.240367114224044e-08,
+      "loss": 1.2679,
+      "step": 13710
+    },
+    {
+      "epoch": 0.98832264110142,
+      "grad_norm": 0.1175918739194208,
+      "learning_rate": 7.151811324117575e-08,
+      "loss": 1.2067,
+      "step": 13711
+    },
+    {
+      "epoch": 0.9883947235637569,
+      "grad_norm": 0.10997525710059071,
+      "learning_rate": 7.063800233636509e-08,
+      "loss": 1.273,
+      "step": 13712
+    },
+    {
+      "epoch": 0.9884668060260938,
+      "grad_norm": 0.1096789685651622,
+      "learning_rate": 6.976333847578121e-08,
+      "loss": 1.2448,
+      "step": 13713
+    },
+    {
+      "epoch": 0.9885388884884307,
+      "grad_norm": 0.12114466023196148,
+      "learning_rate": 6.889412170708598e-08,
+      "loss": 1.3297,
+      "step": 13714
+    },
+    {
+      "epoch": 0.9886109709507677,
+      "grad_norm": 0.12139297580067145,
+      "learning_rate": 6.803035207767483e-08,
+      "loss": 1.1938,
+      "step": 13715
+    },
+    {
+      "epoch": 0.9886830534131046,
+      "grad_norm": 0.13294713102672864,
+      "learning_rate": 6.717202963463232e-08,
+      "loss": 1.2899,
+      "step": 13716
+    },
+    {
+      "epoch": 0.9887551358754415,
+      "grad_norm": 0.1095556646402515,
+      "learning_rate": 6.631915442473213e-08,
+      "loss": 1.3623,
+      "step": 13717
+    },
+    {
+      "epoch": 0.9888272183377784,
+      "grad_norm": 0.12651810206508426,
+      "learning_rate": 6.547172649448152e-08,
+      "loss": 1.2913,
+      "step": 13718
+    },
+    {
+      "epoch": 0.9888993008001153,
+      "grad_norm": 0.11927476674586443,
+      "learning_rate": 6.462974589004355e-08,
+      "loss": 1.2037,
+      "step": 13719
+    },
+    {
+      "epoch": 0.9889713832624523,
+      "grad_norm": 0.1193488554594425,
+      "learning_rate": 6.379321265734817e-08,
+      "loss": 1.339,
+      "step": 13720
+    },
+    {
+      "epoch": 0.9890434657247892,
+      "grad_norm": 0.13189595002676902,
+      "learning_rate": 6.296212684197e-08,
+      "loss": 1.3529,
+      "step": 13721
+    },
+    {
+      "epoch": 0.9891155481871261,
+      "grad_norm": 0.11571785693670014,
+      "learning_rate": 6.213648848921728e-08,
+      "loss": 1.3198,
+      "step": 13722
+    },
+    {
+      "epoch": 0.989187630649463,
+      "grad_norm": 0.1161915419753958,
+      "learning_rate": 6.131629764409841e-08,
+      "loss": 1.3765,
+      "step": 13723
+    },
+    {
+      "epoch": 0.9892597131117999,
+      "grad_norm": 0.11945754413311142,
+      "learning_rate": 6.05015543513221e-08,
+      "loss": 1.2843,
+      "step": 13724
+    },
+    {
+      "epoch": 0.9893317955741369,
+      "grad_norm": 0.12206360361398944,
+      "learning_rate": 5.969225865528615e-08,
+      "loss": 1.2723,
+      "step": 13725
+    },
+    {
+      "epoch": 0.9894038780364737,
+      "grad_norm": 0.11742699200445464,
+      "learning_rate": 5.888841060012196e-08,
+      "loss": 1.2909,
+      "step": 13726
+    },
+    {
+      "epoch": 0.9894759604988106,
+      "grad_norm": 0.132841267424337,
+      "learning_rate": 5.809001022962779e-08,
+      "loss": 1.2685,
+      "step": 13727
+    },
+    {
+      "epoch": 0.9895480429611475,
+      "grad_norm": 0.11975543125860122,
+      "learning_rate": 5.7297057587346604e-08,
+      "loss": 1.1975,
+      "step": 13728
+    },
+    {
+      "epoch": 0.9896201254234844,
+      "grad_norm": 0.1128968227512339,
+      "learning_rate": 5.6509552716477176e-08,
+      "loss": 1.31,
+      "step": 13729
+    },
+    {
+      "epoch": 0.9896922078858214,
+      "grad_norm": 0.1127917480863957,
+      "learning_rate": 5.572749565996294e-08,
+      "loss": 1.2786,
+      "step": 13730
+    },
+    {
+      "epoch": 0.9897642903481583,
+      "grad_norm": 0.13130126563313482,
+      "learning_rate": 5.495088646042534e-08,
+      "loss": 1.1556,
+      "step": 13731
+    },
+    {
+      "epoch": 0.9898363728104952,
+      "grad_norm": 0.121067324012616,
+      "learning_rate": 5.41797251602083e-08,
+      "loss": 1.2764,
+      "step": 13732
+    },
+    {
+      "epoch": 0.9899084552728321,
+      "grad_norm": 0.11522646221472035,
+      "learning_rate": 5.341401180132266e-08,
+      "loss": 1.3109,
+      "step": 13733
+    },
+    {
+      "epoch": 0.989980537735169,
+      "grad_norm": 0.118814677933923,
+      "learning_rate": 5.2653746425535e-08,
+      "loss": 1.3098,
+      "step": 13734
+    },
+    {
+      "epoch": 0.990052620197506,
+      "grad_norm": 0.11035903729147846,
+      "learning_rate": 5.189892907425664e-08,
+      "loss": 1.2836,
+      "step": 13735
+    },
+    {
+      "epoch": 0.9901247026598429,
+      "grad_norm": 0.11365263083980373,
+      "learning_rate": 5.1149559788654655e-08,
+      "loss": 1.1782,
+      "step": 13736
+    },
+    {
+      "epoch": 0.9901967851221798,
+      "grad_norm": 0.1379614516965514,
+      "learning_rate": 5.040563860957415e-08,
+      "loss": 1.416,
+      "step": 13737
+    },
+    {
+      "epoch": 0.9902688675845167,
+      "grad_norm": 0.13585165024795504,
+      "learning_rate": 4.966716557756046e-08,
+      "loss": 1.2032,
+      "step": 13738
+    },
+    {
+      "epoch": 0.9903409500468536,
+      "grad_norm": 0.1323418327633419,
+      "learning_rate": 4.893414073285918e-08,
+      "loss": 1.313,
+      "step": 13739
+    },
+    {
+      "epoch": 0.9904130325091905,
+      "grad_norm": 0.21767130110309438,
+      "learning_rate": 4.820656411543834e-08,
+      "loss": 1.2323,
+      "step": 13740
+    },
+    {
+      "epoch": 0.9904851149715275,
+      "grad_norm": 0.1238139270381781,
+      "learning_rate": 4.7484435764966195e-08,
+      "loss": 1.2926,
+      "step": 13741
+    },
+    {
+      "epoch": 0.9905571974338644,
+      "grad_norm": 0.11625220668672726,
+      "learning_rate": 4.676775572077796e-08,
+      "loss": 1.2734,
+      "step": 13742
+    },
+    {
+      "epoch": 0.9906292798962013,
+      "grad_norm": 0.11883671814491283,
+      "learning_rate": 4.6056524021964584e-08,
+      "loss": 1.2445,
+      "step": 13743
+    },
+    {
+      "epoch": 0.9907013623585381,
+      "grad_norm": 0.11365426562286787,
+      "learning_rate": 4.535074070727285e-08,
+      "loss": 1.1868,
+      "step": 13744
+    },
+    {
+      "epoch": 0.990773444820875,
+      "grad_norm": 0.10674610630306029,
+      "learning_rate": 4.4650405815194196e-08,
+      "loss": 1.2401,
+      "step": 13745
+    },
+    {
+      "epoch": 0.990845527283212,
+      "grad_norm": 0.12262291394390055,
+      "learning_rate": 4.395551938389808e-08,
+      "loss": 1.2977,
+      "step": 13746
+    },
+    {
+      "epoch": 0.9909176097455489,
+      "grad_norm": 0.10800324301964517,
+      "learning_rate": 4.326608145124311e-08,
+      "loss": 1.2754,
+      "step": 13747
+    },
+    {
+      "epoch": 0.9909896922078858,
+      "grad_norm": 0.1265678586589161,
+      "learning_rate": 4.258209205484365e-08,
+      "loss": 1.3459,
+      "step": 13748
+    },
+    {
+      "epoch": 0.9910617746702227,
+      "grad_norm": 0.12177488027306649,
+      "learning_rate": 4.1903551231947666e-08,
+      "loss": 1.3109,
+      "step": 13749
+    },
+    {
+      "epoch": 0.9911338571325596,
+      "grad_norm": 0.12014232234271247,
+      "learning_rate": 4.123045901957001e-08,
+      "loss": 1.196,
+      "step": 13750
+    },
+    {
+      "epoch": 0.9912059395948966,
+      "grad_norm": 0.11481792219958906,
+      "learning_rate": 4.056281545439245e-08,
+      "loss": 1.2745,
+      "step": 13751
+    },
+    {
+      "epoch": 0.9912780220572335,
+      "grad_norm": 0.14349791087208374,
+      "learning_rate": 3.990062057278587e-08,
+      "loss": 1.3133,
+      "step": 13752
+    },
+    {
+      "epoch": 0.9913501045195704,
+      "grad_norm": 0.12823523846215548,
+      "learning_rate": 3.924387441087696e-08,
+      "loss": 1.2206,
+      "step": 13753
+    },
+    {
+      "epoch": 0.9914221869819073,
+      "grad_norm": 0.12173254749900166,
+      "learning_rate": 3.8592577004459285e-08,
+      "loss": 1.2502,
+      "step": 13754
+    },
+    {
+      "epoch": 0.9914942694442442,
+      "grad_norm": 0.12011912759047677,
+      "learning_rate": 3.7946728389015586e-08,
+      "loss": 1.2734,
+      "step": 13755
+    },
+    {
+      "epoch": 0.9915663519065812,
+      "grad_norm": 0.1252284065164756,
+      "learning_rate": 3.730632859976213e-08,
+      "loss": 1.2912,
+      "step": 13756
+    },
+    {
+      "epoch": 0.9916384343689181,
+      "grad_norm": 0.11879726685672089,
+      "learning_rate": 3.667137767160433e-08,
+      "loss": 1.3188,
+      "step": 13757
+    },
+    {
+      "epoch": 0.991710516831255,
+      "grad_norm": 0.12911436431623488,
+      "learning_rate": 3.604187563915895e-08,
+      "loss": 1.2295,
+      "step": 13758
+    },
+    {
+      "epoch": 0.9917825992935919,
+      "grad_norm": 0.12027421492465429,
+      "learning_rate": 3.541782253673187e-08,
+      "loss": 1.2085,
+      "step": 13759
+    },
+    {
+      "epoch": 0.9918546817559288,
+      "grad_norm": 0.10450989850816798,
+      "learning_rate": 3.479921839834033e-08,
+      "loss": 1.2031,
+      "step": 13760
+    },
+    {
+      "epoch": 0.9919267642182656,
+      "grad_norm": 0.1194275962311657,
+      "learning_rate": 3.4186063257701796e-08,
+      "loss": 1.2421,
+      "step": 13761
+    },
+    {
+      "epoch": 0.9919988466806026,
+      "grad_norm": 0.1262981392046082,
+      "learning_rate": 3.357835714825619e-08,
+      "loss": 1.2821,
+      "step": 13762
+    },
+    {
+      "epoch": 0.9920709291429395,
+      "grad_norm": 0.13267308145146176,
+      "learning_rate": 3.297610010311036e-08,
+      "loss": 1.2651,
+      "step": 13763
+    },
+    {
+      "epoch": 0.9921430116052764,
+      "grad_norm": 0.10562196945541831,
+      "learning_rate": 3.237929215509361e-08,
+      "loss": 1.3128,
+      "step": 13764
+    },
+    {
+      "epoch": 0.9922150940676133,
+      "grad_norm": 0.13457397784390193,
+      "learning_rate": 3.178793333673546e-08,
+      "loss": 1.3141,
+      "step": 13765
+    },
+    {
+      "epoch": 0.9922871765299502,
+      "grad_norm": 0.12156505485864252,
+      "learning_rate": 3.120202368028791e-08,
+      "loss": 1.142,
+      "step": 13766
+    },
+    {
+      "epoch": 0.9923592589922872,
+      "grad_norm": 0.12888323084058045,
+      "learning_rate": 3.062156321766985e-08,
+      "loss": 1.2742,
+      "step": 13767
+    },
+    {
+      "epoch": 0.9924313414546241,
+      "grad_norm": 0.11998531517837734,
+      "learning_rate": 3.004655198053374e-08,
+      "loss": 1.215,
+      "step": 13768
+    },
+    {
+      "epoch": 0.992503423916961,
+      "grad_norm": 0.1308955455419151,
+      "learning_rate": 2.947699000021009e-08,
+      "loss": 1.2657,
+      "step": 13769
+    },
+    {
+      "epoch": 0.9925755063792979,
+      "grad_norm": 0.1162269442323337,
+      "learning_rate": 2.8912877307751828e-08,
+      "loss": 1.3044,
+      "step": 13770
+    },
+    {
+      "epoch": 0.9926475888416348,
+      "grad_norm": 0.1313600422403445,
+      "learning_rate": 2.8354213933912133e-08,
+      "loss": 1.2539,
+      "step": 13771
+    },
+    {
+      "epoch": 0.9927196713039718,
+      "grad_norm": 0.11739416066396263,
+      "learning_rate": 2.780099990914442e-08,
+      "loss": 1.2531,
+      "step": 13772
+    },
+    {
+      "epoch": 0.9927917537663087,
+      "grad_norm": 0.13096811663756386,
+      "learning_rate": 2.7253235263591248e-08,
+      "loss": 1.2948,
+      "step": 13773
+    },
+    {
+      "epoch": 0.9928638362286456,
+      "grad_norm": 0.11957347537117186,
+      "learning_rate": 2.6710920027117615e-08,
+      "loss": 1.3164,
+      "step": 13774
+    },
+    {
+      "epoch": 0.9929359186909825,
+      "grad_norm": 0.12846955790759665,
+      "learning_rate": 2.617405422929986e-08,
+      "loss": 1.3376,
+      "step": 13775
+    },
+    {
+      "epoch": 0.9930080011533194,
+      "grad_norm": 0.11306895853996156,
+      "learning_rate": 2.564263789937016e-08,
+      "loss": 1.3216,
+      "step": 13776
+    },
+    {
+      "epoch": 0.9930800836156564,
+      "grad_norm": 0.1287172029585101,
+      "learning_rate": 2.5116671066316433e-08,
+      "loss": 1.3017,
+      "step": 13777
+    },
+    {
+      "epoch": 0.9931521660779932,
+      "grad_norm": 0.12367890303404212,
+      "learning_rate": 2.4596153758815743e-08,
+      "loss": 1.2497,
+      "step": 13778
+    },
+    {
+      "epoch": 0.9932242485403301,
+      "grad_norm": 0.11951330218499927,
+      "learning_rate": 2.4081086005223186e-08,
+      "loss": 1.2575,
+      "step": 13779
+    },
+    {
+      "epoch": 0.993296331002667,
+      "grad_norm": 0.12264099368745751,
+      "learning_rate": 2.3571467833616302e-08,
+      "loss": 1.2793,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9933684134650039,
+      "grad_norm": 0.13173429964025868,
+      "learning_rate": 2.306729927178397e-08,
+      "loss": 1.2624,
+      "step": 13781
+    },
+    {
+      "epoch": 0.9934404959273408,
+      "grad_norm": 0.11710558305172725,
+      "learning_rate": 2.2568580347204215e-08,
+      "loss": 1.2712,
+      "step": 13782
+    },
+    {
+      "epoch": 0.9935125783896778,
+      "grad_norm": 0.1201225495001595,
+      "learning_rate": 2.2075311087066398e-08,
+      "loss": 1.3409,
+      "step": 13783
+    },
+    {
+      "epoch": 0.9935846608520147,
+      "grad_norm": 0.11688051431574803,
+      "learning_rate": 2.1587491518237913e-08,
+      "loss": 1.3066,
+      "step": 13784
+    },
+    {
+      "epoch": 0.9936567433143516,
+      "grad_norm": 0.11719603602641224,
+      "learning_rate": 2.1105121667330806e-08,
+      "loss": 1.3338,
+      "step": 13785
+    },
+    {
+      "epoch": 0.9937288257766885,
+      "grad_norm": 0.13035632703849886,
+      "learning_rate": 2.0628201560635162e-08,
+      "loss": 1.3302,
+      "step": 13786
+    },
+    {
+      "epoch": 0.9938009082390254,
+      "grad_norm": 0.12055187878304159,
+      "learning_rate": 2.01567312241413e-08,
+      "loss": 1.2528,
+      "step": 13787
+    },
+    {
+      "epoch": 0.9938729907013624,
+      "grad_norm": 0.13698476879791865,
+      "learning_rate": 1.9690710683550884e-08,
+      "loss": 1.261,
+      "step": 13788
+    },
+    {
+      "epoch": 0.9939450731636993,
+      "grad_norm": 0.12021523397641647,
+      "learning_rate": 1.9230139964265813e-08,
+      "loss": 1.2794,
+      "step": 13789
+    },
+    {
+      "epoch": 0.9940171556260362,
+      "grad_norm": 0.12093650084453562,
+      "learning_rate": 1.8775019091388235e-08,
+      "loss": 1.3563,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9940892380883731,
+      "grad_norm": 0.11595192850801167,
+      "learning_rate": 1.8325348089731632e-08,
+      "loss": 1.3088,
+      "step": 13791
+    },
+    {
+      "epoch": 0.99416132055071,
+      "grad_norm": 0.11839808369062962,
+      "learning_rate": 1.7881126983809727e-08,
+      "loss": 1.2865,
+      "step": 13792
+    },
+    {
+      "epoch": 0.994233403013047,
+      "grad_norm": 0.12117840329653222,
+      "learning_rate": 1.744235579782538e-08,
+      "loss": 1.3032,
+      "step": 13793
+    },
+    {
+      "epoch": 0.9943054854753839,
+      "grad_norm": 0.11330913282089847,
+      "learning_rate": 1.7009034555703907e-08,
+      "loss": 1.2168,
+      "step": 13794
+    },
+    {
+      "epoch": 0.9943775679377208,
+      "grad_norm": 0.11705072630968541,
+      "learning_rate": 1.6581163281059743e-08,
+      "loss": 1.2193,
+      "step": 13795
+    },
+    {
+      "epoch": 0.9944496504000576,
+      "grad_norm": 0.12002970904416264,
+      "learning_rate": 1.6158741997229775e-08,
+      "loss": 1.3104,
+      "step": 13796
+    },
+    {
+      "epoch": 0.9945217328623945,
+      "grad_norm": 0.3739921750294925,
+      "learning_rate": 1.574177072721783e-08,
+      "loss": 1.1581,
+      "step": 13797
+    },
+    {
+      "epoch": 0.9945938153247315,
+      "grad_norm": 0.11604827719293022,
+      "learning_rate": 1.5330249493772376e-08,
+      "loss": 1.2493,
+      "step": 13798
+    },
+    {
+      "epoch": 0.9946658977870684,
+      "grad_norm": 0.1235772459929892,
+      "learning_rate": 1.4924178319319916e-08,
+      "loss": 1.1982,
+      "step": 13799
+    },
+    {
+      "epoch": 0.9947379802494053,
+      "grad_norm": 0.118235225778353,
+      "learning_rate": 1.4523557225976093e-08,
+      "loss": 1.1227,
+      "step": 13800
+    },
+    {
+      "epoch": 0.9948100627117422,
+      "grad_norm": 0.1266442901566405,
+      "learning_rate": 1.41283862356123e-08,
+      "loss": 1.2973,
+      "step": 13801
+    },
+    {
+      "epoch": 0.9948821451740791,
+      "grad_norm": 0.12111729605355775,
+      "learning_rate": 1.3738665369733561e-08,
+      "loss": 1.3039,
+      "step": 13802
+    },
+    {
+      "epoch": 0.994954227636416,
+      "grad_norm": 0.12334851364730869,
+      "learning_rate": 1.3354394649611745e-08,
+      "loss": 1.3116,
+      "step": 13803
+    },
+    {
+      "epoch": 0.995026310098753,
+      "grad_norm": 0.12804422378446909,
+      "learning_rate": 1.2975574096185661e-08,
+      "loss": 1.2475,
+      "step": 13804
+    },
+    {
+      "epoch": 0.9950983925610899,
+      "grad_norm": 0.12020216215861608,
+      "learning_rate": 1.260220373008325e-08,
+      "loss": 1.2509,
+      "step": 13805
+    },
+    {
+      "epoch": 0.9951704750234268,
+      "grad_norm": 0.11917545480017076,
+      "learning_rate": 1.223428357168821e-08,
+      "loss": 1.2523,
+      "step": 13806
+    },
+    {
+      "epoch": 0.9952425574857637,
+      "grad_norm": 0.12026706669563722,
+      "learning_rate": 1.1871813641040064e-08,
+      "loss": 1.332,
+      "step": 13807
+    },
+    {
+      "epoch": 0.9953146399481007,
+      "grad_norm": 0.13373188369804323,
+      "learning_rate": 1.151479395788968e-08,
+      "loss": 1.262,
+      "step": 13808
+    },
+    {
+      "epoch": 0.9953867224104376,
+      "grad_norm": 0.11690180676991005,
+      "learning_rate": 1.1163224541710371e-08,
+      "loss": 1.2398,
+      "step": 13809
+    },
+    {
+      "epoch": 0.9954588048727745,
+      "grad_norm": 0.12410060668165826,
+      "learning_rate": 1.0817105411664586e-08,
+      "loss": 1.2278,
+      "step": 13810
+    },
+    {
+      "epoch": 0.9955308873351114,
+      "grad_norm": 0.1262544840405068,
+      "learning_rate": 1.047643658660391e-08,
+      "loss": 1.3096,
+      "step": 13811
+    },
+    {
+      "epoch": 0.9956029697974483,
+      "grad_norm": 0.11380549833976848,
+      "learning_rate": 1.014121808511348e-08,
+      "loss": 1.1648,
+      "step": 13812
+    },
+    {
+      "epoch": 0.9956750522597851,
+      "grad_norm": 0.12253388083910419,
+      "learning_rate": 9.811449925467565e-09,
+      "loss": 1.2228,
+      "step": 13813
+    },
+    {
+      "epoch": 0.9957471347221221,
+      "grad_norm": 0.11239345483891389,
+      "learning_rate": 9.487132125629572e-09,
+      "loss": 1.2653,
+      "step": 13814
+    },
+    {
+      "epoch": 0.995819217184459,
+      "grad_norm": 0.12711637560907105,
+      "learning_rate": 9.168264703285356e-09,
+      "loss": 1.3629,
+      "step": 13815
+    },
+    {
+      "epoch": 0.9958912996467959,
+      "grad_norm": 0.12188297008848677,
+      "learning_rate": 8.85484767582101e-09,
+      "loss": 1.2918,
+      "step": 13816
+    },
+    {
+      "epoch": 0.9959633821091328,
+      "grad_norm": 0.11780186034761347,
+      "learning_rate": 8.546881060311762e-09,
+      "loss": 1.3415,
+      "step": 13817
+    },
+    {
+      "epoch": 0.9960354645714697,
+      "grad_norm": 0.13031039458582958,
+      "learning_rate": 8.244364873533084e-09,
+      "loss": 1.2569,
+      "step": 13818
+    },
+    {
+      "epoch": 0.9961075470338067,
+      "grad_norm": 0.12592667233142188,
+      "learning_rate": 7.947299132005093e-09,
+      "loss": 1.2398,
+      "step": 13819
+    },
+    {
+      "epoch": 0.9961796294961436,
+      "grad_norm": 0.11867120610903975,
+      "learning_rate": 7.655683851892637e-09,
+      "loss": 1.2657,
+      "step": 13820
+    },
+    {
+      "epoch": 0.9962517119584805,
+      "grad_norm": 0.12461847534577845,
+      "learning_rate": 7.369519049105211e-09,
+      "loss": 1.2435,
+      "step": 13821
+    },
+    {
+      "epoch": 0.9963237944208174,
+      "grad_norm": 0.11254262660784706,
+      "learning_rate": 7.08880473924145e-09,
+      "loss": 1.2668,
+      "step": 13822
+    },
+    {
+      "epoch": 0.9963958768831543,
+      "grad_norm": 0.1171293783166849,
+      "learning_rate": 6.813540937600227e-09,
+      "loss": 1.2769,
+      "step": 13823
+    },
+    {
+      "epoch": 0.9964679593454913,
+      "grad_norm": 0.11115337170595274,
+      "learning_rate": 6.543727659180654e-09,
+      "loss": 1.2344,
+      "step": 13824
+    },
+    {
+      "epoch": 0.9965400418078282,
+      "grad_norm": 0.11942860816954164,
+      "learning_rate": 6.279364918704289e-09,
+      "loss": 1.3012,
+      "step": 13825
+    },
+    {
+      "epoch": 0.9966121242701651,
+      "grad_norm": 0.12557217871425003,
+      "learning_rate": 6.020452730559623e-09,
+      "loss": 1.3661,
+      "step": 13826
+    },
+    {
+      "epoch": 0.996684206732502,
+      "grad_norm": 0.12026303439464947,
+      "learning_rate": 5.766991108879793e-09,
+      "loss": 1.153,
+      "step": 13827
+    },
+    {
+      "epoch": 0.9967562891948389,
+      "grad_norm": 0.12934767536623804,
+      "learning_rate": 5.518980067475976e-09,
+      "loss": 1.2295,
+      "step": 13828
+    },
+    {
+      "epoch": 0.9968283716571759,
+      "grad_norm": 0.11998315373344066,
+      "learning_rate": 5.276419619848483e-09,
+      "loss": 1.2263,
+      "step": 13829
+    },
+    {
+      "epoch": 0.9969004541195127,
+      "grad_norm": 0.1298776155942104,
+      "learning_rate": 5.039309779253376e-09,
+      "loss": 1.3006,
+      "step": 13830
+    },
+    {
+      "epoch": 0.9969725365818496,
+      "grad_norm": 0.12218520694442914,
+      "learning_rate": 4.8076505585803455e-09,
+      "loss": 1.3257,
+      "step": 13831
+    },
+    {
+      "epoch": 0.9970446190441865,
+      "grad_norm": 0.11811752898470006,
+      "learning_rate": 4.581441970485934e-09,
+      "loss": 1.3471,
+      "step": 13832
+    },
+    {
+      "epoch": 0.9971167015065234,
+      "grad_norm": 0.11677860626391631,
+      "learning_rate": 4.360684027282513e-09,
+      "loss": 1.2319,
+      "step": 13833
+    },
+    {
+      "epoch": 0.9971887839688603,
+      "grad_norm": 0.10698184899146052,
+      "learning_rate": 4.145376741004902e-09,
+      "loss": 1.2275,
+      "step": 13834
+    },
+    {
+      "epoch": 0.9972608664311973,
+      "grad_norm": 0.11874635192435608,
+      "learning_rate": 3.935520123399261e-09,
+      "loss": 1.2682,
+      "step": 13835
+    },
+    {
+      "epoch": 0.9973329488935342,
+      "grad_norm": 0.11852523904700563,
+      "learning_rate": 3.731114185889783e-09,
+      "loss": 1.2601,
+      "step": 13836
+    },
+    {
+      "epoch": 0.9974050313558711,
+      "grad_norm": 0.13079343684311498,
+      "learning_rate": 3.5321589396342114e-09,
+      "loss": 1.2436,
+      "step": 13837
+    },
+    {
+      "epoch": 0.997477113818208,
+      "grad_norm": 0.12321035557132591,
+      "learning_rate": 3.3386543954683212e-09,
+      "loss": 1.2264,
+      "step": 13838
+    },
+    {
+      "epoch": 0.997549196280545,
+      "grad_norm": 0.12155430141100419,
+      "learning_rate": 3.150600563939232e-09,
+      "loss": 1.3645,
+      "step": 13839
+    },
+    {
+      "epoch": 0.9976212787428819,
+      "grad_norm": 0.123805454784621,
+      "learning_rate": 2.9679974553054046e-09,
+      "loss": 1.2663,
+      "step": 13840
+    },
+    {
+      "epoch": 0.9976933612052188,
+      "grad_norm": 0.2374788217293003,
+      "learning_rate": 2.790845079503335e-09,
+      "loss": 1.1722,
+      "step": 13841
+    },
+    {
+      "epoch": 0.9977654436675557,
+      "grad_norm": 0.1225079334978164,
+      "learning_rate": 2.619143446214167e-09,
+      "loss": 1.2027,
+      "step": 13842
+    },
+    {
+      "epoch": 0.9978375261298926,
+      "grad_norm": 0.1293899722826272,
+      "learning_rate": 2.4528925647748778e-09,
+      "loss": 1.3065,
+      "step": 13843
+    },
+    {
+      "epoch": 0.9979096085922295,
+      "grad_norm": 0.11254427272797723,
+      "learning_rate": 2.292092444255989e-09,
+      "loss": 1.2194,
+      "step": 13844
+    },
+    {
+      "epoch": 0.9979816910545665,
+      "grad_norm": 0.11402022462911394,
+      "learning_rate": 2.1367430934282618e-09,
+      "loss": 1.2721,
+      "step": 13845
+    },
+    {
+      "epoch": 0.9980537735169034,
+      "grad_norm": 0.12073495194475363,
+      "learning_rate": 1.986844520751596e-09,
+      "loss": 1.2369,
+      "step": 13846
+    },
+    {
+      "epoch": 0.9981258559792403,
+      "grad_norm": 0.15147985694000696,
+      "learning_rate": 1.842396734408336e-09,
+      "loss": 1.2231,
+      "step": 13847
+    },
+    {
+      "epoch": 0.9981979384415771,
+      "grad_norm": 0.12056319966624258,
+      "learning_rate": 1.7033997422477577e-09,
+      "loss": 1.2197,
+      "step": 13848
+    },
+    {
+      "epoch": 0.998270020903914,
+      "grad_norm": 0.12722116626269336,
+      "learning_rate": 1.5698535518748892e-09,
+      "loss": 1.2474,
+      "step": 13849
+    },
+    {
+      "epoch": 0.998342103366251,
+      "grad_norm": 0.12168475344953539,
+      "learning_rate": 1.441758170550589e-09,
+      "loss": 1.2126,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9984141858285879,
+      "grad_norm": 0.12438285322704931,
+      "learning_rate": 1.319113605269262e-09,
+      "loss": 1.3306,
+      "step": 13851
+    },
+    {
+      "epoch": 0.9984862682909248,
+      "grad_norm": 0.12217103579812648,
+      "learning_rate": 1.201919862703349e-09,
+      "loss": 1.2386,
+      "step": 13852
+    },
+    {
+      "epoch": 0.9985583507532617,
+      "grad_norm": 0.11666811876106041,
+      "learning_rate": 1.0901769492477344e-09,
+      "loss": 1.2281,
+      "step": 13853
+    },
+    {
+      "epoch": 0.9986304332155986,
+      "grad_norm": 0.12189444479424057,
+      "learning_rate": 9.838848709975423e-10,
+      "loss": 1.1465,
+      "step": 13854
+    },
+    {
+      "epoch": 0.9987025156779356,
+      "grad_norm": 0.1194505642996235,
+      "learning_rate": 8.830436337370351e-10,
+      "loss": 1.2753,
+      "step": 13855
+    },
+    {
+      "epoch": 0.9987745981402725,
+      "grad_norm": 0.12601525118205578,
+      "learning_rate": 7.876532429729189e-10,
+      "loss": 1.2916,
+      "step": 13856
+    },
+    {
+      "epoch": 0.9988466806026094,
+      "grad_norm": 0.12123491233399508,
+      "learning_rate": 6.977137039010373e-10,
+      "loss": 1.3048,
+      "step": 13857
+    },
+    {
+      "epoch": 0.9989187630649463,
+      "grad_norm": 0.11592895408223956,
+      "learning_rate": 6.13225021417474e-10,
+      "loss": 1.1967,
+      "step": 13858
+    },
+    {
+      "epoch": 0.9989908455272832,
+      "grad_norm": 0.11389288394496595,
+      "learning_rate": 5.341872001407566e-10,
+      "loss": 1.2029,
+      "step": 13859
+    },
+    {
+      "epoch": 0.9990629279896202,
+      "grad_norm": 0.1270675290029674,
+      "learning_rate": 4.606002443674484e-10,
+      "loss": 1.2292,
+      "step": 13860
+    },
+    {
+      "epoch": 0.9991350104519571,
+      "grad_norm": 0.12537403228942892,
+      "learning_rate": 3.9246415810545446e-10,
+      "loss": 1.3156,
+      "step": 13861
+    },
+    {
+      "epoch": 0.999207092914294,
+      "grad_norm": 0.12228108935360366,
+      "learning_rate": 3.297789450851241e-10,
+      "loss": 1.288,
+      "step": 13862
+    },
+    {
+      "epoch": 0.9992791753766309,
+      "grad_norm": 0.1239400805295967,
+      "learning_rate": 2.72544608714842e-10,
+      "loss": 1.2097,
+      "step": 13863
+    },
+    {
+      "epoch": 0.9993512578389678,
+      "grad_norm": 0.1188071162039046,
+      "learning_rate": 2.207611521143349e-10,
+      "loss": 1.2058,
+      "step": 13864
+    },
+    {
+      "epoch": 0.9994233403013046,
+      "grad_norm": 0.12239142914284858,
+      "learning_rate": 1.7442857810356927e-10,
+      "loss": 1.2147,
+      "step": 13865
+    },
+    {
+      "epoch": 0.9994954227636416,
+      "grad_norm": 0.11819832868421505,
+      "learning_rate": 1.3354688921385362e-10,
+      "loss": 1.3319,
+      "step": 13866
+    },
+    {
+      "epoch": 0.9995675052259785,
+      "grad_norm": 0.13911440495526092,
+      "learning_rate": 9.811608766563396e-11,
+      "loss": 1.1772,
+      "step": 13867
+    },
+    {
+      "epoch": 0.9996395876883154,
+      "grad_norm": 0.12249050866505355,
+      "learning_rate": 6.813617539069839e-11,
+      "loss": 1.1917,
+      "step": 13868
+    },
+    {
+      "epoch": 0.9997116701506523,
+      "grad_norm": 0.13012640784209062,
+      "learning_rate": 4.360715403217697e-11,
+      "loss": 1.3099,
+      "step": 13869
+    },
+    {
+      "epoch": 0.9997837526129892,
+      "grad_norm": 0.11820222545909463,
+      "learning_rate": 2.4529024922337328e-11,
+      "loss": 1.3222,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9998558350753262,
+      "grad_norm": 0.1523890328357535,
+      "learning_rate": 1.0901789104789117e-11,
+      "loss": 1.2473,
+      "step": 13871
+    },
+    {
+      "epoch": 0.9999279175376631,
+      "grad_norm": 0.12085675420851406,
+      "learning_rate": 2.7254473122795276e-12,
+      "loss": 1.1769,
+      "step": 13872
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.12216070208832214,
+      "learning_rate": 0.0,
+      "loss": 1.3194,
+      "step": 13873
+    },
+    {
+      "epoch": 1.0,
+      "step": 13873,
+      "total_flos": 4467075949903872.0,
+      "train_loss": 1.304541949831244,
+      "train_runtime": 80228.9516,
+      "train_samples_per_second": 22.132,
+      "train_steps_per_second": 0.173
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 13873,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4467075949903872.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}