diff --git "a/checkpoint-2212/trainer_state.json" "b/checkpoint-2212/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2212/trainer_state.json"
@@ -0,0 +1,15525 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.10040351293615,
+  "eval_steps": 500,
+  "global_step": 2212,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009494422027059103,
+      "grad_norm": 5.364119052886963,
+      "learning_rate": 6.329113924050633e-07,
+      "loss": 2.9616,
+      "step": 1
+    },
+    {
+      "epoch": 0.0018988844054118206,
+      "grad_norm": 5.426063060760498,
+      "learning_rate": 1.2658227848101265e-06,
+      "loss": 2.9664,
+      "step": 2
+    },
+    {
+      "epoch": 0.0028483266081177306,
+      "grad_norm": 5.507386684417725,
+      "learning_rate": 1.8987341772151901e-06,
+      "loss": 2.96,
+      "step": 3
+    },
+    {
+      "epoch": 0.0037977688108236413,
+      "grad_norm": 5.467552185058594,
+      "learning_rate": 2.531645569620253e-06,
+      "loss": 2.975,
+      "step": 4
+    },
+    {
+      "epoch": 0.004747211013529551,
+      "grad_norm": 5.386384963989258,
+      "learning_rate": 3.1645569620253167e-06,
+      "loss": 2.9473,
+      "step": 5
+    },
+    {
+      "epoch": 0.005696653216235461,
+      "grad_norm": 4.9166951179504395,
+      "learning_rate": 3.7974683544303802e-06,
+      "loss": 2.8597,
+      "step": 6
+    },
+    {
+      "epoch": 0.006646095418941372,
+      "grad_norm": 5.469020843505859,
+      "learning_rate": 4.430379746835443e-06,
+      "loss": 2.9633,
+      "step": 7
+    },
+    {
+      "epoch": 0.0075955376216472826,
+      "grad_norm": 5.380453586578369,
+      "learning_rate": 5.063291139240506e-06,
+      "loss": 2.8931,
+      "step": 8
+    },
+    {
+      "epoch": 0.008544979824353193,
+      "grad_norm": 4.922253131866455,
+      "learning_rate": 5.69620253164557e-06,
+      "loss": 2.7452,
+      "step": 9
+    },
+    {
+      "epoch": 0.009494422027059102,
+      "grad_norm": 5.517508029937744,
+      "learning_rate": 6.329113924050633e-06,
+      "loss": 2.8326,
+      "step": 10
+    },
+    {
+      "epoch": 0.010443864229765013,
+      "grad_norm": 5.810976982116699,
+      "learning_rate": 6.9620253164556965e-06,
+      "loss": 2.7854,
+      "step": 11
+    },
+    {
+      "epoch": 0.011393306432470923,
+      "grad_norm": 5.690661430358887,
+      "learning_rate": 7.5949367088607605e-06,
+      "loss": 2.7069,
+      "step": 12
+    },
+    {
+      "epoch": 0.012342748635176834,
+      "grad_norm": 5.994122505187988,
+      "learning_rate": 8.227848101265822e-06,
+      "loss": 2.5705,
+      "step": 13
+    },
+    {
+      "epoch": 0.013292190837882745,
+      "grad_norm": 5.86803674697876,
+      "learning_rate": 8.860759493670886e-06,
+      "loss": 2.4461,
+      "step": 14
+    },
+    {
+      "epoch": 0.014241633040588654,
+      "grad_norm": 5.448781490325928,
+      "learning_rate": 9.49367088607595e-06,
+      "loss": 2.2408,
+      "step": 15
+    },
+    {
+      "epoch": 0.015191075243294565,
+      "grad_norm": 6.4004902839660645,
+      "learning_rate": 1.0126582278481012e-05,
+      "loss": 2.1205,
+      "step": 16
+    },
+    {
+      "epoch": 0.016140517446000476,
+      "grad_norm": 6.970590591430664,
+      "learning_rate": 1.0759493670886076e-05,
+      "loss": 1.9474,
+      "step": 17
+    },
+    {
+      "epoch": 0.017089959648706386,
+      "grad_norm": 7.423785209655762,
+      "learning_rate": 1.139240506329114e-05,
+      "loss": 1.7348,
+      "step": 18
+    },
+    {
+      "epoch": 0.018039401851412295,
+      "grad_norm": 7.429481029510498,
+      "learning_rate": 1.2025316455696203e-05,
+      "loss": 1.4835,
+      "step": 19
+    },
+    {
+      "epoch": 0.018988844054118204,
+      "grad_norm": 6.7193284034729,
+      "learning_rate": 1.2658227848101267e-05,
+      "loss": 1.2373,
+      "step": 20
+    },
+    {
+      "epoch": 0.019938286256824117,
+      "grad_norm": 4.46099853515625,
+      "learning_rate": 1.3291139240506329e-05,
+      "loss": 1.1095,
+      "step": 21
+    },
+    {
+      "epoch": 0.020887728459530026,
+      "grad_norm": 3.001573085784912,
+      "learning_rate": 1.3924050632911393e-05,
+      "loss": 0.8642,
+      "step": 22
+    },
+    {
+      "epoch": 0.021837170662235936,
+      "grad_norm": 2.197000026702881,
+      "learning_rate": 1.4556962025316457e-05,
+      "loss": 0.7734,
+      "step": 23
+    },
+    {
+      "epoch": 0.022786612864941845,
+      "grad_norm": 1.8113943338394165,
+      "learning_rate": 1.5189873417721521e-05,
+      "loss": 0.7341,
+      "step": 24
+    },
+    {
+      "epoch": 0.023736055067647758,
+      "grad_norm": 1.7461305856704712,
+      "learning_rate": 1.5822784810126583e-05,
+      "loss": 0.6743,
+      "step": 25
+    },
+    {
+      "epoch": 0.024685497270353667,
+      "grad_norm": 1.3315849304199219,
+      "learning_rate": 1.6455696202531644e-05,
+      "loss": 0.5975,
+      "step": 26
+    },
+    {
+      "epoch": 0.025634939473059577,
+      "grad_norm": 0.726314127445221,
+      "learning_rate": 1.7088607594936708e-05,
+      "loss": 0.5659,
+      "step": 27
+    },
+    {
+      "epoch": 0.02658438167576549,
+      "grad_norm": 0.6269010901451111,
+      "learning_rate": 1.7721518987341772e-05,
+      "loss": 0.5226,
+      "step": 28
+    },
+    {
+      "epoch": 0.0275338238784714,
+      "grad_norm": 0.5819966197013855,
+      "learning_rate": 1.8354430379746836e-05,
+      "loss": 0.6132,
+      "step": 29
+    },
+    {
+      "epoch": 0.028483266081177308,
+      "grad_norm": 0.6247850060462952,
+      "learning_rate": 1.89873417721519e-05,
+      "loss": 0.5,
+      "step": 30
+    },
+    {
+      "epoch": 0.029432708283883217,
+      "grad_norm": 0.702621579170227,
+      "learning_rate": 1.962025316455696e-05,
+      "loss": 0.4958,
+      "step": 31
+    },
+    {
+      "epoch": 0.03038215048658913,
+      "grad_norm": 0.6045309901237488,
+      "learning_rate": 2.0253164556962025e-05,
+      "loss": 0.4405,
+      "step": 32
+    },
+    {
+      "epoch": 0.031331592689295036,
+      "grad_norm": 0.5436626076698303,
+      "learning_rate": 2.088607594936709e-05,
+      "loss": 0.5607,
+      "step": 33
+    },
+    {
+      "epoch": 0.03228103489200095,
+      "grad_norm": 0.43146297335624695,
+      "learning_rate": 2.1518987341772153e-05,
+      "loss": 0.3987,
+      "step": 34
+    },
+    {
+      "epoch": 0.03323047709470686,
+      "grad_norm": 0.5124548673629761,
+      "learning_rate": 2.2151898734177217e-05,
+      "loss": 0.5084,
+      "step": 35
+    },
+    {
+      "epoch": 0.03417991929741277,
+      "grad_norm": 0.4466649293899536,
+      "learning_rate": 2.278481012658228e-05,
+      "loss": 0.3761,
+      "step": 36
+    },
+    {
+      "epoch": 0.03512936150011868,
+      "grad_norm": 0.41221529245376587,
+      "learning_rate": 2.341772151898734e-05,
+      "loss": 0.3859,
+      "step": 37
+    },
+    {
+      "epoch": 0.03607880370282459,
+      "grad_norm": 0.3802257180213928,
+      "learning_rate": 2.4050632911392405e-05,
+      "loss": 0.3447,
+      "step": 38
+    },
+    {
+      "epoch": 0.0370282459055305,
+      "grad_norm": 0.47727710008621216,
+      "learning_rate": 2.468354430379747e-05,
+      "loss": 0.3914,
+      "step": 39
+    },
+    {
+      "epoch": 0.03797768810823641,
+      "grad_norm": 0.41048529744148254,
+      "learning_rate": 2.5316455696202533e-05,
+      "loss": 0.2988,
+      "step": 40
+    },
+    {
+      "epoch": 0.038927130310942325,
+      "grad_norm": 0.5019667744636536,
+      "learning_rate": 2.5949367088607597e-05,
+      "loss": 0.2938,
+      "step": 41
+    },
+    {
+      "epoch": 0.039876572513648234,
+      "grad_norm": 0.42121732234954834,
+      "learning_rate": 2.6582278481012658e-05,
+      "loss": 0.2579,
+      "step": 42
+    },
+    {
+      "epoch": 0.04082601471635414,
+      "grad_norm": 0.4193897247314453,
+      "learning_rate": 2.7215189873417722e-05,
+      "loss": 0.3262,
+      "step": 43
+    },
+    {
+      "epoch": 0.04177545691906005,
+      "grad_norm": 0.2978931665420532,
+      "learning_rate": 2.7848101265822786e-05,
+      "loss": 0.2365,
+      "step": 44
+    },
+    {
+      "epoch": 0.04272489912176596,
+      "grad_norm": 0.34771448373794556,
+      "learning_rate": 2.848101265822785e-05,
+      "loss": 0.2364,
+      "step": 45
+    },
+    {
+      "epoch": 0.04367434132447187,
+      "grad_norm": 0.3881576955318451,
+      "learning_rate": 2.9113924050632914e-05,
+      "loss": 0.286,
+      "step": 46
+    },
+    {
+      "epoch": 0.04462378352717778,
+      "grad_norm": 0.33863797783851624,
+      "learning_rate": 2.9746835443037974e-05,
+      "loss": 0.2739,
+      "step": 47
+    },
+    {
+      "epoch": 0.04557322572988369,
+      "grad_norm": 0.2894616723060608,
+      "learning_rate": 3.0379746835443042e-05,
+      "loss": 0.2587,
+      "step": 48
+    },
+    {
+      "epoch": 0.046522667932589606,
+      "grad_norm": 0.22292694449424744,
+      "learning_rate": 3.10126582278481e-05,
+      "loss": 0.1861,
+      "step": 49
+    },
+    {
+      "epoch": 0.047472110135295516,
+      "grad_norm": 0.21907460689544678,
+      "learning_rate": 3.1645569620253167e-05,
+      "loss": 0.1755,
+      "step": 50
+    },
+    {
+      "epoch": 0.048421552338001425,
+      "grad_norm": 0.29593944549560547,
+      "learning_rate": 3.227848101265823e-05,
+      "loss": 0.1856,
+      "step": 51
+    },
+    {
+      "epoch": 0.049370994540707334,
+      "grad_norm": 0.23055657744407654,
+      "learning_rate": 3.291139240506329e-05,
+      "loss": 0.2102,
+      "step": 52
+    },
+    {
+      "epoch": 0.050320436743413244,
+      "grad_norm": 0.18929323554039001,
+      "learning_rate": 3.354430379746836e-05,
+      "loss": 0.1909,
+      "step": 53
+    },
+    {
+      "epoch": 0.05126987894611915,
+      "grad_norm": 0.15004883706569672,
+      "learning_rate": 3.4177215189873416e-05,
+      "loss": 0.1619,
+      "step": 54
+    },
+    {
+      "epoch": 0.05221932114882506,
+      "grad_norm": 0.15621644258499146,
+      "learning_rate": 3.4810126582278487e-05,
+      "loss": 0.1759,
+      "step": 55
+    },
+    {
+      "epoch": 0.05316876335153098,
+      "grad_norm": 0.16266578435897827,
+      "learning_rate": 3.5443037974683544e-05,
+      "loss": 0.1657,
+      "step": 56
+    },
+    {
+      "epoch": 0.05411820555423689,
+      "grad_norm": 0.14417718350887299,
+      "learning_rate": 3.607594936708861e-05,
+      "loss": 0.1698,
+      "step": 57
+    },
+    {
+      "epoch": 0.0550676477569428,
+      "grad_norm": 0.21402889490127563,
+      "learning_rate": 3.670886075949367e-05,
+      "loss": 0.2185,
+      "step": 58
+    },
+    {
+      "epoch": 0.05601708995964871,
+      "grad_norm": 0.1997889280319214,
+      "learning_rate": 3.7341772151898736e-05,
+      "loss": 0.2143,
+      "step": 59
+    },
+    {
+      "epoch": 0.056966532162354616,
+      "grad_norm": 0.13755086064338684,
+      "learning_rate": 3.79746835443038e-05,
+      "loss": 0.1677,
+      "step": 60
+    },
+    {
+      "epoch": 0.057915974365060525,
+      "grad_norm": 0.19304363429546356,
+      "learning_rate": 3.8607594936708864e-05,
+      "loss": 0.2113,
+      "step": 61
+    },
+    {
+      "epoch": 0.058865416567766435,
+      "grad_norm": 0.14066031575202942,
+      "learning_rate": 3.924050632911392e-05,
+      "loss": 0.1612,
+      "step": 62
+    },
+    {
+      "epoch": 0.059814858770472344,
+      "grad_norm": 0.13375213742256165,
+      "learning_rate": 3.987341772151899e-05,
+      "loss": 0.164,
+      "step": 63
+    },
+    {
+      "epoch": 0.06076430097317826,
+      "grad_norm": 0.15216922760009766,
+      "learning_rate": 4.050632911392405e-05,
+      "loss": 0.16,
+      "step": 64
+    },
+    {
+      "epoch": 0.06171374317588417,
+      "grad_norm": 0.16130389273166656,
+      "learning_rate": 4.113924050632912e-05,
+      "loss": 0.1957,
+      "step": 65
+    },
+    {
+      "epoch": 0.06266318537859007,
+      "grad_norm": 0.1791229248046875,
+      "learning_rate": 4.177215189873418e-05,
+      "loss": 0.1993,
+      "step": 66
+    },
+    {
+      "epoch": 0.06361262758129599,
+      "grad_norm": 0.11038907617330551,
+      "learning_rate": 4.240506329113924e-05,
+      "loss": 0.1517,
+      "step": 67
+    },
+    {
+      "epoch": 0.0645620697840019,
+      "grad_norm": 0.13327902555465698,
+      "learning_rate": 4.3037974683544305e-05,
+      "loss": 0.1501,
+      "step": 68
+    },
+    {
+      "epoch": 0.06551151198670781,
+      "grad_norm": 0.13731731474399567,
+      "learning_rate": 4.367088607594937e-05,
+      "loss": 0.1596,
+      "step": 69
+    },
+    {
+      "epoch": 0.06646095418941372,
+      "grad_norm": 0.13924308121204376,
+      "learning_rate": 4.430379746835443e-05,
+      "loss": 0.152,
+      "step": 70
+    },
+    {
+      "epoch": 0.06741039639211963,
+      "grad_norm": 0.1482289433479309,
+      "learning_rate": 4.49367088607595e-05,
+      "loss": 0.1536,
+      "step": 71
+    },
+    {
+      "epoch": 0.06835983859482554,
+      "grad_norm": 0.10759364813566208,
+      "learning_rate": 4.556962025316456e-05,
+      "loss": 0.1543,
+      "step": 72
+    },
+    {
+      "epoch": 0.06930928079753144,
+      "grad_norm": 0.12899678945541382,
+      "learning_rate": 4.6202531645569625e-05,
+      "loss": 0.165,
+      "step": 73
+    },
+    {
+      "epoch": 0.07025872300023736,
+      "grad_norm": 0.11689919233322144,
+      "learning_rate": 4.683544303797468e-05,
+      "loss": 0.1564,
+      "step": 74
+    },
+    {
+      "epoch": 0.07120816520294328,
+      "grad_norm": 0.12697139382362366,
+      "learning_rate": 4.7468354430379746e-05,
+      "loss": 0.162,
+      "step": 75
+    },
+    {
+      "epoch": 0.07215760740564918,
+      "grad_norm": 0.12069376558065414,
+      "learning_rate": 4.810126582278481e-05,
+      "loss": 0.1467,
+      "step": 76
+    },
+    {
+      "epoch": 0.0731070496083551,
+      "grad_norm": 0.10199815034866333,
+      "learning_rate": 4.8734177215189874e-05,
+      "loss": 0.1528,
+      "step": 77
+    },
+    {
+      "epoch": 0.074056491811061,
+      "grad_norm": 0.1142750009894371,
+      "learning_rate": 4.936708860759494e-05,
+      "loss": 0.1574,
+      "step": 78
+    },
+    {
+      "epoch": 0.07500593401376691,
+      "grad_norm": 0.11019093543291092,
+      "learning_rate": 5e-05,
+      "loss": 0.1512,
+      "step": 79
+    },
+    {
+      "epoch": 0.07595537621647282,
+      "grad_norm": 0.09426973015069962,
+      "learning_rate": 5.0632911392405066e-05,
+      "loss": 0.1481,
+      "step": 80
+    },
+    {
+      "epoch": 0.07690481841917873,
+      "grad_norm": 0.09757663309574127,
+      "learning_rate": 5.1265822784810124e-05,
+      "loss": 0.1484,
+      "step": 81
+    },
+    {
+      "epoch": 0.07785426062188465,
+      "grad_norm": 0.10646392405033112,
+      "learning_rate": 5.1898734177215194e-05,
+      "loss": 0.1549,
+      "step": 82
+    },
+    {
+      "epoch": 0.07880370282459055,
+      "grad_norm": 0.12109784036874771,
+      "learning_rate": 5.253164556962026e-05,
+      "loss": 0.1448,
+      "step": 83
+    },
+    {
+      "epoch": 0.07975314502729647,
+      "grad_norm": 0.12039211392402649,
+      "learning_rate": 5.3164556962025316e-05,
+      "loss": 0.1538,
+      "step": 84
+    },
+    {
+      "epoch": 0.08070258723000237,
+      "grad_norm": 0.16873961687088013,
+      "learning_rate": 5.379746835443038e-05,
+      "loss": 0.1971,
+      "step": 85
+    },
+    {
+      "epoch": 0.08165202943270829,
+      "grad_norm": 0.12140022218227386,
+      "learning_rate": 5.4430379746835444e-05,
+      "loss": 0.1497,
+      "step": 86
+    },
+    {
+      "epoch": 0.08260147163541419,
+      "grad_norm": 0.14637599885463715,
+      "learning_rate": 5.5063291139240514e-05,
+      "loss": 0.1958,
+      "step": 87
+    },
+    {
+      "epoch": 0.0835509138381201,
+      "grad_norm": 0.1141396313905716,
+      "learning_rate": 5.569620253164557e-05,
+      "loss": 0.1457,
+      "step": 88
+    },
+    {
+      "epoch": 0.08450035604082601,
+      "grad_norm": 0.2128390371799469,
+      "learning_rate": 5.6329113924050636e-05,
+      "loss": 0.2339,
+      "step": 89
+    },
+    {
+      "epoch": 0.08544979824353192,
+      "grad_norm": 0.18838858604431152,
+      "learning_rate": 5.69620253164557e-05,
+      "loss": 0.2029,
+      "step": 90
+    },
+    {
+      "epoch": 0.08639924044623784,
+      "grad_norm": 0.19592566788196564,
+      "learning_rate": 5.759493670886076e-05,
+      "loss": 0.2276,
+      "step": 91
+    },
+    {
+      "epoch": 0.08734868264894374,
+      "grad_norm": 0.14753012359142303,
+      "learning_rate": 5.822784810126583e-05,
+      "loss": 0.1916,
+      "step": 92
+    },
+    {
+      "epoch": 0.08829812485164966,
+      "grad_norm": 0.1494351178407669,
+      "learning_rate": 5.886075949367089e-05,
+      "loss": 0.1913,
+      "step": 93
+    },
+    {
+      "epoch": 0.08924756705435556,
+      "grad_norm": 0.1173478439450264,
+      "learning_rate": 5.949367088607595e-05,
+      "loss": 0.1438,
+      "step": 94
+    },
+    {
+      "epoch": 0.09019700925706148,
+      "grad_norm": 0.12023188918828964,
+      "learning_rate": 6.012658227848101e-05,
+      "loss": 0.1516,
+      "step": 95
+    },
+    {
+      "epoch": 0.09114645145976738,
+      "grad_norm": 0.1275833547115326,
+      "learning_rate": 6.0759493670886084e-05,
+      "loss": 0.1492,
+      "step": 96
+    },
+    {
+      "epoch": 0.0920958936624733,
+      "grad_norm": 0.1360282599925995,
+      "learning_rate": 6.139240506329115e-05,
+      "loss": 0.1507,
+      "step": 97
+    },
+    {
+      "epoch": 0.09304533586517921,
+      "grad_norm": 0.1586841195821762,
+      "learning_rate": 6.20253164556962e-05,
+      "loss": 0.1956,
+      "step": 98
+    },
+    {
+      "epoch": 0.09399477806788512,
+      "grad_norm": 0.14281995594501495,
+      "learning_rate": 6.265822784810128e-05,
+      "loss": 0.1774,
+      "step": 99
+    },
+    {
+      "epoch": 0.09494422027059103,
+      "grad_norm": 0.12553077936172485,
+      "learning_rate": 6.329113924050633e-05,
+      "loss": 0.148,
+      "step": 100
+    },
+    {
+      "epoch": 0.09589366247329693,
+      "grad_norm": 0.1117570698261261,
+      "learning_rate": 6.392405063291139e-05,
+      "loss": 0.16,
+      "step": 101
+    },
+    {
+      "epoch": 0.09684310467600285,
+      "grad_norm": 0.13955281674861908,
+      "learning_rate": 6.455696202531646e-05,
+      "loss": 0.1464,
+      "step": 102
+    },
+    {
+      "epoch": 0.09779254687870875,
+      "grad_norm": 0.10990285873413086,
+      "learning_rate": 6.518987341772153e-05,
+      "loss": 0.147,
+      "step": 103
+    },
+    {
+      "epoch": 0.09874198908141467,
+      "grad_norm": 0.10545991361141205,
+      "learning_rate": 6.582278481012658e-05,
+      "loss": 0.1436,
+      "step": 104
+    },
+    {
+      "epoch": 0.09969143128412059,
+      "grad_norm": 0.1717437207698822,
+      "learning_rate": 6.645569620253165e-05,
+      "loss": 0.2278,
+      "step": 105
+    },
+    {
+      "epoch": 0.10064087348682649,
+      "grad_norm": 0.10950994491577148,
+      "learning_rate": 6.708860759493672e-05,
+      "loss": 0.1493,
+      "step": 106
+    },
+    {
+      "epoch": 0.1015903156895324,
+      "grad_norm": 0.11200258880853653,
+      "learning_rate": 6.772151898734177e-05,
+      "loss": 0.1536,
+      "step": 107
+    },
+    {
+      "epoch": 0.1025397578922383,
+      "grad_norm": 0.10955105721950531,
+      "learning_rate": 6.835443037974683e-05,
+      "loss": 0.1483,
+      "step": 108
+    },
+    {
+      "epoch": 0.10348920009494422,
+      "grad_norm": 0.11920775473117828,
+      "learning_rate": 6.89873417721519e-05,
+      "loss": 0.1492,
+      "step": 109
+    },
+    {
+      "epoch": 0.10443864229765012,
+      "grad_norm": 0.1390092819929123,
+      "learning_rate": 6.962025316455697e-05,
+      "loss": 0.1849,
+      "step": 110
+    },
+    {
+      "epoch": 0.10538808450035604,
+      "grad_norm": 0.1363140493631363,
+      "learning_rate": 7.025316455696203e-05,
+      "loss": 0.1849,
+      "step": 111
+    },
+    {
+      "epoch": 0.10633752670306196,
+      "grad_norm": 0.09190025180578232,
+      "learning_rate": 7.088607594936709e-05,
+      "loss": 0.1587,
+      "step": 112
+    },
+    {
+      "epoch": 0.10728696890576786,
+      "grad_norm": 0.09020426124334335,
+      "learning_rate": 7.151898734177216e-05,
+      "loss": 0.1377,
+      "step": 113
+    },
+    {
+      "epoch": 0.10823641110847378,
+      "grad_norm": 0.10544883459806442,
+      "learning_rate": 7.215189873417722e-05,
+      "loss": 0.1516,
+      "step": 114
+    },
+    {
+      "epoch": 0.10918585331117968,
+      "grad_norm": 0.12401281297206879,
+      "learning_rate": 7.278481012658229e-05,
+      "loss": 0.154,
+      "step": 115
+    },
+    {
+      "epoch": 0.1101352955138856,
+      "grad_norm": 0.1008707657456398,
+      "learning_rate": 7.341772151898734e-05,
+      "loss": 0.1448,
+      "step": 116
+    },
+    {
+      "epoch": 0.1110847377165915,
+      "grad_norm": 0.10302747040987015,
+      "learning_rate": 7.40506329113924e-05,
+      "loss": 0.1451,
+      "step": 117
+    },
+    {
+      "epoch": 0.11203417991929741,
+      "grad_norm": 0.12748293578624725,
+      "learning_rate": 7.468354430379747e-05,
+      "loss": 0.1829,
+      "step": 118
+    },
+    {
+      "epoch": 0.11298362212200333,
+      "grad_norm": 0.10413361340761185,
+      "learning_rate": 7.531645569620254e-05,
+      "loss": 0.1371,
+      "step": 119
+    },
+    {
+      "epoch": 0.11393306432470923,
+      "grad_norm": 0.1243433803319931,
+      "learning_rate": 7.59493670886076e-05,
+      "loss": 0.1409,
+      "step": 120
+    },
+    {
+      "epoch": 0.11488250652741515,
+      "grad_norm": 0.11630933731794357,
+      "learning_rate": 7.658227848101266e-05,
+      "loss": 0.1372,
+      "step": 121
+    },
+    {
+      "epoch": 0.11583194873012105,
+      "grad_norm": 0.17981529235839844,
+      "learning_rate": 7.721518987341773e-05,
+      "loss": 0.2257,
+      "step": 122
+    },
+    {
+      "epoch": 0.11678139093282697,
+      "grad_norm": 0.14063452184200287,
+      "learning_rate": 7.78481012658228e-05,
+      "loss": 0.1841,
+      "step": 123
+    },
+    {
+      "epoch": 0.11773083313553287,
+      "grad_norm": 0.1264188438653946,
+      "learning_rate": 7.848101265822784e-05,
+      "loss": 0.1471,
+      "step": 124
+    },
+    {
+      "epoch": 0.11868027533823879,
+      "grad_norm": 0.12827955186367035,
+      "learning_rate": 7.911392405063291e-05,
+      "loss": 0.1493,
+      "step": 125
+    },
+    {
+      "epoch": 0.11962971754094469,
+      "grad_norm": 0.09800329059362411,
+      "learning_rate": 7.974683544303798e-05,
+      "loss": 0.1414,
+      "step": 126
+    },
+    {
+      "epoch": 0.1205791597436506,
+      "grad_norm": 0.09902197122573853,
+      "learning_rate": 8.037974683544304e-05,
+      "loss": 0.1462,
+      "step": 127
+    },
+    {
+      "epoch": 0.12152860194635652,
+      "grad_norm": 0.09450504928827286,
+      "learning_rate": 8.10126582278481e-05,
+      "loss": 0.1484,
+      "step": 128
+    },
+    {
+      "epoch": 0.12247804414906242,
+      "grad_norm": 0.11012883484363556,
+      "learning_rate": 8.164556962025317e-05,
+      "loss": 0.1437,
+      "step": 129
+    },
+    {
+      "epoch": 0.12342748635176834,
+      "grad_norm": 0.11717642843723297,
+      "learning_rate": 8.227848101265824e-05,
+      "loss": 0.1478,
+      "step": 130
+    },
+    {
+      "epoch": 0.12437692855447424,
+      "grad_norm": 0.08754123747348785,
+      "learning_rate": 8.29113924050633e-05,
+      "loss": 0.1408,
+      "step": 131
+    },
+    {
+      "epoch": 0.12532637075718014,
+      "grad_norm": 0.10017862170934677,
+      "learning_rate": 8.354430379746835e-05,
+      "loss": 0.1476,
+      "step": 132
+    },
+    {
+      "epoch": 0.12627581295988607,
+      "grad_norm": 0.08994068205356598,
+      "learning_rate": 8.417721518987342e-05,
+      "loss": 0.1478,
+      "step": 133
+    },
+    {
+      "epoch": 0.12722525516259198,
+      "grad_norm": 0.09894968569278717,
+      "learning_rate": 8.481012658227848e-05,
+      "loss": 0.1309,
+      "step": 134
+    },
+    {
+      "epoch": 0.12817469736529788,
+      "grad_norm": 0.10028701275587082,
+      "learning_rate": 8.544303797468355e-05,
+      "loss": 0.1433,
+      "step": 135
+    },
+    {
+      "epoch": 0.1291241395680038,
+      "grad_norm": 0.0897536426782608,
+      "learning_rate": 8.607594936708861e-05,
+      "loss": 0.1459,
+      "step": 136
+    },
+    {
+      "epoch": 0.1300735817707097,
+      "grad_norm": 0.10435349494218826,
+      "learning_rate": 8.670886075949367e-05,
+      "loss": 0.1434,
+      "step": 137
+    },
+    {
+      "epoch": 0.13102302397341561,
+      "grad_norm": 0.11718117445707321,
+      "learning_rate": 8.734177215189874e-05,
+      "loss": 0.1509,
+      "step": 138
+    },
+    {
+      "epoch": 0.13197246617612152,
+      "grad_norm": 0.14426474273204803,
+      "learning_rate": 8.797468354430381e-05,
+      "loss": 0.1373,
+      "step": 139
+    },
+    {
+      "epoch": 0.13292190837882745,
+      "grad_norm": 0.13101965188980103,
+      "learning_rate": 8.860759493670887e-05,
+      "loss": 0.1358,
+      "step": 140
+    },
+    {
+      "epoch": 0.13387135058153335,
+      "grad_norm": 0.11235956102609634,
+      "learning_rate": 8.924050632911392e-05,
+      "loss": 0.1394,
+      "step": 141
+    },
+    {
+      "epoch": 0.13482079278423925,
+      "grad_norm": 0.11327100545167923,
+      "learning_rate": 8.9873417721519e-05,
+      "loss": 0.1443,
+      "step": 142
+    },
+    {
+      "epoch": 0.13577023498694518,
+      "grad_norm": 0.10912016034126282,
+      "learning_rate": 9.050632911392407e-05,
+      "loss": 0.1698,
+      "step": 143
+    },
+    {
+      "epoch": 0.13671967718965108,
+      "grad_norm": 0.16535617411136627,
+      "learning_rate": 9.113924050632912e-05,
+      "loss": 0.2255,
+      "step": 144
+    },
+    {
+      "epoch": 0.137669119392357,
+      "grad_norm": 0.10184327512979507,
+      "learning_rate": 9.177215189873418e-05,
+      "loss": 0.1371,
+      "step": 145
+    },
+    {
+      "epoch": 0.1386185615950629,
+      "grad_norm": 0.10998040437698364,
+      "learning_rate": 9.240506329113925e-05,
+      "loss": 0.1794,
+      "step": 146
+    },
+    {
+      "epoch": 0.13956800379776882,
+      "grad_norm": 0.08974044770002365,
+      "learning_rate": 9.303797468354431e-05,
+      "loss": 0.144,
+      "step": 147
+    },
+    {
+      "epoch": 0.14051744600047472,
+      "grad_norm": 0.12724193930625916,
+      "learning_rate": 9.367088607594936e-05,
+      "loss": 0.1794,
+      "step": 148
+    },
+    {
+      "epoch": 0.14146688820318062,
+      "grad_norm": 0.1079091802239418,
+      "learning_rate": 9.430379746835444e-05,
+      "loss": 0.1399,
+      "step": 149
+    },
+    {
+      "epoch": 0.14241633040588655,
+      "grad_norm": 0.09480807185173035,
+      "learning_rate": 9.493670886075949e-05,
+      "loss": 0.1395,
+      "step": 150
+    },
+    {
+      "epoch": 0.14336577260859246,
+      "grad_norm": 0.08620745688676834,
+      "learning_rate": 9.556962025316456e-05,
+      "loss": 0.1415,
+      "step": 151
+    },
+    {
+      "epoch": 0.14431521481129836,
+      "grad_norm": 0.10517002642154694,
+      "learning_rate": 9.620253164556962e-05,
+      "loss": 0.1723,
+      "step": 152
+    },
+    {
+      "epoch": 0.14526465701400426,
+      "grad_norm": 0.0956311896443367,
+      "learning_rate": 9.683544303797469e-05,
+      "loss": 0.1515,
+      "step": 153
+    },
+    {
+      "epoch": 0.1462140992167102,
+      "grad_norm": 0.08050324022769928,
+      "learning_rate": 9.746835443037975e-05,
+      "loss": 0.1322,
+      "step": 154
+    },
+    {
+      "epoch": 0.1471635414194161,
+      "grad_norm": 0.0853201299905777,
+      "learning_rate": 9.810126582278482e-05,
+      "loss": 0.142,
+      "step": 155
+    },
+    {
+      "epoch": 0.148112983622122,
+      "grad_norm": 0.09991180151700974,
+      "learning_rate": 9.873417721518988e-05,
+      "loss": 0.1348,
+      "step": 156
+    },
+    {
+      "epoch": 0.14906242582482793,
+      "grad_norm": 0.08640603721141815,
+      "learning_rate": 9.936708860759493e-05,
+      "loss": 0.1397,
+      "step": 157
+    },
+    {
+      "epoch": 0.15001186802753383,
+      "grad_norm": 0.09057717025279999,
+      "learning_rate": 0.0001,
+      "loss": 0.1381,
+      "step": 158
+    },
+    {
+      "epoch": 0.15096131023023973,
+      "grad_norm": 0.09916041046380997,
+      "learning_rate": 0.00010063291139240508,
+      "loss": 0.1509,
+      "step": 159
+    },
+    {
+      "epoch": 0.15191075243294563,
+      "grad_norm": 0.09434045851230621,
+      "learning_rate": 0.00010126582278481013,
+      "loss": 0.1388,
+      "step": 160
+    },
+    {
+      "epoch": 0.15286019463565156,
+      "grad_norm": 0.1273377537727356,
+      "learning_rate": 0.0001018987341772152,
+      "loss": 0.1401,
+      "step": 161
+    },
+    {
+      "epoch": 0.15380963683835747,
+      "grad_norm": 0.1297912299633026,
+      "learning_rate": 0.00010253164556962025,
+      "loss": 0.1852,
+      "step": 162
+    },
+    {
+      "epoch": 0.15475907904106337,
+      "grad_norm": 0.1151595488190651,
+      "learning_rate": 0.00010316455696202532,
+      "loss": 0.1848,
+      "step": 163
+    },
+    {
+      "epoch": 0.1557085212437693,
+      "grad_norm": 0.13381290435791016,
+      "learning_rate": 0.00010379746835443039,
+      "loss": 0.1438,
+      "step": 164
+    },
+    {
+      "epoch": 0.1566579634464752,
+      "grad_norm": 0.07880119979381561,
+      "learning_rate": 0.00010443037974683545,
+      "loss": 0.1327,
+      "step": 165
+    },
+    {
+      "epoch": 0.1576074056491811,
+      "grad_norm": 0.0843740776181221,
+      "learning_rate": 0.00010506329113924052,
+      "loss": 0.1398,
+      "step": 166
+    },
+    {
+      "epoch": 0.158556847851887,
+      "grad_norm": 0.0981813594698906,
+      "learning_rate": 0.00010569620253164559,
+      "loss": 0.1409,
+      "step": 167
+    },
+    {
+      "epoch": 0.15950629005459294,
+      "grad_norm": 0.10005304962396622,
+      "learning_rate": 0.00010632911392405063,
+      "loss": 0.1783,
+      "step": 168
+    },
+    {
+      "epoch": 0.16045573225729884,
+      "grad_norm": 0.08365727961063385,
+      "learning_rate": 0.00010696202531645569,
+      "loss": 0.1275,
+      "step": 169
+    },
+    {
+      "epoch": 0.16140517446000474,
+      "grad_norm": 0.1017635315656662,
+      "learning_rate": 0.00010759493670886076,
+      "loss": 0.1792,
+      "step": 170
+    },
+    {
+      "epoch": 0.16235461666271067,
+      "grad_norm": 0.07007888704538345,
+      "learning_rate": 0.00010822784810126583,
+      "loss": 0.1473,
+      "step": 171
+    },
+    {
+      "epoch": 0.16330405886541657,
+      "grad_norm": 0.07718679308891296,
+      "learning_rate": 0.00010886075949367089,
+      "loss": 0.1396,
+      "step": 172
+    },
+    {
+      "epoch": 0.16425350106812248,
+      "grad_norm": 0.07228100299835205,
+      "learning_rate": 0.00010949367088607596,
+      "loss": 0.1398,
+      "step": 173
+    },
+    {
+      "epoch": 0.16520294327082838,
+      "grad_norm": 0.07955378293991089,
+      "learning_rate": 0.00011012658227848103,
+      "loss": 0.1402,
+      "step": 174
+    },
+    {
+      "epoch": 0.1661523854735343,
+      "grad_norm": 0.0816427692770958,
+      "learning_rate": 0.00011075949367088607,
+      "loss": 0.1345,
+      "step": 175
+    },
+    {
+      "epoch": 0.1671018276762402,
+      "grad_norm": 0.07641757279634476,
+      "learning_rate": 0.00011139240506329114,
+      "loss": 0.1373,
+      "step": 176
+    },
+    {
+      "epoch": 0.1680512698789461,
+      "grad_norm": 0.07354450225830078,
+      "learning_rate": 0.0001120253164556962,
+      "loss": 0.1394,
+      "step": 177
+    },
+    {
+      "epoch": 0.16900071208165202,
+      "grad_norm": 0.08322398364543915,
+      "learning_rate": 0.00011265822784810127,
+      "loss": 0.138,
+      "step": 178
+    },
+    {
+      "epoch": 0.16995015428435795,
+      "grad_norm": 0.13528607785701752,
+      "learning_rate": 0.00011329113924050634,
+      "loss": 0.2188,
+      "step": 179
+    },
+    {
+      "epoch": 0.17089959648706385,
+      "grad_norm": 0.10803692042827606,
+      "learning_rate": 0.0001139240506329114,
+      "loss": 0.1782,
+      "step": 180
+    },
+    {
+      "epoch": 0.17184903868976975,
+      "grad_norm": 0.08404573053121567,
+      "learning_rate": 0.00011455696202531647,
+      "loss": 0.1394,
+      "step": 181
+    },
+    {
+      "epoch": 0.17279848089247568,
+      "grad_norm": 0.12790893018245697,
+      "learning_rate": 0.00011518987341772151,
+      "loss": 0.2157,
+      "step": 182
+    },
+    {
+      "epoch": 0.17374792309518158,
+      "grad_norm": 0.09879907220602036,
+      "learning_rate": 0.00011582278481012658,
+      "loss": 0.1693,
+      "step": 183
+    },
+    {
+      "epoch": 0.17469736529788749,
+      "grad_norm": 0.08092228323221207,
+      "learning_rate": 0.00011645569620253166,
+      "loss": 0.136,
+      "step": 184
+    },
+    {
+      "epoch": 0.1756468075005934,
+      "grad_norm": 0.07660632580518723,
+      "learning_rate": 0.00011708860759493671,
+      "loss": 0.1332,
+      "step": 185
+    },
+    {
+      "epoch": 0.17659624970329932,
+      "grad_norm": 0.07474201172590256,
+      "learning_rate": 0.00011772151898734178,
+      "loss": 0.1301,
+      "step": 186
+    },
+    {
+      "epoch": 0.17754569190600522,
+      "grad_norm": 0.09162931889295578,
+      "learning_rate": 0.00011835443037974685,
+      "loss": 0.1407,
+      "step": 187
+    },
+    {
+      "epoch": 0.17849513410871112,
+      "grad_norm": 0.08646775782108307,
+      "learning_rate": 0.0001189873417721519,
+      "loss": 0.139,
+      "step": 188
+    },
+    {
+      "epoch": 0.17944457631141705,
+      "grad_norm": 0.0759253203868866,
+      "learning_rate": 0.00011962025316455696,
+      "loss": 0.1342,
+      "step": 189
+    },
+    {
+      "epoch": 0.18039401851412296,
+      "grad_norm": 0.08292865008115768,
+      "learning_rate": 0.00012025316455696203,
+      "loss": 0.1389,
+      "step": 190
+    },
+    {
+      "epoch": 0.18134346071682886,
+      "grad_norm": 0.12379574030637741,
+      "learning_rate": 0.0001208860759493671,
+      "loss": 0.1795,
+      "step": 191
+    },
+    {
+      "epoch": 0.18229290291953476,
+      "grad_norm": 0.10240278393030167,
+      "learning_rate": 0.00012151898734177217,
+      "loss": 0.1721,
+      "step": 192
+    },
+    {
+      "epoch": 0.1832423451222407,
+      "grad_norm": 0.09666036069393158,
+      "learning_rate": 0.00012215189873417722,
+      "loss": 0.1783,
+      "step": 193
+    },
+    {
+      "epoch": 0.1841917873249466,
+      "grad_norm": 0.08314768224954605,
+      "learning_rate": 0.0001227848101265823,
+      "loss": 0.1429,
+      "step": 194
+    },
+    {
+      "epoch": 0.1851412295276525,
+      "grad_norm": 0.07590368390083313,
+      "learning_rate": 0.00012341772151898734,
+      "loss": 0.1393,
+      "step": 195
+    },
+    {
+      "epoch": 0.18609067173035843,
+      "grad_norm": 0.10585250705480576,
+      "learning_rate": 0.0001240506329113924,
+      "loss": 0.2155,
+      "step": 196
+    },
+    {
+      "epoch": 0.18704011393306433,
+      "grad_norm": 0.06995555013418198,
+      "learning_rate": 0.00012468354430379748,
+      "loss": 0.1374,
+      "step": 197
+    },
+    {
+      "epoch": 0.18798955613577023,
+      "grad_norm": 0.07370735704898834,
+      "learning_rate": 0.00012531645569620255,
+      "loss": 0.1367,
+      "step": 198
+    },
+    {
+      "epoch": 0.18893899833847613,
+      "grad_norm": 0.07194443792104721,
+      "learning_rate": 0.0001259493670886076,
+      "loss": 0.1437,
+      "step": 199
+    },
+    {
+      "epoch": 0.18988844054118206,
+      "grad_norm": 0.06982647627592087,
+      "learning_rate": 0.00012658227848101267,
+      "loss": 0.1358,
+      "step": 200
+    },
+    {
+      "epoch": 0.19083788274388797,
+      "grad_norm": 0.06538347154855728,
+      "learning_rate": 0.0001272151898734177,
+      "loss": 0.1354,
+      "step": 201
+    },
+    {
+      "epoch": 0.19178732494659387,
+      "grad_norm": 0.07789324969053268,
+      "learning_rate": 0.00012784810126582278,
+      "loss": 0.178,
+      "step": 202
+    },
+    {
+      "epoch": 0.1927367671492998,
+      "grad_norm": 0.07376820594072342,
+      "learning_rate": 0.00012848101265822785,
+      "loss": 0.1621,
+      "step": 203
+    },
+    {
+      "epoch": 0.1936862093520057,
+      "grad_norm": 0.0720745250582695,
+      "learning_rate": 0.00012911392405063292,
+      "loss": 0.132,
+      "step": 204
+    },
+    {
+      "epoch": 0.1946356515547116,
+      "grad_norm": 0.06211116537451744,
+      "learning_rate": 0.000129746835443038,
+      "loss": 0.1387,
+      "step": 205
+    },
+    {
+      "epoch": 0.1955850937574175,
+      "grad_norm": 0.06701771914958954,
+      "learning_rate": 0.00013037974683544306,
+      "loss": 0.14,
+      "step": 206
+    },
+    {
+      "epoch": 0.19653453596012344,
+      "grad_norm": 0.07692532986402512,
+      "learning_rate": 0.0001310126582278481,
+      "loss": 0.1322,
+      "step": 207
+    },
+    {
+      "epoch": 0.19748397816282934,
+      "grad_norm": 0.07763269543647766,
+      "learning_rate": 0.00013164556962025315,
+      "loss": 0.1393,
+      "step": 208
+    },
+    {
+      "epoch": 0.19843342036553524,
+      "grad_norm": 0.08769022673368454,
+      "learning_rate": 0.00013227848101265822,
+      "loss": 0.1489,
+      "step": 209
+    },
+    {
+      "epoch": 0.19938286256824117,
+      "grad_norm": 0.08881859481334686,
+      "learning_rate": 0.0001329113924050633,
+      "loss": 0.1765,
+      "step": 210
+    },
+    {
+      "epoch": 0.20033230477094707,
+      "grad_norm": 0.06811822950839996,
+      "learning_rate": 0.00013354430379746836,
+      "loss": 0.1332,
+      "step": 211
+    },
+    {
+      "epoch": 0.20128174697365298,
+      "grad_norm": 0.06390922516584396,
+      "learning_rate": 0.00013417721518987343,
+      "loss": 0.1343,
+      "step": 212
+    },
+    {
+      "epoch": 0.20223118917635888,
+      "grad_norm": 0.06630406528711319,
+      "learning_rate": 0.0001348101265822785,
+      "loss": 0.1329,
+      "step": 213
+    },
+    {
+      "epoch": 0.2031806313790648,
+      "grad_norm": 0.0730772465467453,
+      "learning_rate": 0.00013544303797468355,
+      "loss": 0.1354,
+      "step": 214
+    },
+    {
+      "epoch": 0.2041300735817707,
+      "grad_norm": 0.06487323343753815,
+      "learning_rate": 0.00013607594936708862,
+      "loss": 0.1297,
+      "step": 215
+    },
+    {
+      "epoch": 0.2050795157844766,
+      "grad_norm": 0.06967955082654953,
+      "learning_rate": 0.00013670886075949366,
+      "loss": 0.1398,
+      "step": 216
+    },
+    {
+      "epoch": 0.20602895798718254,
+      "grad_norm": 0.08531820774078369,
+      "learning_rate": 0.00013734177215189873,
+      "loss": 0.1336,
+      "step": 217
+    },
+    {
+      "epoch": 0.20697840018988845,
+      "grad_norm": 0.0757659375667572,
+      "learning_rate": 0.0001379746835443038,
+      "loss": 0.1606,
+      "step": 218
+    },
+    {
+      "epoch": 0.20792784239259435,
+      "grad_norm": 0.060206469148397446,
+      "learning_rate": 0.00013860759493670888,
+      "loss": 0.1337,
+      "step": 219
+    },
+    {
+      "epoch": 0.20887728459530025,
+      "grad_norm": 0.07996556162834167,
+      "learning_rate": 0.00013924050632911395,
+      "loss": 0.1308,
+      "step": 220
+    },
+    {
+      "epoch": 0.20982672679800618,
+      "grad_norm": 0.06206861138343811,
+      "learning_rate": 0.000139873417721519,
+      "loss": 0.1347,
+      "step": 221
+    },
+    {
+      "epoch": 0.21077616900071208,
+      "grad_norm": 0.08736416697502136,
+      "learning_rate": 0.00014050632911392406,
+      "loss": 0.1768,
+      "step": 222
+    },
+    {
+      "epoch": 0.21172561120341798,
+      "grad_norm": 0.06427916139364243,
+      "learning_rate": 0.00014113924050632913,
+      "loss": 0.1374,
+      "step": 223
+    },
+    {
+      "epoch": 0.21267505340612392,
+      "grad_norm": 0.10996536910533905,
+      "learning_rate": 0.00014177215189873418,
+      "loss": 0.222,
+      "step": 224
+    },
+    {
+      "epoch": 0.21362449560882982,
+      "grad_norm": 0.08439125120639801,
+      "learning_rate": 0.00014240506329113925,
+      "loss": 0.1854,
+      "step": 225
+    },
+    {
+      "epoch": 0.21457393781153572,
+      "grad_norm": 0.06892693787813187,
+      "learning_rate": 0.00014303797468354432,
+      "loss": 0.139,
+      "step": 226
+    },
+    {
+      "epoch": 0.21552338001424162,
+      "grad_norm": 0.08241122961044312,
+      "learning_rate": 0.0001436708860759494,
+      "loss": 0.173,
+      "step": 227
+    },
+    {
+      "epoch": 0.21647282221694755,
+      "grad_norm": 0.07911046594381332,
+      "learning_rate": 0.00014430379746835443,
+      "loss": 0.1418,
+      "step": 228
+    },
+    {
+      "epoch": 0.21742226441965345,
+      "grad_norm": 0.06346064805984497,
+      "learning_rate": 0.0001449367088607595,
+      "loss": 0.1406,
+      "step": 229
+    },
+    {
+      "epoch": 0.21837170662235936,
+      "grad_norm": 0.060393668711185455,
+      "learning_rate": 0.00014556962025316457,
+      "loss": 0.1417,
+      "step": 230
+    },
+    {
+      "epoch": 0.2193211488250653,
+      "grad_norm": 0.05912507325410843,
+      "learning_rate": 0.00014620253164556962,
+      "loss": 0.1298,
+      "step": 231
+    },
+    {
+      "epoch": 0.2202705910277712,
+      "grad_norm": 0.07730337232351303,
+      "learning_rate": 0.0001468354430379747,
+      "loss": 0.1769,
+      "step": 232
+    },
+    {
+      "epoch": 0.2212200332304771,
+      "grad_norm": 0.07612381875514984,
+      "learning_rate": 0.00014746835443037976,
+      "loss": 0.1338,
+      "step": 233
+    },
+    {
+      "epoch": 0.222169475433183,
+      "grad_norm": 0.055311791598796844,
+      "learning_rate": 0.0001481012658227848,
+      "loss": 0.1313,
+      "step": 234
+    },
+    {
+      "epoch": 0.22311891763588892,
+      "grad_norm": 0.08492033183574677,
+      "learning_rate": 0.00014873417721518987,
+      "loss": 0.1367,
+      "step": 235
+    },
+    {
+      "epoch": 0.22406835983859483,
+      "grad_norm": 0.07133237272500992,
+      "learning_rate": 0.00014936708860759494,
+      "loss": 0.1308,
+      "step": 236
+    },
+    {
+      "epoch": 0.22501780204130073,
+      "grad_norm": 0.07148605585098267,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 0.133,
+      "step": 237
+    },
+    {
+      "epoch": 0.22596724424400666,
+      "grad_norm": 0.06900472939014435,
+      "learning_rate": 0.00015063291139240508,
+      "loss": 0.138,
+      "step": 238
+    },
+    {
+      "epoch": 0.22691668644671256,
+      "grad_norm": 0.062325432896614075,
+      "learning_rate": 0.00015126582278481013,
+      "loss": 0.1338,
+      "step": 239
+    },
+    {
+      "epoch": 0.22786612864941846,
+      "grad_norm": 0.06719667464494705,
+      "learning_rate": 0.0001518987341772152,
+      "loss": 0.1316,
+      "step": 240
+    },
+    {
+      "epoch": 0.22881557085212437,
+      "grad_norm": 0.07456009089946747,
+      "learning_rate": 0.00015253164556962024,
+      "loss": 0.1412,
+      "step": 241
+    },
+    {
+      "epoch": 0.2297650130548303,
+      "grad_norm": 0.05619575083255768,
+      "learning_rate": 0.00015316455696202531,
+      "loss": 0.1342,
+      "step": 242
+    },
+    {
+      "epoch": 0.2307144552575362,
+      "grad_norm": 0.06157098710536957,
+      "learning_rate": 0.00015379746835443038,
+      "loss": 0.1329,
+      "step": 243
+    },
+    {
+      "epoch": 0.2316638974602421,
+      "grad_norm": 0.06759827584028244,
+      "learning_rate": 0.00015443037974683546,
+      "loss": 0.1411,
+      "step": 244
+    },
+    {
+      "epoch": 0.232613339662948,
+      "grad_norm": 0.06892479956150055,
+      "learning_rate": 0.00015506329113924053,
+      "loss": 0.1484,
+      "step": 245
+    },
+    {
+      "epoch": 0.23356278186565393,
+      "grad_norm": 0.08536699414253235,
+      "learning_rate": 0.0001556962025316456,
+      "loss": 0.1855,
+      "step": 246
+    },
+    {
+      "epoch": 0.23451222406835984,
+      "grad_norm": 0.06800314784049988,
+      "learning_rate": 0.00015632911392405064,
+      "loss": 0.1379,
+      "step": 247
+    },
+    {
+      "epoch": 0.23546166627106574,
+      "grad_norm": 0.0625622496008873,
+      "learning_rate": 0.00015696202531645568,
+      "loss": 0.1344,
+      "step": 248
+    },
+    {
+      "epoch": 0.23641110847377167,
+      "grad_norm": 0.06030593812465668,
+      "learning_rate": 0.00015759493670886075,
+      "loss": 0.1254,
+      "step": 249
+    },
+    {
+      "epoch": 0.23736055067647757,
+      "grad_norm": 0.06694353371858597,
+      "learning_rate": 0.00015822784810126583,
+      "loss": 0.1413,
+      "step": 250
+    },
+    {
+      "epoch": 0.23830999287918347,
+      "grad_norm": 0.06594134122133255,
+      "learning_rate": 0.0001588607594936709,
+      "loss": 0.1394,
+      "step": 251
+    },
+    {
+      "epoch": 0.23925943508188938,
+      "grad_norm": 0.09062930941581726,
+      "learning_rate": 0.00015949367088607597,
+      "loss": 0.1883,
+      "step": 252
+    },
+    {
+      "epoch": 0.2402088772845953,
+      "grad_norm": 0.06029089167714119,
+      "learning_rate": 0.00016012658227848104,
+      "loss": 0.1271,
+      "step": 253
+    },
+    {
+      "epoch": 0.2411583194873012,
+      "grad_norm": 0.08471622318029404,
+      "learning_rate": 0.00016075949367088608,
+      "loss": 0.172,
+      "step": 254
+    },
+    {
+      "epoch": 0.2421077616900071,
+      "grad_norm": 0.061710160225629807,
+      "learning_rate": 0.00016139240506329115,
+      "loss": 0.1348,
+      "step": 255
+    },
+    {
+      "epoch": 0.24305720389271304,
+      "grad_norm": 0.0812671035528183,
+      "learning_rate": 0.0001620253164556962,
+      "loss": 0.1312,
+      "step": 256
+    },
+    {
+      "epoch": 0.24400664609541894,
+      "grad_norm": 0.06917005032300949,
+      "learning_rate": 0.00016265822784810127,
+      "loss": 0.1464,
+      "step": 257
+    },
+    {
+      "epoch": 0.24495608829812485,
+      "grad_norm": 0.0905887708067894,
+      "learning_rate": 0.00016329113924050634,
+      "loss": 0.1759,
+      "step": 258
+    },
+    {
+      "epoch": 0.24590553050083075,
+      "grad_norm": 0.05976787209510803,
+      "learning_rate": 0.0001639240506329114,
+      "loss": 0.1404,
+      "step": 259
+    },
+    {
+      "epoch": 0.24685497270353668,
+      "grad_norm": 0.07545675337314606,
+      "learning_rate": 0.00016455696202531648,
+      "loss": 0.1322,
+      "step": 260
+    },
+    {
+      "epoch": 0.24780441490624258,
+      "grad_norm": 0.07035024464130402,
+      "learning_rate": 0.00016518987341772152,
+      "loss": 0.1378,
+      "step": 261
+    },
+    {
+      "epoch": 0.24875385710894848,
+      "grad_norm": 0.07665737718343735,
+      "learning_rate": 0.0001658227848101266,
+      "loss": 0.1827,
+      "step": 262
+    },
+    {
+      "epoch": 0.24970329931165441,
+      "grad_norm": 0.06619013845920563,
+      "learning_rate": 0.00016645569620253166,
+      "loss": 0.1284,
+      "step": 263
+    },
+    {
+      "epoch": 0.2506527415143603,
+      "grad_norm": 0.0647001713514328,
+      "learning_rate": 0.0001670886075949367,
+      "loss": 0.133,
+      "step": 264
+    },
+    {
+      "epoch": 0.2516021837170662,
+      "grad_norm": 0.060702718794345856,
+      "learning_rate": 0.00016772151898734178,
+      "loss": 0.1335,
+      "step": 265
+    },
+    {
+      "epoch": 0.25255162591977215,
+      "grad_norm": 0.0508468896150589,
+      "learning_rate": 0.00016835443037974685,
+      "loss": 0.1333,
+      "step": 266
+    },
+    {
+      "epoch": 0.253501068122478,
+      "grad_norm": 0.09877864271402359,
+      "learning_rate": 0.0001689873417721519,
+      "loss": 0.2031,
+      "step": 267
+    },
+    {
+      "epoch": 0.25445051032518395,
+      "grad_norm": 0.06673337519168854,
+      "learning_rate": 0.00016962025316455696,
+      "loss": 0.1356,
+      "step": 268
+    },
+    {
+      "epoch": 0.2553999525278899,
+      "grad_norm": 0.10604165494441986,
+      "learning_rate": 0.00017025316455696204,
+      "loss": 0.2517,
+      "step": 269
+    },
+    {
+      "epoch": 0.25634939473059576,
+      "grad_norm": 0.07689858227968216,
+      "learning_rate": 0.0001708860759493671,
+      "loss": 0.1761,
+      "step": 270
+    },
+    {
+      "epoch": 0.2572988369333017,
+      "grad_norm": 0.05482449755072594,
+      "learning_rate": 0.00017151898734177218,
+      "loss": 0.131,
+      "step": 271
+    },
+    {
+      "epoch": 0.2582482791360076,
+      "grad_norm": 0.08622145652770996,
+      "learning_rate": 0.00017215189873417722,
+      "loss": 0.1335,
+      "step": 272
+    },
+    {
+      "epoch": 0.2591977213387135,
+      "grad_norm": 0.0748213455080986,
+      "learning_rate": 0.0001727848101265823,
+      "loss": 0.176,
+      "step": 273
+    },
+    {
+      "epoch": 0.2601471635414194,
+      "grad_norm": 0.06163305416703224,
+      "learning_rate": 0.00017341772151898733,
+      "loss": 0.1381,
+      "step": 274
+    },
+    {
+      "epoch": 0.26109660574412535,
+      "grad_norm": 0.06141841039061546,
+      "learning_rate": 0.0001740506329113924,
+      "loss": 0.1353,
+      "step": 275
+    },
+    {
+      "epoch": 0.26204604794683123,
+      "grad_norm": 0.07326913625001907,
+      "learning_rate": 0.00017468354430379748,
+      "loss": 0.1441,
+      "step": 276
+    },
+    {
+      "epoch": 0.26299549014953716,
+      "grad_norm": 0.05951124057173729,
+      "learning_rate": 0.00017531645569620255,
+      "loss": 0.1321,
+      "step": 277
+    },
+    {
+      "epoch": 0.26394493235224303,
+      "grad_norm": 0.08364073932170868,
+      "learning_rate": 0.00017594936708860762,
+      "loss": 0.187,
+      "step": 278
+    },
+    {
+      "epoch": 0.26489437455494896,
+      "grad_norm": 0.05849132314324379,
+      "learning_rate": 0.00017658227848101266,
+      "loss": 0.1393,
+      "step": 279
+    },
+    {
+      "epoch": 0.2658438167576549,
+      "grad_norm": 0.05452360957860947,
+      "learning_rate": 0.00017721518987341773,
+      "loss": 0.1342,
+      "step": 280
+    },
+    {
+      "epoch": 0.26679325896036077,
+      "grad_norm": 0.04878188297152519,
+      "learning_rate": 0.00017784810126582278,
+      "loss": 0.1445,
+      "step": 281
+    },
+    {
+      "epoch": 0.2677427011630667,
+      "grad_norm": 0.06066753342747688,
+      "learning_rate": 0.00017848101265822785,
+      "loss": 0.1423,
+      "step": 282
+    },
+    {
+      "epoch": 0.26869214336577263,
+      "grad_norm": 0.04918207973241806,
+      "learning_rate": 0.00017911392405063292,
+      "loss": 0.1316,
+      "step": 283
+    },
+    {
+      "epoch": 0.2696415855684785,
+      "grad_norm": 0.05103525519371033,
+      "learning_rate": 0.000179746835443038,
+      "loss": 0.1313,
+      "step": 284
+    },
+    {
+      "epoch": 0.27059102777118443,
+      "grad_norm": 0.05667628347873688,
+      "learning_rate": 0.00018037974683544306,
+      "loss": 0.1434,
+      "step": 285
+    },
+    {
+      "epoch": 0.27154046997389036,
+      "grad_norm": 0.06226016581058502,
+      "learning_rate": 0.00018101265822784813,
+      "loss": 0.1357,
+      "step": 286
+    },
+    {
+      "epoch": 0.27248991217659624,
+      "grad_norm": 0.04695293679833412,
+      "learning_rate": 0.00018164556962025317,
+      "loss": 0.1314,
+      "step": 287
+    },
+    {
+      "epoch": 0.27343935437930217,
+      "grad_norm": 0.05762844532728195,
+      "learning_rate": 0.00018227848101265824,
+      "loss": 0.1349,
+      "step": 288
+    },
+    {
+      "epoch": 0.27438879658200804,
+      "grad_norm": 0.05454534292221069,
+      "learning_rate": 0.0001829113924050633,
+      "loss": 0.1432,
+      "step": 289
+    },
+    {
+      "epoch": 0.275338238784714,
+      "grad_norm": 0.050270579755306244,
+      "learning_rate": 0.00018354430379746836,
+      "loss": 0.1272,
+      "step": 290
+    },
+    {
+      "epoch": 0.2762876809874199,
+      "grad_norm": 0.0688452497124672,
+      "learning_rate": 0.00018417721518987343,
+      "loss": 0.1708,
+      "step": 291
+    },
+    {
+      "epoch": 0.2772371231901258,
+      "grad_norm": 0.06213200092315674,
+      "learning_rate": 0.0001848101265822785,
+      "loss": 0.1674,
+      "step": 292
+    },
+    {
+      "epoch": 0.2781865653928317,
+      "grad_norm": 0.059717319905757904,
+      "learning_rate": 0.00018544303797468354,
+      "loss": 0.169,
+      "step": 293
+    },
+    {
+      "epoch": 0.27913600759553764,
+      "grad_norm": 0.06223325803875923,
+      "learning_rate": 0.00018607594936708861,
+      "loss": 0.1369,
+      "step": 294
+    },
+    {
+      "epoch": 0.2800854497982435,
+      "grad_norm": 0.053163208067417145,
+      "learning_rate": 0.00018670886075949369,
+      "loss": 0.133,
+      "step": 295
+    },
+    {
+      "epoch": 0.28103489200094944,
+      "grad_norm": 0.06647945195436478,
+      "learning_rate": 0.00018734177215189873,
+      "loss": 0.1438,
+      "step": 296
+    },
+    {
+      "epoch": 0.2819843342036554,
+      "grad_norm": 0.0588272288441658,
+      "learning_rate": 0.0001879746835443038,
+      "loss": 0.1338,
+      "step": 297
+    },
+    {
+      "epoch": 0.28293377640636125,
+      "grad_norm": 0.05841274932026863,
+      "learning_rate": 0.00018860759493670887,
+      "loss": 0.1329,
+      "step": 298
+    },
+    {
+      "epoch": 0.2838832186090672,
+      "grad_norm": 0.09033369272947311,
+      "learning_rate": 0.00018924050632911394,
+      "loss": 0.1747,
+      "step": 299
+    },
+    {
+      "epoch": 0.2848326608117731,
+      "grad_norm": 0.052215326577425,
+      "learning_rate": 0.00018987341772151899,
+      "loss": 0.1296,
+      "step": 300
+    },
+    {
+      "epoch": 0.285782103014479,
+      "grad_norm": 0.05880101025104523,
+      "learning_rate": 0.00019050632911392406,
+      "loss": 0.1287,
+      "step": 301
+    },
+    {
+      "epoch": 0.2867315452171849,
+      "grad_norm": 0.0691700354218483,
+      "learning_rate": 0.00019113924050632913,
+      "loss": 0.1676,
+      "step": 302
+    },
+    {
+      "epoch": 0.2876809874198908,
+      "grad_norm": 0.057025909423828125,
+      "learning_rate": 0.0001917721518987342,
+      "loss": 0.1346,
+      "step": 303
+    },
+    {
+      "epoch": 0.2886304296225967,
+      "grad_norm": 0.04936329275369644,
+      "learning_rate": 0.00019240506329113924,
+      "loss": 0.1354,
+      "step": 304
+    },
+    {
+      "epoch": 0.28957987182530265,
+      "grad_norm": 0.0680055245757103,
+      "learning_rate": 0.0001930379746835443,
+      "loss": 0.1344,
+      "step": 305
+    },
+    {
+      "epoch": 0.2905293140280085,
+      "grad_norm": 0.07374466210603714,
+      "learning_rate": 0.00019367088607594938,
+      "loss": 0.1428,
+      "step": 306
+    },
+    {
+      "epoch": 0.29147875623071445,
+      "grad_norm": 0.061204761266708374,
+      "learning_rate": 0.00019430379746835443,
+      "loss": 0.1246,
+      "step": 307
+    },
+    {
+      "epoch": 0.2924281984334204,
+      "grad_norm": 0.053467705845832825,
+      "learning_rate": 0.0001949367088607595,
+      "loss": 0.1342,
+      "step": 308
+    },
+    {
+      "epoch": 0.29337764063612626,
+      "grad_norm": 0.057525087147951126,
+      "learning_rate": 0.00019556962025316457,
+      "loss": 0.1377,
+      "step": 309
+    },
+    {
+      "epoch": 0.2943270828388322,
+      "grad_norm": 0.07857844978570938,
+      "learning_rate": 0.00019620253164556964,
+      "loss": 0.2076,
+      "step": 310
+    },
+    {
+      "epoch": 0.2952765250415381,
+      "grad_norm": 0.05250545218586922,
+      "learning_rate": 0.0001968354430379747,
+      "loss": 0.1432,
+      "step": 311
+    },
+    {
+      "epoch": 0.296225967244244,
+      "grad_norm": 0.07495012134313583,
+      "learning_rate": 0.00019746835443037975,
+      "loss": 0.1766,
+      "step": 312
+    },
+    {
+      "epoch": 0.2971754094469499,
+      "grad_norm": 0.04692578688263893,
+      "learning_rate": 0.0001981012658227848,
+      "loss": 0.1408,
+      "step": 313
+    },
+    {
+      "epoch": 0.29812485164965585,
+      "grad_norm": 0.055666085332632065,
+      "learning_rate": 0.00019873417721518987,
+      "loss": 0.1391,
+      "step": 314
+    },
+    {
+      "epoch": 0.29907429385236173,
+      "grad_norm": 0.050465911626815796,
+      "learning_rate": 0.00019936708860759494,
+      "loss": 0.1415,
+      "step": 315
+    },
+    {
+      "epoch": 0.30002373605506766,
+      "grad_norm": 0.051260240375995636,
+      "learning_rate": 0.0002,
+      "loss": 0.1423,
+      "step": 316
+    },
+    {
+      "epoch": 0.30097317825777353,
+      "grad_norm": 0.0503215529024601,
+      "learning_rate": 0.000199999938945738,
+      "loss": 0.1348,
+      "step": 317
+    },
+    {
+      "epoch": 0.30192262046047946,
+      "grad_norm": 0.04917483776807785,
+      "learning_rate": 0.0001999997557830265,
+      "loss": 0.1342,
+      "step": 318
+    },
+    {
+      "epoch": 0.3028720626631854,
+      "grad_norm": 0.06354209035634995,
+      "learning_rate": 0.00019999945051208916,
+      "loss": 0.1365,
+      "step": 319
+    },
+    {
+      "epoch": 0.30382150486589127,
+      "grad_norm": 0.04878314957022667,
+      "learning_rate": 0.0001999990231332988,
+      "loss": 0.13,
+      "step": 320
+    },
+    {
+      "epoch": 0.3047709470685972,
+      "grad_norm": 0.07046223431825638,
+      "learning_rate": 0.0001999984736471772,
+      "loss": 0.1394,
+      "step": 321
+    },
+    {
+      "epoch": 0.30572038927130313,
+      "grad_norm": 0.04456232488155365,
+      "learning_rate": 0.00019999780205439538,
+      "loss": 0.1278,
+      "step": 322
+    },
+    {
+      "epoch": 0.306669831474009,
+      "grad_norm": 0.06280628591775894,
+      "learning_rate": 0.00019999700835577342,
+      "loss": 0.1715,
+      "step": 323
+    },
+    {
+      "epoch": 0.30761927367671493,
+      "grad_norm": 0.07462131977081299,
+      "learning_rate": 0.00019999609255228046,
+      "loss": 0.1772,
+      "step": 324
+    },
+    {
+      "epoch": 0.30856871587942086,
+      "grad_norm": 0.059642352163791656,
+      "learning_rate": 0.00019999505464503482,
+      "loss": 0.1294,
+      "step": 325
+    },
+    {
+      "epoch": 0.30951815808212674,
+      "grad_norm": 0.06458820402622223,
+      "learning_rate": 0.00019999389463530383,
+      "loss": 0.173,
+      "step": 326
+    },
+    {
+      "epoch": 0.31046760028483267,
+      "grad_norm": 0.05901939421892166,
+      "learning_rate": 0.00019999261252450396,
+      "loss": 0.1419,
+      "step": 327
+    },
+    {
+      "epoch": 0.3114170424875386,
+      "grad_norm": 0.055540215224027634,
+      "learning_rate": 0.00019999120831420083,
+      "loss": 0.1314,
+      "step": 328
+    },
+    {
+      "epoch": 0.3123664846902445,
+      "grad_norm": 0.0546739287674427,
+      "learning_rate": 0.00019998968200610903,
+      "loss": 0.1354,
+      "step": 329
+    },
+    {
+      "epoch": 0.3133159268929504,
+      "grad_norm": 0.0689477026462555,
+      "learning_rate": 0.00019998803360209234,
+      "loss": 0.132,
+      "step": 330
+    },
+    {
+      "epoch": 0.3142653690956563,
+      "grad_norm": 0.05279696360230446,
+      "learning_rate": 0.00019998626310416365,
+      "loss": 0.1424,
+      "step": 331
+    },
+    {
+      "epoch": 0.3152148112983622,
+      "grad_norm": 0.055384278297424316,
+      "learning_rate": 0.00019998437051448482,
+      "loss": 0.141,
+      "step": 332
+    },
+    {
+      "epoch": 0.31616425350106814,
+      "grad_norm": 0.04636182263493538,
+      "learning_rate": 0.0001999823558353669,
+      "loss": 0.1414,
+      "step": 333
+    },
+    {
+      "epoch": 0.317113695703774,
+      "grad_norm": 0.04795726016163826,
+      "learning_rate": 0.00019998021906926993,
+      "loss": 0.1255,
+      "step": 334
+    },
+    {
+      "epoch": 0.31806313790647994,
+      "grad_norm": 0.05326540395617485,
+      "learning_rate": 0.00019997796021880318,
+      "loss": 0.1309,
+      "step": 335
+    },
+    {
+      "epoch": 0.3190125801091859,
+      "grad_norm": 0.0684736892580986,
+      "learning_rate": 0.00019997557928672484,
+      "loss": 0.1825,
+      "step": 336
+    },
+    {
+      "epoch": 0.31996202231189175,
+      "grad_norm": 0.042282164096832275,
+      "learning_rate": 0.0001999730762759422,
+      "loss": 0.12,
+      "step": 337
+    },
+    {
+      "epoch": 0.3209114645145977,
+      "grad_norm": 0.05297423154115677,
+      "learning_rate": 0.00019997045118951175,
+      "loss": 0.1309,
+      "step": 338
+    },
+    {
+      "epoch": 0.3218609067173036,
+      "grad_norm": 0.080621138215065,
+      "learning_rate": 0.00019996770403063883,
+      "loss": 0.2134,
+      "step": 339
+    },
+    {
+      "epoch": 0.3228103489200095,
+      "grad_norm": 0.05552308261394501,
+      "learning_rate": 0.00019996483480267803,
+      "loss": 0.1361,
+      "step": 340
+    },
+    {
+      "epoch": 0.3237597911227154,
+      "grad_norm": 0.05070111155509949,
+      "learning_rate": 0.00019996184350913287,
+      "loss": 0.1314,
+      "step": 341
+    },
+    {
+      "epoch": 0.32470923332542134,
+      "grad_norm": 0.04412266984581947,
+      "learning_rate": 0.00019995873015365601,
+      "loss": 0.1299,
+      "step": 342
+    },
+    {
+      "epoch": 0.3256586755281272,
+      "grad_norm": 0.0445338599383831,
+      "learning_rate": 0.00019995549474004917,
+      "loss": 0.1313,
+      "step": 343
+    },
+    {
+      "epoch": 0.32660811773083315,
+      "grad_norm": 0.08224980533123016,
+      "learning_rate": 0.000199952137272263,
+      "loss": 0.1844,
+      "step": 344
+    },
+    {
+      "epoch": 0.327557559933539,
+      "grad_norm": 0.04331446811556816,
+      "learning_rate": 0.0001999486577543972,
+      "loss": 0.133,
+      "step": 345
+    },
+    {
+      "epoch": 0.32850700213624495,
+      "grad_norm": 0.049314577132463455,
+      "learning_rate": 0.00019994505619070068,
+      "loss": 0.1351,
+      "step": 346
+    },
+    {
+      "epoch": 0.3294564443389509,
+      "grad_norm": 0.0697011798620224,
+      "learning_rate": 0.00019994133258557117,
+      "loss": 0.1709,
+      "step": 347
+    },
+    {
+      "epoch": 0.33040588654165676,
+      "grad_norm": 0.0510990135371685,
+      "learning_rate": 0.00019993748694355557,
+      "loss": 0.1365,
+      "step": 348
+    },
+    {
+      "epoch": 0.3313553287443627,
+      "grad_norm": 0.05100785568356514,
+      "learning_rate": 0.00019993351926934967,
+      "loss": 0.1302,
+      "step": 349
+    },
+    {
+      "epoch": 0.3323047709470686,
+      "grad_norm": 0.08001980185508728,
+      "learning_rate": 0.00019992942956779838,
+      "loss": 0.1736,
+      "step": 350
+    },
+    {
+      "epoch": 0.3332542131497745,
+      "grad_norm": 0.05298507958650589,
+      "learning_rate": 0.00019992521784389559,
+      "loss": 0.159,
+      "step": 351
+    },
+    {
+      "epoch": 0.3342036553524804,
+      "grad_norm": 0.04655485600233078,
+      "learning_rate": 0.00019992088410278414,
+      "loss": 0.1401,
+      "step": 352
+    },
+    {
+      "epoch": 0.33515309755518635,
+      "grad_norm": 0.047509439289569855,
+      "learning_rate": 0.00019991642834975594,
+      "loss": 0.1369,
+      "step": 353
+    },
+    {
+      "epoch": 0.3361025397578922,
+      "grad_norm": 0.046006906777620316,
+      "learning_rate": 0.0001999118505902518,
+      "loss": 0.1384,
+      "step": 354
+    },
+    {
+      "epoch": 0.33705198196059816,
+      "grad_norm": 0.07522892951965332,
+      "learning_rate": 0.00019990715082986155,
+      "loss": 0.2254,
+      "step": 355
+    },
+    {
+      "epoch": 0.33800142416330403,
+      "grad_norm": 0.048646144568920135,
+      "learning_rate": 0.00019990232907432404,
+      "loss": 0.1355,
+      "step": 356
+    },
+    {
+      "epoch": 0.33895086636600996,
+      "grad_norm": 0.03941798582673073,
+      "learning_rate": 0.000199897385329527,
+      "loss": 0.1242,
+      "step": 357
+    },
+    {
+      "epoch": 0.3399003085687159,
+      "grad_norm": 0.04582727700471878,
+      "learning_rate": 0.0001998923196015072,
+      "loss": 0.1347,
+      "step": 358
+    },
+    {
+      "epoch": 0.34084975077142177,
+      "grad_norm": 0.05890033766627312,
+      "learning_rate": 0.00019988713189645027,
+      "loss": 0.1356,
+      "step": 359
+    },
+    {
+      "epoch": 0.3417991929741277,
+      "grad_norm": 0.050398606806993484,
+      "learning_rate": 0.00019988182222069093,
+      "loss": 0.1379,
+      "step": 360
+    },
+    {
+      "epoch": 0.3427486351768336,
+      "grad_norm": 0.053657352924346924,
+      "learning_rate": 0.00019987639058071267,
+      "loss": 0.1417,
+      "step": 361
+    },
+    {
+      "epoch": 0.3436980773795395,
+      "grad_norm": 0.04928993433713913,
+      "learning_rate": 0.00019987083698314804,
+      "loss": 0.1269,
+      "step": 362
+    },
+    {
+      "epoch": 0.34464751958224543,
+      "grad_norm": 0.04932550713419914,
+      "learning_rate": 0.0001998651614347784,
+      "loss": 0.1429,
+      "step": 363
+    },
+    {
+      "epoch": 0.34559696178495136,
+      "grad_norm": 0.0531768873333931,
+      "learning_rate": 0.00019985936394253413,
+      "loss": 0.1367,
+      "step": 364
+    },
+    {
+      "epoch": 0.34654640398765724,
+      "grad_norm": 0.05342009291052818,
+      "learning_rate": 0.00019985344451349443,
+      "loss": 0.1365,
+      "step": 365
+    },
+    {
+      "epoch": 0.34749584619036317,
+      "grad_norm": 0.04960772022604942,
+      "learning_rate": 0.00019984740315488742,
+      "loss": 0.133,
+      "step": 366
+    },
+    {
+      "epoch": 0.3484452883930691,
+      "grad_norm": 0.04490765556693077,
+      "learning_rate": 0.00019984123987409013,
+      "loss": 0.1347,
+      "step": 367
+    },
+    {
+      "epoch": 0.34939473059577497,
+      "grad_norm": 0.05546121671795845,
+      "learning_rate": 0.0001998349546786285,
+      "loss": 0.169,
+      "step": 368
+    },
+    {
+      "epoch": 0.3503441727984809,
+      "grad_norm": 0.04962169751524925,
+      "learning_rate": 0.0001998285475761772,
+      "loss": 0.1325,
+      "step": 369
+    },
+    {
+      "epoch": 0.3512936150011868,
+      "grad_norm": 0.0451858825981617,
+      "learning_rate": 0.00019982201857455988,
+      "loss": 0.1291,
+      "step": 370
+    },
+    {
+      "epoch": 0.3522430572038927,
+      "grad_norm": 0.07738906145095825,
+      "learning_rate": 0.00019981536768174903,
+      "loss": 0.1841,
+      "step": 371
+    },
+    {
+      "epoch": 0.35319249940659864,
+      "grad_norm": 0.05104148015379906,
+      "learning_rate": 0.000199808594905866,
+      "loss": 0.1375,
+      "step": 372
+    },
+    {
+      "epoch": 0.3541419416093045,
+      "grad_norm": 0.04850155860185623,
+      "learning_rate": 0.00019980170025518082,
+      "loss": 0.1335,
+      "step": 373
+    },
+    {
+      "epoch": 0.35509138381201044,
+      "grad_norm": 0.050271324813365936,
+      "learning_rate": 0.00019979468373811248,
+      "loss": 0.1394,
+      "step": 374
+    },
+    {
+      "epoch": 0.35604082601471637,
+      "grad_norm": 0.050799645483493805,
+      "learning_rate": 0.0001997875453632288,
+      "loss": 0.135,
+      "step": 375
+    },
+    {
+      "epoch": 0.35699026821742225,
+      "grad_norm": 0.05703526735305786,
+      "learning_rate": 0.00019978028513924627,
+      "loss": 0.1371,
+      "step": 376
+    },
+    {
+      "epoch": 0.3579397104201282,
+      "grad_norm": 0.06665853410959244,
+      "learning_rate": 0.00019977290307503028,
+      "loss": 0.1837,
+      "step": 377
+    },
+    {
+      "epoch": 0.3588891526228341,
+      "grad_norm": 0.04639972746372223,
+      "learning_rate": 0.000199765399179595,
+      "loss": 0.1315,
+      "step": 378
+    },
+    {
+      "epoch": 0.35983859482554,
+      "grad_norm": 0.07625308632850647,
+      "learning_rate": 0.00019975777346210326,
+      "loss": 0.2064,
+      "step": 379
+    },
+    {
+      "epoch": 0.3607880370282459,
+      "grad_norm": 0.048770248889923096,
+      "learning_rate": 0.00019975002593186674,
+      "loss": 0.1363,
+      "step": 380
+    },
+    {
+      "epoch": 0.36173747923095184,
+      "grad_norm": 0.04932136833667755,
+      "learning_rate": 0.00019974215659834582,
+      "loss": 0.1374,
+      "step": 381
+    },
+    {
+      "epoch": 0.3626869214336577,
+      "grad_norm": 0.03848756104707718,
+      "learning_rate": 0.00019973416547114964,
+      "loss": 0.1333,
+      "step": 382
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.04468891769647598,
+      "learning_rate": 0.00019972605256003605,
+      "loss": 0.129,
+      "step": 383
+    },
+    {
+      "epoch": 0.3645858058390695,
+      "grad_norm": 0.048413511365652084,
+      "learning_rate": 0.0001997178178749116,
+      "loss": 0.1314,
+      "step": 384
+    },
+    {
+      "epoch": 0.36553524804177545,
+      "grad_norm": 0.045054856687784195,
+      "learning_rate": 0.00019970946142583155,
+      "loss": 0.1323,
+      "step": 385
+    },
+    {
+      "epoch": 0.3664846902444814,
+      "grad_norm": 0.05541200935840607,
+      "learning_rate": 0.00019970098322299982,
+      "loss": 0.1342,
+      "step": 386
+    },
+    {
+      "epoch": 0.36743413244718726,
+      "grad_norm": 0.06861472874879837,
+      "learning_rate": 0.00019969238327676906,
+      "loss": 0.1347,
+      "step": 387
+    },
+    {
+      "epoch": 0.3683835746498932,
+      "grad_norm": 0.043996453285217285,
+      "learning_rate": 0.00019968366159764047,
+      "loss": 0.132,
+      "step": 388
+    },
+    {
+      "epoch": 0.3693330168525991,
+      "grad_norm": 0.06562239676713943,
+      "learning_rate": 0.000199674818196264,
+      "loss": 0.1759,
+      "step": 389
+    },
+    {
+      "epoch": 0.370282459055305,
+      "grad_norm": 0.04714899882674217,
+      "learning_rate": 0.00019966585308343822,
+      "loss": 0.1274,
+      "step": 390
+    },
+    {
+      "epoch": 0.3712319012580109,
+      "grad_norm": 0.04736959934234619,
+      "learning_rate": 0.00019965676627011026,
+      "loss": 0.1265,
+      "step": 391
+    },
+    {
+      "epoch": 0.37218134346071685,
+      "grad_norm": 0.056829433888196945,
+      "learning_rate": 0.0001996475577673759,
+      "loss": 0.1402,
+      "step": 392
+    },
+    {
+      "epoch": 0.3731307856634227,
+      "grad_norm": 0.0426231250166893,
+      "learning_rate": 0.00019963822758647953,
+      "loss": 0.1364,
+      "step": 393
+    },
+    {
+      "epoch": 0.37408022786612866,
+      "grad_norm": 0.07376877963542938,
+      "learning_rate": 0.00019962877573881404,
+      "loss": 0.2042,
+      "step": 394
+    },
+    {
+      "epoch": 0.3750296700688346,
+      "grad_norm": 0.043273668736219406,
+      "learning_rate": 0.00019961920223592104,
+      "loss": 0.132,
+      "step": 395
+    },
+    {
+      "epoch": 0.37597911227154046,
+      "grad_norm": 0.044406965374946594,
+      "learning_rate": 0.00019960950708949052,
+      "loss": 0.1344,
+      "step": 396
+    },
+    {
+      "epoch": 0.3769285544742464,
+      "grad_norm": 0.040342606604099274,
+      "learning_rate": 0.00019959969031136106,
+      "loss": 0.1214,
+      "step": 397
+    },
+    {
+      "epoch": 0.37787799667695227,
+      "grad_norm": 0.05118388682603836,
+      "learning_rate": 0.00019958975191351983,
+      "loss": 0.14,
+      "step": 398
+    },
+    {
+      "epoch": 0.3788274388796582,
+      "grad_norm": 0.045876793563365936,
+      "learning_rate": 0.00019957969190810245,
+      "loss": 0.1335,
+      "step": 399
+    },
+    {
+      "epoch": 0.3797768810823641,
+      "grad_norm": 0.0645332932472229,
+      "learning_rate": 0.00019956951030739308,
+      "loss": 0.1702,
+      "step": 400
+    },
+    {
+      "epoch": 0.38072632328507,
+      "grad_norm": 0.05039132386445999,
+      "learning_rate": 0.00019955920712382423,
+      "loss": 0.136,
+      "step": 401
+    },
+    {
+      "epoch": 0.38167576548777593,
+      "grad_norm": 0.052004653960466385,
+      "learning_rate": 0.00019954878236997704,
+      "loss": 0.1386,
+      "step": 402
+    },
+    {
+      "epoch": 0.38262520769048186,
+      "grad_norm": 0.05021458491683006,
+      "learning_rate": 0.00019953823605858105,
+      "loss": 0.1378,
+      "step": 403
+    },
+    {
+      "epoch": 0.38357464989318774,
+      "grad_norm": 0.058653559535741806,
+      "learning_rate": 0.0001995275682025141,
+      "loss": 0.1437,
+      "step": 404
+    },
+    {
+      "epoch": 0.38452409209589367,
+      "grad_norm": 0.04466673359274864,
+      "learning_rate": 0.00019951677881480264,
+      "loss": 0.1334,
+      "step": 405
+    },
+    {
+      "epoch": 0.3854735342985996,
+      "grad_norm": 0.06119415909051895,
+      "learning_rate": 0.00019950586790862138,
+      "loss": 0.1296,
+      "step": 406
+    },
+    {
+      "epoch": 0.38642297650130547,
+      "grad_norm": 0.04749077931046486,
+      "learning_rate": 0.0001994948354972935,
+      "loss": 0.1341,
+      "step": 407
+    },
+    {
+      "epoch": 0.3873724187040114,
+      "grad_norm": 0.037752799689769745,
+      "learning_rate": 0.00019948368159429053,
+      "loss": 0.134,
+      "step": 408
+    },
+    {
+      "epoch": 0.38832186090671733,
+      "grad_norm": 0.08903038501739502,
+      "learning_rate": 0.00019947240621323226,
+      "loss": 0.2155,
+      "step": 409
+    },
+    {
+      "epoch": 0.3892713031094232,
+      "grad_norm": 0.03878140076994896,
+      "learning_rate": 0.00019946100936788698,
+      "loss": 0.1176,
+      "step": 410
+    },
+    {
+      "epoch": 0.39022074531212914,
+      "grad_norm": 0.04927309602499008,
+      "learning_rate": 0.00019944949107217113,
+      "loss": 0.1344,
+      "step": 411
+    },
+    {
+      "epoch": 0.391170187514835,
+      "grad_norm": 0.04933890327811241,
+      "learning_rate": 0.00019943785134014962,
+      "loss": 0.1315,
+      "step": 412
+    },
+    {
+      "epoch": 0.39211962971754094,
+      "grad_norm": 0.06702516227960587,
+      "learning_rate": 0.0001994260901860355,
+      "loss": 0.1826,
+      "step": 413
+    },
+    {
+      "epoch": 0.39306907192024687,
+      "grad_norm": 0.048132237046957016,
+      "learning_rate": 0.00019941420762419014,
+      "loss": 0.1436,
+      "step": 414
+    },
+    {
+      "epoch": 0.39401851412295275,
+      "grad_norm": 0.07756894826889038,
+      "learning_rate": 0.00019940220366912318,
+      "loss": 0.2162,
+      "step": 415
+    },
+    {
+      "epoch": 0.3949679563256587,
+      "grad_norm": 0.04789011925458908,
+      "learning_rate": 0.00019939007833549242,
+      "loss": 0.1295,
+      "step": 416
+    },
+    {
+      "epoch": 0.3959173985283646,
+      "grad_norm": 0.04369444027543068,
+      "learning_rate": 0.000199377831638104,
+      "loss": 0.1322,
+      "step": 417
+    },
+    {
+      "epoch": 0.3968668407310705,
+      "grad_norm": 0.05376122146844864,
+      "learning_rate": 0.00019936546359191216,
+      "loss": 0.1743,
+      "step": 418
+    },
+    {
+      "epoch": 0.3978162829337764,
+      "grad_norm": 0.045930229127407074,
+      "learning_rate": 0.0001993529742120193,
+      "loss": 0.1336,
+      "step": 419
+    },
+    {
+      "epoch": 0.39876572513648234,
+      "grad_norm": 0.039980966597795486,
+      "learning_rate": 0.00019934036351367606,
+      "loss": 0.1349,
+      "step": 420
+    },
+    {
+      "epoch": 0.3997151673391882,
+      "grad_norm": 0.03797341510653496,
+      "learning_rate": 0.00019932763151228115,
+      "loss": 0.1256,
+      "step": 421
+    },
+    {
+      "epoch": 0.40066460954189415,
+      "grad_norm": 0.04779914394021034,
+      "learning_rate": 0.00019931477822338146,
+      "loss": 0.1411,
+      "step": 422
+    },
+    {
+      "epoch": 0.4016140517446,
+      "grad_norm": 0.040458668023347855,
+      "learning_rate": 0.00019930180366267193,
+      "loss": 0.126,
+      "step": 423
+    },
+    {
+      "epoch": 0.40256349394730595,
+      "grad_norm": 0.04114462807774544,
+      "learning_rate": 0.0001992887078459956,
+      "loss": 0.127,
+      "step": 424
+    },
+    {
+      "epoch": 0.4035129361500119,
+      "grad_norm": 0.048119012266397476,
+      "learning_rate": 0.00019927549078934358,
+      "loss": 0.1346,
+      "step": 425
+    },
+    {
+      "epoch": 0.40446237835271776,
+      "grad_norm": 0.0545562319457531,
+      "learning_rate": 0.00019926215250885504,
+      "loss": 0.1387,
+      "step": 426
+    },
+    {
+      "epoch": 0.4054118205554237,
+      "grad_norm": 0.052092909812927246,
+      "learning_rate": 0.00019924869302081715,
+      "loss": 0.1389,
+      "step": 427
+    },
+    {
+      "epoch": 0.4063612627581296,
+      "grad_norm": 0.03847799077630043,
+      "learning_rate": 0.0001992351123416651,
+      "loss": 0.1234,
+      "step": 428
+    },
+    {
+      "epoch": 0.4073107049608355,
+      "grad_norm": 0.0436912477016449,
+      "learning_rate": 0.000199221410487982,
+      "loss": 0.1362,
+      "step": 429
+    },
+    {
+      "epoch": 0.4082601471635414,
+      "grad_norm": 0.04420888423919678,
+      "learning_rate": 0.00019920758747649908,
+      "loss": 0.1243,
+      "step": 430
+    },
+    {
+      "epoch": 0.40920958936624735,
+      "grad_norm": 0.037297070026397705,
+      "learning_rate": 0.00019919364332409535,
+      "loss": 0.1331,
+      "step": 431
+    },
+    {
+      "epoch": 0.4101590315689532,
+      "grad_norm": 0.03854360058903694,
+      "learning_rate": 0.00019917957804779782,
+      "loss": 0.1266,
+      "step": 432
+    },
+    {
+      "epoch": 0.41110847377165916,
+      "grad_norm": 0.04071418195962906,
+      "learning_rate": 0.00019916539166478137,
+      "loss": 0.1292,
+      "step": 433
+    },
+    {
+      "epoch": 0.4120579159743651,
+      "grad_norm": 0.04560808837413788,
+      "learning_rate": 0.00019915108419236882,
+      "loss": 0.1381,
+      "step": 434
+    },
+    {
+      "epoch": 0.41300735817707096,
+      "grad_norm": 0.06313233822584152,
+      "learning_rate": 0.00019913665564803078,
+      "loss": 0.2031,
+      "step": 435
+    },
+    {
+      "epoch": 0.4139568003797769,
+      "grad_norm": 0.04507524147629738,
+      "learning_rate": 0.00019912210604938578,
+      "loss": 0.1277,
+      "step": 436
+    },
+    {
+      "epoch": 0.41490624258248276,
+      "grad_norm": 0.05048058554530144,
+      "learning_rate": 0.00019910743541420007,
+      "loss": 0.1315,
+      "step": 437
+    },
+    {
+      "epoch": 0.4158556847851887,
+      "grad_norm": 0.04872648045420647,
+      "learning_rate": 0.0001990926437603878,
+      "loss": 0.1292,
+      "step": 438
+    },
+    {
+      "epoch": 0.4168051269878946,
+      "grad_norm": 0.04400710016489029,
+      "learning_rate": 0.00019907773110601075,
+      "loss": 0.1236,
+      "step": 439
+    },
+    {
+      "epoch": 0.4177545691906005,
+      "grad_norm": 0.051591627299785614,
+      "learning_rate": 0.00019906269746927863,
+      "loss": 0.1358,
+      "step": 440
+    },
+    {
+      "epoch": 0.41870401139330643,
+      "grad_norm": 0.04288725182414055,
+      "learning_rate": 0.00019904754286854877,
+      "loss": 0.126,
+      "step": 441
+    },
+    {
+      "epoch": 0.41965345359601236,
+      "grad_norm": 0.04984726384282112,
+      "learning_rate": 0.00019903226732232622,
+      "loss": 0.1326,
+      "step": 442
+    },
+    {
+      "epoch": 0.42060289579871823,
+      "grad_norm": 0.041585132479667664,
+      "learning_rate": 0.00019901687084926373,
+      "loss": 0.136,
+      "step": 443
+    },
+    {
+      "epoch": 0.42155233800142417,
+      "grad_norm": 0.05849035084247589,
+      "learning_rate": 0.0001990013534681617,
+      "loss": 0.1727,
+      "step": 444
+    },
+    {
+      "epoch": 0.4225017802041301,
+      "grad_norm": 0.043387994170188904,
+      "learning_rate": 0.00019898571519796817,
+      "loss": 0.1393,
+      "step": 445
+    },
+    {
+      "epoch": 0.42345122240683597,
+      "grad_norm": 0.05867496132850647,
+      "learning_rate": 0.0001989699560577788,
+      "loss": 0.1664,
+      "step": 446
+    },
+    {
+      "epoch": 0.4244006646095419,
+      "grad_norm": 0.07019232958555222,
+      "learning_rate": 0.00019895407606683685,
+      "loss": 0.1653,
+      "step": 447
+    },
+    {
+      "epoch": 0.42535010681224783,
+      "grad_norm": 0.04676515609025955,
+      "learning_rate": 0.00019893807524453314,
+      "loss": 0.1368,
+      "step": 448
+    },
+    {
+      "epoch": 0.4262995490149537,
+      "grad_norm": 0.06640240550041199,
+      "learning_rate": 0.00019892195361040607,
+      "loss": 0.2089,
+      "step": 449
+    },
+    {
+      "epoch": 0.42724899121765963,
+      "grad_norm": 0.044658735394477844,
+      "learning_rate": 0.00019890571118414148,
+      "loss": 0.1298,
+      "step": 450
+    },
+    {
+      "epoch": 0.4281984334203655,
+      "grad_norm": 0.04810122773051262,
+      "learning_rate": 0.00019888934798557278,
+      "loss": 0.1288,
+      "step": 451
+    },
+    {
+      "epoch": 0.42914787562307144,
+      "grad_norm": 0.0425436794757843,
+      "learning_rate": 0.0001988728640346808,
+      "loss": 0.1354,
+      "step": 452
+    },
+    {
+      "epoch": 0.43009731782577737,
+      "grad_norm": 0.04513363912701607,
+      "learning_rate": 0.0001988562593515939,
+      "loss": 0.1346,
+      "step": 453
+    },
+    {
+      "epoch": 0.43104676002848324,
+      "grad_norm": 0.052022870630025864,
+      "learning_rate": 0.0001988395339565878,
+      "loss": 0.1302,
+      "step": 454
+    },
+    {
+      "epoch": 0.4319962022311892,
+      "grad_norm": 0.04852641373872757,
+      "learning_rate": 0.0001988226878700856,
+      "loss": 0.1388,
+      "step": 455
+    },
+    {
+      "epoch": 0.4329456444338951,
+      "grad_norm": 0.04990584775805473,
+      "learning_rate": 0.00019880572111265785,
+      "loss": 0.1552,
+      "step": 456
+    },
+    {
+      "epoch": 0.433895086636601,
+      "grad_norm": 0.052271679043769836,
+      "learning_rate": 0.00019878863370502238,
+      "loss": 0.1404,
+      "step": 457
+    },
+    {
+      "epoch": 0.4348445288393069,
+      "grad_norm": 0.04795520752668381,
+      "learning_rate": 0.00019877142566804436,
+      "loss": 0.1341,
+      "step": 458
+    },
+    {
+      "epoch": 0.43579397104201284,
+      "grad_norm": 0.048165664076805115,
+      "learning_rate": 0.00019875409702273632,
+      "loss": 0.1343,
+      "step": 459
+    },
+    {
+      "epoch": 0.4367434132447187,
+      "grad_norm": 0.04213611036539078,
+      "learning_rate": 0.000198736647790258,
+      "loss": 0.1369,
+      "step": 460
+    },
+    {
+      "epoch": 0.43769285544742464,
+      "grad_norm": 0.05819966271519661,
+      "learning_rate": 0.00019871907799191632,
+      "loss": 0.1615,
+      "step": 461
+    },
+    {
+      "epoch": 0.4386422976501306,
+      "grad_norm": 0.057378821074962616,
+      "learning_rate": 0.00019870138764916558,
+      "loss": 0.175,
+      "step": 462
+    },
+    {
+      "epoch": 0.43959173985283645,
+      "grad_norm": 0.0432853177189827,
+      "learning_rate": 0.00019868357678360724,
+      "loss": 0.1371,
+      "step": 463
+    },
+    {
+      "epoch": 0.4405411820555424,
+      "grad_norm": 0.03890872746706009,
+      "learning_rate": 0.0001986656454169898,
+      "loss": 0.1332,
+      "step": 464
+    },
+    {
+      "epoch": 0.44149062425824825,
+      "grad_norm": 0.04006613418459892,
+      "learning_rate": 0.00019864759357120896,
+      "loss": 0.1342,
+      "step": 465
+    },
+    {
+      "epoch": 0.4424400664609542,
+      "grad_norm": 0.049053166061639786,
+      "learning_rate": 0.00019862942126830767,
+      "loss": 0.1756,
+      "step": 466
+    },
+    {
+      "epoch": 0.4433895086636601,
+      "grad_norm": 0.03966079652309418,
+      "learning_rate": 0.00019861112853047577,
+      "loss": 0.1303,
+      "step": 467
+    },
+    {
+      "epoch": 0.444338950866366,
+      "grad_norm": 0.04506433755159378,
+      "learning_rate": 0.0001985927153800503,
+      "loss": 0.136,
+      "step": 468
+    },
+    {
+      "epoch": 0.4452883930690719,
+      "grad_norm": 0.04392915591597557,
+      "learning_rate": 0.00019857418183951526,
+      "loss": 0.1397,
+      "step": 469
+    },
+    {
+      "epoch": 0.44623783527177785,
+      "grad_norm": 0.038007620722055435,
+      "learning_rate": 0.0001985555279315017,
+      "loss": 0.1246,
+      "step": 470
+    },
+    {
+      "epoch": 0.4471872774744837,
+      "grad_norm": 0.048948097974061966,
+      "learning_rate": 0.00019853675367878764,
+      "loss": 0.1329,
+      "step": 471
+    },
+    {
+      "epoch": 0.44813671967718965,
+      "grad_norm": 0.04174380376935005,
+      "learning_rate": 0.00019851785910429806,
+      "loss": 0.13,
+      "step": 472
+    },
+    {
+      "epoch": 0.4490861618798956,
+      "grad_norm": 0.048575468361377716,
+      "learning_rate": 0.00019849884423110478,
+      "loss": 0.1385,
+      "step": 473
+    },
+    {
+      "epoch": 0.45003560408260146,
+      "grad_norm": 0.05167670175433159,
+      "learning_rate": 0.00019847970908242664,
+      "loss": 0.1684,
+      "step": 474
+    },
+    {
+      "epoch": 0.4509850462853074,
+      "grad_norm": 0.06849198788404465,
+      "learning_rate": 0.00019846045368162923,
+      "loss": 0.1795,
+      "step": 475
+    },
+    {
+      "epoch": 0.4519344884880133,
+      "grad_norm": 0.044273603707551956,
+      "learning_rate": 0.0001984410780522251,
+      "loss": 0.1246,
+      "step": 476
+    },
+    {
+      "epoch": 0.4528839306907192,
+      "grad_norm": 0.048194363713264465,
+      "learning_rate": 0.00019842158221787353,
+      "loss": 0.1366,
+      "step": 477
+    },
+    {
+      "epoch": 0.4538333728934251,
+      "grad_norm": 0.033906418830156326,
+      "learning_rate": 0.00019840196620238057,
+      "loss": 0.1235,
+      "step": 478
+    },
+    {
+      "epoch": 0.454782815096131,
+      "grad_norm": 0.043933141976594925,
+      "learning_rate": 0.00019838223002969905,
+      "loss": 0.1195,
+      "step": 479
+    },
+    {
+      "epoch": 0.45573225729883693,
+      "grad_norm": 0.056823644787073135,
+      "learning_rate": 0.00019836237372392854,
+      "loss": 0.1757,
+      "step": 480
+    },
+    {
+      "epoch": 0.45668169950154286,
+      "grad_norm": 0.07587820291519165,
+      "learning_rate": 0.00019834239730931526,
+      "loss": 0.1784,
+      "step": 481
+    },
+    {
+      "epoch": 0.45763114170424873,
+      "grad_norm": 0.04008018970489502,
+      "learning_rate": 0.0001983223008102521,
+      "loss": 0.1306,
+      "step": 482
+    },
+    {
+      "epoch": 0.45858058390695466,
+      "grad_norm": 0.05180038511753082,
+      "learning_rate": 0.00019830208425127867,
+      "loss": 0.1485,
+      "step": 483
+    },
+    {
+      "epoch": 0.4595300261096606,
+      "grad_norm": 0.0691617876291275,
+      "learning_rate": 0.00019828174765708104,
+      "loss": 0.1249,
+      "step": 484
+    },
+    {
+      "epoch": 0.46047946831236647,
+      "grad_norm": 0.0565367266535759,
+      "learning_rate": 0.00019826129105249195,
+      "loss": 0.1744,
+      "step": 485
+    },
+    {
+      "epoch": 0.4614289105150724,
+      "grad_norm": 0.044927019625902176,
+      "learning_rate": 0.00019824071446249072,
+      "loss": 0.1341,
+      "step": 486
+    },
+    {
+      "epoch": 0.46237835271777833,
+      "grad_norm": 0.04481721669435501,
+      "learning_rate": 0.00019822001791220298,
+      "loss": 0.1354,
+      "step": 487
+    },
+    {
+      "epoch": 0.4633277949204842,
+      "grad_norm": 0.05233500525355339,
+      "learning_rate": 0.0001981992014269011,
+      "loss": 0.1501,
+      "step": 488
+    },
+    {
+      "epoch": 0.46427723712319013,
+      "grad_norm": 0.044350553303956985,
+      "learning_rate": 0.00019817826503200372,
+      "loss": 0.1335,
+      "step": 489
+    },
+    {
+      "epoch": 0.465226679325896,
+      "grad_norm": 0.03551819548010826,
+      "learning_rate": 0.000198157208753076,
+      "loss": 0.1322,
+      "step": 490
+    },
+    {
+      "epoch": 0.46617612152860194,
+      "grad_norm": 0.04409592226147652,
+      "learning_rate": 0.00019813603261582943,
+      "loss": 0.1561,
+      "step": 491
+    },
+    {
+      "epoch": 0.46712556373130787,
+      "grad_norm": 0.04842127487063408,
+      "learning_rate": 0.0001981147366461219,
+      "loss": 0.1296,
+      "step": 492
+    },
+    {
+      "epoch": 0.46807500593401374,
+      "grad_norm": 0.04349881038069725,
+      "learning_rate": 0.00019809332086995757,
+      "loss": 0.1319,
+      "step": 493
+    },
+    {
+      "epoch": 0.4690244481367197,
+      "grad_norm": 0.04413028433918953,
+      "learning_rate": 0.00019807178531348698,
+      "loss": 0.1321,
+      "step": 494
+    },
+    {
+      "epoch": 0.4699738903394256,
+      "grad_norm": 0.03972313553094864,
+      "learning_rate": 0.00019805013000300683,
+      "loss": 0.1358,
+      "step": 495
+    },
+    {
+      "epoch": 0.4709233325421315,
+      "grad_norm": 0.052269116044044495,
+      "learning_rate": 0.00019802835496496012,
+      "loss": 0.1389,
+      "step": 496
+    },
+    {
+      "epoch": 0.4718727747448374,
+      "grad_norm": 0.0379653237760067,
+      "learning_rate": 0.00019800646022593603,
+      "loss": 0.1283,
+      "step": 497
+    },
+    {
+      "epoch": 0.47282221694754334,
+      "grad_norm": 0.04370688647031784,
+      "learning_rate": 0.0001979844458126699,
+      "loss": 0.1278,
+      "step": 498
+    },
+    {
+      "epoch": 0.4737716591502492,
+      "grad_norm": 0.03912369906902313,
+      "learning_rate": 0.0001979623117520432,
+      "loss": 0.1257,
+      "step": 499
+    },
+    {
+      "epoch": 0.47472110135295514,
+      "grad_norm": 0.039594005793333054,
+      "learning_rate": 0.00019794005807108352,
+      "loss": 0.1375,
+      "step": 500
+    },
+    {
+      "epoch": 0.4756705435556611,
+      "grad_norm": 0.03889892250299454,
+      "learning_rate": 0.00019791768479696448,
+      "loss": 0.13,
+      "step": 501
+    },
+    {
+      "epoch": 0.47661998575836695,
+      "grad_norm": 0.03966660797595978,
+      "learning_rate": 0.00019789519195700578,
+      "loss": 0.1268,
+      "step": 502
+    },
+    {
+      "epoch": 0.4775694279610729,
+      "grad_norm": 0.04501716047525406,
+      "learning_rate": 0.00019787257957867306,
+      "loss": 0.1423,
+      "step": 503
+    },
+    {
+      "epoch": 0.47851887016377875,
+      "grad_norm": 0.06255436688661575,
+      "learning_rate": 0.000197849847689578,
+      "loss": 0.1799,
+      "step": 504
+    },
+    {
+      "epoch": 0.4794683123664847,
+      "grad_norm": 0.050308458507061005,
+      "learning_rate": 0.00019782699631747813,
+      "loss": 0.1733,
+      "step": 505
+    },
+    {
+      "epoch": 0.4804177545691906,
+      "grad_norm": 0.0357963964343071,
+      "learning_rate": 0.00019780402549027698,
+      "loss": 0.1268,
+      "step": 506
+    },
+    {
+      "epoch": 0.4813671967718965,
+      "grad_norm": 0.03651968017220497,
+      "learning_rate": 0.00019778093523602384,
+      "loss": 0.1267,
+      "step": 507
+    },
+    {
+      "epoch": 0.4823166389746024,
+      "grad_norm": 0.043042074888944626,
+      "learning_rate": 0.0001977577255829139,
+      "loss": 0.1256,
+      "step": 508
+    },
+    {
+      "epoch": 0.48326608117730835,
+      "grad_norm": 0.07031014561653137,
+      "learning_rate": 0.00019773439655928815,
+      "loss": 0.1796,
+      "step": 509
+    },
+    {
+      "epoch": 0.4842155233800142,
+      "grad_norm": 0.04429268836975098,
+      "learning_rate": 0.00019771094819363326,
+      "loss": 0.1298,
+      "step": 510
+    },
+    {
+      "epoch": 0.48516496558272015,
+      "grad_norm": 0.0373898483812809,
+      "learning_rate": 0.00019768738051458172,
+      "loss": 0.1232,
+      "step": 511
+    },
+    {
+      "epoch": 0.4861144077854261,
+      "grad_norm": 0.05853155627846718,
+      "learning_rate": 0.00019766369355091166,
+      "loss": 0.1694,
+      "step": 512
+    },
+    {
+      "epoch": 0.48706384998813196,
+      "grad_norm": 0.05050895735621452,
+      "learning_rate": 0.00019763988733154686,
+      "loss": 0.1665,
+      "step": 513
+    },
+    {
+      "epoch": 0.4880132921908379,
+      "grad_norm": 0.04074448347091675,
+      "learning_rate": 0.0001976159618855568,
+      "loss": 0.1336,
+      "step": 514
+    },
+    {
+      "epoch": 0.4889627343935438,
+      "grad_norm": 0.03826110064983368,
+      "learning_rate": 0.00019759191724215644,
+      "loss": 0.132,
+      "step": 515
+    },
+    {
+      "epoch": 0.4899121765962497,
+      "grad_norm": 0.04392875358462334,
+      "learning_rate": 0.0001975677534307064,
+      "loss": 0.1204,
+      "step": 516
+    },
+    {
+      "epoch": 0.4908616187989556,
+      "grad_norm": 0.04615531116724014,
+      "learning_rate": 0.0001975434704807127,
+      "loss": 0.1358,
+      "step": 517
+    },
+    {
+      "epoch": 0.4918110610016615,
+      "grad_norm": 0.053060565143823624,
+      "learning_rate": 0.00019751906842182688,
+      "loss": 0.1299,
+      "step": 518
+    },
+    {
+      "epoch": 0.49276050320436743,
+      "grad_norm": 0.04905511438846588,
+      "learning_rate": 0.00019749454728384594,
+      "loss": 0.1284,
+      "step": 519
+    },
+    {
+      "epoch": 0.49370994540707336,
+      "grad_norm": 0.04257996007800102,
+      "learning_rate": 0.00019746990709671234,
+      "loss": 0.1353,
+      "step": 520
+    },
+    {
+      "epoch": 0.49465938760977923,
+      "grad_norm": 0.05581909418106079,
+      "learning_rate": 0.0001974451478905138,
+      "loss": 0.1594,
+      "step": 521
+    },
+    {
+      "epoch": 0.49560882981248516,
+      "grad_norm": 0.04603990167379379,
+      "learning_rate": 0.00019742026969548338,
+      "loss": 0.1383,
+      "step": 522
+    },
+    {
+      "epoch": 0.4965582720151911,
+      "grad_norm": 0.058511972427368164,
+      "learning_rate": 0.00019739527254199958,
+      "loss": 0.1725,
+      "step": 523
+    },
+    {
+      "epoch": 0.49750771421789697,
+      "grad_norm": 0.03875808045268059,
+      "learning_rate": 0.000197370156460586,
+      "loss": 0.1405,
+      "step": 524
+    },
+    {
+      "epoch": 0.4984571564206029,
+      "grad_norm": 0.040860000997781754,
+      "learning_rate": 0.00019734492148191151,
+      "loss": 0.139,
+      "step": 525
+    },
+    {
+      "epoch": 0.49940659862330883,
+      "grad_norm": 0.06110459193587303,
+      "learning_rate": 0.00019731956763679014,
+      "loss": 0.223,
+      "step": 526
+    },
+    {
+      "epoch": 0.5003560408260147,
+      "grad_norm": 0.05238598585128784,
+      "learning_rate": 0.00019729409495618117,
+      "loss": 0.1681,
+      "step": 527
+    },
+    {
+      "epoch": 0.5013054830287206,
+      "grad_norm": 0.05180145800113678,
+      "learning_rate": 0.00019726850347118885,
+      "loss": 0.1743,
+      "step": 528
+    },
+    {
+      "epoch": 0.5022549252314266,
+      "grad_norm": 0.05066410079598427,
+      "learning_rate": 0.00019724279321306262,
+      "loss": 0.1634,
+      "step": 529
+    },
+    {
+      "epoch": 0.5032043674341324,
+      "grad_norm": 0.06856084614992142,
+      "learning_rate": 0.00019721696421319684,
+      "loss": 0.1685,
+      "step": 530
+    },
+    {
+      "epoch": 0.5041538096368383,
+      "grad_norm": 0.045972324907779694,
+      "learning_rate": 0.00019719101650313096,
+      "loss": 0.1245,
+      "step": 531
+    },
+    {
+      "epoch": 0.5051032518395443,
+      "grad_norm": 0.04522623121738434,
+      "learning_rate": 0.00019716495011454934,
+      "loss": 0.1367,
+      "step": 532
+    },
+    {
+      "epoch": 0.5060526940422502,
+      "grad_norm": 0.0780516117811203,
+      "learning_rate": 0.00019713876507928126,
+      "loss": 0.1351,
+      "step": 533
+    },
+    {
+      "epoch": 0.507002136244956,
+      "grad_norm": 0.04264210909605026,
+      "learning_rate": 0.00019711246142930088,
+      "loss": 0.1312,
+      "step": 534
+    },
+    {
+      "epoch": 0.507951578447662,
+      "grad_norm": 0.059501100331544876,
+      "learning_rate": 0.00019708603919672718,
+      "loss": 0.1698,
+      "step": 535
+    },
+    {
+      "epoch": 0.5089010206503679,
+      "grad_norm": 0.060105033218860626,
+      "learning_rate": 0.00019705949841382396,
+      "loss": 0.1303,
+      "step": 536
+    },
+    {
+      "epoch": 0.5098504628530738,
+      "grad_norm": 0.04733967408537865,
+      "learning_rate": 0.00019703283911299982,
+      "loss": 0.1245,
+      "step": 537
+    },
+    {
+      "epoch": 0.5107999050557798,
+      "grad_norm": 0.04254663735628128,
+      "learning_rate": 0.00019700606132680798,
+      "loss": 0.1343,
+      "step": 538
+    },
+    {
+      "epoch": 0.5117493472584856,
+      "grad_norm": 0.06302463263273239,
+      "learning_rate": 0.00019697916508794645,
+      "loss": 0.1831,
+      "step": 539
+    },
+    {
+      "epoch": 0.5126987894611915,
+      "grad_norm": 0.05301344394683838,
+      "learning_rate": 0.0001969521504292578,
+      "loss": 0.1316,
+      "step": 540
+    },
+    {
+      "epoch": 0.5136482316638975,
+      "grad_norm": 0.04151083528995514,
+      "learning_rate": 0.00019692501738372922,
+      "loss": 0.1335,
+      "step": 541
+    },
+    {
+      "epoch": 0.5145976738666034,
+      "grad_norm": 0.05647062510251999,
+      "learning_rate": 0.00019689776598449257,
+      "loss": 0.1688,
+      "step": 542
+    },
+    {
+      "epoch": 0.5155471160693093,
+      "grad_norm": 0.037060294300317764,
+      "learning_rate": 0.000196870396264824,
+      "loss": 0.1339,
+      "step": 543
+    },
+    {
+      "epoch": 0.5164965582720152,
+      "grad_norm": 0.04036247730255127,
+      "learning_rate": 0.0001968429082581443,
+      "loss": 0.1361,
+      "step": 544
+    },
+    {
+      "epoch": 0.5174460004747211,
+      "grad_norm": 0.040889665484428406,
+      "learning_rate": 0.00019681530199801875,
+      "loss": 0.1356,
+      "step": 545
+    },
+    {
+      "epoch": 0.518395442677427,
+      "grad_norm": 0.0538480207324028,
+      "learning_rate": 0.00019678757751815686,
+      "loss": 0.1689,
+      "step": 546
+    },
+    {
+      "epoch": 0.519344884880133,
+      "grad_norm": 0.04074794426560402,
+      "learning_rate": 0.0001967597348524126,
+      "loss": 0.1329,
+      "step": 547
+    },
+    {
+      "epoch": 0.5202943270828388,
+      "grad_norm": 0.03896891698241234,
+      "learning_rate": 0.00019673177403478428,
+      "loss": 0.1356,
+      "step": 548
+    },
+    {
+      "epoch": 0.5212437692855447,
+      "grad_norm": 0.04619259387254715,
+      "learning_rate": 0.00019670369509941442,
+      "loss": 0.163,
+      "step": 549
+    },
+    {
+      "epoch": 0.5221932114882507,
+      "grad_norm": 0.035968657582998276,
+      "learning_rate": 0.00019667549808058976,
+      "loss": 0.1242,
+      "step": 550
+    },
+    {
+      "epoch": 0.5231426536909566,
+      "grad_norm": 0.04564007744193077,
+      "learning_rate": 0.0001966471830127413,
+      "loss": 0.1364,
+      "step": 551
+    },
+    {
+      "epoch": 0.5240920958936625,
+      "grad_norm": 0.03991610184311867,
+      "learning_rate": 0.00019661874993044415,
+      "loss": 0.1312,
+      "step": 552
+    },
+    {
+      "epoch": 0.5250415380963683,
+      "grad_norm": 0.037240512669086456,
+      "learning_rate": 0.00019659019886841752,
+      "loss": 0.1279,
+      "step": 553
+    },
+    {
+      "epoch": 0.5259909802990743,
+      "grad_norm": 0.06598762422800064,
+      "learning_rate": 0.00019656152986152468,
+      "loss": 0.2165,
+      "step": 554
+    },
+    {
+      "epoch": 0.5269404225017802,
+      "grad_norm": 0.03867746889591217,
+      "learning_rate": 0.00019653274294477292,
+      "loss": 0.1233,
+      "step": 555
+    },
+    {
+      "epoch": 0.5278898647044861,
+      "grad_norm": 0.051915477961301804,
+      "learning_rate": 0.00019650383815331357,
+      "loss": 0.168,
+      "step": 556
+    },
+    {
+      "epoch": 0.528839306907192,
+      "grad_norm": 0.054896485060453415,
+      "learning_rate": 0.00019647481552244182,
+      "loss": 0.1678,
+      "step": 557
+    },
+    {
+      "epoch": 0.5297887491098979,
+      "grad_norm": 0.05439051240682602,
+      "learning_rate": 0.00019644567508759675,
+      "loss": 0.1607,
+      "step": 558
+    },
+    {
+      "epoch": 0.5307381913126038,
+      "grad_norm": 0.03601578250527382,
+      "learning_rate": 0.00019641641688436135,
+      "loss": 0.1271,
+      "step": 559
+    },
+    {
+      "epoch": 0.5316876335153098,
+      "grad_norm": 0.06025104597210884,
+      "learning_rate": 0.00019638704094846236,
+      "loss": 0.176,
+      "step": 560
+    },
+    {
+      "epoch": 0.5326370757180157,
+      "grad_norm": 0.04126368835568428,
+      "learning_rate": 0.00019635754731577032,
+      "loss": 0.1319,
+      "step": 561
+    },
+    {
+      "epoch": 0.5335865179207215,
+      "grad_norm": 0.05305393040180206,
+      "learning_rate": 0.00019632793602229943,
+      "loss": 0.1699,
+      "step": 562
+    },
+    {
+      "epoch": 0.5345359601234275,
+      "grad_norm": 0.03538331016898155,
+      "learning_rate": 0.00019629820710420764,
+      "loss": 0.124,
+      "step": 563
+    },
+    {
+      "epoch": 0.5354854023261334,
+      "grad_norm": 0.05861300975084305,
+      "learning_rate": 0.0001962683605977965,
+      "loss": 0.1688,
+      "step": 564
+    },
+    {
+      "epoch": 0.5364348445288393,
+      "grad_norm": 0.040226079523563385,
+      "learning_rate": 0.0001962383965395111,
+      "loss": 0.1334,
+      "step": 565
+    },
+    {
+      "epoch": 0.5373842867315453,
+      "grad_norm": 0.035788875073194504,
+      "learning_rate": 0.00019620831496594017,
+      "loss": 0.1281,
+      "step": 566
+    },
+    {
+      "epoch": 0.5383337289342511,
+      "grad_norm": 0.0334162712097168,
+      "learning_rate": 0.0001961781159138158,
+      "loss": 0.1317,
+      "step": 567
+    },
+    {
+      "epoch": 0.539283171136957,
+      "grad_norm": 0.03352081775665283,
+      "learning_rate": 0.00019614779942001364,
+      "loss": 0.1334,
+      "step": 568
+    },
+    {
+      "epoch": 0.540232613339663,
+      "grad_norm": 0.03684060648083687,
+      "learning_rate": 0.00019611736552155274,
+      "loss": 0.1349,
+      "step": 569
+    },
+    {
+      "epoch": 0.5411820555423689,
+      "grad_norm": 0.03640671446919441,
+      "learning_rate": 0.00019608681425559542,
+      "loss": 0.1278,
+      "step": 570
+    },
+    {
+      "epoch": 0.5421314977450747,
+      "grad_norm": 0.04167250171303749,
+      "learning_rate": 0.00019605614565944748,
+      "loss": 0.1384,
+      "step": 571
+    },
+    {
+      "epoch": 0.5430809399477807,
+      "grad_norm": 0.0416824147105217,
+      "learning_rate": 0.00019602535977055778,
+      "loss": 0.1319,
+      "step": 572
+    },
+    {
+      "epoch": 0.5440303821504866,
+      "grad_norm": 0.03897137567400932,
+      "learning_rate": 0.00019599445662651861,
+      "loss": 0.1389,
+      "step": 573
+    },
+    {
+      "epoch": 0.5449798243531925,
+      "grad_norm": 0.03894896060228348,
+      "learning_rate": 0.00019596343626506526,
+      "loss": 0.1341,
+      "step": 574
+    },
+    {
+      "epoch": 0.5459292665558985,
+      "grad_norm": 0.04211690276861191,
+      "learning_rate": 0.00019593229872407627,
+      "loss": 0.1377,
+      "step": 575
+    },
+    {
+      "epoch": 0.5468787087586043,
+      "grad_norm": 0.04308454692363739,
+      "learning_rate": 0.00019590104404157327,
+      "loss": 0.1268,
+      "step": 576
+    },
+    {
+      "epoch": 0.5478281509613102,
+      "grad_norm": 0.0525001622736454,
+      "learning_rate": 0.00019586967225572086,
+      "loss": 0.1775,
+      "step": 577
+    },
+    {
+      "epoch": 0.5487775931640161,
+      "grad_norm": 0.056315965950489044,
+      "learning_rate": 0.00019583818340482664,
+      "loss": 0.1688,
+      "step": 578
+    },
+    {
+      "epoch": 0.5497270353667221,
+      "grad_norm": 0.03801283985376358,
+      "learning_rate": 0.0001958065775273412,
+      "loss": 0.1309,
+      "step": 579
+    },
+    {
+      "epoch": 0.550676477569428,
+      "grad_norm": 0.03738854080438614,
+      "learning_rate": 0.00019577485466185804,
+      "loss": 0.137,
+      "step": 580
+    },
+    {
+      "epoch": 0.5516259197721338,
+      "grad_norm": 0.03772661089897156,
+      "learning_rate": 0.0001957430148471134,
+      "loss": 0.1276,
+      "step": 581
+    },
+    {
+      "epoch": 0.5525753619748398,
+      "grad_norm": 0.039842378348112106,
+      "learning_rate": 0.00019571105812198652,
+      "loss": 0.1329,
+      "step": 582
+    },
+    {
+      "epoch": 0.5535248041775457,
+      "grad_norm": 0.033689334988594055,
+      "learning_rate": 0.0001956789845254992,
+      "loss": 0.1265,
+      "step": 583
+    },
+    {
+      "epoch": 0.5544742463802516,
+      "grad_norm": 0.046588387340307236,
+      "learning_rate": 0.00019564679409681608,
+      "loss": 0.1645,
+      "step": 584
+    },
+    {
+      "epoch": 0.5554236885829575,
+      "grad_norm": 0.03861064463853836,
+      "learning_rate": 0.0001956144868752444,
+      "loss": 0.1267,
+      "step": 585
+    },
+    {
+      "epoch": 0.5563731307856634,
+      "grad_norm": 0.03467525169253349,
+      "learning_rate": 0.000195582062900234,
+      "loss": 0.1299,
+      "step": 586
+    },
+    {
+      "epoch": 0.5573225729883693,
+      "grad_norm": 0.03659389913082123,
+      "learning_rate": 0.0001955495222113774,
+      "loss": 0.1286,
+      "step": 587
+    },
+    {
+      "epoch": 0.5582720151910753,
+      "grad_norm": 0.03826770931482315,
+      "learning_rate": 0.0001955168648484095,
+      "loss": 0.1313,
+      "step": 588
+    },
+    {
+      "epoch": 0.5592214573937812,
+      "grad_norm": 0.038110729306936264,
+      "learning_rate": 0.00019548409085120772,
+      "loss": 0.137,
+      "step": 589
+    },
+    {
+      "epoch": 0.560170899596487,
+      "grad_norm": 0.03989555314183235,
+      "learning_rate": 0.0001954512002597919,
+      "loss": 0.132,
+      "step": 590
+    },
+    {
+      "epoch": 0.561120341799193,
+      "grad_norm": 0.05395180359482765,
+      "learning_rate": 0.00019541819311432427,
+      "loss": 0.1401,
+      "step": 591
+    },
+    {
+      "epoch": 0.5620697840018989,
+      "grad_norm": 0.05007918179035187,
+      "learning_rate": 0.00019538506945510938,
+      "loss": 0.1584,
+      "step": 592
+    },
+    {
+      "epoch": 0.5630192262046048,
+      "grad_norm": 0.047849785536527634,
+      "learning_rate": 0.00019535182932259404,
+      "loss": 0.1265,
+      "step": 593
+    },
+    {
+      "epoch": 0.5639686684073107,
+      "grad_norm": 0.04303041473031044,
+      "learning_rate": 0.00019531847275736726,
+      "loss": 0.1245,
+      "step": 594
+    },
+    {
+      "epoch": 0.5649181106100166,
+      "grad_norm": 0.04128289222717285,
+      "learning_rate": 0.00019528499980016025,
+      "loss": 0.1317,
+      "step": 595
+    },
+    {
+      "epoch": 0.5658675528127225,
+      "grad_norm": 0.04311414808034897,
+      "learning_rate": 0.00019525141049184637,
+      "loss": 0.1364,
+      "step": 596
+    },
+    {
+      "epoch": 0.5668169950154285,
+      "grad_norm": 0.03765838220715523,
+      "learning_rate": 0.00019521770487344103,
+      "loss": 0.1268,
+      "step": 597
+    },
+    {
+      "epoch": 0.5677664372181344,
+      "grad_norm": 0.03674585744738579,
+      "learning_rate": 0.00019518388298610164,
+      "loss": 0.1297,
+      "step": 598
+    },
+    {
+      "epoch": 0.5687158794208402,
+      "grad_norm": 0.036937762051820755,
+      "learning_rate": 0.0001951499448711276,
+      "loss": 0.1303,
+      "step": 599
+    },
+    {
+      "epoch": 0.5696653216235462,
+      "grad_norm": 0.03748161345720291,
+      "learning_rate": 0.0001951158905699603,
+      "loss": 0.1328,
+      "step": 600
+    },
+    {
+      "epoch": 0.5706147638262521,
+      "grad_norm": 0.04011257737874985,
+      "learning_rate": 0.00019508172012418283,
+      "loss": 0.1346,
+      "step": 601
+    },
+    {
+      "epoch": 0.571564206028958,
+      "grad_norm": 0.03853931650519371,
+      "learning_rate": 0.00019504743357552035,
+      "loss": 0.1279,
+      "step": 602
+    },
+    {
+      "epoch": 0.572513648231664,
+      "grad_norm": 0.03750459849834442,
+      "learning_rate": 0.0001950130309658396,
+      "loss": 0.1227,
+      "step": 603
+    },
+    {
+      "epoch": 0.5734630904343698,
+      "grad_norm": 0.05542079731822014,
+      "learning_rate": 0.00019497851233714908,
+      "loss": 0.1647,
+      "step": 604
+    },
+    {
+      "epoch": 0.5744125326370757,
+      "grad_norm": 0.04472218081355095,
+      "learning_rate": 0.00019494387773159898,
+      "loss": 0.1416,
+      "step": 605
+    },
+    {
+      "epoch": 0.5753619748397816,
+      "grad_norm": 0.052323974668979645,
+      "learning_rate": 0.00019490912719148114,
+      "loss": 0.1367,
+      "step": 606
+    },
+    {
+      "epoch": 0.5763114170424876,
+      "grad_norm": 0.037580832839012146,
+      "learning_rate": 0.00019487426075922893,
+      "loss": 0.131,
+      "step": 607
+    },
+    {
+      "epoch": 0.5772608592451934,
+      "grad_norm": 0.03929577395319939,
+      "learning_rate": 0.0001948392784774172,
+      "loss": 0.128,
+      "step": 608
+    },
+    {
+      "epoch": 0.5782103014478993,
+      "grad_norm": 0.03706606104969978,
+      "learning_rate": 0.0001948041803887623,
+      "loss": 0.1316,
+      "step": 609
+    },
+    {
+      "epoch": 0.5791597436506053,
+      "grad_norm": 0.038938358426094055,
+      "learning_rate": 0.00019476896653612203,
+      "loss": 0.1275,
+      "step": 610
+    },
+    {
+      "epoch": 0.5801091858533112,
+      "grad_norm": 0.04818068817257881,
+      "learning_rate": 0.00019473363696249546,
+      "loss": 0.1662,
+      "step": 611
+    },
+    {
+      "epoch": 0.581058628056017,
+      "grad_norm": 0.03735940158367157,
+      "learning_rate": 0.00019469819171102304,
+      "loss": 0.1361,
+      "step": 612
+    },
+    {
+      "epoch": 0.582008070258723,
+      "grad_norm": 0.03568827733397484,
+      "learning_rate": 0.00019466263082498645,
+      "loss": 0.1216,
+      "step": 613
+    },
+    {
+      "epoch": 0.5829575124614289,
+      "grad_norm": 0.03913251310586929,
+      "learning_rate": 0.0001946269543478085,
+      "loss": 0.1321,
+      "step": 614
+    },
+    {
+      "epoch": 0.5839069546641348,
+      "grad_norm": 0.062009479850530624,
+      "learning_rate": 0.0001945911623230533,
+      "loss": 0.1778,
+      "step": 615
+    },
+    {
+      "epoch": 0.5848563968668408,
+      "grad_norm": 0.039088111370801926,
+      "learning_rate": 0.0001945552547944259,
+      "loss": 0.1352,
+      "step": 616
+    },
+    {
+      "epoch": 0.5858058390695466,
+      "grad_norm": 0.041976600885391235,
+      "learning_rate": 0.0001945192318057725,
+      "loss": 0.1394,
+      "step": 617
+    },
+    {
+      "epoch": 0.5867552812722525,
+      "grad_norm": 0.03723563253879547,
+      "learning_rate": 0.00019448309340108018,
+      "loss": 0.1246,
+      "step": 618
+    },
+    {
+      "epoch": 0.5877047234749585,
+      "grad_norm": 0.0382399819791317,
+      "learning_rate": 0.00019444683962447707,
+      "loss": 0.1232,
+      "step": 619
+    },
+    {
+      "epoch": 0.5886541656776644,
+      "grad_norm": 0.03758077695965767,
+      "learning_rate": 0.0001944104705202321,
+      "loss": 0.1417,
+      "step": 620
+    },
+    {
+      "epoch": 0.5896036078803703,
+      "grad_norm": 0.034823786467313766,
+      "learning_rate": 0.000194373986132755,
+      "loss": 0.1304,
+      "step": 621
+    },
+    {
+      "epoch": 0.5905530500830762,
+      "grad_norm": 0.03755120187997818,
+      "learning_rate": 0.00019433738650659641,
+      "loss": 0.133,
+      "step": 622
+    },
+    {
+      "epoch": 0.5915024922857821,
+      "grad_norm": 0.03759913146495819,
+      "learning_rate": 0.00019430067168644754,
+      "loss": 0.1222,
+      "step": 623
+    },
+    {
+      "epoch": 0.592451934488488,
+      "grad_norm": 0.06232694163918495,
+      "learning_rate": 0.0001942638417171403,
+      "loss": 0.1778,
+      "step": 624
+    },
+    {
+      "epoch": 0.593401376691194,
+      "grad_norm": 0.05642306059598923,
+      "learning_rate": 0.00019422689664364725,
+      "loss": 0.1706,
+      "step": 625
+    },
+    {
+      "epoch": 0.5943508188938998,
+      "grad_norm": 0.0827709287405014,
+      "learning_rate": 0.00019418983651108148,
+      "loss": 0.2371,
+      "step": 626
+    },
+    {
+      "epoch": 0.5953002610966057,
+      "grad_norm": 0.03614366054534912,
+      "learning_rate": 0.00019415266136469652,
+      "loss": 0.1225,
+      "step": 627
+    },
+    {
+      "epoch": 0.5962497032993117,
+      "grad_norm": 0.042416494339704514,
+      "learning_rate": 0.00019411537124988643,
+      "loss": 0.1239,
+      "step": 628
+    },
+    {
+      "epoch": 0.5971991455020176,
+      "grad_norm": 0.037246908992528915,
+      "learning_rate": 0.00019407796621218566,
+      "loss": 0.1292,
+      "step": 629
+    },
+    {
+      "epoch": 0.5981485877047235,
+      "grad_norm": 0.05374092981219292,
+      "learning_rate": 0.00019404044629726887,
+      "loss": 0.1782,
+      "step": 630
+    },
+    {
+      "epoch": 0.5990980299074293,
+      "grad_norm": 0.052854426205158234,
+      "learning_rate": 0.00019400281155095112,
+      "loss": 0.1711,
+      "step": 631
+    },
+    {
+      "epoch": 0.6000474721101353,
+      "grad_norm": 0.038800131529569626,
+      "learning_rate": 0.00019396506201918765,
+      "loss": 0.1285,
+      "step": 632
+    },
+    {
+      "epoch": 0.6009969143128412,
+      "grad_norm": 0.040118250995874405,
+      "learning_rate": 0.0001939271977480738,
+      "loss": 0.1335,
+      "step": 633
+    },
+    {
+      "epoch": 0.6019463565155471,
+      "grad_norm": 0.07007341086864471,
+      "learning_rate": 0.00019388921878384517,
+      "loss": 0.2115,
+      "step": 634
+    },
+    {
+      "epoch": 0.602895798718253,
+      "grad_norm": 0.03245210647583008,
+      "learning_rate": 0.0001938511251728772,
+      "loss": 0.1304,
+      "step": 635
+    },
+    {
+      "epoch": 0.6038452409209589,
+      "grad_norm": 0.03384733200073242,
+      "learning_rate": 0.00019381291696168553,
+      "loss": 0.1297,
+      "step": 636
+    },
+    {
+      "epoch": 0.6047946831236648,
+      "grad_norm": 0.04325825348496437,
+      "learning_rate": 0.0001937745941969256,
+      "loss": 0.1337,
+      "step": 637
+    },
+    {
+      "epoch": 0.6057441253263708,
+      "grad_norm": 0.046986173838377,
+      "learning_rate": 0.00019373615692539275,
+      "loss": 0.1385,
+      "step": 638
+    },
+    {
+      "epoch": 0.6066935675290767,
+      "grad_norm": 0.03726234659552574,
+      "learning_rate": 0.0001936976051940222,
+      "loss": 0.1393,
+      "step": 639
+    },
+    {
+      "epoch": 0.6076430097317825,
+      "grad_norm": 0.05574486404657364,
+      "learning_rate": 0.0001936589390498889,
+      "loss": 0.1698,
+      "step": 640
+    },
+    {
+      "epoch": 0.6085924519344885,
+      "grad_norm": 0.052818477153778076,
+      "learning_rate": 0.0001936201585402075,
+      "loss": 0.1722,
+      "step": 641
+    },
+    {
+      "epoch": 0.6095418941371944,
+      "grad_norm": 0.03535636141896248,
+      "learning_rate": 0.00019358126371233231,
+      "loss": 0.129,
+      "step": 642
+    },
+    {
+      "epoch": 0.6104913363399003,
+      "grad_norm": 0.03453061729669571,
+      "learning_rate": 0.00019354225461375724,
+      "loss": 0.1313,
+      "step": 643
+    },
+    {
+      "epoch": 0.6114407785426063,
+      "grad_norm": 0.030467770993709564,
+      "learning_rate": 0.0001935031312921157,
+      "loss": 0.1241,
+      "step": 644
+    },
+    {
+      "epoch": 0.6123902207453121,
+      "grad_norm": 0.03996508568525314,
+      "learning_rate": 0.0001934638937951806,
+      "loss": 0.1265,
+      "step": 645
+    },
+    {
+      "epoch": 0.613339662948018,
+      "grad_norm": 0.034416794776916504,
+      "learning_rate": 0.00019342454217086429,
+      "loss": 0.1341,
+      "step": 646
+    },
+    {
+      "epoch": 0.614289105150724,
+      "grad_norm": 0.03674698621034622,
+      "learning_rate": 0.00019338507646721845,
+      "loss": 0.1399,
+      "step": 647
+    },
+    {
+      "epoch": 0.6152385473534299,
+      "grad_norm": 0.037850040942430496,
+      "learning_rate": 0.0001933454967324341,
+      "loss": 0.1295,
+      "step": 648
+    },
+    {
+      "epoch": 0.6161879895561357,
+      "grad_norm": 0.037829235196113586,
+      "learning_rate": 0.0001933058030148414,
+      "loss": 0.1302,
+      "step": 649
+    },
+    {
+      "epoch": 0.6171374317588417,
+      "grad_norm": 0.03579702973365784,
+      "learning_rate": 0.00019326599536290983,
+      "loss": 0.1352,
+      "step": 650
+    },
+    {
+      "epoch": 0.6180868739615476,
+      "grad_norm": 0.052539851516485214,
+      "learning_rate": 0.00019322607382524785,
+      "loss": 0.1744,
+      "step": 651
+    },
+    {
+      "epoch": 0.6190363161642535,
+      "grad_norm": 0.03814668953418732,
+      "learning_rate": 0.0001931860384506031,
+      "loss": 0.1303,
+      "step": 652
+    },
+    {
+      "epoch": 0.6199857583669595,
+      "grad_norm": 0.03730069473385811,
+      "learning_rate": 0.00019314588928786224,
+      "loss": 0.1236,
+      "step": 653
+    },
+    {
+      "epoch": 0.6209352005696653,
+      "grad_norm": 0.04081875458359718,
+      "learning_rate": 0.00019310562638605078,
+      "loss": 0.1328,
+      "step": 654
+    },
+    {
+      "epoch": 0.6218846427723712,
+      "grad_norm": 0.03532617911696434,
+      "learning_rate": 0.00019306524979433308,
+      "loss": 0.1238,
+      "step": 655
+    },
+    {
+      "epoch": 0.6228340849750772,
+      "grad_norm": 0.035857025533914566,
+      "learning_rate": 0.00019302475956201254,
+      "loss": 0.1244,
+      "step": 656
+    },
+    {
+      "epoch": 0.6237835271777831,
+      "grad_norm": 0.036031339317560196,
+      "learning_rate": 0.0001929841557385311,
+      "loss": 0.1251,
+      "step": 657
+    },
+    {
+      "epoch": 0.624732969380489,
+      "grad_norm": 0.037832874804735184,
+      "learning_rate": 0.00019294343837346944,
+      "loss": 0.1262,
+      "step": 658
+    },
+    {
+      "epoch": 0.6256824115831948,
+      "grad_norm": 0.03651989623904228,
+      "learning_rate": 0.00019290260751654706,
+      "loss": 0.1239,
+      "step": 659
+    },
+    {
+      "epoch": 0.6266318537859008,
+      "grad_norm": 0.03595907241106033,
+      "learning_rate": 0.00019286166321762184,
+      "loss": 0.1342,
+      "step": 660
+    },
+    {
+      "epoch": 0.6275812959886067,
+      "grad_norm": 0.04714696854352951,
+      "learning_rate": 0.00019282060552669025,
+      "loss": 0.1712,
+      "step": 661
+    },
+    {
+      "epoch": 0.6285307381913126,
+      "grad_norm": 0.0448799654841423,
+      "learning_rate": 0.00019277943449388726,
+      "loss": 0.1601,
+      "step": 662
+    },
+    {
+      "epoch": 0.6294801803940185,
+      "grad_norm": 0.03133920207619667,
+      "learning_rate": 0.0001927381501694862,
+      "loss": 0.127,
+      "step": 663
+    },
+    {
+      "epoch": 0.6304296225967244,
+      "grad_norm": 0.051593225449323654,
+      "learning_rate": 0.00019269675260389876,
+      "loss": 0.1659,
+      "step": 664
+    },
+    {
+      "epoch": 0.6313790647994303,
+      "grad_norm": 0.03713349625468254,
+      "learning_rate": 0.0001926552418476749,
+      "loss": 0.1294,
+      "step": 665
+    },
+    {
+      "epoch": 0.6323285070021363,
+      "grad_norm": 0.03420734405517578,
+      "learning_rate": 0.00019261361795150275,
+      "loss": 0.1376,
+      "step": 666
+    },
+    {
+      "epoch": 0.6332779492048421,
+      "grad_norm": 0.04476429522037506,
+      "learning_rate": 0.00019257188096620867,
+      "loss": 0.1595,
+      "step": 667
+    },
+    {
+      "epoch": 0.634227391407548,
+      "grad_norm": 0.05289504677057266,
+      "learning_rate": 0.00019253003094275707,
+      "loss": 0.1589,
+      "step": 668
+    },
+    {
+      "epoch": 0.635176833610254,
+      "grad_norm": 0.042022526264190674,
+      "learning_rate": 0.0001924880679322504,
+      "loss": 0.1316,
+      "step": 669
+    },
+    {
+      "epoch": 0.6361262758129599,
+      "grad_norm": 0.0408223457634449,
+      "learning_rate": 0.00019244599198592907,
+      "loss": 0.1386,
+      "step": 670
+    },
+    {
+      "epoch": 0.6370757180156658,
+      "grad_norm": 0.03941584751009941,
+      "learning_rate": 0.00019240380315517142,
+      "loss": 0.1325,
+      "step": 671
+    },
+    {
+      "epoch": 0.6380251602183717,
+      "grad_norm": 0.03860325738787651,
+      "learning_rate": 0.00019236150149149357,
+      "loss": 0.1215,
+      "step": 672
+    },
+    {
+      "epoch": 0.6389746024210776,
+      "grad_norm": 0.0342581607401371,
+      "learning_rate": 0.00019231908704654948,
+      "loss": 0.1247,
+      "step": 673
+    },
+    {
+      "epoch": 0.6399240446237835,
+      "grad_norm": 0.04099750518798828,
+      "learning_rate": 0.00019227655987213077,
+      "loss": 0.1335,
+      "step": 674
+    },
+    {
+      "epoch": 0.6408734868264895,
+      "grad_norm": 0.031005796045064926,
+      "learning_rate": 0.00019223392002016678,
+      "loss": 0.1297,
+      "step": 675
+    },
+    {
+      "epoch": 0.6418229290291954,
+      "grad_norm": 0.05248212069272995,
+      "learning_rate": 0.0001921911675427244,
+      "loss": 0.1737,
+      "step": 676
+    },
+    {
+      "epoch": 0.6427723712319012,
+      "grad_norm": 0.04168983921408653,
+      "learning_rate": 0.00019214830249200806,
+      "loss": 0.1373,
+      "step": 677
+    },
+    {
+      "epoch": 0.6437218134346072,
+      "grad_norm": 0.03659060224890709,
+      "learning_rate": 0.0001921053249203596,
+      "loss": 0.1263,
+      "step": 678
+    },
+    {
+      "epoch": 0.6446712556373131,
+      "grad_norm": 0.042256928980350494,
+      "learning_rate": 0.00019206223488025834,
+      "loss": 0.1646,
+      "step": 679
+    },
+    {
+      "epoch": 0.645620697840019,
+      "grad_norm": 0.04420709237456322,
+      "learning_rate": 0.00019201903242432086,
+      "loss": 0.1577,
+      "step": 680
+    },
+    {
+      "epoch": 0.646570140042725,
+      "grad_norm": 0.03781798109412193,
+      "learning_rate": 0.00019197571760530107,
+      "loss": 0.1253,
+      "step": 681
+    },
+    {
+      "epoch": 0.6475195822454308,
+      "grad_norm": 0.03728644549846649,
+      "learning_rate": 0.00019193229047609003,
+      "loss": 0.1423,
+      "step": 682
+    },
+    {
+      "epoch": 0.6484690244481367,
+      "grad_norm": 0.05171523615717888,
+      "learning_rate": 0.00019188875108971598,
+      "loss": 0.177,
+      "step": 683
+    },
+    {
+      "epoch": 0.6494184666508427,
+      "grad_norm": 0.05022161453962326,
+      "learning_rate": 0.0001918450994993442,
+      "loss": 0.1616,
+      "step": 684
+    },
+    {
+      "epoch": 0.6503679088535486,
+      "grad_norm": 0.037774864584207535,
+      "learning_rate": 0.00019180133575827707,
+      "loss": 0.1257,
+      "step": 685
+    },
+    {
+      "epoch": 0.6513173510562544,
+      "grad_norm": 0.056198425590991974,
+      "learning_rate": 0.00019175745991995377,
+      "loss": 0.1751,
+      "step": 686
+    },
+    {
+      "epoch": 0.6522667932589603,
+      "grad_norm": 0.05259314179420471,
+      "learning_rate": 0.0001917134720379505,
+      "loss": 0.1655,
+      "step": 687
+    },
+    {
+      "epoch": 0.6532162354616663,
+      "grad_norm": 0.04018954187631607,
+      "learning_rate": 0.00019166937216598013,
+      "loss": 0.1178,
+      "step": 688
+    },
+    {
+      "epoch": 0.6541656776643722,
+      "grad_norm": 0.057170454412698746,
+      "learning_rate": 0.00019162516035789247,
+      "loss": 0.1744,
+      "step": 689
+    },
+    {
+      "epoch": 0.655115119867078,
+      "grad_norm": 0.04647281393408775,
+      "learning_rate": 0.00019158083666767381,
+      "loss": 0.1343,
+      "step": 690
+    },
+    {
+      "epoch": 0.656064562069784,
+      "grad_norm": 0.056390274316072464,
+      "learning_rate": 0.00019153640114944723,
+      "loss": 0.2029,
+      "step": 691
+    },
+    {
+      "epoch": 0.6570140042724899,
+      "grad_norm": 0.03656432405114174,
+      "learning_rate": 0.00019149185385747224,
+      "loss": 0.1249,
+      "step": 692
+    },
+    {
+      "epoch": 0.6579634464751958,
+      "grad_norm": 0.031422629952430725,
+      "learning_rate": 0.0001914471948461449,
+      "loss": 0.1232,
+      "step": 693
+    },
+    {
+      "epoch": 0.6589128886779018,
+      "grad_norm": 0.0463186614215374,
+      "learning_rate": 0.00019140242416999765,
+      "loss": 0.1675,
+      "step": 694
+    },
+    {
+      "epoch": 0.6598623308806076,
+      "grad_norm": 0.03907819464802742,
+      "learning_rate": 0.0001913575418836993,
+      "loss": 0.1307,
+      "step": 695
+    },
+    {
+      "epoch": 0.6608117730833135,
+      "grad_norm": 0.04354274645447731,
+      "learning_rate": 0.00019131254804205498,
+      "loss": 0.1381,
+      "step": 696
+    },
+    {
+      "epoch": 0.6617612152860195,
+      "grad_norm": 0.0355788990855217,
+      "learning_rate": 0.00019126744270000598,
+      "loss": 0.1273,
+      "step": 697
+    },
+    {
+      "epoch": 0.6627106574887254,
+      "grad_norm": 0.0382835678756237,
+      "learning_rate": 0.0001912222259126298,
+      "loss": 0.1184,
+      "step": 698
+    },
+    {
+      "epoch": 0.6636600996914312,
+      "grad_norm": 0.05007009580731392,
+      "learning_rate": 0.00019117689773513993,
+      "loss": 0.1751,
+      "step": 699
+    },
+    {
+      "epoch": 0.6646095418941372,
+      "grad_norm": 0.05426732823252678,
+      "learning_rate": 0.000191131458222886,
+      "loss": 0.175,
+      "step": 700
+    },
+    {
+      "epoch": 0.6655589840968431,
+      "grad_norm": 0.033966466784477234,
+      "learning_rate": 0.00019108590743135352,
+      "loss": 0.123,
+      "step": 701
+    },
+    {
+      "epoch": 0.666508426299549,
+      "grad_norm": 0.04007060080766678,
+      "learning_rate": 0.00019104024541616386,
+      "loss": 0.1386,
+      "step": 702
+    },
+    {
+      "epoch": 0.667457868502255,
+      "grad_norm": 0.05075724050402641,
+      "learning_rate": 0.00019099447223307423,
+      "loss": 0.1698,
+      "step": 703
+    },
+    {
+      "epoch": 0.6684073107049608,
+      "grad_norm": 0.04677930474281311,
+      "learning_rate": 0.00019094858793797757,
+      "loss": 0.1633,
+      "step": 704
+    },
+    {
+      "epoch": 0.6693567529076667,
+      "grad_norm": 0.04063379392027855,
+      "learning_rate": 0.00019090259258690263,
+      "loss": 0.1414,
+      "step": 705
+    },
+    {
+      "epoch": 0.6703061951103727,
+      "grad_norm": 0.039291396737098694,
+      "learning_rate": 0.00019085648623601352,
+      "loss": 0.1273,
+      "step": 706
+    },
+    {
+      "epoch": 0.6712556373130786,
+      "grad_norm": 0.04960642755031586,
+      "learning_rate": 0.00019081026894161008,
+      "loss": 0.1512,
+      "step": 707
+    },
+    {
+      "epoch": 0.6722050795157845,
+      "grad_norm": 0.04266348108649254,
+      "learning_rate": 0.00019076394076012756,
+      "loss": 0.1352,
+      "step": 708
+    },
+    {
+      "epoch": 0.6731545217184904,
+      "grad_norm": 0.03943296894431114,
+      "learning_rate": 0.00019071750174813663,
+      "loss": 0.1332,
+      "step": 709
+    },
+    {
+      "epoch": 0.6741039639211963,
+      "grad_norm": 0.04927997291088104,
+      "learning_rate": 0.0001906709519623433,
+      "loss": 0.1645,
+      "step": 710
+    },
+    {
+      "epoch": 0.6750534061239022,
+      "grad_norm": 0.0418451763689518,
+      "learning_rate": 0.00019062429145958877,
+      "loss": 0.1279,
+      "step": 711
+    },
+    {
+      "epoch": 0.6760028483266081,
+      "grad_norm": 0.04283139482140541,
+      "learning_rate": 0.0001905775202968495,
+      "loss": 0.1388,
+      "step": 712
+    },
+    {
+      "epoch": 0.676952290529314,
+      "grad_norm": 0.05674710497260094,
+      "learning_rate": 0.00019053063853123714,
+      "loss": 0.171,
+      "step": 713
+    },
+    {
+      "epoch": 0.6779017327320199,
+      "grad_norm": 0.03568726405501366,
+      "learning_rate": 0.00019048364621999825,
+      "loss": 0.1329,
+      "step": 714
+    },
+    {
+      "epoch": 0.6788511749347258,
+      "grad_norm": 0.03796301409602165,
+      "learning_rate": 0.00019043654342051447,
+      "loss": 0.1352,
+      "step": 715
+    },
+    {
+      "epoch": 0.6798006171374318,
+      "grad_norm": 0.03538963943719864,
+      "learning_rate": 0.00019038933019030233,
+      "loss": 0.1328,
+      "step": 716
+    },
+    {
+      "epoch": 0.6807500593401377,
+      "grad_norm": 0.05234035104513168,
+      "learning_rate": 0.00019034200658701322,
+      "loss": 0.1649,
+      "step": 717
+    },
+    {
+      "epoch": 0.6816995015428435,
+      "grad_norm": 0.03719701990485191,
+      "learning_rate": 0.00019029457266843327,
+      "loss": 0.1295,
+      "step": 718
+    },
+    {
+      "epoch": 0.6826489437455495,
+      "grad_norm": 0.03594352304935455,
+      "learning_rate": 0.00019024702849248335,
+      "loss": 0.128,
+      "step": 719
+    },
+    {
+      "epoch": 0.6835983859482554,
+      "grad_norm": 0.04097168892621994,
+      "learning_rate": 0.00019019937411721895,
+      "loss": 0.1331,
+      "step": 720
+    },
+    {
+      "epoch": 0.6845478281509613,
+      "grad_norm": 0.03943239524960518,
+      "learning_rate": 0.00019015160960083013,
+      "loss": 0.1337,
+      "step": 721
+    },
+    {
+      "epoch": 0.6854972703536673,
+      "grad_norm": 0.0411958172917366,
+      "learning_rate": 0.00019010373500164145,
+      "loss": 0.1603,
+      "step": 722
+    },
+    {
+      "epoch": 0.6864467125563731,
+      "grad_norm": 0.05295250564813614,
+      "learning_rate": 0.00019005575037811184,
+      "loss": 0.1644,
+      "step": 723
+    },
+    {
+      "epoch": 0.687396154759079,
+      "grad_norm": 0.03916552662849426,
+      "learning_rate": 0.00019000765578883465,
+      "loss": 0.135,
+      "step": 724
+    },
+    {
+      "epoch": 0.688345596961785,
+      "grad_norm": 0.03871094062924385,
+      "learning_rate": 0.00018995945129253745,
+      "loss": 0.1276,
+      "step": 725
+    },
+    {
+      "epoch": 0.6892950391644909,
+      "grad_norm": 0.03405594825744629,
+      "learning_rate": 0.00018991113694808204,
+      "loss": 0.1327,
+      "step": 726
+    },
+    {
+      "epoch": 0.6902444813671967,
+      "grad_norm": 0.03824371099472046,
+      "learning_rate": 0.00018986271281446436,
+      "loss": 0.1357,
+      "step": 727
+    },
+    {
+      "epoch": 0.6911939235699027,
+      "grad_norm": 0.03813684731721878,
+      "learning_rate": 0.0001898141789508144,
+      "loss": 0.1341,
+      "step": 728
+    },
+    {
+      "epoch": 0.6921433657726086,
+      "grad_norm": 0.03283112868666649,
+      "learning_rate": 0.0001897655354163962,
+      "loss": 0.1299,
+      "step": 729
+    },
+    {
+      "epoch": 0.6930928079753145,
+      "grad_norm": 0.03226768597960472,
+      "learning_rate": 0.00018971678227060757,
+      "loss": 0.1272,
+      "step": 730
+    },
+    {
+      "epoch": 0.6940422501780205,
+      "grad_norm": 0.037317484617233276,
+      "learning_rate": 0.0001896679195729803,
+      "loss": 0.1339,
+      "step": 731
+    },
+    {
+      "epoch": 0.6949916923807263,
+      "grad_norm": 0.05428892746567726,
+      "learning_rate": 0.0001896189473831799,
+      "loss": 0.1667,
+      "step": 732
+    },
+    {
+      "epoch": 0.6959411345834322,
+      "grad_norm": 0.04177982360124588,
+      "learning_rate": 0.0001895698657610056,
+      "loss": 0.1337,
+      "step": 733
+    },
+    {
+      "epoch": 0.6968905767861382,
+      "grad_norm": 0.041572730988264084,
+      "learning_rate": 0.00018952067476639024,
+      "loss": 0.1332,
+      "step": 734
+    },
+    {
+      "epoch": 0.6978400189888441,
+      "grad_norm": 0.03430505469441414,
+      "learning_rate": 0.00018947137445940023,
+      "loss": 0.1265,
+      "step": 735
+    },
+    {
+      "epoch": 0.6987894611915499,
+      "grad_norm": 0.03863980621099472,
+      "learning_rate": 0.00018942196490023542,
+      "loss": 0.1337,
+      "step": 736
+    },
+    {
+      "epoch": 0.6997389033942559,
+      "grad_norm": 0.06445252895355225,
+      "learning_rate": 0.00018937244614922912,
+      "loss": 0.2032,
+      "step": 737
+    },
+    {
+      "epoch": 0.7006883455969618,
+      "grad_norm": 0.03358490392565727,
+      "learning_rate": 0.00018932281826684793,
+      "loss": 0.127,
+      "step": 738
+    },
+    {
+      "epoch": 0.7016377877996677,
+      "grad_norm": 0.034341324120759964,
+      "learning_rate": 0.00018927308131369173,
+      "loss": 0.1303,
+      "step": 739
+    },
+    {
+      "epoch": 0.7025872300023736,
+      "grad_norm": 0.035848621279001236,
+      "learning_rate": 0.00018922323535049354,
+      "loss": 0.1272,
+      "step": 740
+    },
+    {
+      "epoch": 0.7035366722050795,
+      "grad_norm": 0.03865866735577583,
+      "learning_rate": 0.0001891732804381196,
+      "loss": 0.136,
+      "step": 741
+    },
+    {
+      "epoch": 0.7044861144077854,
+      "grad_norm": 0.045944251120090485,
+      "learning_rate": 0.0001891232166375691,
+      "loss": 0.1741,
+      "step": 742
+    },
+    {
+      "epoch": 0.7054355566104913,
+      "grad_norm": 0.04418769106268883,
+      "learning_rate": 0.00018907304400997418,
+      "loss": 0.1504,
+      "step": 743
+    },
+    {
+      "epoch": 0.7063849988131973,
+      "grad_norm": 0.062257930636405945,
+      "learning_rate": 0.0001890227626165999,
+      "loss": 0.1786,
+      "step": 744
+    },
+    {
+      "epoch": 0.7073344410159031,
+      "grad_norm": 0.037457846105098724,
+      "learning_rate": 0.00018897237251884415,
+      "loss": 0.1389,
+      "step": 745
+    },
+    {
+      "epoch": 0.708283883218609,
+      "grad_norm": 0.039091672748327255,
+      "learning_rate": 0.0001889218737782375,
+      "loss": 0.1264,
+      "step": 746
+    },
+    {
+      "epoch": 0.709233325421315,
+      "grad_norm": 0.035011596977710724,
+      "learning_rate": 0.00018887126645644324,
+      "loss": 0.1363,
+      "step": 747
+    },
+    {
+      "epoch": 0.7101827676240209,
+      "grad_norm": 0.104104183614254,
+      "learning_rate": 0.00018882055061525722,
+      "loss": 0.1588,
+      "step": 748
+    },
+    {
+      "epoch": 0.7111322098267268,
+      "grad_norm": 0.03222833201289177,
+      "learning_rate": 0.0001887697263166078,
+      "loss": 0.1259,
+      "step": 749
+    },
+    {
+      "epoch": 0.7120816520294327,
+      "grad_norm": 0.049904145300388336,
+      "learning_rate": 0.0001887187936225558,
+      "loss": 0.1676,
+      "step": 750
+    },
+    {
+      "epoch": 0.7130310942321386,
+      "grad_norm": 0.15150390565395355,
+      "learning_rate": 0.00018866775259529435,
+      "loss": 0.1369,
+      "step": 751
+    },
+    {
+      "epoch": 0.7139805364348445,
+      "grad_norm": 0.03994397446513176,
+      "learning_rate": 0.0001886166032971489,
+      "loss": 0.1294,
+      "step": 752
+    },
+    {
+      "epoch": 0.7149299786375505,
+      "grad_norm": 0.06274881213903427,
+      "learning_rate": 0.00018856534579057713,
+      "loss": 0.1659,
+      "step": 753
+    },
+    {
+      "epoch": 0.7158794208402564,
+      "grad_norm": 0.04001612216234207,
+      "learning_rate": 0.00018851398013816883,
+      "loss": 0.1305,
+      "step": 754
+    },
+    {
+      "epoch": 0.7168288630429622,
+      "grad_norm": 0.03961142525076866,
+      "learning_rate": 0.0001884625064026458,
+      "loss": 0.1265,
+      "step": 755
+    },
+    {
+      "epoch": 0.7177783052456682,
+      "grad_norm": 0.033916253596544266,
+      "learning_rate": 0.00018841092464686186,
+      "loss": 0.1336,
+      "step": 756
+    },
+    {
+      "epoch": 0.7187277474483741,
+      "grad_norm": 0.040992431342601776,
+      "learning_rate": 0.00018835923493380278,
+      "loss": 0.1403,
+      "step": 757
+    },
+    {
+      "epoch": 0.71967718965108,
+      "grad_norm": 0.03410341590642929,
+      "learning_rate": 0.00018830743732658608,
+      "loss": 0.1233,
+      "step": 758
+    },
+    {
+      "epoch": 0.720626631853786,
+      "grad_norm": 0.05984083190560341,
+      "learning_rate": 0.000188255531888461,
+      "loss": 0.1417,
+      "step": 759
+    },
+    {
+      "epoch": 0.7215760740564918,
+      "grad_norm": 0.03874243050813675,
+      "learning_rate": 0.00018820351868280858,
+      "loss": 0.1366,
+      "step": 760
+    },
+    {
+      "epoch": 0.7225255162591977,
+      "grad_norm": 0.05256400629878044,
+      "learning_rate": 0.00018815139777314136,
+      "loss": 0.172,
+      "step": 761
+    },
+    {
+      "epoch": 0.7234749584619037,
+      "grad_norm": 0.039005253463983536,
+      "learning_rate": 0.0001880991692231034,
+      "loss": 0.1312,
+      "step": 762
+    },
+    {
+      "epoch": 0.7244244006646096,
+      "grad_norm": 0.04029637575149536,
+      "learning_rate": 0.0001880468330964702,
+      "loss": 0.1327,
+      "step": 763
+    },
+    {
+      "epoch": 0.7253738428673154,
+      "grad_norm": 0.04493672773241997,
+      "learning_rate": 0.00018799438945714866,
+      "loss": 0.1555,
+      "step": 764
+    },
+    {
+      "epoch": 0.7263232850700213,
+      "grad_norm": 0.03862634301185608,
+      "learning_rate": 0.0001879418383691769,
+      "loss": 0.133,
+      "step": 765
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.03904002159833908,
+      "learning_rate": 0.00018788917989672434,
+      "loss": 0.1259,
+      "step": 766
+    },
+    {
+      "epoch": 0.7282221694754332,
+      "grad_norm": 0.037936531007289886,
+      "learning_rate": 0.0001878364141040914,
+      "loss": 0.1263,
+      "step": 767
+    },
+    {
+      "epoch": 0.729171611678139,
+      "grad_norm": 0.03802201896905899,
+      "learning_rate": 0.0001877835410557096,
+      "loss": 0.134,
+      "step": 768
+    },
+    {
+      "epoch": 0.730121053880845,
+      "grad_norm": 0.03759211301803589,
+      "learning_rate": 0.00018773056081614154,
+      "loss": 0.1383,
+      "step": 769
+    },
+    {
+      "epoch": 0.7310704960835509,
+      "grad_norm": 0.0498163104057312,
+      "learning_rate": 0.0001876774734500805,
+      "loss": 0.1573,
+      "step": 770
+    },
+    {
+      "epoch": 0.7320199382862568,
+      "grad_norm": 0.036126043647527695,
+      "learning_rate": 0.00018762427902235072,
+      "loss": 0.1274,
+      "step": 771
+    },
+    {
+      "epoch": 0.7329693804889628,
+      "grad_norm": 0.044809550046920776,
+      "learning_rate": 0.0001875709775979071,
+      "loss": 0.1703,
+      "step": 772
+    },
+    {
+      "epoch": 0.7339188226916686,
+      "grad_norm": 0.050454337149858475,
+      "learning_rate": 0.0001875175692418353,
+      "loss": 0.1699,
+      "step": 773
+    },
+    {
+      "epoch": 0.7348682648943745,
+      "grad_norm": 0.06160600110888481,
+      "learning_rate": 0.00018746405401935142,
+      "loss": 0.1806,
+      "step": 774
+    },
+    {
+      "epoch": 0.7358177070970805,
+      "grad_norm": 0.05408332124352455,
+      "learning_rate": 0.0001874104319958021,
+      "loss": 0.1681,
+      "step": 775
+    },
+    {
+      "epoch": 0.7367671492997864,
+      "grad_norm": 0.03859655559062958,
+      "learning_rate": 0.00018735670323666442,
+      "loss": 0.1297,
+      "step": 776
+    },
+    {
+      "epoch": 0.7377165915024922,
+      "grad_norm": 0.05268474668264389,
+      "learning_rate": 0.00018730286780754577,
+      "loss": 0.1658,
+      "step": 777
+    },
+    {
+      "epoch": 0.7386660337051982,
+      "grad_norm": 0.06406822055578232,
+      "learning_rate": 0.00018724892577418381,
+      "loss": 0.199,
+      "step": 778
+    },
+    {
+      "epoch": 0.7396154759079041,
+      "grad_norm": 0.05488892272114754,
+      "learning_rate": 0.00018719487720244638,
+      "loss": 0.1669,
+      "step": 779
+    },
+    {
+      "epoch": 0.74056491811061,
+      "grad_norm": 0.03732241317629814,
+      "learning_rate": 0.00018714072215833132,
+      "loss": 0.1337,
+      "step": 780
+    },
+    {
+      "epoch": 0.741514360313316,
+      "grad_norm": 0.05548230558633804,
+      "learning_rate": 0.00018708646070796664,
+      "loss": 0.1652,
+      "step": 781
+    },
+    {
+      "epoch": 0.7424638025160218,
+      "grad_norm": 0.06930623203516006,
+      "learning_rate": 0.0001870320929176101,
+      "loss": 0.1647,
+      "step": 782
+    },
+    {
+      "epoch": 0.7434132447187277,
+      "grad_norm": 0.05485931411385536,
+      "learning_rate": 0.0001869776188536495,
+      "loss": 0.2149,
+      "step": 783
+    },
+    {
+      "epoch": 0.7443626869214337,
+      "grad_norm": 0.03739183023571968,
+      "learning_rate": 0.00018692303858260228,
+      "loss": 0.1257,
+      "step": 784
+    },
+    {
+      "epoch": 0.7453121291241396,
+      "grad_norm": 0.03913332521915436,
+      "learning_rate": 0.00018686835217111557,
+      "loss": 0.1293,
+      "step": 785
+    },
+    {
+      "epoch": 0.7462615713268455,
+      "grad_norm": 0.03580600768327713,
+      "learning_rate": 0.0001868135596859662,
+      "loss": 0.1278,
+      "step": 786
+    },
+    {
+      "epoch": 0.7472110135295514,
+      "grad_norm": 0.03586685657501221,
+      "learning_rate": 0.00018675866119406042,
+      "loss": 0.128,
+      "step": 787
+    },
+    {
+      "epoch": 0.7481604557322573,
+      "grad_norm": 0.04061829298734665,
+      "learning_rate": 0.00018670365676243397,
+      "loss": 0.1256,
+      "step": 788
+    },
+    {
+      "epoch": 0.7491098979349632,
+      "grad_norm": 0.03580275923013687,
+      "learning_rate": 0.000186648546458252,
+      "loss": 0.1265,
+      "step": 789
+    },
+    {
+      "epoch": 0.7500593401376692,
+      "grad_norm": 0.04277309030294418,
+      "learning_rate": 0.00018659333034880884,
+      "loss": 0.1678,
+      "step": 790
+    },
+    {
+      "epoch": 0.751008782340375,
+      "grad_norm": 0.03997024893760681,
+      "learning_rate": 0.00018653800850152808,
+      "loss": 0.1251,
+      "step": 791
+    },
+    {
+      "epoch": 0.7519582245430809,
+      "grad_norm": 0.03809446841478348,
+      "learning_rate": 0.0001864825809839624,
+      "loss": 0.1354,
+      "step": 792
+    },
+    {
+      "epoch": 0.7529076667457868,
+      "grad_norm": 0.05002079904079437,
+      "learning_rate": 0.00018642704786379354,
+      "loss": 0.1492,
+      "step": 793
+    },
+    {
+      "epoch": 0.7538571089484928,
+      "grad_norm": 0.03734049201011658,
+      "learning_rate": 0.00018637140920883217,
+      "loss": 0.1328,
+      "step": 794
+    },
+    {
+      "epoch": 0.7548065511511987,
+      "grad_norm": 0.034287337213754654,
+      "learning_rate": 0.00018631566508701784,
+      "loss": 0.1261,
+      "step": 795
+    },
+    {
+      "epoch": 0.7557559933539045,
+      "grad_norm": 0.0322953499853611,
+      "learning_rate": 0.00018625981556641882,
+      "loss": 0.1251,
+      "step": 796
+    },
+    {
+      "epoch": 0.7567054355566105,
+      "grad_norm": 0.03397887200117111,
+      "learning_rate": 0.00018620386071523218,
+      "loss": 0.1226,
+      "step": 797
+    },
+    {
+      "epoch": 0.7576548777593164,
+      "grad_norm": 0.048685140907764435,
+      "learning_rate": 0.0001861478006017836,
+      "loss": 0.1677,
+      "step": 798
+    },
+    {
+      "epoch": 0.7586043199620223,
+      "grad_norm": 0.06330600380897522,
+      "learning_rate": 0.00018609163529452723,
+      "loss": 0.2012,
+      "step": 799
+    },
+    {
+      "epoch": 0.7595537621647283,
+      "grad_norm": 0.04262509569525719,
+      "learning_rate": 0.00018603536486204564,
+      "loss": 0.1271,
+      "step": 800
+    },
+    {
+      "epoch": 0.7605032043674341,
+      "grad_norm": 0.04021213576197624,
+      "learning_rate": 0.00018597898937304988,
+      "loss": 0.1426,
+      "step": 801
+    },
+    {
+      "epoch": 0.76145264657014,
+      "grad_norm": 0.05070256441831589,
+      "learning_rate": 0.0001859225088963792,
+      "loss": 0.209,
+      "step": 802
+    },
+    {
+      "epoch": 0.762402088772846,
+      "grad_norm": 0.05344654247164726,
+      "learning_rate": 0.00018586592350100113,
+      "loss": 0.2093,
+      "step": 803
+    },
+    {
+      "epoch": 0.7633515309755519,
+      "grad_norm": 0.03695262596011162,
+      "learning_rate": 0.0001858092332560112,
+      "loss": 0.1264,
+      "step": 804
+    },
+    {
+      "epoch": 0.7643009731782577,
+      "grad_norm": 0.041282836347818375,
+      "learning_rate": 0.00018575243823063306,
+      "loss": 0.1275,
+      "step": 805
+    },
+    {
+      "epoch": 0.7652504153809637,
+      "grad_norm": 0.038663093000650406,
+      "learning_rate": 0.00018569553849421828,
+      "loss": 0.1285,
+      "step": 806
+    },
+    {
+      "epoch": 0.7661998575836696,
+      "grad_norm": 0.05324345454573631,
+      "learning_rate": 0.00018563853411624628,
+      "loss": 0.1691,
+      "step": 807
+    },
+    {
+      "epoch": 0.7671492997863755,
+      "grad_norm": 0.0382021889090538,
+      "learning_rate": 0.00018558142516632425,
+      "loss": 0.1299,
+      "step": 808
+    },
+    {
+      "epoch": 0.7680987419890815,
+      "grad_norm": 0.05059641972184181,
+      "learning_rate": 0.00018552421171418712,
+      "loss": 0.1685,
+      "step": 809
+    },
+    {
+      "epoch": 0.7690481841917873,
+      "grad_norm": 0.041547179222106934,
+      "learning_rate": 0.00018546689382969737,
+      "loss": 0.1322,
+      "step": 810
+    },
+    {
+      "epoch": 0.7699976263944932,
+      "grad_norm": 0.047367729246616364,
+      "learning_rate": 0.00018540947158284503,
+      "loss": 0.1662,
+      "step": 811
+    },
+    {
+      "epoch": 0.7709470685971992,
+      "grad_norm": 0.07076044380664825,
+      "learning_rate": 0.00018535194504374754,
+      "loss": 0.1749,
+      "step": 812
+    },
+    {
+      "epoch": 0.7718965107999051,
+      "grad_norm": 0.05194571986794472,
+      "learning_rate": 0.00018529431428264973,
+      "loss": 0.1595,
+      "step": 813
+    },
+    {
+      "epoch": 0.7728459530026109,
+      "grad_norm": 0.034832440316677094,
+      "learning_rate": 0.00018523657936992367,
+      "loss": 0.1279,
+      "step": 814
+    },
+    {
+      "epoch": 0.7737953952053169,
+      "grad_norm": 0.03709466755390167,
+      "learning_rate": 0.00018517874037606862,
+      "loss": 0.1161,
+      "step": 815
+    },
+    {
+      "epoch": 0.7747448374080228,
+      "grad_norm": 0.03341936320066452,
+      "learning_rate": 0.00018512079737171086,
+      "loss": 0.1277,
+      "step": 816
+    },
+    {
+      "epoch": 0.7756942796107287,
+      "grad_norm": 0.0411679781973362,
+      "learning_rate": 0.00018506275042760382,
+      "loss": 0.1284,
+      "step": 817
+    },
+    {
+      "epoch": 0.7766437218134347,
+      "grad_norm": 0.04416754096746445,
+      "learning_rate": 0.00018500459961462773,
+      "loss": 0.1647,
+      "step": 818
+    },
+    {
+      "epoch": 0.7775931640161405,
+      "grad_norm": 0.03680622950196266,
+      "learning_rate": 0.00018494634500378966,
+      "loss": 0.1371,
+      "step": 819
+    },
+    {
+      "epoch": 0.7785426062188464,
+      "grad_norm": 0.037342917174100876,
+      "learning_rate": 0.0001848879866662235,
+      "loss": 0.1308,
+      "step": 820
+    },
+    {
+      "epoch": 0.7794920484215523,
+      "grad_norm": 0.04237838089466095,
+      "learning_rate": 0.00018482952467318976,
+      "loss": 0.1623,
+      "step": 821
+    },
+    {
+      "epoch": 0.7804414906242583,
+      "grad_norm": 0.04467133805155754,
+      "learning_rate": 0.00018477095909607546,
+      "loss": 0.1651,
+      "step": 822
+    },
+    {
+      "epoch": 0.7813909328269641,
+      "grad_norm": 0.04672664403915405,
+      "learning_rate": 0.00018471229000639424,
+      "loss": 0.1735,
+      "step": 823
+    },
+    {
+      "epoch": 0.78234037502967,
+      "grad_norm": 0.03545104339718819,
+      "learning_rate": 0.00018465351747578597,
+      "loss": 0.1342,
+      "step": 824
+    },
+    {
+      "epoch": 0.783289817232376,
+      "grad_norm": 0.04771837964653969,
+      "learning_rate": 0.000184594641576017,
+      "loss": 0.175,
+      "step": 825
+    },
+    {
+      "epoch": 0.7842392594350819,
+      "grad_norm": 0.03531822934746742,
+      "learning_rate": 0.00018453566237897976,
+      "loss": 0.1321,
+      "step": 826
+    },
+    {
+      "epoch": 0.7851887016377878,
+      "grad_norm": 0.04098953306674957,
+      "learning_rate": 0.00018447657995669295,
+      "loss": 0.1372,
+      "step": 827
+    },
+    {
+      "epoch": 0.7861381438404937,
+      "grad_norm": 0.053972020745277405,
+      "learning_rate": 0.00018441739438130114,
+      "loss": 0.1673,
+      "step": 828
+    },
+    {
+      "epoch": 0.7870875860431996,
+      "grad_norm": 0.03818265721201897,
+      "learning_rate": 0.00018435810572507507,
+      "loss": 0.1322,
+      "step": 829
+    },
+    {
+      "epoch": 0.7880370282459055,
+      "grad_norm": 0.033827316015958786,
+      "learning_rate": 0.0001842987140604112,
+      "loss": 0.1445,
+      "step": 830
+    },
+    {
+      "epoch": 0.7889864704486115,
+      "grad_norm": 0.041385356336832047,
+      "learning_rate": 0.00018423921945983179,
+      "loss": 0.1332,
+      "step": 831
+    },
+    {
+      "epoch": 0.7899359126513174,
+      "grad_norm": 0.03948013484477997,
+      "learning_rate": 0.00018417962199598483,
+      "loss": 0.1412,
+      "step": 832
+    },
+    {
+      "epoch": 0.7908853548540232,
+      "grad_norm": 0.044912584125995636,
+      "learning_rate": 0.00018411992174164393,
+      "loss": 0.1684,
+      "step": 833
+    },
+    {
+      "epoch": 0.7918347970567292,
+      "grad_norm": 0.03675195202231407,
+      "learning_rate": 0.0001840601187697082,
+      "loss": 0.1334,
+      "step": 834
+    },
+    {
+      "epoch": 0.7927842392594351,
+      "grad_norm": 0.0349728949368,
+      "learning_rate": 0.0001840002131532021,
+      "loss": 0.1323,
+      "step": 835
+    },
+    {
+      "epoch": 0.793733681462141,
+      "grad_norm": 0.03763123229146004,
+      "learning_rate": 0.0001839402049652755,
+      "loss": 0.1283,
+      "step": 836
+    },
+    {
+      "epoch": 0.794683123664847,
+      "grad_norm": 0.036798711866140366,
+      "learning_rate": 0.00018388009427920362,
+      "loss": 0.1272,
+      "step": 837
+    },
+    {
+      "epoch": 0.7956325658675528,
+      "grad_norm": 0.036771487444639206,
+      "learning_rate": 0.00018381988116838663,
+      "loss": 0.126,
+      "step": 838
+    },
+    {
+      "epoch": 0.7965820080702587,
+      "grad_norm": 0.060571007430553436,
+      "learning_rate": 0.00018375956570634987,
+      "loss": 0.1736,
+      "step": 839
+    },
+    {
+      "epoch": 0.7975314502729647,
+      "grad_norm": 0.0332857109606266,
+      "learning_rate": 0.00018369914796674373,
+      "loss": 0.1301,
+      "step": 840
+    },
+    {
+      "epoch": 0.7984808924756706,
+      "grad_norm": 0.045279379934072495,
+      "learning_rate": 0.00018363862802334334,
+      "loss": 0.1602,
+      "step": 841
+    },
+    {
+      "epoch": 0.7994303346783764,
+      "grad_norm": 0.03676297515630722,
+      "learning_rate": 0.00018357800595004877,
+      "loss": 0.1299,
+      "step": 842
+    },
+    {
+      "epoch": 0.8003797768810824,
+      "grad_norm": 0.05098710209131241,
+      "learning_rate": 0.0001835172818208847,
+      "loss": 0.1289,
+      "step": 843
+    },
+    {
+      "epoch": 0.8013292190837883,
+      "grad_norm": 0.047296855598688126,
+      "learning_rate": 0.00018345645571000052,
+      "loss": 0.1716,
+      "step": 844
+    },
+    {
+      "epoch": 0.8022786612864942,
+      "grad_norm": 0.03570317476987839,
+      "learning_rate": 0.00018339552769167003,
+      "loss": 0.1337,
+      "step": 845
+    },
+    {
+      "epoch": 0.8032281034892,
+      "grad_norm": 0.03380590304732323,
+      "learning_rate": 0.00018333449784029156,
+      "loss": 0.1218,
+      "step": 846
+    },
+    {
+      "epoch": 0.804177545691906,
+      "grad_norm": 0.0340820774435997,
+      "learning_rate": 0.00018327336623038778,
+      "loss": 0.1324,
+      "step": 847
+    },
+    {
+      "epoch": 0.8051269878946119,
+      "grad_norm": 0.03311248868703842,
+      "learning_rate": 0.00018321213293660558,
+      "loss": 0.1308,
+      "step": 848
+    },
+    {
+      "epoch": 0.8060764300973178,
+      "grad_norm": 0.035102471709251404,
+      "learning_rate": 0.00018315079803371605,
+      "loss": 0.1345,
+      "step": 849
+    },
+    {
+      "epoch": 0.8070258723000238,
+      "grad_norm": 0.03358345478773117,
+      "learning_rate": 0.0001830893615966143,
+      "loss": 0.1341,
+      "step": 850
+    },
+    {
+      "epoch": 0.8079753145027296,
+      "grad_norm": 0.06460444629192352,
+      "learning_rate": 0.00018302782370031948,
+      "loss": 0.2051,
+      "step": 851
+    },
+    {
+      "epoch": 0.8089247567054355,
+      "grad_norm": 0.033203575760126114,
+      "learning_rate": 0.0001829661844199746,
+      "loss": 0.1357,
+      "step": 852
+    },
+    {
+      "epoch": 0.8098741989081415,
+      "grad_norm": 0.03588509559631348,
+      "learning_rate": 0.0001829044438308465,
+      "loss": 0.1335,
+      "step": 853
+    },
+    {
+      "epoch": 0.8108236411108474,
+      "grad_norm": 0.04263895004987717,
+      "learning_rate": 0.00018284260200832563,
+      "loss": 0.1739,
+      "step": 854
+    },
+    {
+      "epoch": 0.8117730833135532,
+      "grad_norm": 0.04004021733999252,
+      "learning_rate": 0.00018278065902792618,
+      "loss": 0.131,
+      "step": 855
+    },
+    {
+      "epoch": 0.8127225255162592,
+      "grad_norm": 0.035174135118722916,
+      "learning_rate": 0.00018271861496528584,
+      "loss": 0.1248,
+      "step": 856
+    },
+    {
+      "epoch": 0.8136719677189651,
+      "grad_norm": 0.03610173240303993,
+      "learning_rate": 0.00018265646989616566,
+      "loss": 0.1287,
+      "step": 857
+    },
+    {
+      "epoch": 0.814621409921671,
+      "grad_norm": 0.035818714648485184,
+      "learning_rate": 0.00018259422389645008,
+      "loss": 0.1335,
+      "step": 858
+    },
+    {
+      "epoch": 0.815570852124377,
+      "grad_norm": 0.03248162940144539,
+      "learning_rate": 0.00018253187704214672,
+      "loss": 0.1308,
+      "step": 859
+    },
+    {
+      "epoch": 0.8165202943270828,
+      "grad_norm": 0.031658854335546494,
+      "learning_rate": 0.00018246942940938646,
+      "loss": 0.1339,
+      "step": 860
+    },
+    {
+      "epoch": 0.8174697365297887,
+      "grad_norm": 0.035879503935575485,
+      "learning_rate": 0.0001824068810744232,
+      "loss": 0.1222,
+      "step": 861
+    },
+    {
+      "epoch": 0.8184191787324947,
+      "grad_norm": 0.05258049815893173,
+      "learning_rate": 0.0001823442321136337,
+      "loss": 0.1937,
+      "step": 862
+    },
+    {
+      "epoch": 0.8193686209352006,
+      "grad_norm": 0.03625549003481865,
+      "learning_rate": 0.0001822814826035178,
+      "loss": 0.1268,
+      "step": 863
+    },
+    {
+      "epoch": 0.8203180631379065,
+      "grad_norm": 0.04990942031145096,
+      "learning_rate": 0.00018221863262069793,
+      "loss": 0.1661,
+      "step": 864
+    },
+    {
+      "epoch": 0.8212675053406124,
+      "grad_norm": 0.0631263256072998,
+      "learning_rate": 0.00018215568224191927,
+      "loss": 0.2126,
+      "step": 865
+    },
+    {
+      "epoch": 0.8222169475433183,
+      "grad_norm": 0.03726550564169884,
+      "learning_rate": 0.00018209263154404958,
+      "loss": 0.1334,
+      "step": 866
+    },
+    {
+      "epoch": 0.8231663897460242,
+      "grad_norm": 0.040383536368608475,
+      "learning_rate": 0.0001820294806040792,
+      "loss": 0.1619,
+      "step": 867
+    },
+    {
+      "epoch": 0.8241158319487302,
+      "grad_norm": 0.03525468334555626,
+      "learning_rate": 0.00018196622949912078,
+      "loss": 0.1263,
+      "step": 868
+    },
+    {
+      "epoch": 0.825065274151436,
+      "grad_norm": 0.03585941344499588,
+      "learning_rate": 0.00018190287830640933,
+      "loss": 0.1245,
+      "step": 869
+    },
+    {
+      "epoch": 0.8260147163541419,
+      "grad_norm": 0.03207286074757576,
+      "learning_rate": 0.00018183942710330202,
+      "loss": 0.1262,
+      "step": 870
+    },
+    {
+      "epoch": 0.8269641585568479,
+      "grad_norm": 0.04638965427875519,
+      "learning_rate": 0.00018177587596727822,
+      "loss": 0.1653,
+      "step": 871
+    },
+    {
+      "epoch": 0.8279136007595538,
+      "grad_norm": 0.030705489218235016,
+      "learning_rate": 0.00018171222497593922,
+      "loss": 0.1276,
+      "step": 872
+    },
+    {
+      "epoch": 0.8288630429622597,
+      "grad_norm": 0.03139735013246536,
+      "learning_rate": 0.00018164847420700837,
+      "loss": 0.1344,
+      "step": 873
+    },
+    {
+      "epoch": 0.8298124851649655,
+      "grad_norm": 0.039802953600883484,
+      "learning_rate": 0.00018158462373833078,
+      "loss": 0.1373,
+      "step": 874
+    },
+    {
+      "epoch": 0.8307619273676715,
+      "grad_norm": 0.03284341096878052,
+      "learning_rate": 0.00018152067364787325,
+      "loss": 0.1236,
+      "step": 875
+    },
+    {
+      "epoch": 0.8317113695703774,
+      "grad_norm": 0.056572429835796356,
+      "learning_rate": 0.0001814566240137244,
+      "loss": 0.1665,
+      "step": 876
+    },
+    {
+      "epoch": 0.8326608117730833,
+      "grad_norm": 0.03471997380256653,
+      "learning_rate": 0.00018139247491409424,
+      "loss": 0.13,
+      "step": 877
+    },
+    {
+      "epoch": 0.8336102539757893,
+      "grad_norm": 0.03601829707622528,
+      "learning_rate": 0.00018132822642731426,
+      "loss": 0.127,
+      "step": 878
+    },
+    {
+      "epoch": 0.8345596961784951,
+      "grad_norm": 0.032708846032619476,
+      "learning_rate": 0.00018126387863183737,
+      "loss": 0.1264,
+      "step": 879
+    },
+    {
+      "epoch": 0.835509138381201,
+      "grad_norm": 0.035340629518032074,
+      "learning_rate": 0.00018119943160623773,
+      "loss": 0.1334,
+      "step": 880
+    },
+    {
+      "epoch": 0.836458580583907,
+      "grad_norm": 0.030397990718483925,
+      "learning_rate": 0.00018113488542921061,
+      "loss": 0.1254,
+      "step": 881
+    },
+    {
+      "epoch": 0.8374080227866129,
+      "grad_norm": 0.03871999308466911,
+      "learning_rate": 0.00018107024017957244,
+      "loss": 0.132,
+      "step": 882
+    },
+    {
+      "epoch": 0.8383574649893187,
+      "grad_norm": 0.04331507533788681,
+      "learning_rate": 0.00018100549593626052,
+      "loss": 0.1354,
+      "step": 883
+    },
+    {
+      "epoch": 0.8393069071920247,
+      "grad_norm": 0.03445984423160553,
+      "learning_rate": 0.00018094065277833314,
+      "loss": 0.129,
+      "step": 884
+    },
+    {
+      "epoch": 0.8402563493947306,
+      "grad_norm": 0.03362146392464638,
+      "learning_rate": 0.0001808757107849693,
+      "loss": 0.125,
+      "step": 885
+    },
+    {
+      "epoch": 0.8412057915974365,
+      "grad_norm": 0.041491370648145676,
+      "learning_rate": 0.00018081067003546876,
+      "loss": 0.1314,
+      "step": 886
+    },
+    {
+      "epoch": 0.8421552338001425,
+      "grad_norm": 0.034560974687337875,
+      "learning_rate": 0.00018074553060925175,
+      "loss": 0.126,
+      "step": 887
+    },
+    {
+      "epoch": 0.8431046760028483,
+      "grad_norm": 0.049931105226278305,
+      "learning_rate": 0.0001806802925858591,
+      "loss": 0.1709,
+      "step": 888
+    },
+    {
+      "epoch": 0.8440541182055542,
+      "grad_norm": 0.035841234028339386,
+      "learning_rate": 0.00018061495604495195,
+      "loss": 0.1396,
+      "step": 889
+    },
+    {
+      "epoch": 0.8450035604082602,
+      "grad_norm": 0.03359563648700714,
+      "learning_rate": 0.00018054952106631188,
+      "loss": 0.1323,
+      "step": 890
+    },
+    {
+      "epoch": 0.8459530026109661,
+      "grad_norm": 0.03390706703066826,
+      "learning_rate": 0.00018048398772984046,
+      "loss": 0.1287,
+      "step": 891
+    },
+    {
+      "epoch": 0.8469024448136719,
+      "grad_norm": 0.0474267303943634,
+      "learning_rate": 0.00018041835611555957,
+      "loss": 0.1693,
+      "step": 892
+    },
+    {
+      "epoch": 0.8478518870163779,
+      "grad_norm": 0.0334562286734581,
+      "learning_rate": 0.00018035262630361097,
+      "loss": 0.1295,
+      "step": 893
+    },
+    {
+      "epoch": 0.8488013292190838,
+      "grad_norm": 0.03383705019950867,
+      "learning_rate": 0.00018028679837425634,
+      "loss": 0.1259,
+      "step": 894
+    },
+    {
+      "epoch": 0.8497507714217897,
+      "grad_norm": 0.03384934738278389,
+      "learning_rate": 0.00018022087240787728,
+      "loss": 0.1218,
+      "step": 895
+    },
+    {
+      "epoch": 0.8507002136244957,
+      "grad_norm": 0.04088185727596283,
+      "learning_rate": 0.0001801548484849749,
+      "loss": 0.1343,
+      "step": 896
+    },
+    {
+      "epoch": 0.8516496558272015,
+      "grad_norm": 0.05273745581507683,
+      "learning_rate": 0.00018008872668617013,
+      "loss": 0.1688,
+      "step": 897
+    },
+    {
+      "epoch": 0.8525990980299074,
+      "grad_norm": 0.03253067284822464,
+      "learning_rate": 0.00018002250709220325,
+      "loss": 0.1333,
+      "step": 898
+    },
+    {
+      "epoch": 0.8535485402326133,
+      "grad_norm": 0.03033488616347313,
+      "learning_rate": 0.0001799561897839341,
+      "loss": 0.1292,
+      "step": 899
+    },
+    {
+      "epoch": 0.8544979824353193,
+      "grad_norm": 0.033945854753255844,
+      "learning_rate": 0.00017988977484234174,
+      "loss": 0.1415,
+      "step": 900
+    },
+    {
+      "epoch": 0.8554474246380251,
+      "grad_norm": 0.04456301033496857,
+      "learning_rate": 0.0001798232623485244,
+      "loss": 0.1762,
+      "step": 901
+    },
+    {
+      "epoch": 0.856396866840731,
+      "grad_norm": 0.03912430256605148,
+      "learning_rate": 0.00017975665238369962,
+      "loss": 0.142,
+      "step": 902
+    },
+    {
+      "epoch": 0.857346309043437,
+      "grad_norm": 0.032741378992795944,
+      "learning_rate": 0.0001796899450292038,
+      "loss": 0.1212,
+      "step": 903
+    },
+    {
+      "epoch": 0.8582957512461429,
+      "grad_norm": 0.047262486070394516,
+      "learning_rate": 0.0001796231403664923,
+      "loss": 0.1762,
+      "step": 904
+    },
+    {
+      "epoch": 0.8592451934488488,
+      "grad_norm": 0.03242664784193039,
+      "learning_rate": 0.00017955623847713928,
+      "loss": 0.1323,
+      "step": 905
+    },
+    {
+      "epoch": 0.8601946356515547,
+      "grad_norm": 0.030855266377329826,
+      "learning_rate": 0.0001794892394428377,
+      "loss": 0.1258,
+      "step": 906
+    },
+    {
+      "epoch": 0.8611440778542606,
+      "grad_norm": 0.03360726311802864,
+      "learning_rate": 0.00017942214334539907,
+      "loss": 0.1325,
+      "step": 907
+    },
+    {
+      "epoch": 0.8620935200569665,
+      "grad_norm": 0.032459285110235214,
+      "learning_rate": 0.00017935495026675345,
+      "loss": 0.1267,
+      "step": 908
+    },
+    {
+      "epoch": 0.8630429622596725,
+      "grad_norm": 0.04160567373037338,
+      "learning_rate": 0.00017928766028894928,
+      "loss": 0.1255,
+      "step": 909
+    },
+    {
+      "epoch": 0.8639924044623783,
+      "grad_norm": 0.03851740434765816,
+      "learning_rate": 0.0001792202734941534,
+      "loss": 0.1212,
+      "step": 910
+    },
+    {
+      "epoch": 0.8649418466650842,
+      "grad_norm": 0.03414515405893326,
+      "learning_rate": 0.00017915278996465084,
+      "loss": 0.1239,
+      "step": 911
+    },
+    {
+      "epoch": 0.8658912888677902,
+      "grad_norm": 0.17817381024360657,
+      "learning_rate": 0.0001790852097828447,
+      "loss": 0.1336,
+      "step": 912
+    },
+    {
+      "epoch": 0.8668407310704961,
+      "grad_norm": 0.03545542433857918,
+      "learning_rate": 0.0001790175330312562,
+      "loss": 0.1353,
+      "step": 913
+    },
+    {
+      "epoch": 0.867790173273202,
+      "grad_norm": 0.03207210451364517,
+      "learning_rate": 0.00017894975979252436,
+      "loss": 0.1243,
+      "step": 914
+    },
+    {
+      "epoch": 0.868739615475908,
+      "grad_norm": 0.046145763248205185,
+      "learning_rate": 0.0001788818901494061,
+      "loss": 0.1668,
+      "step": 915
+    },
+    {
+      "epoch": 0.8696890576786138,
+      "grad_norm": 0.03051767125725746,
+      "learning_rate": 0.00017881392418477607,
+      "loss": 0.1311,
+      "step": 916
+    },
+    {
+      "epoch": 0.8706384998813197,
+      "grad_norm": 0.03918071463704109,
+      "learning_rate": 0.00017874586198162647,
+      "loss": 0.1692,
+      "step": 917
+    },
+    {
+      "epoch": 0.8715879420840257,
+      "grad_norm": 0.03229302540421486,
+      "learning_rate": 0.0001786777036230671,
+      "loss": 0.1276,
+      "step": 918
+    },
+    {
+      "epoch": 0.8725373842867316,
+      "grad_norm": 0.032113853842020035,
+      "learning_rate": 0.00017860944919232503,
+      "loss": 0.1256,
+      "step": 919
+    },
+    {
+      "epoch": 0.8734868264894374,
+      "grad_norm": 0.03725959360599518,
+      "learning_rate": 0.00017854109877274484,
+      "loss": 0.1363,
+      "step": 920
+    },
+    {
+      "epoch": 0.8744362686921434,
+      "grad_norm": 0.02805374562740326,
+      "learning_rate": 0.00017847265244778817,
+      "loss": 0.1259,
+      "step": 921
+    },
+    {
+      "epoch": 0.8753857108948493,
+      "grad_norm": 0.03541216999292374,
+      "learning_rate": 0.00017840411030103383,
+      "loss": 0.1288,
+      "step": 922
+    },
+    {
+      "epoch": 0.8763351530975552,
+      "grad_norm": 0.04267534613609314,
+      "learning_rate": 0.0001783354724161776,
+      "loss": 0.1601,
+      "step": 923
+    },
+    {
+      "epoch": 0.8772845953002611,
+      "grad_norm": 0.04881501942873001,
+      "learning_rate": 0.00017826673887703223,
+      "loss": 0.1686,
+      "step": 924
+    },
+    {
+      "epoch": 0.878234037502967,
+      "grad_norm": 0.0337185375392437,
+      "learning_rate": 0.00017819790976752718,
+      "loss": 0.131,
+      "step": 925
+    },
+    {
+      "epoch": 0.8791834797056729,
+      "grad_norm": 0.033597834408283234,
+      "learning_rate": 0.00017812898517170872,
+      "loss": 0.1365,
+      "step": 926
+    },
+    {
+      "epoch": 0.8801329219083788,
+      "grad_norm": 0.047949645668268204,
+      "learning_rate": 0.00017805996517373962,
+      "loss": 0.178,
+      "step": 927
+    },
+    {
+      "epoch": 0.8810823641110848,
+      "grad_norm": 0.03533579409122467,
+      "learning_rate": 0.00017799084985789916,
+      "loss": 0.1281,
+      "step": 928
+    },
+    {
+      "epoch": 0.8820318063137906,
+      "grad_norm": 0.03638564050197601,
+      "learning_rate": 0.0001779216393085831,
+      "loss": 0.136,
+      "step": 929
+    },
+    {
+      "epoch": 0.8829812485164965,
+      "grad_norm": 0.034585777670145035,
+      "learning_rate": 0.00017785233361030333,
+      "loss": 0.1221,
+      "step": 930
+    },
+    {
+      "epoch": 0.8839306907192025,
+      "grad_norm": 0.03344082459807396,
+      "learning_rate": 0.00017778293284768807,
+      "loss": 0.1335,
+      "step": 931
+    },
+    {
+      "epoch": 0.8848801329219084,
+      "grad_norm": 0.029832901433110237,
+      "learning_rate": 0.00017771343710548155,
+      "loss": 0.131,
+      "step": 932
+    },
+    {
+      "epoch": 0.8858295751246142,
+      "grad_norm": 0.030377686023712158,
+      "learning_rate": 0.00017764384646854405,
+      "loss": 0.1216,
+      "step": 933
+    },
+    {
+      "epoch": 0.8867790173273202,
+      "grad_norm": 0.036345310509204865,
+      "learning_rate": 0.0001775741610218516,
+      "loss": 0.1289,
+      "step": 934
+    },
+    {
+      "epoch": 0.8877284595300261,
+      "grad_norm": 0.04609441012144089,
+      "learning_rate": 0.00017750438085049606,
+      "loss": 0.1598,
+      "step": 935
+    },
+    {
+      "epoch": 0.888677901732732,
+      "grad_norm": 0.03439109027385712,
+      "learning_rate": 0.00017743450603968506,
+      "loss": 0.1316,
+      "step": 936
+    },
+    {
+      "epoch": 0.889627343935438,
+      "grad_norm": 0.07119124382734299,
+      "learning_rate": 0.0001773645366747416,
+      "loss": 0.1664,
+      "step": 937
+    },
+    {
+      "epoch": 0.8905767861381438,
+      "grad_norm": 0.03385334461927414,
+      "learning_rate": 0.0001772944728411043,
+      "loss": 0.1294,
+      "step": 938
+    },
+    {
+      "epoch": 0.8915262283408497,
+      "grad_norm": 0.033481206744909286,
+      "learning_rate": 0.00017722431462432705,
+      "loss": 0.1218,
+      "step": 939
+    },
+    {
+      "epoch": 0.8924756705435557,
+      "grad_norm": 0.03365306556224823,
+      "learning_rate": 0.00017715406211007902,
+      "loss": 0.1295,
+      "step": 940
+    },
+    {
+      "epoch": 0.8934251127462616,
+      "grad_norm": 0.03675035014748573,
+      "learning_rate": 0.0001770837153841445,
+      "loss": 0.1237,
+      "step": 941
+    },
+    {
+      "epoch": 0.8943745549489674,
+      "grad_norm": 0.03245026618242264,
+      "learning_rate": 0.00017701327453242284,
+      "loss": 0.1304,
+      "step": 942
+    },
+    {
+      "epoch": 0.8953239971516734,
+      "grad_norm": 0.03346354141831398,
+      "learning_rate": 0.00017694273964092837,
+      "loss": 0.1274,
+      "step": 943
+    },
+    {
+      "epoch": 0.8962734393543793,
+      "grad_norm": 0.048563096672296524,
+      "learning_rate": 0.00017687211079579017,
+      "loss": 0.1719,
+      "step": 944
+    },
+    {
+      "epoch": 0.8972228815570852,
+      "grad_norm": 0.04709222912788391,
+      "learning_rate": 0.0001768013880832521,
+      "loss": 0.1281,
+      "step": 945
+    },
+    {
+      "epoch": 0.8981723237597912,
+      "grad_norm": 0.030402177944779396,
+      "learning_rate": 0.00017673057158967254,
+      "loss": 0.1229,
+      "step": 946
+    },
+    {
+      "epoch": 0.899121765962497,
+      "grad_norm": 0.03577994927763939,
+      "learning_rate": 0.00017665966140152458,
+      "loss": 0.1255,
+      "step": 947
+    },
+    {
+      "epoch": 0.9000712081652029,
+      "grad_norm": 0.04566454887390137,
+      "learning_rate": 0.00017658865760539552,
+      "loss": 0.1617,
+      "step": 948
+    },
+    {
+      "epoch": 0.9010206503679089,
+      "grad_norm": 0.04077988117933273,
+      "learning_rate": 0.00017651756028798713,
+      "loss": 0.1619,
+      "step": 949
+    },
+    {
+      "epoch": 0.9019700925706148,
+      "grad_norm": 0.045764826238155365,
+      "learning_rate": 0.00017644636953611522,
+      "loss": 0.1608,
+      "step": 950
+    },
+    {
+      "epoch": 0.9029195347733207,
+      "grad_norm": 0.035656195133924484,
+      "learning_rate": 0.0001763750854367098,
+      "loss": 0.1288,
+      "step": 951
+    },
+    {
+      "epoch": 0.9038689769760266,
+      "grad_norm": 0.04220154508948326,
+      "learning_rate": 0.0001763037080768148,
+      "loss": 0.1688,
+      "step": 952
+    },
+    {
+      "epoch": 0.9048184191787325,
+      "grad_norm": 0.03406943380832672,
+      "learning_rate": 0.0001762322375435881,
+      "loss": 0.1314,
+      "step": 953
+    },
+    {
+      "epoch": 0.9057678613814384,
+      "grad_norm": 0.037942539900541306,
+      "learning_rate": 0.00017616067392430126,
+      "loss": 0.1342,
+      "step": 954
+    },
+    {
+      "epoch": 0.9067173035841443,
+      "grad_norm": 0.06412187963724136,
+      "learning_rate": 0.00017608901730633964,
+      "loss": 0.2207,
+      "step": 955
+    },
+    {
+      "epoch": 0.9076667457868502,
+      "grad_norm": 0.0313476026058197,
+      "learning_rate": 0.00017601726777720202,
+      "loss": 0.1249,
+      "step": 956
+    },
+    {
+      "epoch": 0.9086161879895561,
+      "grad_norm": 0.0276046060025692,
+      "learning_rate": 0.00017594542542450072,
+      "loss": 0.1212,
+      "step": 957
+    },
+    {
+      "epoch": 0.909565630192262,
+      "grad_norm": 0.032439909875392914,
+      "learning_rate": 0.00017587349033596134,
+      "loss": 0.1277,
+      "step": 958
+    },
+    {
+      "epoch": 0.910515072394968,
+      "grad_norm": 0.039732351899147034,
+      "learning_rate": 0.00017580146259942278,
+      "loss": 0.1222,
+      "step": 959
+    },
+    {
+      "epoch": 0.9114645145976739,
+      "grad_norm": 0.033820103853940964,
+      "learning_rate": 0.00017572934230283707,
+      "loss": 0.1246,
+      "step": 960
+    },
+    {
+      "epoch": 0.9124139568003797,
+      "grad_norm": 0.03361973166465759,
+      "learning_rate": 0.00017565712953426918,
+      "loss": 0.1328,
+      "step": 961
+    },
+    {
+      "epoch": 0.9133633990030857,
+      "grad_norm": 0.0338444709777832,
+      "learning_rate": 0.00017558482438189712,
+      "loss": 0.1306,
+      "step": 962
+    },
+    {
+      "epoch": 0.9143128412057916,
+      "grad_norm": 0.04851710423827171,
+      "learning_rate": 0.0001755124269340116,
+      "loss": 0.1765,
+      "step": 963
+    },
+    {
+      "epoch": 0.9152622834084975,
+      "grad_norm": 0.03290700539946556,
+      "learning_rate": 0.0001754399372790161,
+      "loss": 0.1386,
+      "step": 964
+    },
+    {
+      "epoch": 0.9162117256112035,
+      "grad_norm": 0.034565720707178116,
+      "learning_rate": 0.00017536735550542661,
+      "loss": 0.1212,
+      "step": 965
+    },
+    {
+      "epoch": 0.9171611678139093,
+      "grad_norm": 0.04606771841645241,
+      "learning_rate": 0.00017529468170187176,
+      "loss": 0.1567,
+      "step": 966
+    },
+    {
+      "epoch": 0.9181106100166152,
+      "grad_norm": 0.03279464691877365,
+      "learning_rate": 0.00017522191595709238,
+      "loss": 0.1214,
+      "step": 967
+    },
+    {
+      "epoch": 0.9190600522193212,
+      "grad_norm": 0.036700885742902756,
+      "learning_rate": 0.00017514905835994168,
+      "loss": 0.1314,
+      "step": 968
+    },
+    {
+      "epoch": 0.9200094944220271,
+      "grad_norm": 0.04098424315452576,
+      "learning_rate": 0.00017507610899938501,
+      "loss": 0.164,
+      "step": 969
+    },
+    {
+      "epoch": 0.9209589366247329,
+      "grad_norm": 0.033782679587602615,
+      "learning_rate": 0.0001750030679644997,
+      "loss": 0.1376,
+      "step": 970
+    },
+    {
+      "epoch": 0.9219083788274389,
+      "grad_norm": 0.03304159641265869,
+      "learning_rate": 0.00017492993534447515,
+      "loss": 0.1244,
+      "step": 971
+    },
+    {
+      "epoch": 0.9228578210301448,
+      "grad_norm": 0.03158386051654816,
+      "learning_rate": 0.0001748567112286125,
+      "loss": 0.1345,
+      "step": 972
+    },
+    {
+      "epoch": 0.9238072632328507,
+      "grad_norm": 0.03615015745162964,
+      "learning_rate": 0.00017478339570632458,
+      "loss": 0.1434,
+      "step": 973
+    },
+    {
+      "epoch": 0.9247567054355567,
+      "grad_norm": 0.033553823828697205,
+      "learning_rate": 0.00017470998886713596,
+      "loss": 0.1292,
+      "step": 974
+    },
+    {
+      "epoch": 0.9257061476382625,
+      "grad_norm": 0.03953874111175537,
+      "learning_rate": 0.00017463649080068266,
+      "loss": 0.1621,
+      "step": 975
+    },
+    {
+      "epoch": 0.9266555898409684,
+      "grad_norm": 0.03288433700799942,
+      "learning_rate": 0.00017456290159671202,
+      "loss": 0.1357,
+      "step": 976
+    },
+    {
+      "epoch": 0.9276050320436744,
+      "grad_norm": 0.03154657408595085,
+      "learning_rate": 0.00017448922134508275,
+      "loss": 0.1322,
+      "step": 977
+    },
+    {
+      "epoch": 0.9285544742463803,
+      "grad_norm": 0.05669796094298363,
+      "learning_rate": 0.00017441545013576477,
+      "loss": 0.1761,
+      "step": 978
+    },
+    {
+      "epoch": 0.9295039164490861,
+      "grad_norm": 0.026679178699851036,
+      "learning_rate": 0.00017434158805883896,
+      "loss": 0.1295,
+      "step": 979
+    },
+    {
+      "epoch": 0.930453358651792,
+      "grad_norm": 0.03597673401236534,
+      "learning_rate": 0.00017426763520449721,
+      "loss": 0.1265,
+      "step": 980
+    },
+    {
+      "epoch": 0.931402800854498,
+      "grad_norm": 0.03097674809396267,
+      "learning_rate": 0.0001741935916630423,
+      "loss": 0.1339,
+      "step": 981
+    },
+    {
+      "epoch": 0.9323522430572039,
+      "grad_norm": 0.030252935364842415,
+      "learning_rate": 0.00017411945752488766,
+      "loss": 0.1247,
+      "step": 982
+    },
+    {
+      "epoch": 0.9333016852599098,
+      "grad_norm": 0.03460918739438057,
+      "learning_rate": 0.00017404523288055743,
+      "loss": 0.1323,
+      "step": 983
+    },
+    {
+      "epoch": 0.9342511274626157,
+      "grad_norm": 0.035575591027736664,
+      "learning_rate": 0.00017397091782068622,
+      "loss": 0.1258,
+      "step": 984
+    },
+    {
+      "epoch": 0.9352005696653216,
+      "grad_norm": 0.05128021538257599,
+      "learning_rate": 0.00017389651243601904,
+      "loss": 0.1364,
+      "step": 985
+    },
+    {
+      "epoch": 0.9361500118680275,
+      "grad_norm": 0.04355672374367714,
+      "learning_rate": 0.00017382201681741122,
+      "loss": 0.1656,
+      "step": 986
+    },
+    {
+      "epoch": 0.9370994540707335,
+      "grad_norm": 0.03357682749629021,
+      "learning_rate": 0.0001737474310558282,
+      "loss": 0.1285,
+      "step": 987
+    },
+    {
+      "epoch": 0.9380488962734393,
+      "grad_norm": 0.10623644292354584,
+      "learning_rate": 0.00017367275524234565,
+      "loss": 0.1726,
+      "step": 988
+    },
+    {
+      "epoch": 0.9389983384761452,
+      "grad_norm": 0.03605256229639053,
+      "learning_rate": 0.00017359798946814907,
+      "loss": 0.1358,
+      "step": 989
+    },
+    {
+      "epoch": 0.9399477806788512,
+      "grad_norm": 0.039663393050432205,
+      "learning_rate": 0.00017352313382453378,
+      "loss": 0.1299,
+      "step": 990
+    },
+    {
+      "epoch": 0.9408972228815571,
+      "grad_norm": 0.12416961044073105,
+      "learning_rate": 0.000173448188402905,
+      "loss": 0.1666,
+      "step": 991
+    },
+    {
+      "epoch": 0.941846665084263,
+      "grad_norm": 0.045010216534137726,
+      "learning_rate": 0.00017337315329477742,
+      "loss": 0.1733,
+      "step": 992
+    },
+    {
+      "epoch": 0.9427961072869689,
+      "grad_norm": 0.03456486761569977,
+      "learning_rate": 0.0001732980285917753,
+      "loss": 0.1312,
+      "step": 993
+    },
+    {
+      "epoch": 0.9437455494896748,
+      "grad_norm": 0.039561979472637177,
+      "learning_rate": 0.00017322281438563234,
+      "loss": 0.1354,
+      "step": 994
+    },
+    {
+      "epoch": 0.9446949916923807,
+      "grad_norm": 0.043275121599435806,
+      "learning_rate": 0.00017314751076819146,
+      "loss": 0.1651,
+      "step": 995
+    },
+    {
+      "epoch": 0.9456444338950867,
+      "grad_norm": 0.0392397940158844,
+      "learning_rate": 0.00017307211783140482,
+      "loss": 0.1647,
+      "step": 996
+    },
+    {
+      "epoch": 0.9465938760977926,
+      "grad_norm": 0.03428703919053078,
+      "learning_rate": 0.0001729966356673336,
+      "loss": 0.128,
+      "step": 997
+    },
+    {
+      "epoch": 0.9475433183004984,
+      "grad_norm": 0.03511650487780571,
+      "learning_rate": 0.000172921064368148,
+      "loss": 0.1297,
+      "step": 998
+    },
+    {
+      "epoch": 0.9484927605032044,
+      "grad_norm": 0.030319994315505028,
+      "learning_rate": 0.00017284540402612696,
+      "loss": 0.1269,
+      "step": 999
+    },
+    {
+      "epoch": 0.9494422027059103,
+      "grad_norm": 0.03071141429245472,
+      "learning_rate": 0.00017276965473365827,
+      "loss": 0.1224,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9503916449086162,
+      "grad_norm": 0.04097789525985718,
+      "learning_rate": 0.00017269381658323822,
+      "loss": 0.1597,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9513410871113221,
+      "grad_norm": 0.03407077491283417,
+      "learning_rate": 0.00017261788966747168,
+      "loss": 0.1268,
+      "step": 1002
+    },
+    {
+      "epoch": 0.952290529314028,
+      "grad_norm": 0.035802800208330154,
+      "learning_rate": 0.00017254187407907189,
+      "loss": 0.1338,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9532399715167339,
+      "grad_norm": 0.030097633600234985,
+      "learning_rate": 0.00017246576991086034,
+      "loss": 0.1222,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9541894137194399,
+      "grad_norm": 0.047994308173656464,
+      "learning_rate": 0.0001723895772557667,
+      "loss": 0.1632,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9551388559221458,
+      "grad_norm": 0.03451845049858093,
+      "learning_rate": 0.00017231329620682876,
+      "loss": 0.1278,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9560882981248516,
+      "grad_norm": 0.036820750683546066,
+      "learning_rate": 0.00017223692685719213,
+      "loss": 0.1355,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9570377403275575,
+      "grad_norm": 0.03521284461021423,
+      "learning_rate": 0.0001721604693001103,
+      "loss": 0.1383,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9579871825302635,
+      "grad_norm": 0.036953702569007874,
+      "learning_rate": 0.00017208392362894447,
+      "loss": 0.1352,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9589366247329694,
+      "grad_norm": 0.031185979023575783,
+      "learning_rate": 0.00017200728993716345,
+      "loss": 0.1262,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9598860669356752,
+      "grad_norm": 0.030822455883026123,
+      "learning_rate": 0.00017193056831834346,
+      "loss": 0.1211,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9608355091383812,
+      "grad_norm": 0.031467005610466,
+      "learning_rate": 0.0001718537588661682,
+      "loss": 0.1271,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9617849513410871,
+      "grad_norm": 0.03788928687572479,
+      "learning_rate": 0.0001717768616744285,
+      "loss": 0.1413,
+      "step": 1013
+    },
+    {
+      "epoch": 0.962734393543793,
+      "grad_norm": 0.03359632566571236,
+      "learning_rate": 0.00017169987683702243,
+      "loss": 0.1276,
+      "step": 1014
+    },
+    {
+      "epoch": 0.963683835746499,
+      "grad_norm": 0.03274601325392723,
+      "learning_rate": 0.000171622804447955,
+      "loss": 0.1308,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9646332779492048,
+      "grad_norm": 0.03634633496403694,
+      "learning_rate": 0.0001715456446013382,
+      "loss": 0.1384,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9655827201519107,
+      "grad_norm": 0.02978476695716381,
+      "learning_rate": 0.00017146839739139077,
+      "loss": 0.1301,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9665321623546167,
+      "grad_norm": 0.03389682248234749,
+      "learning_rate": 0.0001713910629124381,
+      "loss": 0.1264,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9674816045573226,
+      "grad_norm": 0.03452256694436073,
+      "learning_rate": 0.00017131364125891224,
+      "loss": 0.1317,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9684310467600284,
+      "grad_norm": 0.03967840224504471,
+      "learning_rate": 0.00017123613252535163,
+      "loss": 0.1308,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9693804889627344,
+      "grad_norm": 0.04021480306982994,
+      "learning_rate": 0.00017115853680640098,
+      "loss": 0.1637,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9703299311654403,
+      "grad_norm": 0.02766057476401329,
+      "learning_rate": 0.00017108085419681132,
+      "loss": 0.1239,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9712793733681462,
+      "grad_norm": 0.029945319518446922,
+      "learning_rate": 0.00017100308479143974,
+      "loss": 0.1236,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9722288155708522,
+      "grad_norm": 0.03135136887431145,
+      "learning_rate": 0.00017092522868524928,
+      "loss": 0.1203,
+      "step": 1024
+    },
+    {
+      "epoch": 0.973178257773558,
+      "grad_norm": 0.04876153543591499,
+      "learning_rate": 0.00017084728597330893,
+      "loss": 0.1802,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9741276999762639,
+      "grad_norm": 0.042958084493875504,
+      "learning_rate": 0.00017076925675079335,
+      "loss": 0.1656,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9750771421789699,
+      "grad_norm": 0.04739035665988922,
+      "learning_rate": 0.00017069114111298287,
+      "loss": 0.167,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9760265843816758,
+      "grad_norm": 0.042968571186065674,
+      "learning_rate": 0.00017061293915526335,
+      "loss": 0.173,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9769760265843817,
+      "grad_norm": 0.07628528028726578,
+      "learning_rate": 0.00017053465097312606,
+      "loss": 0.1351,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9779254687870876,
+      "grad_norm": 0.032479528337717056,
+      "learning_rate": 0.00017045627666216755,
+      "loss": 0.1294,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9788749109897935,
+      "grad_norm": 0.029842333868145943,
+      "learning_rate": 0.0001703778163180895,
+      "loss": 0.1264,
+      "step": 1031
+    },
+    {
+      "epoch": 0.9798243531924994,
+      "grad_norm": 0.03622937202453613,
+      "learning_rate": 0.00017029927003669868,
+      "loss": 0.1287,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9807737953952054,
+      "grad_norm": 0.05245399475097656,
+      "learning_rate": 0.00017022063791390684,
+      "loss": 0.1923,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9817232375979112,
+      "grad_norm": 0.03335704281926155,
+      "learning_rate": 0.00017014192004573047,
+      "loss": 0.1241,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9826726798006171,
+      "grad_norm": 0.03645642474293709,
+      "learning_rate": 0.0001700631165282908,
+      "loss": 0.1321,
+      "step": 1035
+    },
+    {
+      "epoch": 0.983622122003323,
+      "grad_norm": 0.05331774801015854,
+      "learning_rate": 0.00016998422745781363,
+      "loss": 0.169,
+      "step": 1036
+    },
+    {
+      "epoch": 0.984571564206029,
+      "grad_norm": 0.04615236446261406,
+      "learning_rate": 0.00016990525293062927,
+      "loss": 0.1623,
+      "step": 1037
+    },
+    {
+      "epoch": 0.9855210064087349,
+      "grad_norm": 0.047434594482183456,
+      "learning_rate": 0.00016982619304317233,
+      "loss": 0.1303,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9864704486114407,
+      "grad_norm": 0.03144746273756027,
+      "learning_rate": 0.00016974704789198168,
+      "loss": 0.1203,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9874198908141467,
+      "grad_norm": 0.04035501554608345,
+      "learning_rate": 0.00016966781757370028,
+      "loss": 0.1246,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9883693330168526,
+      "grad_norm": 0.03864790499210358,
+      "learning_rate": 0.0001695885021850751,
+      "loss": 0.1305,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9893187752195585,
+      "grad_norm": 0.03547806292772293,
+      "learning_rate": 0.00016950910182295705,
+      "loss": 0.1319,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9902682174222645,
+      "grad_norm": 0.03442002460360527,
+      "learning_rate": 0.0001694296165843007,
+      "loss": 0.1344,
+      "step": 1043
+    },
+    {
+      "epoch": 0.9912176596249703,
+      "grad_norm": 0.0333750881254673,
+      "learning_rate": 0.00016935004656616425,
+      "loss": 0.1278,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9921671018276762,
+      "grad_norm": 0.03143637254834175,
+      "learning_rate": 0.00016927039186570954,
+      "loss": 0.1237,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9931165440303822,
+      "grad_norm": 0.03841651603579521,
+      "learning_rate": 0.0001691906525802017,
+      "loss": 0.1395,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9940659862330881,
+      "grad_norm": 0.03443494066596031,
+      "learning_rate": 0.00016911082880700926,
+      "loss": 0.1422,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9950154284357939,
+      "grad_norm": 0.027661804109811783,
+      "learning_rate": 0.0001690309206436038,
+      "loss": 0.1251,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9959648706384999,
+      "grad_norm": 0.036862559616565704,
+      "learning_rate": 0.00016895092818756006,
+      "loss": 0.1337,
+      "step": 1049
+    },
+    {
+      "epoch": 0.9969143128412058,
+      "grad_norm": 0.035234466195106506,
+      "learning_rate": 0.00016887085153655554,
+      "loss": 0.1267,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9978637550439117,
+      "grad_norm": 0.032372791320085526,
+      "learning_rate": 0.00016879069078837075,
+      "loss": 0.1254,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9988131972466177,
+      "grad_norm": 0.037299785763025284,
+      "learning_rate": 0.00016871044604088877,
+      "loss": 0.1324,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9997626394493235,
+      "grad_norm": 0.03843718767166138,
+      "learning_rate": 0.00016863011739209527,
+      "loss": 0.1328,
+      "step": 1053
+    },
+    {
+      "epoch": 1.0007120816520294,
+      "grad_norm": 0.03160862624645233,
+      "learning_rate": 0.00016854970494007836,
+      "loss": 0.1284,
+      "step": 1054
+    },
+    {
+      "epoch": 1.0016615238547353,
+      "grad_norm": 0.05188068002462387,
+      "learning_rate": 0.00016846920878302852,
+      "loss": 0.1775,
+      "step": 1055
+    },
+    {
+      "epoch": 1.0026109660574412,
+      "grad_norm": 0.04362662881612778,
+      "learning_rate": 0.00016838862901923842,
+      "loss": 0.1577,
+      "step": 1056
+    },
+    {
+      "epoch": 1.0035604082601473,
+      "grad_norm": 0.033426132053136826,
+      "learning_rate": 0.00016830796574710284,
+      "loss": 0.1252,
+      "step": 1057
+    },
+    {
+      "epoch": 1.0045098504628531,
+      "grad_norm": 0.06085265800356865,
+      "learning_rate": 0.00016822721906511844,
+      "loss": 0.1769,
+      "step": 1058
+    },
+    {
+      "epoch": 1.005459292665559,
+      "grad_norm": 0.03222273662686348,
+      "learning_rate": 0.00016814638907188388,
+      "loss": 0.1239,
+      "step": 1059
+    },
+    {
+      "epoch": 1.0064087348682649,
+      "grad_norm": 0.032014038413763046,
+      "learning_rate": 0.00016806547586609947,
+      "loss": 0.1191,
+      "step": 1060
+    },
+    {
+      "epoch": 1.0073581770709708,
+      "grad_norm": 0.03323471546173096,
+      "learning_rate": 0.00016798447954656707,
+      "loss": 0.1334,
+      "step": 1061
+    },
+    {
+      "epoch": 1.0083076192736766,
+      "grad_norm": 0.04325219243764877,
+      "learning_rate": 0.0001679034002121901,
+      "loss": 0.1623,
+      "step": 1062
+    },
+    {
+      "epoch": 1.0092570614763827,
+      "grad_norm": 0.029746338725090027,
+      "learning_rate": 0.0001678222379619734,
+      "loss": 0.1292,
+      "step": 1063
+    },
+    {
+      "epoch": 1.0102065036790886,
+      "grad_norm": 0.03265037387609482,
+      "learning_rate": 0.00016774099289502297,
+      "loss": 0.1271,
+      "step": 1064
+    },
+    {
+      "epoch": 1.0111559458817945,
+      "grad_norm": 0.04023383557796478,
+      "learning_rate": 0.0001676596651105459,
+      "loss": 0.1537,
+      "step": 1065
+    },
+    {
+      "epoch": 1.0121053880845003,
+      "grad_norm": 0.036106862127780914,
+      "learning_rate": 0.00016757825470785042,
+      "loss": 0.1237,
+      "step": 1066
+    },
+    {
+      "epoch": 1.0130548302872062,
+      "grad_norm": 0.04061293974518776,
+      "learning_rate": 0.00016749676178634556,
+      "loss": 0.1324,
+      "step": 1067
+    },
+    {
+      "epoch": 1.014004272489912,
+      "grad_norm": 0.050820399075746536,
+      "learning_rate": 0.0001674151864455411,
+      "loss": 0.1705,
+      "step": 1068
+    },
+    {
+      "epoch": 1.0149537146926182,
+      "grad_norm": 0.037347592413425446,
+      "learning_rate": 0.00016733352878504752,
+      "loss": 0.1248,
+      "step": 1069
+    },
+    {
+      "epoch": 1.015903156895324,
+      "grad_norm": 0.04108656942844391,
+      "learning_rate": 0.00016725178890457571,
+      "loss": 0.1201,
+      "step": 1070
+    },
+    {
+      "epoch": 1.01685259909803,
+      "grad_norm": 0.051215577870607376,
+      "learning_rate": 0.00016716996690393715,
+      "loss": 0.1705,
+      "step": 1071
+    },
+    {
+      "epoch": 1.0178020413007358,
+      "grad_norm": 0.05008477717638016,
+      "learning_rate": 0.00016708806288304336,
+      "loss": 0.1606,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0187514835034417,
+      "grad_norm": 0.05916628614068031,
+      "learning_rate": 0.00016700607694190617,
+      "loss": 0.1824,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0197009257061476,
+      "grad_norm": 0.03331366181373596,
+      "learning_rate": 0.00016692400918063744,
+      "loss": 0.1256,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0206503679088534,
+      "grad_norm": 0.03364944830536842,
+      "learning_rate": 0.00016684185969944885,
+      "loss": 0.1273,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0215998101115595,
+      "grad_norm": 0.02990981563925743,
+      "learning_rate": 0.000166759628598652,
+      "loss": 0.1284,
+      "step": 1076
+    },
+    {
+      "epoch": 1.0225492523142654,
+      "grad_norm": 0.03323819115757942,
+      "learning_rate": 0.00016667731597865796,
+      "loss": 0.1258,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0234986945169713,
+      "grad_norm": 0.03008713200688362,
+      "learning_rate": 0.0001665949219399775,
+      "loss": 0.1244,
+      "step": 1078
+    },
+    {
+      "epoch": 1.0244481367196772,
+      "grad_norm": 0.04623178020119667,
+      "learning_rate": 0.00016651244658322085,
+      "loss": 0.1537,
+      "step": 1079
+    },
+    {
+      "epoch": 1.025397578922383,
+      "grad_norm": 0.034869614988565445,
+      "learning_rate": 0.00016642989000909732,
+      "loss": 0.1272,
+      "step": 1080
+    },
+    {
+      "epoch": 1.026347021125089,
+      "grad_norm": 0.03236447647213936,
+      "learning_rate": 0.0001663472523184156,
+      "loss": 0.1299,
+      "step": 1081
+    },
+    {
+      "epoch": 1.027296463327795,
+      "grad_norm": 0.02806561440229416,
+      "learning_rate": 0.00016626453361208335,
+      "loss": 0.1198,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0282459055305009,
+      "grad_norm": 0.04762514680624008,
+      "learning_rate": 0.0001661817339911071,
+      "loss": 0.1695,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0291953477332068,
+      "grad_norm": 0.039319079369306564,
+      "learning_rate": 0.00016609885355659234,
+      "loss": 0.1612,
+      "step": 1084
+    },
+    {
+      "epoch": 1.0301447899359126,
+      "grad_norm": 0.030540715903043747,
+      "learning_rate": 0.0001660158924097431,
+      "loss": 0.1251,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0310942321386185,
+      "grad_norm": 0.029828663915395737,
+      "learning_rate": 0.000165932850651862,
+      "loss": 0.1287,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0320436743413244,
+      "grad_norm": 0.030012918636202812,
+      "learning_rate": 0.0001658497283843501,
+      "loss": 0.132,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0329931165440305,
+      "grad_norm": 0.03255194425582886,
+      "learning_rate": 0.0001657665257087068,
+      "loss": 0.1257,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0339425587467364,
+      "grad_norm": 0.040951523929834366,
+      "learning_rate": 0.00016568324272652965,
+      "loss": 0.1507,
+      "step": 1089
+    },
+    {
+      "epoch": 1.0348920009494422,
+      "grad_norm": 0.027678990736603737,
+      "learning_rate": 0.00016559987953951427,
+      "loss": 0.1217,
+      "step": 1090
+    },
+    {
+      "epoch": 1.035841443152148,
+      "grad_norm": 0.03241724148392677,
+      "learning_rate": 0.0001655164362494542,
+      "loss": 0.1298,
+      "step": 1091
+    },
+    {
+      "epoch": 1.036790885354854,
+      "grad_norm": 0.038156237453222275,
+      "learning_rate": 0.00016543291295824085,
+      "loss": 0.1252,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0377403275575598,
+      "grad_norm": 0.029806343838572502,
+      "learning_rate": 0.00016534930976786323,
+      "loss": 0.1265,
+      "step": 1093
+    },
+    {
+      "epoch": 1.038689769760266,
+      "grad_norm": 0.035036977380514145,
+      "learning_rate": 0.00016526562678040804,
+      "loss": 0.1247,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0396392119629718,
+      "grad_norm": 0.032223109155893326,
+      "learning_rate": 0.00016518186409805922,
+      "loss": 0.1326,
+      "step": 1095
+    },
+    {
+      "epoch": 1.0405886541656777,
+      "grad_norm": 0.03192323073744774,
+      "learning_rate": 0.0001650980218230982,
+      "loss": 0.1186,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0415380963683836,
+      "grad_norm": 0.031004801392555237,
+      "learning_rate": 0.00016501410005790362,
+      "loss": 0.1292,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0424875385710894,
+      "grad_norm": 0.03421878442168236,
+      "learning_rate": 0.00016493009890495102,
+      "loss": 0.1362,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0434369807737953,
+      "grad_norm": 0.03153158724308014,
+      "learning_rate": 0.00016484601846681297,
+      "loss": 0.1182,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0443864229765012,
+      "grad_norm": 0.03977439925074577,
+      "learning_rate": 0.0001647618588461589,
+      "loss": 0.1327,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0453358651792073,
+      "grad_norm": 0.03982316702604294,
+      "learning_rate": 0.00016467762014575485,
+      "loss": 0.1582,
+      "step": 1101
+    },
+    {
+      "epoch": 1.0462853073819132,
+      "grad_norm": 0.034796085208654404,
+      "learning_rate": 0.00016459330246846348,
+      "loss": 0.1258,
+      "step": 1102
+    },
+    {
+      "epoch": 1.047234749584619,
+      "grad_norm": 0.039261046797037125,
+      "learning_rate": 0.0001645089059172438,
+      "loss": 0.1321,
+      "step": 1103
+    },
+    {
+      "epoch": 1.048184191787325,
+      "grad_norm": 0.04305882379412651,
+      "learning_rate": 0.00016442443059515126,
+      "loss": 0.1406,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0491336339900308,
+      "grad_norm": 0.03491320461034775,
+      "learning_rate": 0.00016433987660533742,
+      "loss": 0.1312,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0500830761927367,
+      "grad_norm": 0.04404641315340996,
+      "learning_rate": 0.00016425524405104986,
+      "loss": 0.1267,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0510325183954428,
+      "grad_norm": 0.034407854080200195,
+      "learning_rate": 0.0001641705330356322,
+      "loss": 0.1268,
+      "step": 1107
+    },
+    {
+      "epoch": 1.0519819605981486,
+      "grad_norm": 0.04843935742974281,
+      "learning_rate": 0.00016408574366252374,
+      "loss": 0.1601,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0529314028008545,
+      "grad_norm": 0.03394000977277756,
+      "learning_rate": 0.0001640008760352596,
+      "loss": 0.13,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0538808450035604,
+      "grad_norm": 0.027672087773680687,
+      "learning_rate": 0.00016391593025747038,
+      "loss": 0.1202,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0548302872062663,
+      "grad_norm": 0.03761329874396324,
+      "learning_rate": 0.0001638309064328821,
+      "loss": 0.1562,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0557797294089721,
+      "grad_norm": 0.048850156366825104,
+      "learning_rate": 0.0001637458046653161,
+      "loss": 0.1707,
+      "step": 1112
+    },
+    {
+      "epoch": 1.0567291716116782,
+      "grad_norm": 0.027066387236118317,
+      "learning_rate": 0.00016366062505868888,
+      "loss": 0.1204,
+      "step": 1113
+    },
+    {
+      "epoch": 1.057678613814384,
+      "grad_norm": 0.034062668681144714,
+      "learning_rate": 0.00016357536771701198,
+      "loss": 0.1378,
+      "step": 1114
+    },
+    {
+      "epoch": 1.05862805601709,
+      "grad_norm": 0.0422850139439106,
+      "learning_rate": 0.00016349003274439194,
+      "loss": 0.1583,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0595774982197959,
+      "grad_norm": 0.0334283784031868,
+      "learning_rate": 0.00016340462024503,
+      "loss": 0.1276,
+      "step": 1116
+    },
+    {
+      "epoch": 1.0605269404225017,
+      "grad_norm": 0.03338415175676346,
+      "learning_rate": 0.00016331913032322212,
+      "loss": 0.1229,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0614763826252076,
+      "grad_norm": 0.03128555044531822,
+      "learning_rate": 0.00016323356308335876,
+      "loss": 0.1167,
+      "step": 1118
+    },
+    {
+      "epoch": 1.0624258248279137,
+      "grad_norm": 0.033790841698646545,
+      "learning_rate": 0.00016314791862992486,
+      "loss": 0.1236,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0633752670306196,
+      "grad_norm": 0.03544427454471588,
+      "learning_rate": 0.00016306219706749953,
+      "loss": 0.1319,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0643247092333254,
+      "grad_norm": 0.03969413787126541,
+      "learning_rate": 0.0001629763985007561,
+      "loss": 0.1612,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0652741514360313,
+      "grad_norm": 0.042924992740154266,
+      "learning_rate": 0.00016289052303446202,
+      "loss": 0.1659,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0662235936387372,
+      "grad_norm": 0.04624541476368904,
+      "learning_rate": 0.00016280457077347848,
+      "loss": 0.1617,
+      "step": 1123
+    },
+    {
+      "epoch": 1.067173035841443,
+      "grad_norm": 0.034341566264629364,
+      "learning_rate": 0.00016271854182276058,
+      "loss": 0.1188,
+      "step": 1124
+    },
+    {
+      "epoch": 1.068122478044149,
+      "grad_norm": 0.03228682279586792,
+      "learning_rate": 0.00016263243628735695,
+      "loss": 0.129,
+      "step": 1125
+    },
+    {
+      "epoch": 1.069071920246855,
+      "grad_norm": 0.036037541925907135,
+      "learning_rate": 0.00016254625427240978,
+      "loss": 0.1309,
+      "step": 1126
+    },
+    {
+      "epoch": 1.070021362449561,
+      "grad_norm": 0.027421532198786736,
+      "learning_rate": 0.0001624599958831547,
+      "loss": 0.1176,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0709708046522668,
+      "grad_norm": 0.030262261629104614,
+      "learning_rate": 0.00016237366122492052,
+      "loss": 0.1195,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0719202468549727,
+      "grad_norm": 0.041230421513319016,
+      "learning_rate": 0.00016228725040312925,
+      "loss": 0.1562,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0728696890576785,
+      "grad_norm": 0.03141395375132561,
+      "learning_rate": 0.00016220076352329582,
+      "loss": 0.13,
+      "step": 1130
+    },
+    {
+      "epoch": 1.0738191312603846,
+      "grad_norm": 0.0343187153339386,
+      "learning_rate": 0.00016211420069102815,
+      "loss": 0.134,
+      "step": 1131
+    },
+    {
+      "epoch": 1.0747685734630905,
+      "grad_norm": 0.04862483590841293,
+      "learning_rate": 0.0001620275620120268,
+      "loss": 0.1574,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0757180156657964,
+      "grad_norm": 0.04204042628407478,
+      "learning_rate": 0.00016194084759208494,
+      "loss": 0.162,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0766674578685023,
+      "grad_norm": 0.03309663385152817,
+      "learning_rate": 0.00016185405753708833,
+      "loss": 0.1251,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0776169000712081,
+      "grad_norm": 0.03191671893000603,
+      "learning_rate": 0.00016176719195301503,
+      "loss": 0.125,
+      "step": 1135
+    },
+    {
+      "epoch": 1.078566342273914,
+      "grad_norm": 0.036822427064180374,
+      "learning_rate": 0.0001616802509459353,
+      "loss": 0.1484,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0795157844766199,
+      "grad_norm": 0.029125772416591644,
+      "learning_rate": 0.00016159323462201149,
+      "loss": 0.1192,
+      "step": 1137
+    },
+    {
+      "epoch": 1.080465226679326,
+      "grad_norm": 0.034059688448905945,
+      "learning_rate": 0.000161506143087498,
+      "loss": 0.1309,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0814146688820319,
+      "grad_norm": 0.03434915095567703,
+      "learning_rate": 0.00016141897644874096,
+      "loss": 0.1336,
+      "step": 1139
+    },
+    {
+      "epoch": 1.0823641110847377,
+      "grad_norm": 0.0348944216966629,
+      "learning_rate": 0.00016133173481217833,
+      "loss": 0.1317,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0833135532874436,
+      "grad_norm": 0.033639729022979736,
+      "learning_rate": 0.00016124441828433957,
+      "loss": 0.1187,
+      "step": 1141
+    },
+    {
+      "epoch": 1.0842629954901495,
+      "grad_norm": 0.03063533827662468,
+      "learning_rate": 0.00016115702697184556,
+      "loss": 0.1332,
+      "step": 1142
+    },
+    {
+      "epoch": 1.0852124376928554,
+      "grad_norm": 0.03273540362715721,
+      "learning_rate": 0.00016106956098140858,
+      "loss": 0.1284,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0861618798955615,
+      "grad_norm": 0.05293993651866913,
+      "learning_rate": 0.00016098202041983206,
+      "loss": 0.1687,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0871113220982673,
+      "grad_norm": 0.03251373767852783,
+      "learning_rate": 0.00016089440539401046,
+      "loss": 0.1252,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0880607643009732,
+      "grad_norm": 0.0367170013487339,
+      "learning_rate": 0.00016080671601092922,
+      "loss": 0.1419,
+      "step": 1146
+    },
+    {
+      "epoch": 1.089010206503679,
+      "grad_norm": 0.030752060934901237,
+      "learning_rate": 0.00016071895237766457,
+      "loss": 0.1257,
+      "step": 1147
+    },
+    {
+      "epoch": 1.089959648706385,
+      "grad_norm": 0.035168685019016266,
+      "learning_rate": 0.00016063111460138334,
+      "loss": 0.1385,
+      "step": 1148
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 0.03252134099602699,
+      "learning_rate": 0.00016054320278934296,
+      "loss": 0.1232,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0918585331117967,
+      "grad_norm": 0.028666459023952484,
+      "learning_rate": 0.00016045521704889128,
+      "loss": 0.1242,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0928079753145028,
+      "grad_norm": 0.047707218676805496,
+      "learning_rate": 0.00016036715748746634,
+      "loss": 0.1643,
+      "step": 1151
+    },
+    {
+      "epoch": 1.0937574175172087,
+      "grad_norm": 0.035980336368083954,
+      "learning_rate": 0.00016027902421259638,
+      "loss": 0.1329,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0947068597199145,
+      "grad_norm": 0.04506576433777809,
+      "learning_rate": 0.00016019081733189967,
+      "loss": 0.1631,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0956563019226204,
+      "grad_norm": 0.030268298462033272,
+      "learning_rate": 0.0001601025369530843,
+      "loss": 0.1319,
+      "step": 1154
+    },
+    {
+      "epoch": 1.0966057441253263,
+      "grad_norm": 0.056095585227012634,
+      "learning_rate": 0.00016001418318394817,
+      "loss": 0.1529,
+      "step": 1155
+    },
+    {
+      "epoch": 1.0975551863280324,
+      "grad_norm": 0.029666630551218987,
+      "learning_rate": 0.0001599257561323787,
+      "loss": 0.126,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0985046285307383,
+      "grad_norm": 0.03648681938648224,
+      "learning_rate": 0.00015983725590635293,
+      "loss": 0.1378,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0994540707334441,
+      "grad_norm": 0.03170529007911682,
+      "learning_rate": 0.00015974868261393714,
+      "loss": 0.1238,
+      "step": 1158
+    },
+    {
+      "epoch": 1.10040351293615,
+      "grad_norm": 0.032316990196704865,
+      "learning_rate": 0.0001596600363632869,
+      "loss": 0.1305,
+      "step": 1159
+    },
+    {
+      "epoch": 1.101352955138856,
+      "grad_norm": 0.03148328512907028,
+      "learning_rate": 0.00015957131726264677,
+      "loss": 0.1303,
+      "step": 1160
+    },
+    {
+      "epoch": 1.1023023973415618,
+      "grad_norm": 0.03739064186811447,
+      "learning_rate": 0.00015948252542035042,
+      "loss": 0.16,
+      "step": 1161
+    },
+    {
+      "epoch": 1.1032518395442676,
+      "grad_norm": 0.034856993705034256,
+      "learning_rate": 0.00015939366094482025,
+      "loss": 0.1273,
+      "step": 1162
+    },
+    {
+      "epoch": 1.1042012817469737,
+      "grad_norm": 0.03102080523967743,
+      "learning_rate": 0.0001593047239445673,
+      "loss": 0.1331,
+      "step": 1163
+    },
+    {
+      "epoch": 1.1051507239496796,
+      "grad_norm": 0.026448125019669533,
+      "learning_rate": 0.00015921571452819127,
+      "loss": 0.1241,
+      "step": 1164
+    },
+    {
+      "epoch": 1.1061001661523855,
+      "grad_norm": 0.034301795065402985,
+      "learning_rate": 0.0001591266328043802,
+      "loss": 0.1283,
+      "step": 1165
+    },
+    {
+      "epoch": 1.1070496083550914,
+      "grad_norm": 0.03346949443221092,
+      "learning_rate": 0.00015903747888191053,
+      "loss": 0.1355,
+      "step": 1166
+    },
+    {
+      "epoch": 1.1079990505577972,
+      "grad_norm": 0.0324571467936039,
+      "learning_rate": 0.00015894825286964675,
+      "loss": 0.1354,
+      "step": 1167
+    },
+    {
+      "epoch": 1.1089484927605031,
+      "grad_norm": 0.05366596579551697,
+      "learning_rate": 0.00015885895487654147,
+      "loss": 0.2099,
+      "step": 1168
+    },
+    {
+      "epoch": 1.1098979349632092,
+      "grad_norm": 0.04298697039484978,
+      "learning_rate": 0.00015876958501163512,
+      "loss": 0.1709,
+      "step": 1169
+    },
+    {
+      "epoch": 1.110847377165915,
+      "grad_norm": 0.02922794409096241,
+      "learning_rate": 0.00015868014338405592,
+      "loss": 0.1201,
+      "step": 1170
+    },
+    {
+      "epoch": 1.111796819368621,
+      "grad_norm": 0.02963380515575409,
+      "learning_rate": 0.00015859063010301974,
+      "loss": 0.119,
+      "step": 1171
+    },
+    {
+      "epoch": 1.1127462615713268,
+      "grad_norm": 0.05272309482097626,
+      "learning_rate": 0.0001585010452778299,
+      "loss": 0.1785,
+      "step": 1172
+    },
+    {
+      "epoch": 1.1136957037740327,
+      "grad_norm": 0.031892240047454834,
+      "learning_rate": 0.00015841138901787714,
+      "loss": 0.1292,
+      "step": 1173
+    },
+    {
+      "epoch": 1.1146451459767386,
+      "grad_norm": 0.02971399575471878,
+      "learning_rate": 0.0001583216614326394,
+      "loss": 0.124,
+      "step": 1174
+    },
+    {
+      "epoch": 1.1155945881794447,
+      "grad_norm": 0.03037869744002819,
+      "learning_rate": 0.00015823186263168169,
+      "loss": 0.1303,
+      "step": 1175
+    },
+    {
+      "epoch": 1.1165440303821506,
+      "grad_norm": 0.02748207375407219,
+      "learning_rate": 0.000158141992724656,
+      "loss": 0.1243,
+      "step": 1176
+    },
+    {
+      "epoch": 1.1174934725848564,
+      "grad_norm": 0.033940836787223816,
+      "learning_rate": 0.00015805205182130113,
+      "loss": 0.1254,
+      "step": 1177
+    },
+    {
+      "epoch": 1.1184429147875623,
+      "grad_norm": 0.03295721858739853,
+      "learning_rate": 0.00015796204003144264,
+      "loss": 0.1235,
+      "step": 1178
+    },
+    {
+      "epoch": 1.1193923569902682,
+      "grad_norm": 0.031388405710458755,
+      "learning_rate": 0.00015787195746499254,
+      "loss": 0.1221,
+      "step": 1179
+    },
+    {
+      "epoch": 1.120341799192974,
+      "grad_norm": 0.033072203397750854,
+      "learning_rate": 0.00015778180423194936,
+      "loss": 0.1284,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1212912413956801,
+      "grad_norm": 0.03310628607869148,
+      "learning_rate": 0.00015769158044239787,
+      "loss": 0.136,
+      "step": 1181
+    },
+    {
+      "epoch": 1.122240683598386,
+      "grad_norm": 0.030244866386055946,
+      "learning_rate": 0.000157601286206509,
+      "loss": 0.1255,
+      "step": 1182
+    },
+    {
+      "epoch": 1.123190125801092,
+      "grad_norm": 0.04400714859366417,
+      "learning_rate": 0.0001575109216345397,
+      "loss": 0.1706,
+      "step": 1183
+    },
+    {
+      "epoch": 1.1241395680037978,
+      "grad_norm": 0.03094104304909706,
+      "learning_rate": 0.00015742048683683288,
+      "loss": 0.1261,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1250890102065036,
+      "grad_norm": 0.03327153995633125,
+      "learning_rate": 0.00015732998192381707,
+      "loss": 0.1334,
+      "step": 1185
+    },
+    {
+      "epoch": 1.1260384524092095,
+      "grad_norm": 0.03229563683271408,
+      "learning_rate": 0.0001572394070060065,
+      "loss": 0.1168,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1269878946119154,
+      "grad_norm": 0.03267960995435715,
+      "learning_rate": 0.0001571487621940009,
+      "loss": 0.1331,
+      "step": 1187
+    },
+    {
+      "epoch": 1.1279373368146215,
+      "grad_norm": 0.02902175299823284,
+      "learning_rate": 0.00015705804759848523,
+      "loss": 0.1302,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1288867790173274,
+      "grad_norm": 0.030303161591291428,
+      "learning_rate": 0.00015696726333022984,
+      "loss": 0.1267,
+      "step": 1189
+    },
+    {
+      "epoch": 1.1298362212200332,
+      "grad_norm": 0.031242702156305313,
+      "learning_rate": 0.00015687640950009,
+      "loss": 0.125,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1307856634227391,
+      "grad_norm": 0.02697862684726715,
+      "learning_rate": 0.00015678548621900597,
+      "loss": 0.1207,
+      "step": 1191
+    },
+    {
+      "epoch": 1.131735105625445,
+      "grad_norm": 0.03510352969169617,
+      "learning_rate": 0.0001566944935980029,
+      "loss": 0.132,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1326845478281509,
+      "grad_norm": 0.03274201229214668,
+      "learning_rate": 0.00015660343174819045,
+      "loss": 0.1262,
+      "step": 1193
+    },
+    {
+      "epoch": 1.133633990030857,
+      "grad_norm": 0.03378736600279808,
+      "learning_rate": 0.00015651230078076296,
+      "loss": 0.1318,
+      "step": 1194
+    },
+    {
+      "epoch": 1.1345834322335628,
+      "grad_norm": 0.032314665615558624,
+      "learning_rate": 0.00015642110080699907,
+      "loss": 0.117,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1355328744362687,
+      "grad_norm": 0.04177004471421242,
+      "learning_rate": 0.00015632983193826174,
+      "loss": 0.1527,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1364823166389746,
+      "grad_norm": 0.05647768825292587,
+      "learning_rate": 0.00015623849428599804,
+      "loss": 0.2019,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1374317588416805,
+      "grad_norm": 0.03124004229903221,
+      "learning_rate": 0.00015614708796173906,
+      "loss": 0.1228,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1383812010443863,
+      "grad_norm": 0.029155779629945755,
+      "learning_rate": 0.00015605561307709964,
+      "loss": 0.126,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1393306432470922,
+      "grad_norm": 0.030367044731974602,
+      "learning_rate": 0.0001559640697437785,
+      "loss": 0.1296,
+      "step": 1200
+    },
+    {
+      "epoch": 1.1402800854497983,
+      "grad_norm": 0.06225927174091339,
+      "learning_rate": 0.00015587245807355778,
+      "loss": 0.2258,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1412295276525042,
+      "grad_norm": 0.03878935053944588,
+      "learning_rate": 0.00015578077817830313,
+      "loss": 0.1322,
+      "step": 1202
+    },
+    {
+      "epoch": 1.14217896985521,
+      "grad_norm": 0.030811108648777008,
+      "learning_rate": 0.0001556890301699636,
+      "loss": 0.127,
+      "step": 1203
+    },
+    {
+      "epoch": 1.143128412057916,
+      "grad_norm": 0.0561293326318264,
+      "learning_rate": 0.00015559721416057127,
+      "loss": 0.1689,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1440778542606218,
+      "grad_norm": 0.05256973206996918,
+      "learning_rate": 0.0001555053302622413,
+      "loss": 0.1735,
+      "step": 1205
+    },
+    {
+      "epoch": 1.145027296463328,
+      "grad_norm": 0.037241023033857346,
+      "learning_rate": 0.0001554133785871718,
+      "loss": 0.1304,
+      "step": 1206
+    },
+    {
+      "epoch": 1.1459767386660338,
+      "grad_norm": 0.04502008110284805,
+      "learning_rate": 0.00015532135924764358,
+      "loss": 0.1594,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1469261808687397,
+      "grad_norm": 0.053607337176799774,
+      "learning_rate": 0.00015522927235602014,
+      "loss": 0.1683,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1478756230714455,
+      "grad_norm": 0.029219908639788628,
+      "learning_rate": 0.00015513711802474735,
+      "loss": 0.1267,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1488250652741514,
+      "grad_norm": 0.03088328242301941,
+      "learning_rate": 0.0001550448963663536,
+      "loss": 0.1318,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1497745074768573,
+      "grad_norm": 0.03560802713036537,
+      "learning_rate": 0.00015495260749344932,
+      "loss": 0.1433,
+      "step": 1211
+    },
+    {
+      "epoch": 1.1507239496795632,
+      "grad_norm": 0.033212240785360336,
+      "learning_rate": 0.00015486025151872706,
+      "loss": 0.1222,
+      "step": 1212
+    },
+    {
+      "epoch": 1.1516733918822692,
+      "grad_norm": 0.032883357256650925,
+      "learning_rate": 0.00015476782855496145,
+      "loss": 0.1317,
+      "step": 1213
+    },
+    {
+      "epoch": 1.1526228340849751,
+      "grad_norm": 0.028118513524532318,
+      "learning_rate": 0.0001546753387150087,
+      "loss": 0.1306,
+      "step": 1214
+    },
+    {
+      "epoch": 1.153572276287681,
+      "grad_norm": 0.040216926485300064,
+      "learning_rate": 0.00015458278211180688,
+      "loss": 0.1482,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1545217184903869,
+      "grad_norm": 0.029117384925484657,
+      "learning_rate": 0.00015449015885837542,
+      "loss": 0.1287,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1554711606930927,
+      "grad_norm": 0.08133453875780106,
+      "learning_rate": 0.00015439746906781524,
+      "loss": 0.2109,
+      "step": 1217
+    },
+    {
+      "epoch": 1.1564206028957986,
+      "grad_norm": 0.03554106503725052,
+      "learning_rate": 0.00015430471285330846,
+      "loss": 0.1297,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1573700450985047,
+      "grad_norm": 0.037721745669841766,
+      "learning_rate": 0.00015421189032811835,
+      "loss": 0.1364,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1583194873012106,
+      "grad_norm": 0.025993864983320236,
+      "learning_rate": 0.00015411900160558912,
+      "loss": 0.1198,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1592689295039165,
+      "grad_norm": 0.03330320492386818,
+      "learning_rate": 0.00015402604679914575,
+      "loss": 0.1253,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1602183717066223,
+      "grad_norm": 0.03463476151227951,
+      "learning_rate": 0.00015393302602229408,
+      "loss": 0.1235,
+      "step": 1222
+    },
+    {
+      "epoch": 1.1611678139093282,
+      "grad_norm": 0.03458210453391075,
+      "learning_rate": 0.00015383993938862037,
+      "loss": 0.1238,
+      "step": 1223
+    },
+    {
+      "epoch": 1.162117256112034,
+      "grad_norm": 0.03344335779547691,
+      "learning_rate": 0.00015374678701179134,
+      "loss": 0.1267,
+      "step": 1224
+    },
+    {
+      "epoch": 1.16306669831474,
+      "grad_norm": 0.027585169300436974,
+      "learning_rate": 0.00015365356900555395,
+      "loss": 0.1192,
+      "step": 1225
+    },
+    {
+      "epoch": 1.164016140517446,
+      "grad_norm": 0.03154926374554634,
+      "learning_rate": 0.00015356028548373538,
+      "loss": 0.1288,
+      "step": 1226
+    },
+    {
+      "epoch": 1.164965582720152,
+      "grad_norm": 0.030677665024995804,
+      "learning_rate": 0.00015346693656024271,
+      "loss": 0.1292,
+      "step": 1227
+    },
+    {
+      "epoch": 1.1659150249228578,
+      "grad_norm": 0.030145341530442238,
+      "learning_rate": 0.00015337352234906298,
+      "loss": 0.1331,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1668644671255637,
+      "grad_norm": 0.03683342784643173,
+      "learning_rate": 0.00015328004296426287,
+      "loss": 0.125,
+      "step": 1229
+    },
+    {
+      "epoch": 1.1678139093282696,
+      "grad_norm": 0.029177436605095863,
+      "learning_rate": 0.0001531864985199887,
+      "loss": 0.1313,
+      "step": 1230
+    },
+    {
+      "epoch": 1.1687633515309757,
+      "grad_norm": 0.0415952131152153,
+      "learning_rate": 0.0001530928891304662,
+      "loss": 0.1642,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1697127937336815,
+      "grad_norm": 0.027380244806408882,
+      "learning_rate": 0.00015299921491000043,
+      "loss": 0.1254,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1706622359363874,
+      "grad_norm": 0.03130066767334938,
+      "learning_rate": 0.00015290547597297555,
+      "loss": 0.1291,
+      "step": 1233
+    },
+    {
+      "epoch": 1.1716116781390933,
+      "grad_norm": 0.04524728283286095,
+      "learning_rate": 0.00015281167243385484,
+      "loss": 0.1627,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1725611203417992,
+      "grad_norm": 0.04698526859283447,
+      "learning_rate": 0.0001527178044071804,
+      "loss": 0.1892,
+      "step": 1235
+    },
+    {
+      "epoch": 1.173510562544505,
+      "grad_norm": 0.041941914707422256,
+      "learning_rate": 0.00015262387200757314,
+      "loss": 0.1603,
+      "step": 1236
+    },
+    {
+      "epoch": 1.174460004747211,
+      "grad_norm": 0.03519544377923012,
+      "learning_rate": 0.0001525298753497324,
+      "loss": 0.1306,
+      "step": 1237
+    },
+    {
+      "epoch": 1.175409446949917,
+      "grad_norm": 0.036025673151016235,
+      "learning_rate": 0.00015243581454843624,
+      "loss": 0.1315,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1763588891526229,
+      "grad_norm": 0.05133717134594917,
+      "learning_rate": 0.0001523416897185409,
+      "loss": 0.1661,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1773083313553288,
+      "grad_norm": 0.03367958217859268,
+      "learning_rate": 0.00015224750097498073,
+      "loss": 0.1311,
+      "step": 1240
+    },
+    {
+      "epoch": 1.1782577735580346,
+      "grad_norm": 0.035372741520404816,
+      "learning_rate": 0.0001521532484327683,
+      "loss": 0.1324,
+      "step": 1241
+    },
+    {
+      "epoch": 1.1792072157607405,
+      "grad_norm": 0.048746585845947266,
+      "learning_rate": 0.000152058932206994,
+      "loss": 0.1756,
+      "step": 1242
+    },
+    {
+      "epoch": 1.1801566579634466,
+      "grad_norm": 0.03578799590468407,
+      "learning_rate": 0.00015196455241282592,
+      "loss": 0.1344,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1811061001661525,
+      "grad_norm": 0.030654437839984894,
+      "learning_rate": 0.00015187010916550988,
+      "loss": 0.1268,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1820555423688583,
+      "grad_norm": 0.02881826087832451,
+      "learning_rate": 0.0001517756025803691,
+      "loss": 0.1149,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1830049845715642,
+      "grad_norm": 0.031242484226822853,
+      "learning_rate": 0.00015168103277280422,
+      "loss": 0.1338,
+      "step": 1246
+    },
+    {
+      "epoch": 1.18395442677427,
+      "grad_norm": 0.028528152033686638,
+      "learning_rate": 0.000151586399858293,
+      "loss": 0.1203,
+      "step": 1247
+    },
+    {
+      "epoch": 1.184903868976976,
+      "grad_norm": 0.028410421684384346,
+      "learning_rate": 0.00015149170395239035,
+      "loss": 0.1296,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1858533111796818,
+      "grad_norm": 0.029383866116404533,
+      "learning_rate": 0.00015139694517072796,
+      "loss": 0.1284,
+      "step": 1249
+    },
+    {
+      "epoch": 1.186802753382388,
+      "grad_norm": 0.031084850430488586,
+      "learning_rate": 0.00015130212362901447,
+      "loss": 0.1272,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1877521955850938,
+      "grad_norm": 0.035449109971523285,
+      "learning_rate": 0.00015120723944303497,
+      "loss": 0.1293,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1887016377877997,
+      "grad_norm": 0.030890950933098793,
+      "learning_rate": 0.0001511122927286512,
+      "loss": 0.1221,
+      "step": 1252
+    },
+    {
+      "epoch": 1.1896510799905056,
+      "grad_norm": 0.06967780739068985,
+      "learning_rate": 0.0001510172836018012,
+      "loss": 0.1277,
+      "step": 1253
+    },
+    {
+      "epoch": 1.1906005221932114,
+      "grad_norm": 0.03075585328042507,
+      "learning_rate": 0.00015092221217849917,
+      "loss": 0.1278,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1915499643959173,
+      "grad_norm": 0.040119290351867676,
+      "learning_rate": 0.00015082707857483544,
+      "loss": 0.1546,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1924994065986234,
+      "grad_norm": 0.03376394882798195,
+      "learning_rate": 0.0001507318829069763,
+      "loss": 0.1325,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1934488488013293,
+      "grad_norm": 0.034935321658849716,
+      "learning_rate": 0.00015063662529116368,
+      "loss": 0.1361,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1943982910040352,
+      "grad_norm": 0.043972812592983246,
+      "learning_rate": 0.00015054130584371528,
+      "loss": 0.1292,
+      "step": 1258
+    },
+    {
+      "epoch": 1.195347733206741,
+      "grad_norm": 0.06039128825068474,
+      "learning_rate": 0.0001504459246810243,
+      "loss": 0.1958,
+      "step": 1259
+    },
+    {
+      "epoch": 1.196297175409447,
+      "grad_norm": 0.030998772010207176,
+      "learning_rate": 0.00015035048191955927,
+      "loss": 0.1166,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1972466176121528,
+      "grad_norm": 0.0286384467035532,
+      "learning_rate": 0.00015025497767586393,
+      "loss": 0.1225,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1981960598148587,
+      "grad_norm": 0.03200898319482803,
+      "learning_rate": 0.0001501594120665571,
+      "loss": 0.1244,
+      "step": 1262
+    },
+    {
+      "epoch": 1.1991455020175648,
+      "grad_norm": 0.032870370894670486,
+      "learning_rate": 0.00015006378520833252,
+      "loss": 0.126,
+      "step": 1263
+    },
+    {
+      "epoch": 1.2000949442202706,
+      "grad_norm": 0.034849826246500015,
+      "learning_rate": 0.00014996809721795872,
+      "loss": 0.1263,
+      "step": 1264
+    },
+    {
+      "epoch": 1.2010443864229765,
+      "grad_norm": 0.045324552804231644,
+      "learning_rate": 0.00014987234821227898,
+      "loss": 0.1668,
+      "step": 1265
+    },
+    {
+      "epoch": 1.2019938286256824,
+      "grad_norm": 0.036612797528505325,
+      "learning_rate": 0.0001497765383082109,
+      "loss": 0.1595,
+      "step": 1266
+    },
+    {
+      "epoch": 1.2029432708283883,
+      "grad_norm": 0.03746375814080238,
+      "learning_rate": 0.00014968066762274657,
+      "loss": 0.1644,
+      "step": 1267
+    },
+    {
+      "epoch": 1.2038927130310944,
+      "grad_norm": 0.03137432038784027,
+      "learning_rate": 0.0001495847362729523,
+      "loss": 0.1239,
+      "step": 1268
+    },
+    {
+      "epoch": 1.2048421552338002,
+      "grad_norm": 0.0314825214445591,
+      "learning_rate": 0.0001494887443759684,
+      "loss": 0.1258,
+      "step": 1269
+    },
+    {
+      "epoch": 1.205791597436506,
+      "grad_norm": 0.032157186418771744,
+      "learning_rate": 0.00014939269204900917,
+      "loss": 0.1233,
+      "step": 1270
+    },
+    {
+      "epoch": 1.206741039639212,
+      "grad_norm": 0.0410330593585968,
+      "learning_rate": 0.0001492965794093627,
+      "loss": 0.153,
+      "step": 1271
+    },
+    {
+      "epoch": 1.2076904818419179,
+      "grad_norm": 0.0325077585875988,
+      "learning_rate": 0.0001492004065743907,
+      "loss": 0.1241,
+      "step": 1272
+    },
+    {
+      "epoch": 1.2086399240446237,
+      "grad_norm": 0.033166393637657166,
+      "learning_rate": 0.00014910417366152844,
+      "loss": 0.1292,
+      "step": 1273
+    },
+    {
+      "epoch": 1.2095893662473296,
+      "grad_norm": 0.02926860749721527,
+      "learning_rate": 0.0001490078807882845,
+      "loss": 0.1242,
+      "step": 1274
+    },
+    {
+      "epoch": 1.2105388084500357,
+      "grad_norm": 0.04637501388788223,
+      "learning_rate": 0.00014891152807224066,
+      "loss": 0.1404,
+      "step": 1275
+    },
+    {
+      "epoch": 1.2114882506527416,
+      "grad_norm": 0.035617321729660034,
+      "learning_rate": 0.0001488151156310518,
+      "loss": 0.1292,
+      "step": 1276
+    },
+    {
+      "epoch": 1.2124376928554474,
+      "grad_norm": 0.036330446600914,
+      "learning_rate": 0.00014871864358244574,
+      "loss": 0.1326,
+      "step": 1277
+    },
+    {
+      "epoch": 1.2133871350581533,
+      "grad_norm": 0.034302353858947754,
+      "learning_rate": 0.00014862211204422305,
+      "loss": 0.1296,
+      "step": 1278
+    },
+    {
+      "epoch": 1.2143365772608592,
+      "grad_norm": 0.027070587500929832,
+      "learning_rate": 0.00014852552113425702,
+      "loss": 0.1227,
+      "step": 1279
+    },
+    {
+      "epoch": 1.215286019463565,
+      "grad_norm": 0.029872050508856773,
+      "learning_rate": 0.00014842887097049333,
+      "loss": 0.1333,
+      "step": 1280
+    },
+    {
+      "epoch": 1.2162354616662712,
+      "grad_norm": 0.0336853489279747,
+      "learning_rate": 0.0001483321616709501,
+      "loss": 0.1264,
+      "step": 1281
+    },
+    {
+      "epoch": 1.217184903868977,
+      "grad_norm": 0.03892628103494644,
+      "learning_rate": 0.00014823539335371763,
+      "loss": 0.1516,
+      "step": 1282
+    },
+    {
+      "epoch": 1.218134346071683,
+      "grad_norm": 0.03041486255824566,
+      "learning_rate": 0.00014813856613695825,
+      "loss": 0.1303,
+      "step": 1283
+    },
+    {
+      "epoch": 1.2190837882743888,
+      "grad_norm": 0.042988792061805725,
+      "learning_rate": 0.00014804168013890628,
+      "loss": 0.1697,
+      "step": 1284
+    },
+    {
+      "epoch": 1.2200332304770947,
+      "grad_norm": 0.03176325559616089,
+      "learning_rate": 0.00014794473547786777,
+      "loss": 0.1309,
+      "step": 1285
+    },
+    {
+      "epoch": 1.2209826726798005,
+      "grad_norm": 0.032539233565330505,
+      "learning_rate": 0.00014784773227222042,
+      "loss": 0.1336,
+      "step": 1286
+    },
+    {
+      "epoch": 1.2219321148825064,
+      "grad_norm": 0.03153330832719803,
+      "learning_rate": 0.00014775067064041341,
+      "loss": 0.1244,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2228815570852125,
+      "grad_norm": 0.03229093924164772,
+      "learning_rate": 0.00014765355070096728,
+      "loss": 0.1331,
+      "step": 1288
+    },
+    {
+      "epoch": 1.2238309992879184,
+      "grad_norm": 0.03116600587964058,
+      "learning_rate": 0.0001475563725724737,
+      "loss": 0.1263,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2247804414906243,
+      "grad_norm": 0.027543194591999054,
+      "learning_rate": 0.0001474591363735955,
+      "loss": 0.1319,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2257298836933301,
+      "grad_norm": 0.031299810856580734,
+      "learning_rate": 0.00014736184222306637,
+      "loss": 0.1235,
+      "step": 1291
+    },
+    {
+      "epoch": 1.226679325896036,
+      "grad_norm": 0.030574094504117966,
+      "learning_rate": 0.00014726449023969073,
+      "loss": 0.1337,
+      "step": 1292
+    },
+    {
+      "epoch": 1.227628768098742,
+      "grad_norm": 0.04210914671421051,
+      "learning_rate": 0.0001471670805423437,
+      "loss": 0.1439,
+      "step": 1293
+    },
+    {
+      "epoch": 1.228578210301448,
+      "grad_norm": 0.03234979510307312,
+      "learning_rate": 0.00014706961324997077,
+      "loss": 0.1339,
+      "step": 1294
+    },
+    {
+      "epoch": 1.2295276525041539,
+      "grad_norm": 0.028707873076200485,
+      "learning_rate": 0.00014697208848158782,
+      "loss": 0.1271,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2304770947068597,
+      "grad_norm": 0.03162172809243202,
+      "learning_rate": 0.0001468745063562809,
+      "loss": 0.1291,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2314265369095656,
+      "grad_norm": 0.032575272023677826,
+      "learning_rate": 0.00014677686699320614,
+      "loss": 0.1345,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2323759791122715,
+      "grad_norm": 0.02751932106912136,
+      "learning_rate": 0.0001466791705115895,
+      "loss": 0.12,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2333254213149774,
+      "grad_norm": 0.026365652680397034,
+      "learning_rate": 0.00014658141703072675,
+      "loss": 0.1147,
+      "step": 1299
+    },
+    {
+      "epoch": 1.2342748635176835,
+      "grad_norm": 0.03227808326482773,
+      "learning_rate": 0.00014648360666998314,
+      "loss": 0.1332,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2352243057203893,
+      "grad_norm": 0.057433344423770905,
+      "learning_rate": 0.00014638573954879356,
+      "loss": 0.2349,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2361737479230952,
+      "grad_norm": 0.0376310870051384,
+      "learning_rate": 0.000146287815786662,
+      "loss": 0.1588,
+      "step": 1302
+    },
+    {
+      "epoch": 1.237123190125801,
+      "grad_norm": 0.04079489782452583,
+      "learning_rate": 0.00014618983550316182,
+      "loss": 0.1625,
+      "step": 1303
+    },
+    {
+      "epoch": 1.238072632328507,
+      "grad_norm": 0.03367112949490547,
+      "learning_rate": 0.00014609179881793524,
+      "loss": 0.1266,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2390220745312128,
+      "grad_norm": 0.030257422477006912,
+      "learning_rate": 0.0001459937058506934,
+      "loss": 0.132,
+      "step": 1305
+    },
+    {
+      "epoch": 1.239971516733919,
+      "grad_norm": 0.03830450400710106,
+      "learning_rate": 0.00014589555672121622,
+      "loss": 0.1349,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2409209589366248,
+      "grad_norm": 0.030318403616547585,
+      "learning_rate": 0.00014579735154935213,
+      "loss": 0.1346,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2418704011393307,
+      "grad_norm": 0.033780913800001144,
+      "learning_rate": 0.000145699090455018,
+      "loss": 0.1373,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2428198433420365,
+      "grad_norm": 0.03642027825117111,
+      "learning_rate": 0.00014560077355819904,
+      "loss": 0.1279,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2437692855447424,
+      "grad_norm": 0.027665693312883377,
+      "learning_rate": 0.00014550240097894852,
+      "loss": 0.1177,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2447187277474483,
+      "grad_norm": 0.030852187424898148,
+      "learning_rate": 0.00014540397283738777,
+      "loss": 0.1373,
+      "step": 1311
+    },
+    {
+      "epoch": 1.2456681699501542,
+      "grad_norm": 0.03137564659118652,
+      "learning_rate": 0.00014530548925370594,
+      "loss": 0.129,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2466176121528603,
+      "grad_norm": 0.032402969896793365,
+      "learning_rate": 0.0001452069503481599,
+      "loss": 0.1394,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2475670543555661,
+      "grad_norm": 0.049039825797080994,
+      "learning_rate": 0.00014510835624107396,
+      "loss": 0.1593,
+      "step": 1314
+    },
+    {
+      "epoch": 1.248516496558272,
+      "grad_norm": 0.0339655838906765,
+      "learning_rate": 0.00014500970705284006,
+      "loss": 0.1367,
+      "step": 1315
+    },
+    {
+      "epoch": 1.249465938760978,
+      "grad_norm": 0.040319304913282394,
+      "learning_rate": 0.00014491100290391716,
+      "loss": 0.1679,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2504153809636838,
+      "grad_norm": 0.024075858294963837,
+      "learning_rate": 0.00014481224391483152,
+      "loss": 0.1273,
+      "step": 1317
+    },
+    {
+      "epoch": 1.2513648231663899,
+      "grad_norm": 0.02570619434118271,
+      "learning_rate": 0.00014471343020617625,
+      "loss": 0.1277,
+      "step": 1318
+    },
+    {
+      "epoch": 1.2523142653690957,
+      "grad_norm": 0.028086170554161072,
+      "learning_rate": 0.00014461456189861132,
+      "loss": 0.1246,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2532637075718016,
+      "grad_norm": 0.02902062050998211,
+      "learning_rate": 0.0001445156391128633,
+      "loss": 0.1305,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2542131497745075,
+      "grad_norm": 0.030352482572197914,
+      "learning_rate": 0.00014441666196972542,
+      "loss": 0.1329,
+      "step": 1321
+    },
+    {
+      "epoch": 1.2551625919772134,
+      "grad_norm": 0.029998816549777985,
+      "learning_rate": 0.00014431763059005718,
+      "loss": 0.131,
+      "step": 1322
+    },
+    {
+      "epoch": 1.2561120341799192,
+      "grad_norm": 0.028111204504966736,
+      "learning_rate": 0.00014421854509478435,
+      "loss": 0.124,
+      "step": 1323
+    },
+    {
+      "epoch": 1.257061476382625,
+      "grad_norm": 0.02926759235560894,
+      "learning_rate": 0.00014411940560489877,
+      "loss": 0.1215,
+      "step": 1324
+    },
+    {
+      "epoch": 1.258010918585331,
+      "grad_norm": 0.03321680426597595,
+      "learning_rate": 0.00014402021224145815,
+      "loss": 0.1216,
+      "step": 1325
+    },
+    {
+      "epoch": 1.258960360788037,
+      "grad_norm": 0.03386010602116585,
+      "learning_rate": 0.00014392096512558613,
+      "loss": 0.1335,
+      "step": 1326
+    },
+    {
+      "epoch": 1.259909802990743,
+      "grad_norm": 0.03989921137690544,
+      "learning_rate": 0.0001438216643784718,
+      "loss": 0.1481,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2608592451934488,
+      "grad_norm": 0.030915161594748497,
+      "learning_rate": 0.00014372231012136995,
+      "loss": 0.1254,
+      "step": 1328
+    },
+    {
+      "epoch": 1.2618086873961547,
+      "grad_norm": 0.04395739734172821,
+      "learning_rate": 0.00014362290247560053,
+      "loss": 0.1537,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2627581295988608,
+      "grad_norm": 0.02942941151559353,
+      "learning_rate": 0.00014352344156254873,
+      "loss": 0.1248,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2637075718015667,
+      "grad_norm": 0.02858722023665905,
+      "learning_rate": 0.00014342392750366485,
+      "loss": 0.1236,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2646570140042726,
+      "grad_norm": 0.029218707233667374,
+      "learning_rate": 0.000143324360420464,
+      "loss": 0.1227,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2656064562069784,
+      "grad_norm": 0.03079938143491745,
+      "learning_rate": 0.0001432247404345261,
+      "loss": 0.1251,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2665558984096843,
+      "grad_norm": 0.03626713901758194,
+      "learning_rate": 0.00014312506766749563,
+      "loss": 0.1407,
+      "step": 1334
+    },
+    {
+      "epoch": 1.2675053406123902,
+      "grad_norm": 0.026556458324193954,
+      "learning_rate": 0.00014302534224108152,
+      "loss": 0.1235,
+      "step": 1335
+    },
+    {
+      "epoch": 1.268454782815096,
+      "grad_norm": 0.033421531319618225,
+      "learning_rate": 0.00014292556427705706,
+      "loss": 0.1324,
+      "step": 1336
+    },
+    {
+      "epoch": 1.269404225017802,
+      "grad_norm": 0.0425841398537159,
+      "learning_rate": 0.00014282573389725966,
+      "loss": 0.1674,
+      "step": 1337
+    },
+    {
+      "epoch": 1.270353667220508,
+      "grad_norm": 0.03258546441793442,
+      "learning_rate": 0.00014272585122359068,
+      "loss": 0.131,
+      "step": 1338
+    },
+    {
+      "epoch": 1.271303109423214,
+      "grad_norm": 0.03566194325685501,
+      "learning_rate": 0.00014262591637801536,
+      "loss": 0.128,
+      "step": 1339
+    },
+    {
+      "epoch": 1.2722525516259198,
+      "grad_norm": 0.03155380114912987,
+      "learning_rate": 0.0001425259294825627,
+      "loss": 0.1277,
+      "step": 1340
+    },
+    {
+      "epoch": 1.2732019938286256,
+      "grad_norm": 0.04435742273926735,
+      "learning_rate": 0.00014242589065932524,
+      "loss": 0.1594,
+      "step": 1341
+    },
+    {
+      "epoch": 1.2741514360313315,
+      "grad_norm": 0.03839895501732826,
+      "learning_rate": 0.0001423258000304589,
+      "loss": 0.1598,
+      "step": 1342
+    },
+    {
+      "epoch": 1.2751008782340376,
+      "grad_norm": 0.037625472992658615,
+      "learning_rate": 0.00014222565771818282,
+      "loss": 0.1276,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2760503204367435,
+      "grad_norm": 0.029852135106921196,
+      "learning_rate": 0.00014212546384477934,
+      "loss": 0.1272,
+      "step": 1344
+    },
+    {
+      "epoch": 1.2769997626394494,
+      "grad_norm": 0.04018719121813774,
+      "learning_rate": 0.00014202521853259368,
+      "loss": 0.153,
+      "step": 1345
+    },
+    {
+      "epoch": 1.2779492048421552,
+      "grad_norm": 0.028893720358610153,
+      "learning_rate": 0.00014192492190403402,
+      "loss": 0.1245,
+      "step": 1346
+    },
+    {
+      "epoch": 1.2788986470448611,
+      "grad_norm": 0.029052307829260826,
+      "learning_rate": 0.000141824574081571,
+      "loss": 0.1272,
+      "step": 1347
+    },
+    {
+      "epoch": 1.279848089247567,
+      "grad_norm": 0.03023959882557392,
+      "learning_rate": 0.00014172417518773788,
+      "loss": 0.1259,
+      "step": 1348
+    },
+    {
+      "epoch": 1.2807975314502729,
+      "grad_norm": 0.06486006826162338,
+      "learning_rate": 0.00014162372534513027,
+      "loss": 0.2279,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2817469736529787,
+      "grad_norm": 0.034129489213228226,
+      "learning_rate": 0.00014152322467640599,
+      "loss": 0.138,
+      "step": 1350
+    },
+    {
+      "epoch": 1.2826964158556848,
+      "grad_norm": 0.034559451043605804,
+      "learning_rate": 0.0001414226733042849,
+      "loss": 0.1321,
+      "step": 1351
+    },
+    {
+      "epoch": 1.2836458580583907,
+      "grad_norm": 0.03269064798951149,
+      "learning_rate": 0.0001413220713515489,
+      "loss": 0.1297,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2845953002610966,
+      "grad_norm": 0.030022764578461647,
+      "learning_rate": 0.0001412214189410414,
+      "loss": 0.1278,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2855447424638025,
+      "grad_norm": 0.05017710104584694,
+      "learning_rate": 0.00014112071619566766,
+      "loss": 0.1572,
+      "step": 1354
+    },
+    {
+      "epoch": 1.2864941846665086,
+      "grad_norm": 0.035493746399879456,
+      "learning_rate": 0.00014101996323839433,
+      "loss": 0.1277,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2874436268692144,
+      "grad_norm": 0.03152285888791084,
+      "learning_rate": 0.0001409191601922493,
+      "loss": 0.1321,
+      "step": 1356
+    },
+    {
+      "epoch": 1.2883930690719203,
+      "grad_norm": 0.029683001339435577,
+      "learning_rate": 0.00014081830718032175,
+      "loss": 0.1224,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2893425112746262,
+      "grad_norm": 0.03202647715806961,
+      "learning_rate": 0.0001407174043257617,
+      "loss": 0.13,
+      "step": 1358
+    },
+    {
+      "epoch": 1.290291953477332,
+      "grad_norm": 0.026956327259540558,
+      "learning_rate": 0.00014061645175178025,
+      "loss": 0.1225,
+      "step": 1359
+    },
+    {
+      "epoch": 1.291241395680038,
+      "grad_norm": 0.03426060825586319,
+      "learning_rate": 0.00014051544958164903,
+      "loss": 0.1345,
+      "step": 1360
+    },
+    {
+      "epoch": 1.2921908378827438,
+      "grad_norm": 0.031120451167225838,
+      "learning_rate": 0.00014041439793870036,
+      "loss": 0.1246,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2931402800854497,
+      "grad_norm": 0.02656574547290802,
+      "learning_rate": 0.00014031329694632683,
+      "loss": 0.1297,
+      "step": 1362
+    },
+    {
+      "epoch": 1.2940897222881558,
+      "grad_norm": 0.02752675488591194,
+      "learning_rate": 0.00014021214672798143,
+      "loss": 0.1294,
+      "step": 1363
+    },
+    {
+      "epoch": 1.2950391644908616,
+      "grad_norm": 0.02884535677731037,
+      "learning_rate": 0.00014011094740717714,
+      "loss": 0.126,
+      "step": 1364
+    },
+    {
+      "epoch": 1.2959886066935675,
+      "grad_norm": 0.03029620461165905,
+      "learning_rate": 0.00014000969910748704,
+      "loss": 0.1338,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2969380488962734,
+      "grad_norm": 0.04302069917321205,
+      "learning_rate": 0.00013990840195254384,
+      "loss": 0.1653,
+      "step": 1366
+    },
+    {
+      "epoch": 1.2978874910989793,
+      "grad_norm": 0.048259928822517395,
+      "learning_rate": 0.00013980705606604011,
+      "loss": 0.1269,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2988369333016854,
+      "grad_norm": 0.029876641929149628,
+      "learning_rate": 0.00013970566157172774,
+      "loss": 0.1282,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2997863755043912,
+      "grad_norm": 0.35130763053894043,
+      "learning_rate": 0.00013960421859341804,
+      "loss": 0.1434,
+      "step": 1369
+    },
+    {
+      "epoch": 1.3007358177070971,
+      "grad_norm": 0.23888662457466125,
+      "learning_rate": 0.00013950272725498156,
+      "loss": 0.186,
+      "step": 1370
+    },
+    {
+      "epoch": 1.301685259909803,
+      "grad_norm": 0.24867364764213562,
+      "learning_rate": 0.00013940118768034792,
+      "loss": 0.1585,
+      "step": 1371
+    },
+    {
+      "epoch": 1.3026347021125089,
+      "grad_norm": 0.10958488285541534,
+      "learning_rate": 0.0001392995999935055,
+      "loss": 0.1448,
+      "step": 1372
+    },
+    {
+      "epoch": 1.3035841443152147,
+      "grad_norm": 0.05493846908211708,
+      "learning_rate": 0.0001391979643185016,
+      "loss": 0.1292,
+      "step": 1373
+    },
+    {
+      "epoch": 1.3045335865179206,
+      "grad_norm": 0.0429663360118866,
+      "learning_rate": 0.000139096280779442,
+      "loss": 0.1243,
+      "step": 1374
+    },
+    {
+      "epoch": 1.3054830287206267,
+      "grad_norm": 0.02995472215116024,
+      "learning_rate": 0.000138994549500491,
+      "loss": 0.1216,
+      "step": 1375
+    },
+    {
+      "epoch": 1.3064324709233326,
+      "grad_norm": 0.04113904386758804,
+      "learning_rate": 0.00013889277060587119,
+      "loss": 0.1586,
+      "step": 1376
+    },
+    {
+      "epoch": 1.3073819131260385,
+      "grad_norm": 0.030523164197802544,
+      "learning_rate": 0.0001387909442198632,
+      "loss": 0.1212,
+      "step": 1377
+    },
+    {
+      "epoch": 1.3083313553287443,
+      "grad_norm": 0.04093822091817856,
+      "learning_rate": 0.00013868907046680576,
+      "loss": 0.1254,
+      "step": 1378
+    },
+    {
+      "epoch": 1.3092807975314502,
+      "grad_norm": 0.04895343258976936,
+      "learning_rate": 0.0001385871494710954,
+      "loss": 0.1636,
+      "step": 1379
+    },
+    {
+      "epoch": 1.3102302397341563,
+      "grad_norm": 0.09062381833791733,
+      "learning_rate": 0.0001384851813571864,
+      "loss": 0.167,
+      "step": 1380
+    },
+    {
+      "epoch": 1.3111796819368622,
+      "grad_norm": 0.034514930099248886,
+      "learning_rate": 0.00013838316624959044,
+      "loss": 0.1186,
+      "step": 1381
+    },
+    {
+      "epoch": 1.312129124139568,
+      "grad_norm": 0.052746132016181946,
+      "learning_rate": 0.0001382811042728767,
+      "loss": 0.1289,
+      "step": 1382
+    },
+    {
+      "epoch": 1.313078566342274,
+      "grad_norm": 0.06090299040079117,
+      "learning_rate": 0.00013817899555167154,
+      "loss": 0.1599,
+      "step": 1383
+    },
+    {
+      "epoch": 1.3140280085449798,
+      "grad_norm": 0.036167677491903305,
+      "learning_rate": 0.00013807684021065842,
+      "loss": 0.1294,
+      "step": 1384
+    },
+    {
+      "epoch": 1.3149774507476857,
+      "grad_norm": 0.05916972458362579,
+      "learning_rate": 0.00013797463837457775,
+      "loss": 0.1263,
+      "step": 1385
+    },
+    {
+      "epoch": 1.3159268929503916,
+      "grad_norm": 0.03488500416278839,
+      "learning_rate": 0.00013787239016822662,
+      "loss": 0.1347,
+      "step": 1386
+    },
+    {
+      "epoch": 1.3168763351530974,
+      "grad_norm": 0.038088779896497726,
+      "learning_rate": 0.00013777009571645885,
+      "loss": 0.1302,
+      "step": 1387
+    },
+    {
+      "epoch": 1.3178257773558035,
+      "grad_norm": 0.05069038271903992,
+      "learning_rate": 0.00013766775514418469,
+      "loss": 0.1553,
+      "step": 1388
+    },
+    {
+      "epoch": 1.3187752195585094,
+      "grad_norm": 0.06475794315338135,
+      "learning_rate": 0.00013756536857637065,
+      "loss": 0.2018,
+      "step": 1389
+    },
+    {
+      "epoch": 1.3197246617612153,
+      "grad_norm": 0.03393110632896423,
+      "learning_rate": 0.00013746293613803952,
+      "loss": 0.1238,
+      "step": 1390
+    },
+    {
+      "epoch": 1.3206741039639212,
+      "grad_norm": 0.04623769596219063,
+      "learning_rate": 0.00013736045795427002,
+      "loss": 0.1603,
+      "step": 1391
+    },
+    {
+      "epoch": 1.321623546166627,
+      "grad_norm": 0.03696979209780693,
+      "learning_rate": 0.0001372579341501967,
+      "loss": 0.1291,
+      "step": 1392
+    },
+    {
+      "epoch": 1.3225729883693331,
+      "grad_norm": 0.041281502693891525,
+      "learning_rate": 0.00013715536485100994,
+      "loss": 0.1344,
+      "step": 1393
+    },
+    {
+      "epoch": 1.323522430572039,
+      "grad_norm": 0.03636416420340538,
+      "learning_rate": 0.00013705275018195557,
+      "loss": 0.1334,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3244718727747449,
+      "grad_norm": 0.037941355258226395,
+      "learning_rate": 0.0001369500902683348,
+      "loss": 0.1256,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3254213149774507,
+      "grad_norm": 0.04853476956486702,
+      "learning_rate": 0.0001368473852355042,
+      "loss": 0.1604,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3263707571801566,
+      "grad_norm": 0.034059878438711166,
+      "learning_rate": 0.00013674463520887533,
+      "loss": 0.1308,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3273201993828625,
+      "grad_norm": 0.03482759743928909,
+      "learning_rate": 0.00013664184031391473,
+      "loss": 0.128,
+      "step": 1398
+    },
+    {
+      "epoch": 1.3282696415855684,
+      "grad_norm": 0.032961416989564896,
+      "learning_rate": 0.00013653900067614377,
+      "loss": 0.1276,
+      "step": 1399
+    },
+    {
+      "epoch": 1.3292190837882745,
+      "grad_norm": 0.03538922220468521,
+      "learning_rate": 0.00013643611642113842,
+      "loss": 0.1215,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3301685259909803,
+      "grad_norm": 0.032544538378715515,
+      "learning_rate": 0.00013633318767452903,
+      "loss": 0.1235,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3311179681936862,
+      "grad_norm": 0.04236935079097748,
+      "learning_rate": 0.00013623021456200048,
+      "loss": 0.1663,
+      "step": 1402
+    },
+    {
+      "epoch": 1.332067410396392,
+      "grad_norm": 0.04283679649233818,
+      "learning_rate": 0.00013612719720929164,
+      "loss": 0.1622,
+      "step": 1403
+    },
+    {
+      "epoch": 1.333016852599098,
+      "grad_norm": 0.03691123425960541,
+      "learning_rate": 0.00013602413574219553,
+      "loss": 0.1247,
+      "step": 1404
+    },
+    {
+      "epoch": 1.333966294801804,
+      "grad_norm": 0.028608130291104317,
+      "learning_rate": 0.000135921030286559,
+      "loss": 0.1153,
+      "step": 1405
+    },
+    {
+      "epoch": 1.33491573700451,
+      "grad_norm": 0.03310587257146835,
+      "learning_rate": 0.00013581788096828253,
+      "loss": 0.1305,
+      "step": 1406
+    },
+    {
+      "epoch": 1.3358651792072158,
+      "grad_norm": 0.03368659317493439,
+      "learning_rate": 0.00013571468791332024,
+      "loss": 0.128,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3368146214099217,
+      "grad_norm": 0.04785076901316643,
+      "learning_rate": 0.00013561145124767968,
+      "loss": 0.1715,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3377640636126276,
+      "grad_norm": 0.03625485301017761,
+      "learning_rate": 0.0001355081710974217,
+      "loss": 0.1305,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3387135058153334,
+      "grad_norm": 0.03318242356181145,
+      "learning_rate": 0.00013540484758866,
+      "loss": 0.1244,
+      "step": 1410
+    },
+    {
+      "epoch": 1.3396629480180393,
+      "grad_norm": 0.03148429095745087,
+      "learning_rate": 0.0001353014808475615,
+      "loss": 0.1311,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3406123902207452,
+      "grad_norm": 0.03518190607428551,
+      "learning_rate": 0.00013519807100034577,
+      "loss": 0.1276,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3415618324234513,
+      "grad_norm": 0.031286850571632385,
+      "learning_rate": 0.00013509461817328507,
+      "loss": 0.1252,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3425112746261572,
+      "grad_norm": 0.04668812453746796,
+      "learning_rate": 0.00013499112249270407,
+      "loss": 0.1639,
+      "step": 1414
+    },
+    {
+      "epoch": 1.343460716828863,
+      "grad_norm": 0.03220203518867493,
+      "learning_rate": 0.00013488758408497988,
+      "loss": 0.1254,
+      "step": 1415
+    },
+    {
+      "epoch": 1.344410159031569,
+      "grad_norm": 0.03599967062473297,
+      "learning_rate": 0.0001347840030765417,
+      "loss": 0.1307,
+      "step": 1416
+    },
+    {
+      "epoch": 1.3453596012342748,
+      "grad_norm": 0.03225992992520332,
+      "learning_rate": 0.00013468037959387075,
+      "loss": 0.12,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3463090434369809,
+      "grad_norm": 0.0338221937417984,
+      "learning_rate": 0.00013457671376350012,
+      "loss": 0.1199,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3472584856396868,
+      "grad_norm": 0.046623844653367996,
+      "learning_rate": 0.00013447300571201468,
+      "loss": 0.1695,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3482079278423926,
+      "grad_norm": 0.029645482078194618,
+      "learning_rate": 0.00013436925556605078,
+      "loss": 0.127,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3491573700450985,
+      "grad_norm": 0.03167693316936493,
+      "learning_rate": 0.00013426546345229618,
+      "loss": 0.1268,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3501068122478044,
+      "grad_norm": 0.04903126880526543,
+      "learning_rate": 0.0001341616294974899,
+      "loss": 0.1616,
+      "step": 1422
+    },
+    {
+      "epoch": 1.3510562544505103,
+      "grad_norm": 0.03323996067047119,
+      "learning_rate": 0.00013405775382842206,
+      "loss": 0.1345,
+      "step": 1423
+    },
+    {
+      "epoch": 1.3520056966532161,
+      "grad_norm": 0.03047449141740799,
+      "learning_rate": 0.0001339538365719337,
+      "loss": 0.1286,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3529551388559222,
+      "grad_norm": 0.03170877322554588,
+      "learning_rate": 0.00013384987785491665,
+      "loss": 0.1264,
+      "step": 1425
+    },
+    {
+      "epoch": 1.353904581058628,
+      "grad_norm": 0.03209366276860237,
+      "learning_rate": 0.00013374587780431337,
+      "loss": 0.1256,
+      "step": 1426
+    },
+    {
+      "epoch": 1.354854023261334,
+      "grad_norm": 0.04508209601044655,
+      "learning_rate": 0.00013364183654711678,
+      "loss": 0.1675,
+      "step": 1427
+    },
+    {
+      "epoch": 1.3558034654640398,
+      "grad_norm": 0.0423690564930439,
+      "learning_rate": 0.00013353775421037008,
+      "loss": 0.1578,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3567529076667457,
+      "grad_norm": 0.03641896694898605,
+      "learning_rate": 0.0001334336309211668,
+      "loss": 0.1321,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3577023498694518,
+      "grad_norm": 0.03118027374148369,
+      "learning_rate": 0.00013332946680665023,
+      "loss": 0.1284,
+      "step": 1430
+    },
+    {
+      "epoch": 1.3586517920721577,
+      "grad_norm": 0.029232513159513474,
+      "learning_rate": 0.00013322526199401367,
+      "loss": 0.1251,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3596012342748636,
+      "grad_norm": 0.028796685859560966,
+      "learning_rate": 0.00013312101661050007,
+      "loss": 0.1266,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3605506764775694,
+      "grad_norm": 0.04318710416555405,
+      "learning_rate": 0.00013301673078340196,
+      "loss": 0.162,
+      "step": 1433
+    },
+    {
+      "epoch": 1.3615001186802753,
+      "grad_norm": 0.0319884791970253,
+      "learning_rate": 0.00013291240464006118,
+      "loss": 0.1248,
+      "step": 1434
+    },
+    {
+      "epoch": 1.3624495608829812,
+      "grad_norm": 0.034977950155735016,
+      "learning_rate": 0.0001328080383078689,
+      "loss": 0.1325,
+      "step": 1435
+    },
+    {
+      "epoch": 1.363399003085687,
+      "grad_norm": 0.026079673320055008,
+      "learning_rate": 0.00013270363191426524,
+      "loss": 0.1266,
+      "step": 1436
+    },
+    {
+      "epoch": 1.364348445288393,
+      "grad_norm": 0.03158127889037132,
+      "learning_rate": 0.0001325991855867394,
+      "loss": 0.1326,
+      "step": 1437
+    },
+    {
+      "epoch": 1.365297887491099,
+      "grad_norm": 0.030608203262090683,
+      "learning_rate": 0.00013249469945282916,
+      "loss": 0.1358,
+      "step": 1438
+    },
+    {
+      "epoch": 1.366247329693805,
+      "grad_norm": 0.028404321521520615,
+      "learning_rate": 0.00013239017364012105,
+      "loss": 0.1273,
+      "step": 1439
+    },
+    {
+      "epoch": 1.3671967718965108,
+      "grad_norm": 0.03177287429571152,
+      "learning_rate": 0.00013228560827624995,
+      "loss": 0.1241,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3681462140992167,
+      "grad_norm": 0.030803440138697624,
+      "learning_rate": 0.00013218100348889912,
+      "loss": 0.1271,
+      "step": 1441
+    },
+    {
+      "epoch": 1.3690956563019228,
+      "grad_norm": 0.038757532835006714,
+      "learning_rate": 0.0001320763594057999,
+      "loss": 0.1302,
+      "step": 1442
+    },
+    {
+      "epoch": 1.3700450985046286,
+      "grad_norm": 0.030192391946911812,
+      "learning_rate": 0.00013197167615473164,
+      "loss": 0.1246,
+      "step": 1443
+    },
+    {
+      "epoch": 1.3709945407073345,
+      "grad_norm": 0.056465089321136475,
+      "learning_rate": 0.00013186695386352158,
+      "loss": 0.186,
+      "step": 1444
+    },
+    {
+      "epoch": 1.3719439829100404,
+      "grad_norm": 0.029975950717926025,
+      "learning_rate": 0.00013176219266004442,
+      "loss": 0.1238,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3728934251127463,
+      "grad_norm": 0.03526061028242111,
+      "learning_rate": 0.00013165739267222262,
+      "loss": 0.1198,
+      "step": 1446
+    },
+    {
+      "epoch": 1.3738428673154521,
+      "grad_norm": 0.031407009810209274,
+      "learning_rate": 0.0001315525540280259,
+      "loss": 0.1283,
+      "step": 1447
+    },
+    {
+      "epoch": 1.374792309518158,
+      "grad_norm": 0.05062803998589516,
+      "learning_rate": 0.0001314476768554712,
+      "loss": 0.1598,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3757417517208639,
+      "grad_norm": 0.05837153270840645,
+      "learning_rate": 0.0001313427612826224,
+      "loss": 0.1706,
+      "step": 1449
+    },
+    {
+      "epoch": 1.37669119392357,
+      "grad_norm": 0.032810915261507034,
+      "learning_rate": 0.0001312378074375904,
+      "loss": 0.1228,
+      "step": 1450
+    },
+    {
+      "epoch": 1.3776406361262759,
+      "grad_norm": 0.052471403032541275,
+      "learning_rate": 0.0001311328154485328,
+      "loss": 0.2028,
+      "step": 1451
+    },
+    {
+      "epoch": 1.3785900783289817,
+      "grad_norm": 0.062009330838918686,
+      "learning_rate": 0.00013102778544365378,
+      "loss": 0.1659,
+      "step": 1452
+    },
+    {
+      "epoch": 1.3795395205316876,
+      "grad_norm": 0.03157106414437294,
+      "learning_rate": 0.00013092271755120392,
+      "loss": 0.1103,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3804889627343935,
+      "grad_norm": 0.03269607201218605,
+      "learning_rate": 0.00013081761189948006,
+      "loss": 0.1274,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3814384049370996,
+      "grad_norm": 0.032491281628608704,
+      "learning_rate": 0.00013071246861682515,
+      "loss": 0.1299,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3823878471398054,
+      "grad_norm": 0.04476369544863701,
+      "learning_rate": 0.00013060728783162814,
+      "loss": 0.1543,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3833372893425113,
+      "grad_norm": 0.03646747022867203,
+      "learning_rate": 0.0001305020696723237,
+      "loss": 0.1401,
+      "step": 1457
+    },
+    {
+      "epoch": 1.3842867315452172,
+      "grad_norm": 0.03056688979268074,
+      "learning_rate": 0.0001303968142673922,
+      "loss": 0.1318,
+      "step": 1458
+    },
+    {
+      "epoch": 1.385236173747923,
+      "grad_norm": 0.0505138523876667,
+      "learning_rate": 0.00013029152174535942,
+      "loss": 0.1702,
+      "step": 1459
+    },
+    {
+      "epoch": 1.386185615950629,
+      "grad_norm": 0.03478003665804863,
+      "learning_rate": 0.00013018619223479654,
+      "loss": 0.136,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3871350581533348,
+      "grad_norm": 0.03196396306157112,
+      "learning_rate": 0.00013008082586431983,
+      "loss": 0.1235,
+      "step": 1461
+    },
+    {
+      "epoch": 1.3880845003560407,
+      "grad_norm": 0.030580265447497368,
+      "learning_rate": 0.0001299754227625907,
+      "loss": 0.1274,
+      "step": 1462
+    },
+    {
+      "epoch": 1.3890339425587468,
+      "grad_norm": 0.043844275176525116,
+      "learning_rate": 0.00012986998305831524,
+      "loss": 0.172,
+      "step": 1463
+    },
+    {
+      "epoch": 1.3899833847614527,
+      "grad_norm": 0.031638097018003464,
+      "learning_rate": 0.00012976450688024433,
+      "loss": 0.1221,
+      "step": 1464
+    },
+    {
+      "epoch": 1.3909328269641585,
+      "grad_norm": 0.030004551634192467,
+      "learning_rate": 0.00012965899435717337,
+      "loss": 0.1321,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3918822691668644,
+      "grad_norm": 0.031170252710580826,
+      "learning_rate": 0.00012955344561794218,
+      "loss": 0.1327,
+      "step": 1466
+    },
+    {
+      "epoch": 1.3928317113695705,
+      "grad_norm": 0.02902391366660595,
+      "learning_rate": 0.00012944786079143472,
+      "loss": 0.1266,
+      "step": 1467
+    },
+    {
+      "epoch": 1.3937811535722764,
+      "grad_norm": 0.031269557774066925,
+      "learning_rate": 0.00012934224000657913,
+      "loss": 0.1237,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3947305957749823,
+      "grad_norm": 0.04232865571975708,
+      "learning_rate": 0.0001292365833923473,
+      "loss": 0.1529,
+      "step": 1469
+    },
+    {
+      "epoch": 1.3956800379776881,
+      "grad_norm": 0.03645455837249756,
+      "learning_rate": 0.00012913089107775502,
+      "loss": 0.1594,
+      "step": 1470
+    },
+    {
+      "epoch": 1.396629480180394,
+      "grad_norm": 0.029367268085479736,
+      "learning_rate": 0.00012902516319186161,
+      "loss": 0.121,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3975789223830999,
+      "grad_norm": 0.03407928720116615,
+      "learning_rate": 0.00012891939986376985,
+      "loss": 0.1289,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3985283645858058,
+      "grad_norm": 0.034839022904634476,
+      "learning_rate": 0.00012881360122262575,
+      "loss": 0.1321,
+      "step": 1473
+    },
+    {
+      "epoch": 1.3994778067885116,
+      "grad_norm": 0.04005248472094536,
+      "learning_rate": 0.00012870776739761847,
+      "loss": 0.17,
+      "step": 1474
+    },
+    {
+      "epoch": 1.4004272489912177,
+      "grad_norm": 0.045347243547439575,
+      "learning_rate": 0.00012860189851798012,
+      "loss": 0.16,
+      "step": 1475
+    },
+    {
+      "epoch": 1.4013766911939236,
+      "grad_norm": 0.031093263998627663,
+      "learning_rate": 0.00012849599471298565,
+      "loss": 0.1268,
+      "step": 1476
+    },
+    {
+      "epoch": 1.4023261333966295,
+      "grad_norm": 0.025807669386267662,
+      "learning_rate": 0.00012839005611195269,
+      "loss": 0.1155,
+      "step": 1477
+    },
+    {
+      "epoch": 1.4032755755993354,
+      "grad_norm": 0.04496198520064354,
+      "learning_rate": 0.00012828408284424117,
+      "loss": 0.1696,
+      "step": 1478
+    },
+    {
+      "epoch": 1.4042250178020412,
+      "grad_norm": 0.028986521065235138,
+      "learning_rate": 0.00012817807503925357,
+      "loss": 0.1244,
+      "step": 1479
+    },
+    {
+      "epoch": 1.4051744600047473,
+      "grad_norm": 0.043852079659700394,
+      "learning_rate": 0.00012807203282643443,
+      "loss": 0.1562,
+      "step": 1480
+    },
+    {
+      "epoch": 1.4061239022074532,
+      "grad_norm": 0.028449110686779022,
+      "learning_rate": 0.00012796595633527032,
+      "loss": 0.1276,
+      "step": 1481
+    },
+    {
+      "epoch": 1.407073344410159,
+      "grad_norm": 0.03772464022040367,
+      "learning_rate": 0.00012785984569528975,
+      "loss": 0.163,
+      "step": 1482
+    },
+    {
+      "epoch": 1.408022786612865,
+      "grad_norm": 0.03227540850639343,
+      "learning_rate": 0.00012775370103606276,
+      "loss": 0.1272,
+      "step": 1483
+    },
+    {
+      "epoch": 1.4089722288155708,
+      "grad_norm": 0.03001963160932064,
+      "learning_rate": 0.0001276475224872011,
+      "loss": 0.1256,
+      "step": 1484
+    },
+    {
+      "epoch": 1.4099216710182767,
+      "grad_norm": 0.0357728935778141,
+      "learning_rate": 0.00012754131017835777,
+      "loss": 0.1355,
+      "step": 1485
+    },
+    {
+      "epoch": 1.4108711132209826,
+      "grad_norm": 0.03219794109463692,
+      "learning_rate": 0.0001274350642392271,
+      "loss": 0.1304,
+      "step": 1486
+    },
+    {
+      "epoch": 1.4118205554236885,
+      "grad_norm": 0.04606242850422859,
+      "learning_rate": 0.00012732878479954445,
+      "loss": 0.1629,
+      "step": 1487
+    },
+    {
+      "epoch": 1.4127699976263945,
+      "grad_norm": 0.04288827255368233,
+      "learning_rate": 0.000127222471989086,
+      "loss": 0.1667,
+      "step": 1488
+    },
+    {
+      "epoch": 1.4137194398291004,
+      "grad_norm": 0.031533095985651016,
+      "learning_rate": 0.0001271161259376688,
+      "loss": 0.1266,
+      "step": 1489
+    },
+    {
+      "epoch": 1.4146688820318063,
+      "grad_norm": 0.03418329730629921,
+      "learning_rate": 0.00012700974677515046,
+      "loss": 0.1441,
+      "step": 1490
+    },
+    {
+      "epoch": 1.4156183242345122,
+      "grad_norm": 0.028918685391545296,
+      "learning_rate": 0.00012690333463142897,
+      "loss": 0.117,
+      "step": 1491
+    },
+    {
+      "epoch": 1.4165677664372183,
+      "grad_norm": 0.04630662500858307,
+      "learning_rate": 0.00012679688963644265,
+      "loss": 0.1694,
+      "step": 1492
+    },
+    {
+      "epoch": 1.4175172086399241,
+      "grad_norm": 0.028670761734247208,
+      "learning_rate": 0.00012669041192016993,
+      "loss": 0.1218,
+      "step": 1493
+    },
+    {
+      "epoch": 1.41846665084263,
+      "grad_norm": 0.03250902146100998,
+      "learning_rate": 0.0001265839016126291,
+      "loss": 0.1353,
+      "step": 1494
+    },
+    {
+      "epoch": 1.419416093045336,
+      "grad_norm": 0.03904202580451965,
+      "learning_rate": 0.00012647735884387842,
+      "loss": 0.1566,
+      "step": 1495
+    },
+    {
+      "epoch": 1.4203655352480418,
+      "grad_norm": 0.030610278248786926,
+      "learning_rate": 0.00012637078374401568,
+      "loss": 0.1248,
+      "step": 1496
+    },
+    {
+      "epoch": 1.4213149774507476,
+      "grad_norm": 0.0320439413189888,
+      "learning_rate": 0.00012626417644317808,
+      "loss": 0.1341,
+      "step": 1497
+    },
+    {
+      "epoch": 1.4222644196534535,
+      "grad_norm": 0.03740748390555382,
+      "learning_rate": 0.0001261575370715423,
+      "loss": 0.1374,
+      "step": 1498
+    },
+    {
+      "epoch": 1.4232138618561594,
+      "grad_norm": 0.041164278984069824,
+      "learning_rate": 0.00012605086575932407,
+      "loss": 0.1242,
+      "step": 1499
+    },
+    {
+      "epoch": 1.4241633040588655,
+      "grad_norm": 0.02908271551132202,
+      "learning_rate": 0.00012594416263677816,
+      "loss": 0.1224,
+      "step": 1500
+    },
+    {
+      "epoch": 1.4251127462615714,
+      "grad_norm": 0.030539128929376602,
+      "learning_rate": 0.0001258374278341982,
+      "loss": 0.1236,
+      "step": 1501
+    },
+    {
+      "epoch": 1.4260621884642772,
+      "grad_norm": 0.027197500690817833,
+      "learning_rate": 0.00012573066148191647,
+      "loss": 0.1254,
+      "step": 1502
+    },
+    {
+      "epoch": 1.427011630666983,
+      "grad_norm": 0.031813718378543854,
+      "learning_rate": 0.00012562386371030377,
+      "loss": 0.1294,
+      "step": 1503
+    },
+    {
+      "epoch": 1.427961072869689,
+      "grad_norm": 0.04184641316533089,
+      "learning_rate": 0.00012551703464976928,
+      "loss": 0.1615,
+      "step": 1504
+    },
+    {
+      "epoch": 1.428910515072395,
+      "grad_norm": 0.03790717199444771,
+      "learning_rate": 0.00012541017443076042,
+      "loss": 0.1638,
+      "step": 1505
+    },
+    {
+      "epoch": 1.429859957275101,
+      "grad_norm": 0.03084125556051731,
+      "learning_rate": 0.00012530328318376258,
+      "loss": 0.1292,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4308093994778068,
+      "grad_norm": 0.042278289794921875,
+      "learning_rate": 0.00012519636103929912,
+      "loss": 0.1691,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4317588416805127,
+      "grad_norm": 0.02734595723450184,
+      "learning_rate": 0.0001250894081279311,
+      "loss": 0.1248,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4327082838832186,
+      "grad_norm": 0.02997264452278614,
+      "learning_rate": 0.00012498242458025712,
+      "loss": 0.124,
+      "step": 1509
+    },
+    {
+      "epoch": 1.4336577260859245,
+      "grad_norm": 0.031008126214146614,
+      "learning_rate": 0.00012487541052691323,
+      "loss": 0.1335,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4346071682886303,
+      "grad_norm": 0.042471520602703094,
+      "learning_rate": 0.0001247683660985727,
+      "loss": 0.1589,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4355566104913362,
+      "grad_norm": 0.027912134304642677,
+      "learning_rate": 0.00012466129142594588,
+      "loss": 0.1208,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4365060526940423,
+      "grad_norm": 0.03753120079636574,
+      "learning_rate": 0.0001245541866397801,
+      "loss": 0.1626,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4374554948967482,
+      "grad_norm": 0.02756452187895775,
+      "learning_rate": 0.0001244470518708594,
+      "loss": 0.1215,
+      "step": 1514
+    },
+    {
+      "epoch": 1.438404937099454,
+      "grad_norm": 0.03357706964015961,
+      "learning_rate": 0.0001243398872500045,
+      "loss": 0.1376,
+      "step": 1515
+    },
+    {
+      "epoch": 1.43935437930216,
+      "grad_norm": 0.032955266535282135,
+      "learning_rate": 0.00012423269290807258,
+      "loss": 0.1364,
+      "step": 1516
+    },
+    {
+      "epoch": 1.440303821504866,
+      "grad_norm": 0.03405732661485672,
+      "learning_rate": 0.000124125468975957,
+      "loss": 0.126,
+      "step": 1517
+    },
+    {
+      "epoch": 1.441253263707572,
+      "grad_norm": 0.03659766912460327,
+      "learning_rate": 0.00012401821558458728,
+      "loss": 0.1561,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4422027059102778,
+      "grad_norm": 0.030413653701543808,
+      "learning_rate": 0.00012391093286492905,
+      "loss": 0.1253,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4431521481129836,
+      "grad_norm": 0.027322586625814438,
+      "learning_rate": 0.00012380362094798362,
+      "loss": 0.1217,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4441015903156895,
+      "grad_norm": 0.037558842450380325,
+      "learning_rate": 0.00012369627996478797,
+      "loss": 0.1348,
+      "step": 1521
+    },
+    {
+      "epoch": 1.4450510325183954,
+      "grad_norm": 0.028927726671099663,
+      "learning_rate": 0.0001235889100464146,
+      "loss": 0.1184,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4460004747211013,
+      "grad_norm": 0.028258686885237694,
+      "learning_rate": 0.00012348151132397133,
+      "loss": 0.1276,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4469499169238071,
+      "grad_norm": 0.027749182656407356,
+      "learning_rate": 0.00012337408392860117,
+      "loss": 0.1235,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4478993591265132,
+      "grad_norm": 0.03684193268418312,
+      "learning_rate": 0.0001232666279914821,
+      "loss": 0.1421,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4488488013292191,
+      "grad_norm": 0.033882539719343185,
+      "learning_rate": 0.00012315914364382705,
+      "loss": 0.1236,
+      "step": 1526
+    },
+    {
+      "epoch": 1.449798243531925,
+      "grad_norm": 0.03675851225852966,
+      "learning_rate": 0.00012305163101688352,
+      "loss": 0.1618,
+      "step": 1527
+    },
+    {
+      "epoch": 1.4507476857346309,
+      "grad_norm": 0.030739063397049904,
+      "learning_rate": 0.00012294409024193355,
+      "loss": 0.127,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4516971279373367,
+      "grad_norm": 0.026678606867790222,
+      "learning_rate": 0.00012283652145029362,
+      "loss": 0.1191,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4526465701400428,
+      "grad_norm": 0.028862981125712395,
+      "learning_rate": 0.0001227289247733144,
+      "loss": 0.1255,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4535960123427487,
+      "grad_norm": 0.02993926964700222,
+      "learning_rate": 0.0001226213003423807,
+      "loss": 0.1235,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.02765621617436409,
+      "learning_rate": 0.00012251364828891093,
+      "loss": 0.127,
+      "step": 1532
+    },
+    {
+      "epoch": 1.4554948967481605,
+      "grad_norm": 0.03139231353998184,
+      "learning_rate": 0.00012240596874435756,
+      "loss": 0.1225,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4564443389508663,
+      "grad_norm": 0.030977755784988403,
+      "learning_rate": 0.00012229826184020649,
+      "loss": 0.1249,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4573937811535722,
+      "grad_norm": 0.03158799931406975,
+      "learning_rate": 0.000122190527707977,
+      "loss": 0.1254,
+      "step": 1535
+    },
+    {
+      "epoch": 1.458343223356278,
+      "grad_norm": 0.03206062689423561,
+      "learning_rate": 0.00012208276647922162,
+      "loss": 0.1332,
+      "step": 1536
+    },
+    {
+      "epoch": 1.459292665558984,
+      "grad_norm": 0.028509238734841347,
+      "learning_rate": 0.00012197497828552601,
+      "loss": 0.1196,
+      "step": 1537
+    },
+    {
+      "epoch": 1.46024210776169,
+      "grad_norm": 0.03207945078611374,
+      "learning_rate": 0.0001218671632585088,
+      "loss": 0.1241,
+      "step": 1538
+    },
+    {
+      "epoch": 1.461191549964396,
+      "grad_norm": 0.026839956641197205,
+      "learning_rate": 0.00012175932152982125,
+      "loss": 0.1246,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4621409921671018,
+      "grad_norm": 0.03129103407263756,
+      "learning_rate": 0.0001216514532311474,
+      "loss": 0.1192,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4630904343698077,
+      "grad_norm": 0.027821926400065422,
+      "learning_rate": 0.00012154355849420353,
+      "loss": 0.1221,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4640398765725138,
+      "grad_norm": 0.028374498710036278,
+      "learning_rate": 0.00012143563745073836,
+      "loss": 0.1253,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4649893187752197,
+      "grad_norm": 0.05225376784801483,
+      "learning_rate": 0.0001213276902325327,
+      "loss": 0.1194,
+      "step": 1543
+    },
+    {
+      "epoch": 1.4659387609779255,
+      "grad_norm": 0.04315062612295151,
+      "learning_rate": 0.00012121971697139926,
+      "loss": 0.1613,
+      "step": 1544
+    },
+    {
+      "epoch": 1.4668882031806314,
+      "grad_norm": 0.027962563559412956,
+      "learning_rate": 0.00012111171779918264,
+      "loss": 0.1284,
+      "step": 1545
+    },
+    {
+      "epoch": 1.4678376453833373,
+      "grad_norm": 0.0343998521566391,
+      "learning_rate": 0.000121003692847759,
+      "loss": 0.1208,
+      "step": 1546
+    },
+    {
+      "epoch": 1.4687870875860431,
+      "grad_norm": 0.031139155849814415,
+      "learning_rate": 0.00012089564224903607,
+      "loss": 0.1323,
+      "step": 1547
+    },
+    {
+      "epoch": 1.469736529788749,
+      "grad_norm": 0.028015103191137314,
+      "learning_rate": 0.00012078756613495277,
+      "loss": 0.1266,
+      "step": 1548
+    },
+    {
+      "epoch": 1.470685971991455,
+      "grad_norm": 0.04571033641695976,
+      "learning_rate": 0.00012067946463747928,
+      "loss": 0.1561,
+      "step": 1549
+    },
+    {
+      "epoch": 1.471635414194161,
+      "grad_norm": 0.027564501389861107,
+      "learning_rate": 0.00012057133788861677,
+      "loss": 0.1214,
+      "step": 1550
+    },
+    {
+      "epoch": 1.4725848563968669,
+      "grad_norm": 0.027498599141836166,
+      "learning_rate": 0.00012046318602039717,
+      "loss": 0.1233,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4735342985995727,
+      "grad_norm": 0.040304671972990036,
+      "learning_rate": 0.00012035500916488316,
+      "loss": 0.1636,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4744837408022786,
+      "grad_norm": 0.045689020305871964,
+      "learning_rate": 0.00012024680745416787,
+      "loss": 0.1696,
+      "step": 1553
+    },
+    {
+      "epoch": 1.4754331830049845,
+      "grad_norm": 0.02754673734307289,
+      "learning_rate": 0.00012013858102037485,
+      "loss": 0.1265,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4763826252076906,
+      "grad_norm": 0.04088296741247177,
+      "learning_rate": 0.0001200303299956578,
+      "loss": 0.1675,
+      "step": 1555
+    },
+    {
+      "epoch": 1.4773320674103965,
+      "grad_norm": 0.03440406173467636,
+      "learning_rate": 0.0001199220545122004,
+      "loss": 0.1245,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4782815096131023,
+      "grad_norm": 0.04637245833873749,
+      "learning_rate": 0.00011981375470221628,
+      "loss": 0.1668,
+      "step": 1557
+    },
+    {
+      "epoch": 1.4792309518158082,
+      "grad_norm": 0.03825761005282402,
+      "learning_rate": 0.00011970543069794875,
+      "loss": 0.1599,
+      "step": 1558
+    },
+    {
+      "epoch": 1.480180394018514,
+      "grad_norm": 0.031823255121707916,
+      "learning_rate": 0.00011959708263167067,
+      "loss": 0.1232,
+      "step": 1559
+    },
+    {
+      "epoch": 1.48112983622122,
+      "grad_norm": 0.032104745507240295,
+      "learning_rate": 0.00011948871063568419,
+      "loss": 0.1237,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4820792784239258,
+      "grad_norm": 0.05664534121751785,
+      "learning_rate": 0.00011938031484232079,
+      "loss": 0.1617,
+      "step": 1561
+    },
+    {
+      "epoch": 1.483028720626632,
+      "grad_norm": 0.02819441631436348,
+      "learning_rate": 0.00011927189538394101,
+      "loss": 0.1198,
+      "step": 1562
+    },
+    {
+      "epoch": 1.4839781628293378,
+      "grad_norm": 0.030618941411376,
+      "learning_rate": 0.00011916345239293423,
+      "loss": 0.1291,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4849276050320437,
+      "grad_norm": 0.027746165171265602,
+      "learning_rate": 0.00011905498600171859,
+      "loss": 0.1287,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4858770472347496,
+      "grad_norm": 0.05675269663333893,
+      "learning_rate": 0.00011894649634274075,
+      "loss": 0.1988,
+      "step": 1565
+    },
+    {
+      "epoch": 1.4868264894374554,
+      "grad_norm": 0.044594231992959976,
+      "learning_rate": 0.00011883798354847589,
+      "loss": 0.1611,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4877759316401615,
+      "grad_norm": 0.050554268062114716,
+      "learning_rate": 0.00011872944775142736,
+      "loss": 0.1762,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4887253738428674,
+      "grad_norm": 0.04185627028346062,
+      "learning_rate": 0.0001186208890841266,
+      "loss": 0.1593,
+      "step": 1568
+    },
+    {
+      "epoch": 1.4896748160455733,
+      "grad_norm": 0.030089320614933968,
+      "learning_rate": 0.00011851230767913303,
+      "loss": 0.1217,
+      "step": 1569
+    },
+    {
+      "epoch": 1.4906242582482792,
+      "grad_norm": 0.033343639224767685,
+      "learning_rate": 0.00011840370366903382,
+      "loss": 0.1284,
+      "step": 1570
+    },
+    {
+      "epoch": 1.491573700450985,
+      "grad_norm": 0.030400337651371956,
+      "learning_rate": 0.00011829507718644366,
+      "loss": 0.1315,
+      "step": 1571
+    },
+    {
+      "epoch": 1.492523142653691,
+      "grad_norm": 0.026966162025928497,
+      "learning_rate": 0.00011818642836400477,
+      "loss": 0.1206,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4934725848563968,
+      "grad_norm": 0.030626913532614708,
+      "learning_rate": 0.00011807775733438664,
+      "loss": 0.1228,
+      "step": 1573
+    },
+    {
+      "epoch": 1.4944220270591027,
+      "grad_norm": 0.03993997722864151,
+      "learning_rate": 0.00011796906423028588,
+      "loss": 0.1488,
+      "step": 1574
+    },
+    {
+      "epoch": 1.4953714692618088,
+      "grad_norm": 0.030942171812057495,
+      "learning_rate": 0.00011786034918442596,
+      "loss": 0.1284,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4963209114645146,
+      "grad_norm": 0.027237005531787872,
+      "learning_rate": 0.00011775161232955729,
+      "loss": 0.1278,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4972703536672205,
+      "grad_norm": 0.030200913548469543,
+      "learning_rate": 0.0001176428537984568,
+      "loss": 0.1264,
+      "step": 1577
+    },
+    {
+      "epoch": 1.4982197958699264,
+      "grad_norm": 0.03261629864573479,
+      "learning_rate": 0.00011753407372392795,
+      "loss": 0.128,
+      "step": 1578
+    },
+    {
+      "epoch": 1.4991692380726322,
+      "grad_norm": 0.033451877534389496,
+      "learning_rate": 0.0001174252722388005,
+      "loss": 0.1345,
+      "step": 1579
+    },
+    {
+      "epoch": 1.5001186802753383,
+      "grad_norm": 0.045393262058496475,
+      "learning_rate": 0.00011731644947593026,
+      "loss": 0.1627,
+      "step": 1580
+    },
+    {
+      "epoch": 1.5010681224780442,
+      "grad_norm": 0.05379907041788101,
+      "learning_rate": 0.00011720760556819916,
+      "loss": 0.1867,
+      "step": 1581
+    },
+    {
+      "epoch": 1.50201756468075,
+      "grad_norm": 0.025756366550922394,
+      "learning_rate": 0.00011709874064851487,
+      "loss": 0.1229,
+      "step": 1582
+    },
+    {
+      "epoch": 1.502967006883456,
+      "grad_norm": 0.03149167448282242,
+      "learning_rate": 0.00011698985484981077,
+      "loss": 0.124,
+      "step": 1583
+    },
+    {
+      "epoch": 1.5039164490861618,
+      "grad_norm": 0.044556062668561935,
+      "learning_rate": 0.00011688094830504566,
+      "loss": 0.1522,
+      "step": 1584
+    },
+    {
+      "epoch": 1.5048658912888677,
+      "grad_norm": 0.029473567381501198,
+      "learning_rate": 0.00011677202114720374,
+      "loss": 0.1231,
+      "step": 1585
+    },
+    {
+      "epoch": 1.5058153334915736,
+      "grad_norm": 0.02735304646193981,
+      "learning_rate": 0.00011666307350929435,
+      "loss": 0.1242,
+      "step": 1586
+    },
+    {
+      "epoch": 1.5067647756942795,
+      "grad_norm": 0.04760657250881195,
+      "learning_rate": 0.00011655410552435184,
+      "loss": 0.1574,
+      "step": 1587
+    },
+    {
+      "epoch": 1.5077142178969856,
+      "grad_norm": 0.029798056930303574,
+      "learning_rate": 0.00011644511732543542,
+      "loss": 0.1289,
+      "step": 1588
+    },
+    {
+      "epoch": 1.5086636600996914,
+      "grad_norm": 0.0329173281788826,
+      "learning_rate": 0.00011633610904562892,
+      "loss": 0.1251,
+      "step": 1589
+    },
+    {
+      "epoch": 1.5096131023023973,
+      "grad_norm": 0.030921783298254013,
+      "learning_rate": 0.00011622708081804081,
+      "loss": 0.1257,
+      "step": 1590
+    },
+    {
+      "epoch": 1.5105625445051034,
+      "grad_norm": 0.031381431967020035,
+      "learning_rate": 0.0001161180327758038,
+      "loss": 0.1258,
+      "step": 1591
+    },
+    {
+      "epoch": 1.5115119867078093,
+      "grad_norm": 0.050078343600034714,
+      "learning_rate": 0.00011600896505207488,
+      "loss": 0.1764,
+      "step": 1592
+    },
+    {
+      "epoch": 1.5124614289105152,
+      "grad_norm": 0.04285876452922821,
+      "learning_rate": 0.00011589987778003501,
+      "loss": 0.1607,
+      "step": 1593
+    },
+    {
+      "epoch": 1.513410871113221,
+      "grad_norm": 0.02959441766142845,
+      "learning_rate": 0.00011579077109288907,
+      "loss": 0.1266,
+      "step": 1594
+    },
+    {
+      "epoch": 1.514360313315927,
+      "grad_norm": 0.03024190105497837,
+      "learning_rate": 0.00011568164512386559,
+      "loss": 0.1306,
+      "step": 1595
+    },
+    {
+      "epoch": 1.5153097555186328,
+      "grad_norm": 0.031076421961188316,
+      "learning_rate": 0.00011557250000621674,
+      "loss": 0.1326,
+      "step": 1596
+    },
+    {
+      "epoch": 1.5162591977213387,
+      "grad_norm": 0.10080650448799133,
+      "learning_rate": 0.00011546333587321795,
+      "loss": 0.1596,
+      "step": 1597
+    },
+    {
+      "epoch": 1.5172086399240445,
+      "grad_norm": 0.029658254235982895,
+      "learning_rate": 0.000115354152858168,
+      "loss": 0.1306,
+      "step": 1598
+    },
+    {
+      "epoch": 1.5181580821267504,
+      "grad_norm": 0.025450505316257477,
+      "learning_rate": 0.00011524495109438857,
+      "loss": 0.1199,
+      "step": 1599
+    },
+    {
+      "epoch": 1.5191075243294563,
+      "grad_norm": 0.0450344942510128,
+      "learning_rate": 0.00011513573071522439,
+      "loss": 0.1671,
+      "step": 1600
+    },
+    {
+      "epoch": 1.5200569665321624,
+      "grad_norm": 0.02928159199655056,
+      "learning_rate": 0.00011502649185404281,
+      "loss": 0.1247,
+      "step": 1601
+    },
+    {
+      "epoch": 1.5210064087348683,
+      "grad_norm": 0.04587600752711296,
+      "learning_rate": 0.00011491723464423385,
+      "loss": 0.1746,
+      "step": 1602
+    },
+    {
+      "epoch": 1.5219558509375741,
+      "grad_norm": 0.03615221753716469,
+      "learning_rate": 0.00011480795921920984,
+      "loss": 0.153,
+      "step": 1603
+    },
+    {
+      "epoch": 1.5229052931402802,
+      "grad_norm": 0.026470355689525604,
+      "learning_rate": 0.00011469866571240535,
+      "loss": 0.1204,
+      "step": 1604
+    },
+    {
+      "epoch": 1.523854735342986,
+      "grad_norm": 0.0324716791510582,
+      "learning_rate": 0.00011458935425727713,
+      "loss": 0.127,
+      "step": 1605
+    },
+    {
+      "epoch": 1.524804177545692,
+      "grad_norm": 0.028570136055350304,
+      "learning_rate": 0.00011448002498730375,
+      "loss": 0.1244,
+      "step": 1606
+    },
+    {
+      "epoch": 1.5257536197483978,
+      "grad_norm": 0.03165844827890396,
+      "learning_rate": 0.00011437067803598558,
+      "loss": 0.1286,
+      "step": 1607
+    },
+    {
+      "epoch": 1.5267030619511037,
+      "grad_norm": 0.029820239171385765,
+      "learning_rate": 0.00011426131353684457,
+      "loss": 0.1232,
+      "step": 1608
+    },
+    {
+      "epoch": 1.5276525041538096,
+      "grad_norm": 0.03704296052455902,
+      "learning_rate": 0.00011415193162342407,
+      "loss": 0.1262,
+      "step": 1609
+    },
+    {
+      "epoch": 1.5286019463565155,
+      "grad_norm": 0.029848331585526466,
+      "learning_rate": 0.00011404253242928877,
+      "loss": 0.1345,
+      "step": 1610
+    },
+    {
+      "epoch": 1.5295513885592213,
+      "grad_norm": 0.02494947612285614,
+      "learning_rate": 0.00011393311608802437,
+      "loss": 0.1247,
+      "step": 1611
+    },
+    {
+      "epoch": 1.5305008307619272,
+      "grad_norm": 0.024150602519512177,
+      "learning_rate": 0.0001138236827332376,
+      "loss": 0.1195,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5314502729646333,
+      "grad_norm": 0.028073903173208237,
+      "learning_rate": 0.00011371423249855584,
+      "loss": 0.1298,
+      "step": 1613
+    },
+    {
+      "epoch": 1.5323997151673392,
+      "grad_norm": 0.03696022182703018,
+      "learning_rate": 0.0001136047655176272,
+      "loss": 0.1658,
+      "step": 1614
+    },
+    {
+      "epoch": 1.533349157370045,
+      "grad_norm": 0.026254741474986076,
+      "learning_rate": 0.00011349528192412018,
+      "loss": 0.1209,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5342985995727512,
+      "grad_norm": 0.04131542891263962,
+      "learning_rate": 0.0001133857818517236,
+      "loss": 0.1566,
+      "step": 1616
+    },
+    {
+      "epoch": 1.535248041775457,
+      "grad_norm": 0.028996463865041733,
+      "learning_rate": 0.00011327626543414636,
+      "loss": 0.1287,
+      "step": 1617
+    },
+    {
+      "epoch": 1.536197483978163,
+      "grad_norm": 0.029789695516228676,
+      "learning_rate": 0.00011316673280511738,
+      "loss": 0.1303,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5371469261808688,
+      "grad_norm": 0.026316309347748756,
+      "learning_rate": 0.00011305718409838528,
+      "loss": 0.1224,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5380963683835747,
+      "grad_norm": 0.029919691383838654,
+      "learning_rate": 0.0001129476194477184,
+      "loss": 0.1297,
+      "step": 1620
+    },
+    {
+      "epoch": 1.5390458105862805,
+      "grad_norm": 0.029496189206838608,
+      "learning_rate": 0.0001128380389869045,
+      "loss": 0.1259,
+      "step": 1621
+    },
+    {
+      "epoch": 1.5399952527889864,
+      "grad_norm": 0.0289089847356081,
+      "learning_rate": 0.00011272844284975072,
+      "loss": 0.1138,
+      "step": 1622
+    },
+    {
+      "epoch": 1.5409446949916923,
+      "grad_norm": 0.03120460920035839,
+      "learning_rate": 0.00011261883117008321,
+      "loss": 0.1255,
+      "step": 1623
+    },
+    {
+      "epoch": 1.5418941371943982,
+      "grad_norm": 0.030355585739016533,
+      "learning_rate": 0.0001125092040817472,
+      "loss": 0.1234,
+      "step": 1624
+    },
+    {
+      "epoch": 1.542843579397104,
+      "grad_norm": 0.026462506502866745,
+      "learning_rate": 0.00011239956171860675,
+      "loss": 0.1194,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5437930215998101,
+      "grad_norm": 0.04788212105631828,
+      "learning_rate": 0.00011228990421454449,
+      "loss": 0.1721,
+      "step": 1626
+    },
+    {
+      "epoch": 1.544742463802516,
+      "grad_norm": 0.04839539900422096,
+      "learning_rate": 0.00011218023170346159,
+      "loss": 0.1616,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5456919060052219,
+      "grad_norm": 0.029102666303515434,
+      "learning_rate": 0.00011207054431927752,
+      "loss": 0.1244,
+      "step": 1628
+    },
+    {
+      "epoch": 1.546641348207928,
+      "grad_norm": 0.030667860060930252,
+      "learning_rate": 0.00011196084219592994,
+      "loss": 0.1211,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5475907904106339,
+      "grad_norm": 0.029911190271377563,
+      "learning_rate": 0.00011185112546737451,
+      "loss": 0.1209,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5485402326133397,
+      "grad_norm": 0.026976125314831734,
+      "learning_rate": 0.00011174139426758466,
+      "loss": 0.1201,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5494896748160456,
+      "grad_norm": 0.04256618767976761,
+      "learning_rate": 0.00011163164873055158,
+      "loss": 0.1716,
+      "step": 1632
+    },
+    {
+      "epoch": 1.5504391170187515,
+      "grad_norm": 0.04431037977337837,
+      "learning_rate": 0.00011152188899028393,
+      "loss": 0.1535,
+      "step": 1633
+    },
+    {
+      "epoch": 1.5513885592214574,
+      "grad_norm": 0.040324628353118896,
+      "learning_rate": 0.00011141211518080768,
+      "loss": 0.1559,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5523380014241632,
+      "grad_norm": 0.02631363458931446,
+      "learning_rate": 0.00011130232743616602,
+      "loss": 0.1254,
+      "step": 1635
+    },
+    {
+      "epoch": 1.553287443626869,
+      "grad_norm": 0.031095106154680252,
+      "learning_rate": 0.00011119252589041917,
+      "loss": 0.1296,
+      "step": 1636
+    },
+    {
+      "epoch": 1.554236885829575,
+      "grad_norm": 0.029540112242102623,
+      "learning_rate": 0.00011108271067764413,
+      "loss": 0.1237,
+      "step": 1637
+    },
+    {
+      "epoch": 1.555186328032281,
+      "grad_norm": 0.032384589314460754,
+      "learning_rate": 0.00011097288193193465,
+      "loss": 0.122,
+      "step": 1638
+    },
+    {
+      "epoch": 1.556135770234987,
+      "grad_norm": 0.041656699031591415,
+      "learning_rate": 0.00011086303978740102,
+      "loss": 0.1627,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5570852124376928,
+      "grad_norm": 0.02969949133694172,
+      "learning_rate": 0.00011075318437816981,
+      "loss": 0.1278,
+      "step": 1640
+    },
+    {
+      "epoch": 1.558034654640399,
+      "grad_norm": 0.029392141848802567,
+      "learning_rate": 0.00011064331583838389,
+      "loss": 0.1222,
+      "step": 1641
+    },
+    {
+      "epoch": 1.5589840968431048,
+      "grad_norm": 0.03221196308732033,
+      "learning_rate": 0.0001105334343022021,
+      "loss": 0.1301,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5599335390458107,
+      "grad_norm": 0.04113316535949707,
+      "learning_rate": 0.00011042353990379917,
+      "loss": 0.1545,
+      "step": 1643
+    },
+    {
+      "epoch": 1.5608829812485165,
+      "grad_norm": 0.04606552794575691,
+      "learning_rate": 0.00011031363277736546,
+      "loss": 0.1582,
+      "step": 1644
+    },
+    {
+      "epoch": 1.5618324234512224,
+      "grad_norm": 0.05224507302045822,
+      "learning_rate": 0.00011020371305710701,
+      "loss": 0.1702,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5627818656539283,
+      "grad_norm": 0.03016183339059353,
+      "learning_rate": 0.00011009378087724518,
+      "loss": 0.1299,
+      "step": 1646
+    },
+    {
+      "epoch": 1.5637313078566342,
+      "grad_norm": 0.02981068380177021,
+      "learning_rate": 0.00010998383637201648,
+      "loss": 0.1258,
+      "step": 1647
+    },
+    {
+      "epoch": 1.56468075005934,
+      "grad_norm": 0.027657071128487587,
+      "learning_rate": 0.00010987387967567252,
+      "loss": 0.1338,
+      "step": 1648
+    },
+    {
+      "epoch": 1.565630192262046,
+      "grad_norm": 0.030992772430181503,
+      "learning_rate": 0.00010976391092247986,
+      "loss": 0.1249,
+      "step": 1649
+    },
+    {
+      "epoch": 1.566579634464752,
+      "grad_norm": 0.039394572377204895,
+      "learning_rate": 0.00010965393024671966,
+      "loss": 0.1598,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5675290766674579,
+      "grad_norm": 0.042735736817121506,
+      "learning_rate": 0.00010954393778268777,
+      "loss": 0.1566,
+      "step": 1651
+    },
+    {
+      "epoch": 1.5684785188701638,
+      "grad_norm": 0.03833623602986336,
+      "learning_rate": 0.00010943393366469427,
+      "loss": 0.1557,
+      "step": 1652
+    },
+    {
+      "epoch": 1.5694279610728696,
+      "grad_norm": 0.028264719992876053,
+      "learning_rate": 0.00010932391802706363,
+      "loss": 0.1295,
+      "step": 1653
+    },
+    {
+      "epoch": 1.5703774032755757,
+      "grad_norm": 0.030619991943240166,
+      "learning_rate": 0.00010921389100413428,
+      "loss": 0.1294,
+      "step": 1654
+    },
+    {
+      "epoch": 1.5713268454782816,
+      "grad_norm": 0.0320441797375679,
+      "learning_rate": 0.00010910385273025865,
+      "loss": 0.1266,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5722762876809875,
+      "grad_norm": 0.028138084337115288,
+      "learning_rate": 0.00010899380333980282,
+      "loss": 0.1177,
+      "step": 1656
+    },
+    {
+      "epoch": 1.5732257298836934,
+      "grad_norm": 0.030276020988821983,
+      "learning_rate": 0.00010888374296714644,
+      "loss": 0.1258,
+      "step": 1657
+    },
+    {
+      "epoch": 1.5741751720863992,
+      "grad_norm": 0.04478145390748978,
+      "learning_rate": 0.00010877367174668269,
+      "loss": 0.1555,
+      "step": 1658
+    },
+    {
+      "epoch": 1.575124614289105,
+      "grad_norm": 0.025850724428892136,
+      "learning_rate": 0.00010866358981281783,
+      "loss": 0.1186,
+      "step": 1659
+    },
+    {
+      "epoch": 1.576074056491811,
+      "grad_norm": 0.03069223277270794,
+      "learning_rate": 0.00010855349729997135,
+      "loss": 0.1314,
+      "step": 1660
+    },
+    {
+      "epoch": 1.5770234986945169,
+      "grad_norm": 0.050107646733522415,
+      "learning_rate": 0.00010844339434257558,
+      "loss": 0.1632,
+      "step": 1661
+    },
+    {
+      "epoch": 1.5779729408972227,
+      "grad_norm": 0.04577158764004707,
+      "learning_rate": 0.00010833328107507556,
+      "loss": 0.1777,
+      "step": 1662
+    },
+    {
+      "epoch": 1.5789223830999288,
+      "grad_norm": 0.030337292701005936,
+      "learning_rate": 0.00010822315763192903,
+      "loss": 0.1277,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5798718253026347,
+      "grad_norm": 0.05210689827799797,
+      "learning_rate": 0.00010811302414760609,
+      "loss": 0.1583,
+      "step": 1664
+    },
+    {
+      "epoch": 1.5808212675053406,
+      "grad_norm": 0.056502800434827805,
+      "learning_rate": 0.00010800288075658911,
+      "loss": 0.1533,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5817707097080467,
+      "grad_norm": 0.03072887845337391,
+      "learning_rate": 0.00010789272759337257,
+      "loss": 0.1349,
+      "step": 1666
+    },
+    {
+      "epoch": 1.5827201519107525,
+      "grad_norm": 0.030606022104620934,
+      "learning_rate": 0.00010778256479246283,
+      "loss": 0.1288,
+      "step": 1667
+    },
+    {
+      "epoch": 1.5836695941134584,
+      "grad_norm": 0.027298327535390854,
+      "learning_rate": 0.00010767239248837811,
+      "loss": 0.1256,
+      "step": 1668
+    },
+    {
+      "epoch": 1.5846190363161643,
+      "grad_norm": 0.044606730341911316,
+      "learning_rate": 0.00010756221081564813,
+      "loss": 0.1673,
+      "step": 1669
+    },
+    {
+      "epoch": 1.5855684785188702,
+      "grad_norm": 0.04783207178115845,
+      "learning_rate": 0.00010745201990881417,
+      "loss": 0.1539,
+      "step": 1670
+    },
+    {
+      "epoch": 1.586517920721576,
+      "grad_norm": 0.03072645701467991,
+      "learning_rate": 0.00010734181990242868,
+      "loss": 0.1255,
+      "step": 1671
+    },
+    {
+      "epoch": 1.587467362924282,
+      "grad_norm": 0.028861412778496742,
+      "learning_rate": 0.00010723161093105527,
+      "loss": 0.1252,
+      "step": 1672
+    },
+    {
+      "epoch": 1.5884168051269878,
+      "grad_norm": 0.06253904849290848,
+      "learning_rate": 0.0001071213931292685,
+      "loss": 0.208,
+      "step": 1673
+    },
+    {
+      "epoch": 1.5893662473296937,
+      "grad_norm": 0.02638799510896206,
+      "learning_rate": 0.00010701116663165368,
+      "loss": 0.1307,
+      "step": 1674
+    },
+    {
+      "epoch": 1.5903156895323998,
+      "grad_norm": 0.02984490990638733,
+      "learning_rate": 0.00010690093157280678,
+      "loss": 0.1305,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5912651317351056,
+      "grad_norm": 0.029151547700166702,
+      "learning_rate": 0.0001067906880873342,
+      "loss": 0.1278,
+      "step": 1676
+    },
+    {
+      "epoch": 1.5922145739378115,
+      "grad_norm": 0.03524734824895859,
+      "learning_rate": 0.00010668043630985259,
+      "loss": 0.1426,
+      "step": 1677
+    },
+    {
+      "epoch": 1.5931640161405174,
+      "grad_norm": 0.03428010269999504,
+      "learning_rate": 0.00010657017637498881,
+      "loss": 0.125,
+      "step": 1678
+    },
+    {
+      "epoch": 1.5941134583432235,
+      "grad_norm": 0.05122271180152893,
+      "learning_rate": 0.00010645990841737965,
+      "loss": 0.1569,
+      "step": 1679
+    },
+    {
+      "epoch": 1.5950629005459294,
+      "grad_norm": 0.035184647887945175,
+      "learning_rate": 0.00010634963257167167,
+      "loss": 0.1358,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5960123427486352,
+      "grad_norm": 0.03079393319785595,
+      "learning_rate": 0.00010623934897252106,
+      "loss": 0.1268,
+      "step": 1681
+    },
+    {
+      "epoch": 1.5969617849513411,
+      "grad_norm": 0.03128993511199951,
+      "learning_rate": 0.00010612905775459349,
+      "loss": 0.1223,
+      "step": 1682
+    },
+    {
+      "epoch": 1.597911227154047,
+      "grad_norm": 0.02892274223268032,
+      "learning_rate": 0.00010601875905256398,
+      "loss": 0.1293,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5988606693567529,
+      "grad_norm": 0.030298085883259773,
+      "learning_rate": 0.00010590845300111663,
+      "loss": 0.1293,
+      "step": 1684
+    },
+    {
+      "epoch": 1.5998101115594587,
+      "grad_norm": 0.028357641771435738,
+      "learning_rate": 0.00010579813973494454,
+      "loss": 0.1269,
+      "step": 1685
+    },
+    {
+      "epoch": 1.6007595537621646,
+      "grad_norm": 0.027196258306503296,
+      "learning_rate": 0.00010568781938874959,
+      "loss": 0.1296,
+      "step": 1686
+    },
+    {
+      "epoch": 1.6017089959648705,
+      "grad_norm": 0.05107175186276436,
+      "learning_rate": 0.00010557749209724233,
+      "loss": 0.1604,
+      "step": 1687
+    },
+    {
+      "epoch": 1.6026584381675766,
+      "grad_norm": 0.050482697784900665,
+      "learning_rate": 0.00010546715799514178,
+      "loss": 0.1902,
+      "step": 1688
+    },
+    {
+      "epoch": 1.6036078803702825,
+      "grad_norm": 0.026253553107380867,
+      "learning_rate": 0.00010535681721717529,
+      "loss": 0.1226,
+      "step": 1689
+    },
+    {
+      "epoch": 1.6045573225729883,
+      "grad_norm": 0.03308340907096863,
+      "learning_rate": 0.0001052464698980784,
+      "loss": 0.1384,
+      "step": 1690
+    },
+    {
+      "epoch": 1.6055067647756944,
+      "grad_norm": 0.03511514514684677,
+      "learning_rate": 0.00010513611617259454,
+      "loss": 0.1577,
+      "step": 1691
+    },
+    {
+      "epoch": 1.6064562069784003,
+      "grad_norm": 0.0325862281024456,
+      "learning_rate": 0.00010502575617547501,
+      "loss": 0.1343,
+      "step": 1692
+    },
+    {
+      "epoch": 1.6074056491811062,
+      "grad_norm": 0.029174668714404106,
+      "learning_rate": 0.00010491539004147879,
+      "loss": 0.1222,
+      "step": 1693
+    },
+    {
+      "epoch": 1.608355091383812,
+      "grad_norm": 0.028774891048669815,
+      "learning_rate": 0.00010480501790537236,
+      "loss": 0.1237,
+      "step": 1694
+    },
+    {
+      "epoch": 1.609304533586518,
+      "grad_norm": 0.030504655092954636,
+      "learning_rate": 0.00010469463990192947,
+      "loss": 0.1228,
+      "step": 1695
+    },
+    {
+      "epoch": 1.6102539757892238,
+      "grad_norm": 0.02889893390238285,
+      "learning_rate": 0.0001045842561659311,
+      "loss": 0.1208,
+      "step": 1696
+    },
+    {
+      "epoch": 1.6112034179919297,
+      "grad_norm": 0.028366010636091232,
+      "learning_rate": 0.00010447386683216518,
+      "loss": 0.1193,
+      "step": 1697
+    },
+    {
+      "epoch": 1.6121528601946356,
+      "grad_norm": 0.028841307386755943,
+      "learning_rate": 0.0001043634720354265,
+      "loss": 0.1287,
+      "step": 1698
+    },
+    {
+      "epoch": 1.6131023023973414,
+      "grad_norm": 0.03739466145634651,
+      "learning_rate": 0.00010425307191051654,
+      "loss": 0.1369,
+      "step": 1699
+    },
+    {
+      "epoch": 1.6140517446000475,
+      "grad_norm": 0.031514909118413925,
+      "learning_rate": 0.00010414266659224323,
+      "loss": 0.1301,
+      "step": 1700
+    },
+    {
+      "epoch": 1.6150011868027534,
+      "grad_norm": 0.04335467517375946,
+      "learning_rate": 0.00010403225621542089,
+      "loss": 0.1543,
+      "step": 1701
+    },
+    {
+      "epoch": 1.6159506290054593,
+      "grad_norm": 0.026403702795505524,
+      "learning_rate": 0.00010392184091487,
+      "loss": 0.1229,
+      "step": 1702
+    },
+    {
+      "epoch": 1.6169000712081654,
+      "grad_norm": 0.06492079049348831,
+      "learning_rate": 0.00010381142082541706,
+      "loss": 0.1526,
+      "step": 1703
+    },
+    {
+      "epoch": 1.6178495134108712,
+      "grad_norm": 0.031183555722236633,
+      "learning_rate": 0.00010370099608189439,
+      "loss": 0.1255,
+      "step": 1704
+    },
+    {
+      "epoch": 1.6187989556135771,
+      "grad_norm": 0.02694527618587017,
+      "learning_rate": 0.00010359056681914006,
+      "loss": 0.1178,
+      "step": 1705
+    },
+    {
+      "epoch": 1.619748397816283,
+      "grad_norm": 0.031590498983860016,
+      "learning_rate": 0.00010348013317199756,
+      "loss": 0.1311,
+      "step": 1706
+    },
+    {
+      "epoch": 1.6206978400189889,
+      "grad_norm": 0.034355372190475464,
+      "learning_rate": 0.00010336969527531577,
+      "loss": 0.1363,
+      "step": 1707
+    },
+    {
+      "epoch": 1.6216472822216947,
+      "grad_norm": 0.04307783022522926,
+      "learning_rate": 0.00010325925326394886,
+      "loss": 0.1659,
+      "step": 1708
+    },
+    {
+      "epoch": 1.6225967244244006,
+      "grad_norm": 0.05346130579710007,
+      "learning_rate": 0.00010314880727275591,
+      "loss": 0.2022,
+      "step": 1709
+    },
+    {
+      "epoch": 1.6235461666271065,
+      "grad_norm": 0.028053171932697296,
+      "learning_rate": 0.00010303835743660086,
+      "loss": 0.1251,
+      "step": 1710
+    },
+    {
+      "epoch": 1.6244956088298124,
+      "grad_norm": 0.03092275932431221,
+      "learning_rate": 0.00010292790389035239,
+      "loss": 0.1299,
+      "step": 1711
+    },
+    {
+      "epoch": 1.6254450510325182,
+      "grad_norm": 0.03486338630318642,
+      "learning_rate": 0.00010281744676888368,
+      "loss": 0.1374,
+      "step": 1712
+    },
+    {
+      "epoch": 1.6263944932352243,
+      "grad_norm": 0.030160879716277122,
+      "learning_rate": 0.00010270698620707231,
+      "loss": 0.1381,
+      "step": 1713
+    },
+    {
+      "epoch": 1.6273439354379302,
+      "grad_norm": 0.03339090943336487,
+      "learning_rate": 0.00010259652233980007,
+      "loss": 0.1316,
+      "step": 1714
+    },
+    {
+      "epoch": 1.628293377640636,
+      "grad_norm": 0.027464497834444046,
+      "learning_rate": 0.00010248605530195268,
+      "loss": 0.1204,
+      "step": 1715
+    },
+    {
+      "epoch": 1.6292428198433422,
+      "grad_norm": 0.02733561024069786,
+      "learning_rate": 0.00010237558522841985,
+      "loss": 0.1259,
+      "step": 1716
+    },
+    {
+      "epoch": 1.630192262046048,
+      "grad_norm": 0.029772773385047913,
+      "learning_rate": 0.00010226511225409499,
+      "loss": 0.1252,
+      "step": 1717
+    },
+    {
+      "epoch": 1.631141704248754,
+      "grad_norm": 0.04465902969241142,
+      "learning_rate": 0.00010215463651387499,
+      "loss": 0.151,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6320911464514598,
+      "grad_norm": 0.028140738606452942,
+      "learning_rate": 0.0001020441581426601,
+      "loss": 0.1221,
+      "step": 1719
+    },
+    {
+      "epoch": 1.6330405886541657,
+      "grad_norm": 0.026517389342188835,
+      "learning_rate": 0.00010193367727535392,
+      "loss": 0.1253,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6339900308568716,
+      "grad_norm": 0.06271334737539291,
+      "learning_rate": 0.00010182319404686293,
+      "loss": 0.2072,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6349394730595774,
+      "grad_norm": 0.051109135150909424,
+      "learning_rate": 0.00010171270859209662,
+      "loss": 0.1923,
+      "step": 1722
+    },
+    {
+      "epoch": 1.6358889152622833,
+      "grad_norm": 0.02858104184269905,
+      "learning_rate": 0.00010160222104596716,
+      "loss": 0.1317,
+      "step": 1723
+    },
+    {
+      "epoch": 1.6368383574649892,
+      "grad_norm": 0.031905338168144226,
+      "learning_rate": 0.00010149173154338917,
+      "loss": 0.1334,
+      "step": 1724
+    },
+    {
+      "epoch": 1.6377877996676953,
+      "grad_norm": 0.035984478890895844,
+      "learning_rate": 0.00010138124021927984,
+      "loss": 0.1308,
+      "step": 1725
+    },
+    {
+      "epoch": 1.6387372418704012,
+      "grad_norm": 0.029515955597162247,
+      "learning_rate": 0.00010127074720855845,
+      "loss": 0.1232,
+      "step": 1726
+    },
+    {
+      "epoch": 1.639686684073107,
+      "grad_norm": 0.03353870287537575,
+      "learning_rate": 0.0001011602526461464,
+      "loss": 0.1342,
+      "step": 1727
+    },
+    {
+      "epoch": 1.6406361262758131,
+      "grad_norm": 0.02787208929657936,
+      "learning_rate": 0.00010104975666696697,
+      "loss": 0.1216,
+      "step": 1728
+    },
+    {
+      "epoch": 1.641585568478519,
+      "grad_norm": 0.061213839799165726,
+      "learning_rate": 0.0001009392594059452,
+      "loss": 0.2093,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6425350106812249,
+      "grad_norm": 0.034235142171382904,
+      "learning_rate": 0.0001008287609980076,
+      "loss": 0.1329,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6434844528839307,
+      "grad_norm": 0.026360424235463142,
+      "learning_rate": 0.00010071826157808217,
+      "loss": 0.1239,
+      "step": 1731
+    },
+    {
+      "epoch": 1.6444338950866366,
+      "grad_norm": 0.026264041662216187,
+      "learning_rate": 0.00010060776128109812,
+      "loss": 0.12,
+      "step": 1732
+    },
+    {
+      "epoch": 1.6453833372893425,
+      "grad_norm": 0.02740940824151039,
+      "learning_rate": 0.00010049726024198578,
+      "loss": 0.1314,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6463327794920484,
+      "grad_norm": 0.04096614569425583,
+      "learning_rate": 0.00010038675859567628,
+      "loss": 0.1681,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6472822216947542,
+      "grad_norm": 0.04552573338150978,
+      "learning_rate": 0.00010027625647710155,
+      "loss": 0.16,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6482316638974601,
+      "grad_norm": 0.034032173454761505,
+      "learning_rate": 0.00010016575402119413,
+      "loss": 0.1326,
+      "step": 1736
+    },
+    {
+      "epoch": 1.649181106100166,
+      "grad_norm": 0.03644052520394325,
+      "learning_rate": 0.00010005525136288692,
+      "loss": 0.146,
+      "step": 1737
+    },
+    {
+      "epoch": 1.650130548302872,
+      "grad_norm": 0.04277161881327629,
+      "learning_rate": 9.994474863711311e-05,
+      "loss": 0.1719,
+      "step": 1738
+    },
+    {
+      "epoch": 1.651079990505578,
+      "grad_norm": 0.027901561930775642,
+      "learning_rate": 9.98342459788059e-05,
+      "loss": 0.1238,
+      "step": 1739
+    },
+    {
+      "epoch": 1.6520294327082838,
+      "grad_norm": 0.030957000330090523,
+      "learning_rate": 9.972374352289848e-05,
+      "loss": 0.1315,
+      "step": 1740
+    },
+    {
+      "epoch": 1.65297887491099,
+      "grad_norm": 0.029299341142177582,
+      "learning_rate": 9.961324140432376e-05,
+      "loss": 0.1247,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6539283171136958,
+      "grad_norm": 0.0292718093842268,
+      "learning_rate": 9.950273975801424e-05,
+      "loss": 0.1296,
+      "step": 1742
+    },
+    {
+      "epoch": 1.6548777593164017,
+      "grad_norm": 0.03113977424800396,
+      "learning_rate": 9.93922387189019e-05,
+      "loss": 0.1294,
+      "step": 1743
+    },
+    {
+      "epoch": 1.6558272015191076,
+      "grad_norm": 0.05127384141087532,
+      "learning_rate": 9.928173842191786e-05,
+      "loss": 0.1623,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6567766437218134,
+      "grad_norm": 0.03058856725692749,
+      "learning_rate": 9.917123900199245e-05,
+      "loss": 0.1251,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6577260859245193,
+      "grad_norm": 0.043525367975234985,
+      "learning_rate": 9.906074059405486e-05,
+      "loss": 0.1584,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6586755281272252,
+      "grad_norm": 0.02724611759185791,
+      "learning_rate": 9.895024333303305e-05,
+      "loss": 0.1273,
+      "step": 1747
+    },
+    {
+      "epoch": 1.659624970329931,
+      "grad_norm": 0.026182804256677628,
+      "learning_rate": 9.883974735385361e-05,
+      "loss": 0.1165,
+      "step": 1748
+    },
+    {
+      "epoch": 1.660574412532637,
+      "grad_norm": 0.030495688319206238,
+      "learning_rate": 9.87292527914416e-05,
+      "loss": 0.1251,
+      "step": 1749
+    },
+    {
+      "epoch": 1.661523854735343,
+      "grad_norm": 0.03013971447944641,
+      "learning_rate": 9.861875978072017e-05,
+      "loss": 0.1278,
+      "step": 1750
+    },
+    {
+      "epoch": 1.662473296938049,
+      "grad_norm": 0.03999912738800049,
+      "learning_rate": 9.850826845661082e-05,
+      "loss": 0.1519,
+      "step": 1751
+    },
+    {
+      "epoch": 1.6634227391407548,
+      "grad_norm": 0.029559755697846413,
+      "learning_rate": 9.839777895403287e-05,
+      "loss": 0.1293,
+      "step": 1752
+    },
+    {
+      "epoch": 1.6643721813434609,
+      "grad_norm": 0.04213762283325195,
+      "learning_rate": 9.828729140790337e-05,
+      "loss": 0.1696,
+      "step": 1753
+    },
+    {
+      "epoch": 1.6653216235461668,
+      "grad_norm": 0.029974251985549927,
+      "learning_rate": 9.817680595313705e-05,
+      "loss": 0.1182,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6662710657488726,
+      "grad_norm": 0.03835977986454964,
+      "learning_rate": 9.806632272464607e-05,
+      "loss": 0.1467,
+      "step": 1755
+    },
+    {
+      "epoch": 1.6672205079515785,
+      "grad_norm": 0.031473349779844284,
+      "learning_rate": 9.795584185733988e-05,
+      "loss": 0.1305,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6681699501542844,
+      "grad_norm": 0.02675897814333439,
+      "learning_rate": 9.784536348612504e-05,
+      "loss": 0.1196,
+      "step": 1757
+    },
+    {
+      "epoch": 1.6691193923569903,
+      "grad_norm": 0.0419435016810894,
+      "learning_rate": 9.773488774590504e-05,
+      "loss": 0.1558,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6700688345596961,
+      "grad_norm": 0.027311773970723152,
+      "learning_rate": 9.762441477158016e-05,
+      "loss": 0.1236,
+      "step": 1759
+    },
+    {
+      "epoch": 1.671018276762402,
+      "grad_norm": 0.05605548992753029,
+      "learning_rate": 9.751394469804734e-05,
+      "loss": 0.1787,
+      "step": 1760
+    },
+    {
+      "epoch": 1.6719677189651079,
+      "grad_norm": 0.025175364688038826,
+      "learning_rate": 9.740347766019997e-05,
+      "loss": 0.1275,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6729171611678137,
+      "grad_norm": 0.04951293021440506,
+      "learning_rate": 9.729301379292773e-05,
+      "loss": 0.2023,
+      "step": 1762
+    },
+    {
+      "epoch": 1.6738666033705198,
+      "grad_norm": 0.02842806465923786,
+      "learning_rate": 9.718255323111635e-05,
+      "loss": 0.1238,
+      "step": 1763
+    },
+    {
+      "epoch": 1.6748160455732257,
+      "grad_norm": 0.029241712763905525,
+      "learning_rate": 9.707209610964765e-05,
+      "loss": 0.121,
+      "step": 1764
+    },
+    {
+      "epoch": 1.6757654877759316,
+      "grad_norm": 0.03337705507874489,
+      "learning_rate": 9.696164256339917e-05,
+      "loss": 0.1354,
+      "step": 1765
+    },
+    {
+      "epoch": 1.6767149299786377,
+      "grad_norm": 0.030520187690854073,
+      "learning_rate": 9.685119272724411e-05,
+      "loss": 0.1256,
+      "step": 1766
+    },
+    {
+      "epoch": 1.6776643721813436,
+      "grad_norm": 0.0318867489695549,
+      "learning_rate": 9.674074673605115e-05,
+      "loss": 0.1286,
+      "step": 1767
+    },
+    {
+      "epoch": 1.6786138143840494,
+      "grad_norm": 0.02671106904745102,
+      "learning_rate": 9.663030472468424e-05,
+      "loss": 0.1297,
+      "step": 1768
+    },
+    {
+      "epoch": 1.6795632565867553,
+      "grad_norm": 0.041651055216789246,
+      "learning_rate": 9.651986682800249e-05,
+      "loss": 0.1618,
+      "step": 1769
+    },
+    {
+      "epoch": 1.6805126987894612,
+      "grad_norm": 0.03195889666676521,
+      "learning_rate": 9.640943318085999e-05,
+      "loss": 0.1279,
+      "step": 1770
+    },
+    {
+      "epoch": 1.681462140992167,
+      "grad_norm": 0.04086165875196457,
+      "learning_rate": 9.629900391810563e-05,
+      "loss": 0.1678,
+      "step": 1771
+    },
+    {
+      "epoch": 1.682411583194873,
+      "grad_norm": 0.025256391614675522,
+      "learning_rate": 9.618857917458298e-05,
+      "loss": 0.1197,
+      "step": 1772
+    },
+    {
+      "epoch": 1.6833610253975788,
+      "grad_norm": 0.03352576494216919,
+      "learning_rate": 9.607815908513005e-05,
+      "loss": 0.1345,
+      "step": 1773
+    },
+    {
+      "epoch": 1.6843104676002847,
+      "grad_norm": 0.06082432344555855,
+      "learning_rate": 9.596774378457916e-05,
+      "loss": 0.1639,
+      "step": 1774
+    },
+    {
+      "epoch": 1.6852599098029908,
+      "grad_norm": 0.029191186651587486,
+      "learning_rate": 9.585733340775677e-05,
+      "loss": 0.1305,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6862093520056967,
+      "grad_norm": 0.029343895614147186,
+      "learning_rate": 9.574692808948348e-05,
+      "loss": 0.1265,
+      "step": 1776
+    },
+    {
+      "epoch": 1.6871587942084025,
+      "grad_norm": 0.02953837811946869,
+      "learning_rate": 9.56365279645735e-05,
+      "loss": 0.1281,
+      "step": 1777
+    },
+    {
+      "epoch": 1.6881082364111086,
+      "grad_norm": 0.028798846527934074,
+      "learning_rate": 9.552613316783483e-05,
+      "loss": 0.1257,
+      "step": 1778
+    },
+    {
+      "epoch": 1.6890576786138145,
+      "grad_norm": 0.02905990555882454,
+      "learning_rate": 9.54157438340689e-05,
+      "loss": 0.1308,
+      "step": 1779
+    },
+    {
+      "epoch": 1.6900071208165204,
+      "grad_norm": 0.02965502068400383,
+      "learning_rate": 9.530536009807053e-05,
+      "loss": 0.1296,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6909565630192263,
+      "grad_norm": 0.029197504743933678,
+      "learning_rate": 9.519498209462766e-05,
+      "loss": 0.1204,
+      "step": 1781
+    },
+    {
+      "epoch": 1.6919060052219321,
+      "grad_norm": 0.04930657148361206,
+      "learning_rate": 9.508460995852122e-05,
+      "loss": 0.1522,
+      "step": 1782
+    },
+    {
+      "epoch": 1.692855447424638,
+      "grad_norm": 0.027312377467751503,
+      "learning_rate": 9.497424382452501e-05,
+      "loss": 0.1203,
+      "step": 1783
+    },
+    {
+      "epoch": 1.6938048896273439,
+      "grad_norm": 0.03260885551571846,
+      "learning_rate": 9.486388382740548e-05,
+      "loss": 0.1334,
+      "step": 1784
+    },
+    {
+      "epoch": 1.6947543318300498,
+      "grad_norm": 0.052055153995752335,
+      "learning_rate": 9.475353010192162e-05,
+      "loss": 0.2113,
+      "step": 1785
+    },
+    {
+      "epoch": 1.6957037740327556,
+      "grad_norm": 0.038476429879665375,
+      "learning_rate": 9.464318278282472e-05,
+      "loss": 0.1669,
+      "step": 1786
+    },
+    {
+      "epoch": 1.6966532162354615,
+      "grad_norm": 0.03111964277923107,
+      "learning_rate": 9.453284200485825e-05,
+      "loss": 0.1255,
+      "step": 1787
+    },
+    {
+      "epoch": 1.6976026584381676,
+      "grad_norm": 0.02803085930645466,
+      "learning_rate": 9.44225079027577e-05,
+      "loss": 0.1297,
+      "step": 1788
+    },
+    {
+      "epoch": 1.6985521006408735,
+      "grad_norm": 0.029160544276237488,
+      "learning_rate": 9.431218061125044e-05,
+      "loss": 0.1263,
+      "step": 1789
+    },
+    {
+      "epoch": 1.6995015428435793,
+      "grad_norm": 0.03121958300471306,
+      "learning_rate": 9.420186026505548e-05,
+      "loss": 0.1277,
+      "step": 1790
+    },
+    {
+      "epoch": 1.7004509850462854,
+      "grad_norm": 0.027002684772014618,
+      "learning_rate": 9.40915469988834e-05,
+      "loss": 0.1236,
+      "step": 1791
+    },
+    {
+      "epoch": 1.7014004272489913,
+      "grad_norm": 0.027172109112143517,
+      "learning_rate": 9.398124094743604e-05,
+      "loss": 0.1241,
+      "step": 1792
+    },
+    {
+      "epoch": 1.7023498694516972,
+      "grad_norm": 0.02658463642001152,
+      "learning_rate": 9.387094224540653e-05,
+      "loss": 0.1268,
+      "step": 1793
+    },
+    {
+      "epoch": 1.703299311654403,
+      "grad_norm": 0.028862452134490013,
+      "learning_rate": 9.376065102747898e-05,
+      "loss": 0.1317,
+      "step": 1794
+    },
+    {
+      "epoch": 1.704248753857109,
+      "grad_norm": 0.025097506120800972,
+      "learning_rate": 9.365036742832838e-05,
+      "loss": 0.1164,
+      "step": 1795
+    },
+    {
+      "epoch": 1.7051981960598148,
+      "grad_norm": 0.04557095095515251,
+      "learning_rate": 9.354009158262038e-05,
+      "loss": 0.1522,
+      "step": 1796
+    },
+    {
+      "epoch": 1.7061476382625207,
+      "grad_norm": 0.03487172722816467,
+      "learning_rate": 9.342982362501123e-05,
+      "loss": 0.1398,
+      "step": 1797
+    },
+    {
+      "epoch": 1.7070970804652266,
+      "grad_norm": 0.02509194053709507,
+      "learning_rate": 9.331956369014746e-05,
+      "loss": 0.1166,
+      "step": 1798
+    },
+    {
+      "epoch": 1.7080465226679324,
+      "grad_norm": 0.030363403260707855,
+      "learning_rate": 9.320931191266587e-05,
+      "loss": 0.1191,
+      "step": 1799
+    },
+    {
+      "epoch": 1.7089959648706385,
+      "grad_norm": 0.08008571714162827,
+      "learning_rate": 9.309906842719323e-05,
+      "loss": 0.1494,
+      "step": 1800
+    },
+    {
+      "epoch": 1.7099454070733444,
+      "grad_norm": 0.027857676148414612,
+      "learning_rate": 9.298883336834633e-05,
+      "loss": 0.1307,
+      "step": 1801
+    },
+    {
+      "epoch": 1.7108948492760503,
+      "grad_norm": 0.02744341269135475,
+      "learning_rate": 9.28786068707315e-05,
+      "loss": 0.1217,
+      "step": 1802
+    },
+    {
+      "epoch": 1.7118442914787564,
+      "grad_norm": 0.0324774868786335,
+      "learning_rate": 9.276838906894472e-05,
+      "loss": 0.1311,
+      "step": 1803
+    },
+    {
+      "epoch": 1.7127937336814623,
+      "grad_norm": 0.028761887922883034,
+      "learning_rate": 9.265818009757132e-05,
+      "loss": 0.1275,
+      "step": 1804
+    },
+    {
+      "epoch": 1.7137431758841681,
+      "grad_norm": 0.02950756810605526,
+      "learning_rate": 9.254798009118584e-05,
+      "loss": 0.1262,
+      "step": 1805
+    },
+    {
+      "epoch": 1.714692618086874,
+      "grad_norm": 0.027881214395165443,
+      "learning_rate": 9.243778918435187e-05,
+      "loss": 0.1266,
+      "step": 1806
+    },
+    {
+      "epoch": 1.7156420602895799,
+      "grad_norm": 0.05155957117676735,
+      "learning_rate": 9.232760751162193e-05,
+      "loss": 0.1936,
+      "step": 1807
+    },
+    {
+      "epoch": 1.7165915024922858,
+      "grad_norm": 0.029041916131973267,
+      "learning_rate": 9.221743520753719e-05,
+      "loss": 0.1204,
+      "step": 1808
+    },
+    {
+      "epoch": 1.7175409446949916,
+      "grad_norm": 0.030144108459353447,
+      "learning_rate": 9.210727240662747e-05,
+      "loss": 0.1285,
+      "step": 1809
+    },
+    {
+      "epoch": 1.7184903868976975,
+      "grad_norm": 0.028103960677981377,
+      "learning_rate": 9.199711924341093e-05,
+      "loss": 0.125,
+      "step": 1810
+    },
+    {
+      "epoch": 1.7194398291004034,
+      "grad_norm": 0.03844073414802551,
+      "learning_rate": 9.188697585239394e-05,
+      "loss": 0.1525,
+      "step": 1811
+    },
+    {
+      "epoch": 1.7203892713031093,
+      "grad_norm": 0.04454744979739189,
+      "learning_rate": 9.177684236807099e-05,
+      "loss": 0.1616,
+      "step": 1812
+    },
+    {
+      "epoch": 1.7213387135058154,
+      "grad_norm": 0.027989163994789124,
+      "learning_rate": 9.166671892492446e-05,
+      "loss": 0.1265,
+      "step": 1813
+    },
+    {
+      "epoch": 1.7222881557085212,
+      "grad_norm": 0.04422546178102493,
+      "learning_rate": 9.155660565742444e-05,
+      "loss": 0.159,
+      "step": 1814
+    },
+    {
+      "epoch": 1.723237597911227,
+      "grad_norm": 0.027917252853512764,
+      "learning_rate": 9.144650270002866e-05,
+      "loss": 0.1229,
+      "step": 1815
+    },
+    {
+      "epoch": 1.7241870401139332,
+      "grad_norm": 0.05252804979681969,
+      "learning_rate": 9.133641018718217e-05,
+      "loss": 0.1955,
+      "step": 1816
+    },
+    {
+      "epoch": 1.725136482316639,
+      "grad_norm": 0.029228439554572105,
+      "learning_rate": 9.122632825331733e-05,
+      "loss": 0.1197,
+      "step": 1817
+    },
+    {
+      "epoch": 1.726085924519345,
+      "grad_norm": 0.02810599096119404,
+      "learning_rate": 9.111625703285356e-05,
+      "loss": 0.1284,
+      "step": 1818
+    },
+    {
+      "epoch": 1.7270353667220508,
+      "grad_norm": 0.02618074230849743,
+      "learning_rate": 9.10061966601972e-05,
+      "loss": 0.1239,
+      "step": 1819
+    },
+    {
+      "epoch": 1.7279848089247567,
+      "grad_norm": 0.026649268344044685,
+      "learning_rate": 9.089614726974137e-05,
+      "loss": 0.1218,
+      "step": 1820
+    },
+    {
+      "epoch": 1.7289342511274626,
+      "grad_norm": 0.02857782505452633,
+      "learning_rate": 9.078610899586575e-05,
+      "loss": 0.133,
+      "step": 1821
+    },
+    {
+      "epoch": 1.7298836933301684,
+      "grad_norm": 0.026309454813599586,
+      "learning_rate": 9.067608197293642e-05,
+      "loss": 0.1175,
+      "step": 1822
+    },
+    {
+      "epoch": 1.7308331355328743,
+      "grad_norm": 0.02791914902627468,
+      "learning_rate": 9.056606633530578e-05,
+      "loss": 0.12,
+      "step": 1823
+    },
+    {
+      "epoch": 1.7317825777355802,
+      "grad_norm": 0.030874181538820267,
+      "learning_rate": 9.045606221731229e-05,
+      "loss": 0.1307,
+      "step": 1824
+    },
+    {
+      "epoch": 1.7327320199382863,
+      "grad_norm": 0.030806539580225945,
+      "learning_rate": 9.034606975328033e-05,
+      "loss": 0.1188,
+      "step": 1825
+    },
+    {
+      "epoch": 1.7336814621409922,
+      "grad_norm": 0.028665419667959213,
+      "learning_rate": 9.023608907752015e-05,
+      "loss": 0.1311,
+      "step": 1826
+    },
+    {
+      "epoch": 1.734630904343698,
+      "grad_norm": 0.029301505535840988,
+      "learning_rate": 9.012612032432747e-05,
+      "loss": 0.1325,
+      "step": 1827
+    },
+    {
+      "epoch": 1.7355803465464041,
+      "grad_norm": 0.0394410640001297,
+      "learning_rate": 9.001616362798353e-05,
+      "loss": 0.164,
+      "step": 1828
+    },
+    {
+      "epoch": 1.73652978874911,
+      "grad_norm": 0.02754109725356102,
+      "learning_rate": 8.990621912275484e-05,
+      "loss": 0.12,
+      "step": 1829
+    },
+    {
+      "epoch": 1.737479230951816,
+      "grad_norm": 0.02719545178115368,
+      "learning_rate": 8.9796286942893e-05,
+      "loss": 0.1203,
+      "step": 1830
+    },
+    {
+      "epoch": 1.7384286731545218,
+      "grad_norm": 0.02480783686041832,
+      "learning_rate": 8.968636722263455e-05,
+      "loss": 0.1225,
+      "step": 1831
+    },
+    {
+      "epoch": 1.7393781153572276,
+      "grad_norm": 0.025418803095817566,
+      "learning_rate": 8.957646009620085e-05,
+      "loss": 0.125,
+      "step": 1832
+    },
+    {
+      "epoch": 1.7403275575599335,
+      "grad_norm": 0.024165470153093338,
+      "learning_rate": 8.94665656977979e-05,
+      "loss": 0.125,
+      "step": 1833
+    },
+    {
+      "epoch": 1.7412769997626394,
+      "grad_norm": 0.03011813573539257,
+      "learning_rate": 8.935668416161612e-05,
+      "loss": 0.1337,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7422264419653453,
+      "grad_norm": 0.038413502275943756,
+      "learning_rate": 8.92468156218302e-05,
+      "loss": 0.1715,
+      "step": 1835
+    },
+    {
+      "epoch": 1.7431758841680511,
+      "grad_norm": 0.023849591612815857,
+      "learning_rate": 8.9136960212599e-05,
+      "loss": 0.1148,
+      "step": 1836
+    },
+    {
+      "epoch": 1.7441253263707572,
+      "grad_norm": 0.027159664779901505,
+      "learning_rate": 8.902711806806536e-05,
+      "loss": 0.1255,
+      "step": 1837
+    },
+    {
+      "epoch": 1.745074768573463,
+      "grad_norm": 0.030395383015275,
+      "learning_rate": 8.89172893223559e-05,
+      "loss": 0.1267,
+      "step": 1838
+    },
+    {
+      "epoch": 1.746024210776169,
+      "grad_norm": 0.025772254914045334,
+      "learning_rate": 8.880747410958085e-05,
+      "loss": 0.1212,
+      "step": 1839
+    },
+    {
+      "epoch": 1.7469736529788749,
+      "grad_norm": 0.04073212668299675,
+      "learning_rate": 8.8697672563834e-05,
+      "loss": 0.1229,
+      "step": 1840
+    },
+    {
+      "epoch": 1.747923095181581,
+      "grad_norm": 0.048602957278490067,
+      "learning_rate": 8.858788481919235e-05,
+      "loss": 0.1587,
+      "step": 1841
+    },
+    {
+      "epoch": 1.7488725373842868,
+      "grad_norm": 0.030672159045934677,
+      "learning_rate": 8.84781110097161e-05,
+      "loss": 0.1276,
+      "step": 1842
+    },
+    {
+      "epoch": 1.7498219795869927,
+      "grad_norm": 0.029867272824048996,
+      "learning_rate": 8.836835126944843e-05,
+      "loss": 0.1316,
+      "step": 1843
+    },
+    {
+      "epoch": 1.7507714217896986,
+      "grad_norm": 0.03122364915907383,
+      "learning_rate": 8.825860573241535e-05,
+      "loss": 0.1276,
+      "step": 1844
+    },
+    {
+      "epoch": 1.7517208639924045,
+      "grad_norm": 0.03530842810869217,
+      "learning_rate": 8.814887453262555e-05,
+      "loss": 0.1272,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7526703061951103,
+      "grad_norm": 0.028104711323976517,
+      "learning_rate": 8.803915780407009e-05,
+      "loss": 0.1277,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7536197483978162,
+      "grad_norm": 0.02434263750910759,
+      "learning_rate": 8.792945568072252e-05,
+      "loss": 0.1136,
+      "step": 1847
+    },
+    {
+      "epoch": 1.754569190600522,
+      "grad_norm": 0.027843188494443893,
+      "learning_rate": 8.781976829653846e-05,
+      "loss": 0.1199,
+      "step": 1848
+    },
+    {
+      "epoch": 1.755518632803228,
+      "grad_norm": 0.03688850998878479,
+      "learning_rate": 8.771009578545553e-05,
+      "loss": 0.1345,
+      "step": 1849
+    },
+    {
+      "epoch": 1.756468075005934,
+      "grad_norm": 0.027186516672372818,
+      "learning_rate": 8.760043828139325e-05,
+      "loss": 0.1149,
+      "step": 1850
+    },
+    {
+      "epoch": 1.75741751720864,
+      "grad_norm": 0.04049715772271156,
+      "learning_rate": 8.749079591825278e-05,
+      "loss": 0.1585,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7583669594113458,
+      "grad_norm": 0.02956775203347206,
+      "learning_rate": 8.738116882991679e-05,
+      "loss": 0.1303,
+      "step": 1852
+    },
+    {
+      "epoch": 1.759316401614052,
+      "grad_norm": 0.026160147041082382,
+      "learning_rate": 8.72715571502493e-05,
+      "loss": 0.1237,
+      "step": 1853
+    },
+    {
+      "epoch": 1.7602658438167578,
+      "grad_norm": 0.023719090968370438,
+      "learning_rate": 8.71619610130955e-05,
+      "loss": 0.1172,
+      "step": 1854
+    },
+    {
+      "epoch": 1.7612152860194636,
+      "grad_norm": 0.024884294718503952,
+      "learning_rate": 8.705238055228161e-05,
+      "loss": 0.123,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7621647282221695,
+      "grad_norm": 0.028241394087672234,
+      "learning_rate": 8.694281590161474e-05,
+      "loss": 0.129,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7631141704248754,
+      "grad_norm": 0.028791090473532677,
+      "learning_rate": 8.683326719488263e-05,
+      "loss": 0.121,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7640636126275813,
+      "grad_norm": 0.046369921416044235,
+      "learning_rate": 8.672373456585365e-05,
+      "loss": 0.1666,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7650130548302871,
+      "grad_norm": 0.025271739810705185,
+      "learning_rate": 8.661421814827641e-05,
+      "loss": 0.1196,
+      "step": 1859
+    },
+    {
+      "epoch": 1.765962497032993,
+      "grad_norm": 0.02569795772433281,
+      "learning_rate": 8.650471807587983e-05,
+      "loss": 0.1235,
+      "step": 1860
+    },
+    {
+      "epoch": 1.766911939235699,
+      "grad_norm": 0.03638843819499016,
+      "learning_rate": 8.639523448237282e-05,
+      "loss": 0.1523,
+      "step": 1861
+    },
+    {
+      "epoch": 1.767861381438405,
+      "grad_norm": 0.03260574862360954,
+      "learning_rate": 8.628576750144419e-05,
+      "loss": 0.1328,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7688108236411109,
+      "grad_norm": 0.02770201303064823,
+      "learning_rate": 8.617631726676243e-05,
+      "loss": 0.1256,
+      "step": 1863
+    },
+    {
+      "epoch": 1.7697602658438167,
+      "grad_norm": 0.02869422361254692,
+      "learning_rate": 8.606688391197564e-05,
+      "loss": 0.1261,
+      "step": 1864
+    },
+    {
+      "epoch": 1.7707097080465226,
+      "grad_norm": 0.02792002633213997,
+      "learning_rate": 8.595746757071125e-05,
+      "loss": 0.1277,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7716591502492287,
+      "grad_norm": 0.025662843137979507,
+      "learning_rate": 8.584806837657594e-05,
+      "loss": 0.1163,
+      "step": 1866
+    },
+    {
+      "epoch": 1.7726085924519346,
+      "grad_norm": 0.027771448716521263,
+      "learning_rate": 8.573868646315546e-05,
+      "loss": 0.1273,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7735580346546405,
+      "grad_norm": 0.026355689391493797,
+      "learning_rate": 8.562932196401444e-05,
+      "loss": 0.1241,
+      "step": 1868
+    },
+    {
+      "epoch": 1.7745074768573463,
+      "grad_norm": 0.028244782239198685,
+      "learning_rate": 8.551997501269629e-05,
+      "loss": 0.1319,
+      "step": 1869
+    },
+    {
+      "epoch": 1.7754569190600522,
+      "grad_norm": 0.027661755681037903,
+      "learning_rate": 8.541064574272292e-05,
+      "loss": 0.134,
+      "step": 1870
+    },
+    {
+      "epoch": 1.776406361262758,
+      "grad_norm": 0.026287924498319626,
+      "learning_rate": 8.530133428759468e-05,
+      "loss": 0.1215,
+      "step": 1871
+    },
+    {
+      "epoch": 1.777355803465464,
+      "grad_norm": 0.040449049323797226,
+      "learning_rate": 8.519204078079021e-05,
+      "loss": 0.179,
+      "step": 1872
+    },
+    {
+      "epoch": 1.7783052456681698,
+      "grad_norm": 0.022792836651206017,
+      "learning_rate": 8.508276535576619e-05,
+      "loss": 0.1208,
+      "step": 1873
+    },
+    {
+      "epoch": 1.7792546878708757,
+      "grad_norm": 0.042618922889232635,
+      "learning_rate": 8.497350814595721e-05,
+      "loss": 0.1666,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7802041300735818,
+      "grad_norm": 0.0521213673055172,
+      "learning_rate": 8.486426928477561e-05,
+      "loss": 0.1858,
+      "step": 1875
+    },
+    {
+      "epoch": 1.7811535722762877,
+      "grad_norm": 0.04682205617427826,
+      "learning_rate": 8.475504890561142e-05,
+      "loss": 0.2037,
+      "step": 1876
+    },
+    {
+      "epoch": 1.7821030144789936,
+      "grad_norm": 0.041265182197093964,
+      "learning_rate": 8.464584714183204e-05,
+      "loss": 0.1775,
+      "step": 1877
+    },
+    {
+      "epoch": 1.7830524566816996,
+      "grad_norm": 0.027602121233940125,
+      "learning_rate": 8.453666412678206e-05,
+      "loss": 0.1186,
+      "step": 1878
+    },
+    {
+      "epoch": 1.7840018988844055,
+      "grad_norm": 0.025885846465826035,
+      "learning_rate": 8.442749999378327e-05,
+      "loss": 0.1275,
+      "step": 1879
+    },
+    {
+      "epoch": 1.7849513410871114,
+      "grad_norm": 0.05111463740468025,
+      "learning_rate": 8.43183548761344e-05,
+      "loss": 0.1535,
+      "step": 1880
+    },
+    {
+      "epoch": 1.7859007832898173,
+      "grad_norm": 0.026447484269738197,
+      "learning_rate": 8.420922890711094e-05,
+      "loss": 0.1244,
+      "step": 1881
+    },
+    {
+      "epoch": 1.7868502254925231,
+      "grad_norm": 0.046114444732666016,
+      "learning_rate": 8.410012221996502e-05,
+      "loss": 0.1549,
+      "step": 1882
+    },
+    {
+      "epoch": 1.787799667695229,
+      "grad_norm": 0.027883663773536682,
+      "learning_rate": 8.399103494792514e-05,
+      "loss": 0.1186,
+      "step": 1883
+    },
+    {
+      "epoch": 1.788749109897935,
+      "grad_norm": 0.02667239122092724,
+      "learning_rate": 8.388196722419621e-05,
+      "loss": 0.1367,
+      "step": 1884
+    },
+    {
+      "epoch": 1.7896985521006408,
+      "grad_norm": 0.028317047283053398,
+      "learning_rate": 8.377291918195922e-05,
+      "loss": 0.1293,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7906479943033466,
+      "grad_norm": 0.044224657118320465,
+      "learning_rate": 8.36638909543711e-05,
+      "loss": 0.1722,
+      "step": 1886
+    },
+    {
+      "epoch": 1.7915974365060527,
+      "grad_norm": 0.029220551252365112,
+      "learning_rate": 8.35548826745646e-05,
+      "loss": 0.1315,
+      "step": 1887
+    },
+    {
+      "epoch": 1.7925468787087586,
+      "grad_norm": 0.028357302770018578,
+      "learning_rate": 8.344589447564818e-05,
+      "loss": 0.1271,
+      "step": 1888
+    },
+    {
+      "epoch": 1.7934963209114645,
+      "grad_norm": 0.027882913127541542,
+      "learning_rate": 8.333692649070568e-05,
+      "loss": 0.1311,
+      "step": 1889
+    },
+    {
+      "epoch": 1.7944457631141706,
+      "grad_norm": 0.029344897717237473,
+      "learning_rate": 8.322797885279627e-05,
+      "loss": 0.1231,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7953952053168765,
+      "grad_norm": 0.039409589022397995,
+      "learning_rate": 8.311905169495435e-05,
+      "loss": 0.1651,
+      "step": 1891
+    },
+    {
+      "epoch": 1.7963446475195823,
+      "grad_norm": 0.025551313534379005,
+      "learning_rate": 8.301014515018925e-05,
+      "loss": 0.1162,
+      "step": 1892
+    },
+    {
+      "epoch": 1.7972940897222882,
+      "grad_norm": 0.027775781229138374,
+      "learning_rate": 8.290125935148516e-05,
+      "loss": 0.1254,
+      "step": 1893
+    },
+    {
+      "epoch": 1.798243531924994,
+      "grad_norm": 0.025555282831192017,
+      "learning_rate": 8.279239443180088e-05,
+      "loss": 0.1173,
+      "step": 1894
+    },
+    {
+      "epoch": 1.7991929741277,
+      "grad_norm": 0.027120131999254227,
+      "learning_rate": 8.268355052406978e-05,
+      "loss": 0.123,
+      "step": 1895
+    },
+    {
+      "epoch": 1.8001424163304058,
+      "grad_norm": 0.027624819427728653,
+      "learning_rate": 8.257472776119957e-05,
+      "loss": 0.1313,
+      "step": 1896
+    },
+    {
+      "epoch": 1.8010918585331117,
+      "grad_norm": 0.04341182857751846,
+      "learning_rate": 8.246592627607208e-05,
+      "loss": 0.1695,
+      "step": 1897
+    },
+    {
+      "epoch": 1.8020413007358176,
+      "grad_norm": 0.049969110637903214,
+      "learning_rate": 8.235714620154323e-05,
+      "loss": 0.1609,
+      "step": 1898
+    },
+    {
+      "epoch": 1.8029907429385235,
+      "grad_norm": 0.033960528671741486,
+      "learning_rate": 8.224838767044275e-05,
+      "loss": 0.1299,
+      "step": 1899
+    },
+    {
+      "epoch": 1.8039401851412296,
+      "grad_norm": 0.045656926929950714,
+      "learning_rate": 8.213965081557402e-05,
+      "loss": 0.1613,
+      "step": 1900
+    },
+    {
+      "epoch": 1.8048896273439354,
+      "grad_norm": 0.028559250757098198,
+      "learning_rate": 8.203093576971414e-05,
+      "loss": 0.1222,
+      "step": 1901
+    },
+    {
+      "epoch": 1.8058390695466413,
+      "grad_norm": 0.026110464707016945,
+      "learning_rate": 8.192224266561336e-05,
+      "loss": 0.1204,
+      "step": 1902
+    },
+    {
+      "epoch": 1.8067885117493474,
+      "grad_norm": 0.023995952680706978,
+      "learning_rate": 8.181357163599522e-05,
+      "loss": 0.1164,
+      "step": 1903
+    },
+    {
+      "epoch": 1.8077379539520533,
+      "grad_norm": 0.054921507835388184,
+      "learning_rate": 8.170492281355635e-05,
+      "loss": 0.1535,
+      "step": 1904
+    },
+    {
+      "epoch": 1.8086873961547592,
+      "grad_norm": 0.03065885417163372,
+      "learning_rate": 8.159629633096619e-05,
+      "loss": 0.1325,
+      "step": 1905
+    },
+    {
+      "epoch": 1.809636838357465,
+      "grad_norm": 0.02722746506333351,
+      "learning_rate": 8.148769232086698e-05,
+      "loss": 0.1201,
+      "step": 1906
+    },
+    {
+      "epoch": 1.810586280560171,
+      "grad_norm": 0.026619885116815567,
+      "learning_rate": 8.13791109158734e-05,
+      "loss": 0.125,
+      "step": 1907
+    },
+    {
+      "epoch": 1.8115357227628768,
+      "grad_norm": 0.02867223508656025,
+      "learning_rate": 8.127055224857266e-05,
+      "loss": 0.1279,
+      "step": 1908
+    },
+    {
+      "epoch": 1.8124851649655827,
+      "grad_norm": 0.029073316603899002,
+      "learning_rate": 8.116201645152412e-05,
+      "loss": 0.1195,
+      "step": 1909
+    },
+    {
+      "epoch": 1.8134346071682885,
+      "grad_norm": 0.04955434426665306,
+      "learning_rate": 8.105350365725926e-05,
+      "loss": 0.1657,
+      "step": 1910
+    },
+    {
+      "epoch": 1.8143840493709944,
+      "grad_norm": 0.03038748912513256,
+      "learning_rate": 8.094501399828143e-05,
+      "loss": 0.1244,
+      "step": 1911
+    },
+    {
+      "epoch": 1.8153334915737005,
+      "grad_norm": 0.08984460681676865,
+      "learning_rate": 8.08365476070658e-05,
+      "loss": 0.1602,
+      "step": 1912
+    },
+    {
+      "epoch": 1.8162829337764064,
+      "grad_norm": 0.028402511030435562,
+      "learning_rate": 8.0728104616059e-05,
+      "loss": 0.1162,
+      "step": 1913
+    },
+    {
+      "epoch": 1.8172323759791122,
+      "grad_norm": 0.03234838321805,
+      "learning_rate": 8.061968515767922e-05,
+      "loss": 0.1271,
+      "step": 1914
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 0.031242968514561653,
+      "learning_rate": 8.051128936431584e-05,
+      "loss": 0.1266,
+      "step": 1915
+    },
+    {
+      "epoch": 1.8191312603845242,
+      "grad_norm": 0.030700111761689186,
+      "learning_rate": 8.040291736832937e-05,
+      "loss": 0.1294,
+      "step": 1916
+    },
+    {
+      "epoch": 1.82008070258723,
+      "grad_norm": 0.031359221786260605,
+      "learning_rate": 8.029456930205128e-05,
+      "loss": 0.1316,
+      "step": 1917
+    },
+    {
+      "epoch": 1.821030144789936,
+      "grad_norm": 0.03048795275390148,
+      "learning_rate": 8.018624529778375e-05,
+      "loss": 0.127,
+      "step": 1918
+    },
+    {
+      "epoch": 1.8219795869926418,
+      "grad_norm": 0.03642559424042702,
+      "learning_rate": 8.007794548779964e-05,
+      "loss": 0.1577,
+      "step": 1919
+    },
+    {
+      "epoch": 1.8229290291953477,
+      "grad_norm": 0.029665078967809677,
+      "learning_rate": 7.996967000434224e-05,
+      "loss": 0.1266,
+      "step": 1920
+    },
+    {
+      "epoch": 1.8238784713980536,
+      "grad_norm": 0.027418775483965874,
+      "learning_rate": 7.986141897962518e-05,
+      "loss": 0.1319,
+      "step": 1921
+    },
+    {
+      "epoch": 1.8248279136007595,
+      "grad_norm": 0.03040868602693081,
+      "learning_rate": 7.975319254583216e-05,
+      "loss": 0.1263,
+      "step": 1922
+    },
+    {
+      "epoch": 1.8257773558034653,
+      "grad_norm": 0.030304348096251488,
+      "learning_rate": 7.96449908351169e-05,
+      "loss": 0.1357,
+      "step": 1923
+    },
+    {
+      "epoch": 1.8267267980061712,
+      "grad_norm": 0.029055261984467506,
+      "learning_rate": 7.953681397960287e-05,
+      "loss": 0.1293,
+      "step": 1924
+    },
+    {
+      "epoch": 1.8276762402088773,
+      "grad_norm": 0.030226033180952072,
+      "learning_rate": 7.942866211138324e-05,
+      "loss": 0.1335,
+      "step": 1925
+    },
+    {
+      "epoch": 1.8286256824115832,
+      "grad_norm": 0.02857894077897072,
+      "learning_rate": 7.93205353625207e-05,
+      "loss": 0.1307,
+      "step": 1926
+    },
+    {
+      "epoch": 1.829575124614289,
+      "grad_norm": 0.031932733952999115,
+      "learning_rate": 7.921243386504723e-05,
+      "loss": 0.1345,
+      "step": 1927
+    },
+    {
+      "epoch": 1.8305245668169952,
+      "grad_norm": 0.030729882419109344,
+      "learning_rate": 7.910435775096394e-05,
+      "loss": 0.1256,
+      "step": 1928
+    },
+    {
+      "epoch": 1.831474009019701,
+      "grad_norm": 0.056101903319358826,
+      "learning_rate": 7.899630715224098e-05,
+      "loss": 0.1858,
+      "step": 1929
+    },
+    {
+      "epoch": 1.832423451222407,
+      "grad_norm": 0.02785342186689377,
+      "learning_rate": 7.888828220081738e-05,
+      "loss": 0.1244,
+      "step": 1930
+    },
+    {
+      "epoch": 1.8333728934251128,
+      "grad_norm": 0.031731851398944855,
+      "learning_rate": 7.878028302860076e-05,
+      "loss": 0.1339,
+      "step": 1931
+    },
+    {
+      "epoch": 1.8343223356278187,
+      "grad_norm": 0.02918057143688202,
+      "learning_rate": 7.867230976746733e-05,
+      "loss": 0.1309,
+      "step": 1932
+    },
+    {
+      "epoch": 1.8352717778305245,
+      "grad_norm": 0.029841450974345207,
+      "learning_rate": 7.856436254926165e-05,
+      "loss": 0.1244,
+      "step": 1933
+    },
+    {
+      "epoch": 1.8362212200332304,
+      "grad_norm": 0.02823001891374588,
+      "learning_rate": 7.845644150579649e-05,
+      "loss": 0.1253,
+      "step": 1934
+    },
+    {
+      "epoch": 1.8371706622359363,
+      "grad_norm": 0.027422424405813217,
+      "learning_rate": 7.834854676885262e-05,
+      "loss": 0.1141,
+      "step": 1935
+    },
+    {
+      "epoch": 1.8381201044386422,
+      "grad_norm": 0.02671034075319767,
+      "learning_rate": 7.824067847017876e-05,
+      "loss": 0.1204,
+      "step": 1936
+    },
+    {
+      "epoch": 1.8390695466413483,
+      "grad_norm": 0.02999771386384964,
+      "learning_rate": 7.813283674149123e-05,
+      "loss": 0.1258,
+      "step": 1937
+    },
+    {
+      "epoch": 1.8400189888440541,
+      "grad_norm": 0.026322634890675545,
+      "learning_rate": 7.8025021714474e-05,
+      "loss": 0.117,
+      "step": 1938
+    },
+    {
+      "epoch": 1.84096843104676,
+      "grad_norm": 0.02665727399289608,
+      "learning_rate": 7.791723352077842e-05,
+      "loss": 0.1168,
+      "step": 1939
+    },
+    {
+      "epoch": 1.841917873249466,
+      "grad_norm": 0.043415650725364685,
+      "learning_rate": 7.780947229202305e-05,
+      "loss": 0.1666,
+      "step": 1940
+    },
+    {
+      "epoch": 1.842867315452172,
+      "grad_norm": 0.03161248564720154,
+      "learning_rate": 7.770173815979356e-05,
+      "loss": 0.1307,
+      "step": 1941
+    },
+    {
+      "epoch": 1.8438167576548778,
+      "grad_norm": 0.028511585667729378,
+      "learning_rate": 7.759403125564246e-05,
+      "loss": 0.1329,
+      "step": 1942
+    },
+    {
+      "epoch": 1.8447661998575837,
+      "grad_norm": 0.03340164199471474,
+      "learning_rate": 7.74863517110891e-05,
+      "loss": 0.1233,
+      "step": 1943
+    },
+    {
+      "epoch": 1.8457156420602896,
+      "grad_norm": 0.027353493496775627,
+      "learning_rate": 7.737869965761937e-05,
+      "loss": 0.1227,
+      "step": 1944
+    },
+    {
+      "epoch": 1.8466650842629955,
+      "grad_norm": 0.04435974359512329,
+      "learning_rate": 7.72710752266856e-05,
+      "loss": 0.1586,
+      "step": 1945
+    },
+    {
+      "epoch": 1.8476145264657013,
+      "grad_norm": 0.03443425893783569,
+      "learning_rate": 7.716347854970642e-05,
+      "loss": 0.1285,
+      "step": 1946
+    },
+    {
+      "epoch": 1.8485639686684072,
+      "grad_norm": 0.02941983938217163,
+      "learning_rate": 7.705590975806652e-05,
+      "loss": 0.1227,
+      "step": 1947
+    },
+    {
+      "epoch": 1.849513410871113,
+      "grad_norm": 0.031541094183921814,
+      "learning_rate": 7.694836898311654e-05,
+      "loss": 0.1307,
+      "step": 1948
+    },
+    {
+      "epoch": 1.850462853073819,
+      "grad_norm": 0.030199352651834488,
+      "learning_rate": 7.684085635617297e-05,
+      "loss": 0.126,
+      "step": 1949
+    },
+    {
+      "epoch": 1.851412295276525,
+      "grad_norm": 0.024474412202835083,
+      "learning_rate": 7.673337200851787e-05,
+      "loss": 0.1222,
+      "step": 1950
+    },
+    {
+      "epoch": 1.852361737479231,
+      "grad_norm": 0.02779853343963623,
+      "learning_rate": 7.662591607139882e-05,
+      "loss": 0.1242,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8533111796819368,
+      "grad_norm": 0.036591142416000366,
+      "learning_rate": 7.651848867602867e-05,
+      "loss": 0.1593,
+      "step": 1952
+    },
+    {
+      "epoch": 1.854260621884643,
+      "grad_norm": 0.029311561957001686,
+      "learning_rate": 7.641108995358542e-05,
+      "loss": 0.1237,
+      "step": 1953
+    },
+    {
+      "epoch": 1.8552100640873488,
+      "grad_norm": 0.029321955516934395,
+      "learning_rate": 7.630372003521204e-05,
+      "loss": 0.1289,
+      "step": 1954
+    },
+    {
+      "epoch": 1.8561595062900547,
+      "grad_norm": 0.034341856837272644,
+      "learning_rate": 7.619637905201642e-05,
+      "loss": 0.1323,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8571089484927605,
+      "grad_norm": 0.02771337330341339,
+      "learning_rate": 7.608906713507098e-05,
+      "loss": 0.133,
+      "step": 1956
+    },
+    {
+      "epoch": 1.8580583906954664,
+      "grad_norm": 0.03146693855524063,
+      "learning_rate": 7.598178441541274e-05,
+      "loss": 0.1346,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8590078328981723,
+      "grad_norm": 0.028372354805469513,
+      "learning_rate": 7.587453102404306e-05,
+      "loss": 0.1194,
+      "step": 1958
+    },
+    {
+      "epoch": 1.8599572751008782,
+      "grad_norm": 0.027766333892941475,
+      "learning_rate": 7.576730709192744e-05,
+      "loss": 0.1241,
+      "step": 1959
+    },
+    {
+      "epoch": 1.860906717303584,
+      "grad_norm": 0.026262789964675903,
+      "learning_rate": 7.566011274999549e-05,
+      "loss": 0.1191,
+      "step": 1960
+    },
+    {
+      "epoch": 1.86185615950629,
+      "grad_norm": 0.028471313416957855,
+      "learning_rate": 7.555294812914061e-05,
+      "loss": 0.1208,
+      "step": 1961
+    },
+    {
+      "epoch": 1.862805601708996,
+      "grad_norm": 0.0403280183672905,
+      "learning_rate": 7.544581336021994e-05,
+      "loss": 0.169,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8637550439117019,
+      "grad_norm": 0.029336489737033844,
+      "learning_rate": 7.533870857405414e-05,
+      "loss": 0.1275,
+      "step": 1963
+    },
+    {
+      "epoch": 1.8647044861144078,
+      "grad_norm": 0.05861514061689377,
+      "learning_rate": 7.523163390142732e-05,
+      "loss": 0.1984,
+      "step": 1964
+    },
+    {
+      "epoch": 1.8656539283171139,
+      "grad_norm": 0.026019204407930374,
+      "learning_rate": 7.51245894730868e-05,
+      "loss": 0.1194,
+      "step": 1965
+    },
+    {
+      "epoch": 1.8666033705198197,
+      "grad_norm": 0.043394673615694046,
+      "learning_rate": 7.501757541974289e-05,
+      "loss": 0.1598,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8675528127225256,
+      "grad_norm": 0.025404971092939377,
+      "learning_rate": 7.49105918720689e-05,
+      "loss": 0.1159,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8685022549252315,
+      "grad_norm": 0.028354499489068985,
+      "learning_rate": 7.480363896070089e-05,
+      "loss": 0.1216,
+      "step": 1968
+    },
+    {
+      "epoch": 1.8694516971279374,
+      "grad_norm": 0.06748262792825699,
+      "learning_rate": 7.469671681623742e-05,
+      "loss": 0.1888,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8704011393306432,
+      "grad_norm": 0.027621906250715256,
+      "learning_rate": 7.458982556923963e-05,
+      "loss": 0.1196,
+      "step": 1970
+    },
+    {
+      "epoch": 1.871350581533349,
+      "grad_norm": 0.031015096232295036,
+      "learning_rate": 7.448296535023077e-05,
+      "loss": 0.1266,
+      "step": 1971
+    },
+    {
+      "epoch": 1.872300023736055,
+      "grad_norm": 0.02895331382751465,
+      "learning_rate": 7.437613628969627e-05,
+      "loss": 0.1284,
+      "step": 1972
+    },
+    {
+      "epoch": 1.8732494659387608,
+      "grad_norm": 0.04110453650355339,
+      "learning_rate": 7.426933851808355e-05,
+      "loss": 0.1545,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8741989081414667,
+      "grad_norm": 0.030046746134757996,
+      "learning_rate": 7.416257216580181e-05,
+      "loss": 0.1269,
+      "step": 1974
+    },
+    {
+      "epoch": 1.8751483503441728,
+      "grad_norm": 0.04101106524467468,
+      "learning_rate": 7.405583736322182e-05,
+      "loss": 0.1621,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8760977925468787,
+      "grad_norm": 0.04438061639666557,
+      "learning_rate": 7.394913424067591e-05,
+      "loss": 0.1693,
+      "step": 1976
+    },
+    {
+      "epoch": 1.8770472347495846,
+      "grad_norm": 0.028554782271385193,
+      "learning_rate": 7.38424629284577e-05,
+      "loss": 0.1265,
+      "step": 1977
+    },
+    {
+      "epoch": 1.8779966769522907,
+      "grad_norm": 0.028575632721185684,
+      "learning_rate": 7.373582355682191e-05,
+      "loss": 0.12,
+      "step": 1978
+    },
+    {
+      "epoch": 1.8789461191549965,
+      "grad_norm": 0.027870802208781242,
+      "learning_rate": 7.362921625598436e-05,
+      "loss": 0.1301,
+      "step": 1979
+    },
+    {
+      "epoch": 1.8798955613577024,
+      "grad_norm": 0.03442731872200966,
+      "learning_rate": 7.352264115612158e-05,
+      "loss": 0.1537,
+      "step": 1980
+    },
+    {
+      "epoch": 1.8808450035604083,
+      "grad_norm": 0.03244437277317047,
+      "learning_rate": 7.341609838737089e-05,
+      "loss": 0.1356,
+      "step": 1981
+    },
+    {
+      "epoch": 1.8817944457631142,
+      "grad_norm": 0.0298260897397995,
+      "learning_rate": 7.330958807983011e-05,
+      "loss": 0.1238,
+      "step": 1982
+    },
+    {
+      "epoch": 1.88274388796582,
+      "grad_norm": 0.02521882764995098,
+      "learning_rate": 7.320311036355736e-05,
+      "loss": 0.1185,
+      "step": 1983
+    },
+    {
+      "epoch": 1.883693330168526,
+      "grad_norm": 0.02781338430941105,
+      "learning_rate": 7.309666536857106e-05,
+      "loss": 0.1214,
+      "step": 1984
+    },
+    {
+      "epoch": 1.8846427723712318,
+      "grad_norm": 0.05946779251098633,
+      "learning_rate": 7.299025322484958e-05,
+      "loss": 0.1897,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8855922145739377,
+      "grad_norm": 0.028507012873888016,
+      "learning_rate": 7.288387406233122e-05,
+      "loss": 0.1282,
+      "step": 1986
+    },
+    {
+      "epoch": 1.8865416567766438,
+      "grad_norm": 0.029610810801386833,
+      "learning_rate": 7.277752801091404e-05,
+      "loss": 0.1302,
+      "step": 1987
+    },
+    {
+      "epoch": 1.8874910989793496,
+      "grad_norm": 0.030304808169603348,
+      "learning_rate": 7.267121520045558e-05,
+      "loss": 0.132,
+      "step": 1988
+    },
+    {
+      "epoch": 1.8884405411820555,
+      "grad_norm": 0.028647607192397118,
+      "learning_rate": 7.256493576077292e-05,
+      "loss": 0.1309,
+      "step": 1989
+    },
+    {
+      "epoch": 1.8893899833847616,
+      "grad_norm": 0.04364948347210884,
+      "learning_rate": 7.245868982164226e-05,
+      "loss": 0.1628,
+      "step": 1990
+    },
+    {
+      "epoch": 1.8903394255874675,
+      "grad_norm": 0.029457390308380127,
+      "learning_rate": 7.235247751279893e-05,
+      "loss": 0.1163,
+      "step": 1991
+    },
+    {
+      "epoch": 1.8912888677901734,
+      "grad_norm": 0.059156183153390884,
+      "learning_rate": 7.224629896393726e-05,
+      "loss": 0.2033,
+      "step": 1992
+    },
+    {
+      "epoch": 1.8922383099928792,
+      "grad_norm": 0.02888781949877739,
+      "learning_rate": 7.214015430471028e-05,
+      "loss": 0.1242,
+      "step": 1993
+    },
+    {
+      "epoch": 1.893187752195585,
+      "grad_norm": 0.03040069155395031,
+      "learning_rate": 7.20340436647297e-05,
+      "loss": 0.1211,
+      "step": 1994
+    },
+    {
+      "epoch": 1.894137194398291,
+      "grad_norm": 0.028204258531332016,
+      "learning_rate": 7.192796717356562e-05,
+      "loss": 0.1267,
+      "step": 1995
+    },
+    {
+      "epoch": 1.8950866366009969,
+      "grad_norm": 0.030367571860551834,
+      "learning_rate": 7.182192496074648e-05,
+      "loss": 0.1232,
+      "step": 1996
+    },
+    {
+      "epoch": 1.8960360788037027,
+      "grad_norm": 0.02587362751364708,
+      "learning_rate": 7.171591715575888e-05,
+      "loss": 0.1261,
+      "step": 1997
+    },
+    {
+      "epoch": 1.8969855210064086,
+      "grad_norm": 0.028903882950544357,
+      "learning_rate": 7.160994388804736e-05,
+      "loss": 0.1318,
+      "step": 1998
+    },
+    {
+      "epoch": 1.8979349632091147,
+      "grad_norm": 0.025526562705636024,
+      "learning_rate": 7.150400528701436e-05,
+      "loss": 0.1205,
+      "step": 1999
+    },
+    {
+      "epoch": 1.8988844054118206,
+      "grad_norm": 0.029438691213726997,
+      "learning_rate": 7.139810148201987e-05,
+      "loss": 0.131,
+      "step": 2000
+    },
+    {
+      "epoch": 1.8998338476145264,
+      "grad_norm": 0.026545461267232895,
+      "learning_rate": 7.129223260238154e-05,
+      "loss": 0.1219,
+      "step": 2001
+    },
+    {
+      "epoch": 1.9007832898172323,
+      "grad_norm": 0.03137153759598732,
+      "learning_rate": 7.118639877737425e-05,
+      "loss": 0.1389,
+      "step": 2002
+    },
+    {
+      "epoch": 1.9017327320199384,
+      "grad_norm": 0.03655494004487991,
+      "learning_rate": 7.108060013623017e-05,
+      "loss": 0.1592,
+      "step": 2003
+    },
+    {
+      "epoch": 1.9026821742226443,
+      "grad_norm": 0.0271841399371624,
+      "learning_rate": 7.09748368081384e-05,
+      "loss": 0.1309,
+      "step": 2004
+    },
+    {
+      "epoch": 1.9036316164253502,
+      "grad_norm": 0.028577405959367752,
+      "learning_rate": 7.086910892224499e-05,
+      "loss": 0.1261,
+      "step": 2005
+    },
+    {
+      "epoch": 1.904581058628056,
+      "grad_norm": 0.028025876730680466,
+      "learning_rate": 7.076341660765271e-05,
+      "loss": 0.1323,
+      "step": 2006
+    },
+    {
+      "epoch": 1.905530500830762,
+      "grad_norm": 0.03332342579960823,
+      "learning_rate": 7.065775999342091e-05,
+      "loss": 0.1327,
+      "step": 2007
+    },
+    {
+      "epoch": 1.9064799430334678,
+      "grad_norm": 0.043180011212825775,
+      "learning_rate": 7.055213920856529e-05,
+      "loss": 0.1613,
+      "step": 2008
+    },
+    {
+      "epoch": 1.9074293852361737,
+      "grad_norm": 0.04228482022881508,
+      "learning_rate": 7.044655438205785e-05,
+      "loss": 0.1594,
+      "step": 2009
+    },
+    {
+      "epoch": 1.9083788274388795,
+      "grad_norm": 0.029172802343964577,
+      "learning_rate": 7.034100564282664e-05,
+      "loss": 0.1258,
+      "step": 2010
+    },
+    {
+      "epoch": 1.9093282696415854,
+      "grad_norm": 0.0426810160279274,
+      "learning_rate": 7.02354931197557e-05,
+      "loss": 0.16,
+      "step": 2011
+    },
+    {
+      "epoch": 1.9102777118442915,
+      "grad_norm": 0.025085503235459328,
+      "learning_rate": 7.013001694168478e-05,
+      "loss": 0.1233,
+      "step": 2012
+    },
+    {
+      "epoch": 1.9112271540469974,
+      "grad_norm": 0.0266293715685606,
+      "learning_rate": 7.002457723740934e-05,
+      "loss": 0.1214,
+      "step": 2013
+    },
+    {
+      "epoch": 1.9121765962497033,
+      "grad_norm": 0.03064984828233719,
+      "learning_rate": 6.991917413568017e-05,
+      "loss": 0.1186,
+      "step": 2014
+    },
+    {
+      "epoch": 1.9131260384524094,
+      "grad_norm": 0.026003271341323853,
+      "learning_rate": 6.981380776520348e-05,
+      "loss": 0.1228,
+      "step": 2015
+    },
+    {
+      "epoch": 1.9140754806551152,
+      "grad_norm": 0.045436155050992966,
+      "learning_rate": 6.970847825464059e-05,
+      "loss": 0.174,
+      "step": 2016
+    },
+    {
+      "epoch": 1.915024922857821,
+      "grad_norm": 0.029938362538814545,
+      "learning_rate": 6.960318573260783e-05,
+      "loss": 0.1201,
+      "step": 2017
+    },
+    {
+      "epoch": 1.915974365060527,
+      "grad_norm": 0.026935014873743057,
+      "learning_rate": 6.949793032767634e-05,
+      "loss": 0.1165,
+      "step": 2018
+    },
+    {
+      "epoch": 1.9169238072632329,
+      "grad_norm": 0.02809876948595047,
+      "learning_rate": 6.93927121683719e-05,
+      "loss": 0.1248,
+      "step": 2019
+    },
+    {
+      "epoch": 1.9178732494659387,
+      "grad_norm": 0.03932083770632744,
+      "learning_rate": 6.928753138317488e-05,
+      "loss": 0.1607,
+      "step": 2020
+    },
+    {
+      "epoch": 1.9188226916686446,
+      "grad_norm": 0.029043098911643028,
+      "learning_rate": 6.918238810051999e-05,
+      "loss": 0.1292,
+      "step": 2021
+    },
+    {
+      "epoch": 1.9197721338713505,
+      "grad_norm": 0.03849990293383598,
+      "learning_rate": 6.907728244879611e-05,
+      "loss": 0.1611,
+      "step": 2022
+    },
+    {
+      "epoch": 1.9207215760740564,
+      "grad_norm": 0.028439447283744812,
+      "learning_rate": 6.897221455634624e-05,
+      "loss": 0.1265,
+      "step": 2023
+    },
+    {
+      "epoch": 1.9216710182767625,
+      "grad_norm": 0.028611112385988235,
+      "learning_rate": 6.886718455146724e-05,
+      "loss": 0.1312,
+      "step": 2024
+    },
+    {
+      "epoch": 1.9226204604794683,
+      "grad_norm": 0.02605103701353073,
+      "learning_rate": 6.87621925624096e-05,
+      "loss": 0.1241,
+      "step": 2025
+    },
+    {
+      "epoch": 1.9235699026821742,
+      "grad_norm": 0.06604333966970444,
+      "learning_rate": 6.865723871737762e-05,
+      "loss": 0.2016,
+      "step": 2026
+    },
+    {
+      "epoch": 1.92451934488488,
+      "grad_norm": 0.044974714517593384,
+      "learning_rate": 6.855232314452884e-05,
+      "loss": 0.1778,
+      "step": 2027
+    },
+    {
+      "epoch": 1.9254687870875862,
+      "grad_norm": 0.03168616443872452,
+      "learning_rate": 6.844744597197409e-05,
+      "loss": 0.1327,
+      "step": 2028
+    },
+    {
+      "epoch": 1.926418229290292,
+      "grad_norm": 0.029546428471803665,
+      "learning_rate": 6.834260732777736e-05,
+      "loss": 0.1302,
+      "step": 2029
+    },
+    {
+      "epoch": 1.927367671492998,
+      "grad_norm": 0.05021713301539421,
+      "learning_rate": 6.823780733995557e-05,
+      "loss": 0.1863,
+      "step": 2030
+    },
+    {
+      "epoch": 1.9283171136957038,
+      "grad_norm": 0.030026502907276154,
+      "learning_rate": 6.813304613647845e-05,
+      "loss": 0.1349,
+      "step": 2031
+    },
+    {
+      "epoch": 1.9292665558984097,
+      "grad_norm": 0.03538592904806137,
+      "learning_rate": 6.802832384526836e-05,
+      "loss": 0.1374,
+      "step": 2032
+    },
+    {
+      "epoch": 1.9302159981011155,
+      "grad_norm": 0.027488164603710175,
+      "learning_rate": 6.792364059420012e-05,
+      "loss": 0.1237,
+      "step": 2033
+    },
+    {
+      "epoch": 1.9311654403038214,
+      "grad_norm": 0.03553836792707443,
+      "learning_rate": 6.781899651110091e-05,
+      "loss": 0.1522,
+      "step": 2034
+    },
+    {
+      "epoch": 1.9321148825065273,
+      "grad_norm": 0.029753949493169785,
+      "learning_rate": 6.771439172375007e-05,
+      "loss": 0.1222,
+      "step": 2035
+    },
+    {
+      "epoch": 1.9330643247092332,
+      "grad_norm": 0.03108718991279602,
+      "learning_rate": 6.760982635987899e-05,
+      "loss": 0.1186,
+      "step": 2036
+    },
+    {
+      "epoch": 1.9340137669119393,
+      "grad_norm": 0.02662482298910618,
+      "learning_rate": 6.750530054717088e-05,
+      "loss": 0.1189,
+      "step": 2037
+    },
+    {
+      "epoch": 1.9349632091146451,
+      "grad_norm": 0.029288165271282196,
+      "learning_rate": 6.740081441326062e-05,
+      "loss": 0.1179,
+      "step": 2038
+    },
+    {
+      "epoch": 1.935912651317351,
+      "grad_norm": 0.03478897735476494,
+      "learning_rate": 6.729636808573476e-05,
+      "loss": 0.1249,
+      "step": 2039
+    },
+    {
+      "epoch": 1.9368620935200571,
+      "grad_norm": 0.03969739004969597,
+      "learning_rate": 6.719196169213114e-05,
+      "loss": 0.1579,
+      "step": 2040
+    },
+    {
+      "epoch": 1.937811535722763,
+      "grad_norm": 0.030195200815796852,
+      "learning_rate": 6.708759535993884e-05,
+      "loss": 0.115,
+      "step": 2041
+    },
+    {
+      "epoch": 1.9387609779254689,
+      "grad_norm": 0.03426138311624527,
+      "learning_rate": 6.698326921659808e-05,
+      "loss": 0.1266,
+      "step": 2042
+    },
+    {
+      "epoch": 1.9397104201281747,
+      "grad_norm": 0.05202037841081619,
+      "learning_rate": 6.687898338949998e-05,
+      "loss": 0.193,
+      "step": 2043
+    },
+    {
+      "epoch": 1.9406598623308806,
+      "grad_norm": 0.027649085968732834,
+      "learning_rate": 6.67747380059864e-05,
+      "loss": 0.1222,
+      "step": 2044
+    },
+    {
+      "epoch": 1.9416093045335865,
+      "grad_norm": 0.026928169652819633,
+      "learning_rate": 6.667053319334982e-05,
+      "loss": 0.1204,
+      "step": 2045
+    },
+    {
+      "epoch": 1.9425587467362924,
+      "grad_norm": 0.056547269225120544,
+      "learning_rate": 6.656636907883325e-05,
+      "loss": 0.1602,
+      "step": 2046
+    },
+    {
+      "epoch": 1.9435081889389982,
+      "grad_norm": 0.026589645072817802,
+      "learning_rate": 6.646224578962993e-05,
+      "loss": 0.1214,
+      "step": 2047
+    },
+    {
+      "epoch": 1.9444576311417041,
+      "grad_norm": 0.02858765795826912,
+      "learning_rate": 6.635816345288329e-05,
+      "loss": 0.1242,
+      "step": 2048
+    },
+    {
+      "epoch": 1.9454070733444102,
+      "grad_norm": 0.04160701856017113,
+      "learning_rate": 6.625412219568668e-05,
+      "loss": 0.1606,
+      "step": 2049
+    },
+    {
+      "epoch": 1.946356515547116,
+      "grad_norm": 0.03329680487513542,
+      "learning_rate": 6.615012214508336e-05,
+      "loss": 0.1346,
+      "step": 2050
+    },
+    {
+      "epoch": 1.947305957749822,
+      "grad_norm": 0.041767850518226624,
+      "learning_rate": 6.604616342806632e-05,
+      "loss": 0.1566,
+      "step": 2051
+    },
+    {
+      "epoch": 1.948255399952528,
+      "grad_norm": 0.027340400964021683,
+      "learning_rate": 6.594224617157795e-05,
+      "loss": 0.1253,
+      "step": 2052
+    },
+    {
+      "epoch": 1.949204842155234,
+      "grad_norm": 0.06383645534515381,
+      "learning_rate": 6.583837050251012e-05,
+      "loss": 0.1518,
+      "step": 2053
+    },
+    {
+      "epoch": 1.9501542843579398,
+      "grad_norm": 0.051231034100055695,
+      "learning_rate": 6.573453654770383e-05,
+      "loss": 0.1565,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9501542843579398,
+      "eval_loss": 0.37301480770111084,
+      "eval_runtime": 38.0432,
+      "eval_samples_per_second": 2.261,
+      "eval_steps_per_second": 2.261,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9513410871113221,
+      "grad_norm": 0.4340180456638336,
+      "learning_rate": 0.00013378114170405474,
+      "loss": 0.3769,
+      "step": 2055
+    },
+    {
+      "epoch": 1.952290529314028,
+      "grad_norm": 0.2337399125099182,
+      "learning_rate": 0.00013372497405242763,
+      "loss": 0.3148,
+      "step": 2056
+    },
+    {
+      "epoch": 1.953239971516734,
+      "grad_norm": 0.15890651941299438,
+      "learning_rate": 0.00013366879439324493,
+      "loss": 0.3167,
+      "step": 2057
+    },
+    {
+      "epoch": 1.95418941371944,
+      "grad_norm": 0.14962317049503326,
+      "learning_rate": 0.00013361260274650906,
+      "loss": 0.3146,
+      "step": 2058
+    },
+    {
+      "epoch": 1.9551388559221459,
+      "grad_norm": 5.296742916107178,
+      "learning_rate": 0.00013355639913222668,
+      "loss": 0.4622,
+      "step": 2059
+    },
+    {
+      "epoch": 1.9560882981248517,
+      "grad_norm": 7.226221084594727,
+      "learning_rate": 0.0001335001835704087,
+      "loss": 1.4115,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9570377403275576,
+      "grad_norm": 3.240274667739868,
+      "learning_rate": 0.00013344395608107031,
+      "loss": 1.2552,
+      "step": 2061
+    },
+    {
+      "epoch": 1.9579871825302635,
+      "grad_norm": 2.299501657485962,
+      "learning_rate": 0.00013338771668423095,
+      "loss": 0.3784,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9589366247329694,
+      "grad_norm": 2.4971210956573486,
+      "learning_rate": 0.00013333146539991431,
+      "loss": 0.6146,
+      "step": 2063
+    },
+    {
+      "epoch": 1.9598860669356752,
+      "grad_norm": 1.7239331007003784,
+      "learning_rate": 0.00013327520224814822,
+      "loss": 0.4257,
+      "step": 2064
+    },
+    {
+      "epoch": 1.9608355091383811,
+      "grad_norm": 0.29740026593208313,
+      "learning_rate": 0.00013321892724896484,
+      "loss": 0.3187,
+      "step": 2065
+    },
+    {
+      "epoch": 1.961784951341087,
+      "grad_norm": 8.102334022521973,
+      "learning_rate": 0.0001331626404224005,
+      "loss": 0.481,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9627343935437929,
+      "grad_norm": 0.29957181215286255,
+      "learning_rate": 0.0001331063417884958,
+      "loss": 0.3117,
+      "step": 2067
+    },
+    {
+      "epoch": 1.963683835746499,
+      "grad_norm": 2.238389730453491,
+      "learning_rate": 0.00013305003136729552,
+      "loss": 0.3736,
+      "step": 2068
+    },
+    {
+      "epoch": 1.9646332779492048,
+      "grad_norm": 0.37475112080574036,
+      "learning_rate": 0.0001329937091788485,
+      "loss": 0.3093,
+      "step": 2069
+    },
+    {
+      "epoch": 1.9655827201519107,
+      "grad_norm": 1.1860514879226685,
+      "learning_rate": 0.00013293737524320797,
+      "loss": 0.3951,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9665321623546168,
+      "grad_norm": 0.2817871868610382,
+      "learning_rate": 0.00013288102958043126,
+      "loss": 0.3127,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9674816045573227,
+      "grad_norm": 0.3843158483505249,
+      "learning_rate": 0.00013282467221057984,
+      "loss": 0.2984,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9684310467600286,
+      "grad_norm": 0.40091991424560547,
+      "learning_rate": 0.0001327683031537194,
+      "loss": 0.2966,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9693804889627344,
+      "grad_norm": 0.34588465094566345,
+      "learning_rate": 0.00013271192242991976,
+      "loss": 0.3163,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9703299311654403,
+      "grad_norm": 0.19917060434818268,
+      "learning_rate": 0.00013265553005925492,
+      "loss": 0.3001,
+      "step": 2075
+    },
+    {
+      "epoch": 1.9712793733681462,
+      "grad_norm": 0.17843176424503326,
+      "learning_rate": 0.00013259912606180301,
+      "loss": 0.3018,
+      "step": 2076
+    },
+    {
+      "epoch": 1.972228815570852,
+      "grad_norm": 0.10518278181552887,
+      "learning_rate": 0.00013254271045764636,
+      "loss": 0.2883,
+      "step": 2077
+    },
+    {
+      "epoch": 1.973178257773558,
+      "grad_norm": 0.16444529592990875,
+      "learning_rate": 0.00013248628326687124,
+      "loss": 0.3041,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9741276999762638,
+      "grad_norm": 0.1926691085100174,
+      "learning_rate": 0.00013242984450956828,
+      "loss": 0.2763,
+      "step": 2079
+    },
+    {
+      "epoch": 1.97507714217897,
+      "grad_norm": 0.24896161258220673,
+      "learning_rate": 0.00013237339420583212,
+      "loss": 0.2895,
+      "step": 2080
+    },
+    {
+      "epoch": 1.9760265843816758,
+      "grad_norm": 0.23915739357471466,
+      "learning_rate": 0.00013231693237576148,
+      "loss": 0.2901,
+      "step": 2081
+    },
+    {
+      "epoch": 1.9769760265843817,
+      "grad_norm": 0.08436968922615051,
+      "learning_rate": 0.00013226045903945926,
+      "loss": 0.278,
+      "step": 2082
+    },
+    {
+      "epoch": 1.9779254687870877,
+      "grad_norm": 0.9301303625106812,
+      "learning_rate": 0.00013220397421703247,
+      "loss": 0.316,
+      "step": 2083
+    },
+    {
+      "epoch": 1.9788749109897936,
+      "grad_norm": 1.2519832849502563,
+      "learning_rate": 0.00013214747792859201,
+      "loss": 0.2931,
+      "step": 2084
+    },
+    {
+      "epoch": 1.9798243531924995,
+      "grad_norm": 0.4805239140987396,
+      "learning_rate": 0.00013209097019425316,
+      "loss": 0.4146,
+      "step": 2085
+    },
+    {
+      "epoch": 1.9807737953952054,
+      "grad_norm": 0.21979232132434845,
+      "learning_rate": 0.00013203445103413507,
+      "loss": 0.3,
+      "step": 2086
+    },
+    {
+      "epoch": 1.9817232375979112,
+      "grad_norm": 0.1640891432762146,
+      "learning_rate": 0.000131977920468361,
+      "loss": 0.2969,
+      "step": 2087
+    },
+    {
+      "epoch": 1.9826726798006171,
+      "grad_norm": 0.17019522190093994,
+      "learning_rate": 0.0001319213785170583,
+      "loss": 0.2914,
+      "step": 2088
+    },
+    {
+      "epoch": 1.983622122003323,
+      "grad_norm": 0.09475825726985931,
+      "learning_rate": 0.00013186482520035839,
+      "loss": 0.297,
+      "step": 2089
+    },
+    {
+      "epoch": 1.9845715642060289,
+      "grad_norm": 0.09213607013225555,
+      "learning_rate": 0.00013180826053839668,
+      "loss": 0.288,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9855210064087347,
+      "grad_norm": 0.11374935507774353,
+      "learning_rate": 0.00013175168455131263,
+      "loss": 0.2796,
+      "step": 2091
+    },
+    {
+      "epoch": 1.9864704486114406,
+      "grad_norm": 0.10812429338693619,
+      "learning_rate": 0.0001316950972592498,
+      "loss": 0.3057,
+      "step": 2092
+    },
+    {
+      "epoch": 1.9874198908141467,
+      "grad_norm": 0.07910951226949692,
+      "learning_rate": 0.00013163849868235564,
+      "loss": 0.2877,
+      "step": 2093
+    },
+    {
+      "epoch": 1.9883693330168526,
+      "grad_norm": 0.09240693598985672,
+      "learning_rate": 0.00013158188884078182,
+      "loss": 0.2906,
+      "step": 2094
+    },
+    {
+      "epoch": 1.9893187752195585,
+      "grad_norm": 0.097608283162117,
+      "learning_rate": 0.00013152526775468378,
+      "loss": 0.2906,
+      "step": 2095
+    },
+    {
+      "epoch": 1.9902682174222646,
+      "grad_norm": 0.2190292775630951,
+      "learning_rate": 0.00013146863544422118,
+      "loss": 0.2835,
+      "step": 2096
+    },
+    {
+      "epoch": 1.9912176596249704,
+      "grad_norm": 0.07066213339567184,
+      "learning_rate": 0.00013141199192955751,
+      "loss": 0.2856,
+      "step": 2097
+    },
+    {
+      "epoch": 1.9921671018276763,
+      "grad_norm": 0.10716898739337921,
+      "learning_rate": 0.0001313553372308604,
+      "loss": 0.3012,
+      "step": 2098
+    },
+    {
+      "epoch": 1.9931165440303822,
+      "grad_norm": 0.07971798628568649,
+      "learning_rate": 0.00013129867136830127,
+      "loss": 0.2678,
+      "step": 2099
+    },
+    {
+      "epoch": 1.994065986233088,
+      "grad_norm": 0.11225918680429459,
+      "learning_rate": 0.00013124199436205576,
+      "loss": 0.2799,
+      "step": 2100
+    },
+    {
+      "epoch": 1.995015428435794,
+      "grad_norm": 0.08741844445466995,
+      "learning_rate": 0.00013118530623230327,
+      "loss": 0.284,
+      "step": 2101
+    },
+    {
+      "epoch": 1.9959648706384998,
+      "grad_norm": 0.07644308358430862,
+      "learning_rate": 0.00013112860699922722,
+      "loss": 0.2988,
+      "step": 2102
+    },
+    {
+      "epoch": 1.9969143128412057,
+      "grad_norm": 0.07610399276018143,
+      "learning_rate": 0.00013107189668301508,
+      "loss": 0.2813,
+      "step": 2103
+    },
+    {
+      "epoch": 1.9978637550439116,
+      "grad_norm": 0.1364275962114334,
+      "learning_rate": 0.0001310151753038581,
+      "loss": 0.3006,
+      "step": 2104
+    },
+    {
+      "epoch": 1.9988131972466177,
+      "grad_norm": 0.06598393619060516,
+      "learning_rate": 0.0001309584428819516,
+      "loss": 0.2822,
+      "step": 2105
+    },
+    {
+      "epoch": 1.9997626394493235,
+      "grad_norm": 0.08182472735643387,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.2757,
+      "step": 2106
+    },
+    {
+      "epoch": 2.0007120816520296,
+      "grad_norm": 0.06756250560283661,
+      "learning_rate": 0.0001308449449906907,
+      "loss": 0.2619,
+      "step": 2107
+    },
+    {
+      "epoch": 2.0016615238547355,
+      "grad_norm": 0.05981763079762459,
+      "learning_rate": 0.00013078817956174656,
+      "loss": 0.2856,
+      "step": 2108
+    },
+    {
+      "epoch": 2.0026109660574414,
+      "grad_norm": 0.0790615975856781,
+      "learning_rate": 0.0001307314031708732,
+      "loss": 0.2875,
+      "step": 2109
+    },
+    {
+      "epoch": 2.0035604082601473,
+      "grad_norm": 0.06421328336000443,
+      "learning_rate": 0.00013067461583828553,
+      "loss": 0.2683,
+      "step": 2110
+    },
+    {
+      "epoch": 2.004509850462853,
+      "grad_norm": 0.06607569754123688,
+      "learning_rate": 0.0001306178175842023,
+      "loss": 0.271,
+      "step": 2111
+    },
+    {
+      "epoch": 2.005459292665559,
+      "grad_norm": 0.06524945050477982,
+      "learning_rate": 0.00013056100842884612,
+      "loss": 0.2796,
+      "step": 2112
+    },
+    {
+      "epoch": 2.006408734868265,
+      "grad_norm": 0.05927155539393425,
+      "learning_rate": 0.00013050418839244355,
+      "loss": 0.2755,
+      "step": 2113
+    },
+    {
+      "epoch": 2.0073581770709708,
+      "grad_norm": 0.06408464163541794,
+      "learning_rate": 0.000130447357495225,
+      "loss": 0.2748,
+      "step": 2114
+    },
+    {
+      "epoch": 2.0083076192736766,
+      "grad_norm": 0.05964144691824913,
+      "learning_rate": 0.0001303905157574247,
+      "loss": 0.2772,
+      "step": 2115
+    },
+    {
+      "epoch": 2.0092570614763825,
+      "grad_norm": 0.05294380709528923,
+      "learning_rate": 0.00013033366319928079,
+      "loss": 0.2543,
+      "step": 2116
+    },
+    {
+      "epoch": 2.0102065036790884,
+      "grad_norm": 0.06316480785608292,
+      "learning_rate": 0.00013027679984103528,
+      "loss": 0.2659,
+      "step": 2117
+    },
+    {
+      "epoch": 2.0111559458817942,
+      "grad_norm": 0.0780426636338234,
+      "learning_rate": 0.000130219925702934,
+      "loss": 0.2809,
+      "step": 2118
+    },
+    {
+      "epoch": 2.0121053880845006,
+      "grad_norm": 0.05921616032719612,
+      "learning_rate": 0.00013016304080522656,
+      "loss": 0.2651,
+      "step": 2119
+    },
+    {
+      "epoch": 2.0130548302872064,
+      "grad_norm": 0.063509002327919,
+      "learning_rate": 0.0001301061451681665,
+      "loss": 0.2766,
+      "step": 2120
+    },
+    {
+      "epoch": 2.0140042724899123,
+      "grad_norm": 0.06251564621925354,
+      "learning_rate": 0.0001300492388120111,
+      "loss": 0.2826,
+      "step": 2121
+    },
+    {
+      "epoch": 2.014953714692618,
+      "grad_norm": 0.07721933722496033,
+      "learning_rate": 0.0001299923217570215,
+      "loss": 0.2876,
+      "step": 2122
+    },
+    {
+      "epoch": 2.015903156895324,
+      "grad_norm": 0.22655093669891357,
+      "learning_rate": 0.0001299353940234627,
+      "loss": 0.3023,
+      "step": 2123
+    },
+    {
+      "epoch": 2.01685259909803,
+      "grad_norm": 0.16343270242214203,
+      "learning_rate": 0.0001298784556316034,
+      "loss": 0.2902,
+      "step": 2124
+    },
+    {
+      "epoch": 2.017802041300736,
+      "grad_norm": 0.0674663856625557,
+      "learning_rate": 0.00012982150660171613,
+      "loss": 0.2639,
+      "step": 2125
+    },
+    {
+      "epoch": 2.0187514835034417,
+      "grad_norm": 0.06698331236839294,
+      "learning_rate": 0.00012976454695407723,
+      "loss": 0.2918,
+      "step": 2126
+    },
+    {
+      "epoch": 2.0197009257061476,
+      "grad_norm": 0.05850343778729439,
+      "learning_rate": 0.00012970757670896683,
+      "loss": 0.2691,
+      "step": 2127
+    },
+    {
+      "epoch": 2.0206503679088534,
+      "grad_norm": 0.05704069882631302,
+      "learning_rate": 0.0001296505958866688,
+      "loss": 0.2688,
+      "step": 2128
+    },
+    {
+      "epoch": 2.0215998101115593,
+      "grad_norm": 0.060891758650541306,
+      "learning_rate": 0.00012959360450747075,
+      "loss": 0.2652,
+      "step": 2129
+    },
+    {
+      "epoch": 2.022549252314265,
+      "grad_norm": 0.061691515147686005,
+      "learning_rate": 0.00012953660259166412,
+      "loss": 0.2756,
+      "step": 2130
+    },
+    {
+      "epoch": 2.023498694516971,
+      "grad_norm": 0.059189558029174805,
+      "learning_rate": 0.00012947959015954406,
+      "loss": 0.2759,
+      "step": 2131
+    },
+    {
+      "epoch": 2.0244481367196774,
+      "grad_norm": 0.06447713077068329,
+      "learning_rate": 0.00012942256723140952,
+      "loss": 0.2773,
+      "step": 2132
+    },
+    {
+      "epoch": 2.0253975789223833,
+      "grad_norm": 0.06263953447341919,
+      "learning_rate": 0.0001293655338275631,
+      "loss": 0.2867,
+      "step": 2133
+    },
+    {
+      "epoch": 2.026347021125089,
+      "grad_norm": 0.0576293058693409,
+      "learning_rate": 0.00012930848996831114,
+      "loss": 0.2776,
+      "step": 2134
+    },
+    {
+      "epoch": 2.027296463327795,
+      "grad_norm": 0.05699608847498894,
+      "learning_rate": 0.00012925143567396374,
+      "loss": 0.2757,
+      "step": 2135
+    },
+    {
+      "epoch": 2.028245905530501,
+      "grad_norm": 0.052561014890670776,
+      "learning_rate": 0.00012919437096483476,
+      "loss": 0.2555,
+      "step": 2136
+    },
+    {
+      "epoch": 2.0291953477332068,
+      "grad_norm": 0.053198445588350296,
+      "learning_rate": 0.00012913729586124165,
+      "loss": 0.2676,
+      "step": 2137
+    },
+    {
+      "epoch": 2.0301447899359126,
+      "grad_norm": 0.09329196810722351,
+      "learning_rate": 0.00012908021038350568,
+      "loss": 0.2796,
+      "step": 2138
+    },
+    {
+      "epoch": 2.0310942321386185,
+      "grad_norm": 0.07239534705877304,
+      "learning_rate": 0.00012902311455195172,
+      "loss": 0.2809,
+      "step": 2139
+    },
+    {
+      "epoch": 2.0320436743413244,
+      "grad_norm": 0.06299670785665512,
+      "learning_rate": 0.00012896600838690838,
+      "loss": 0.2672,
+      "step": 2140
+    },
+    {
+      "epoch": 2.0329931165440303,
+      "grad_norm": 0.05467437952756882,
+      "learning_rate": 0.00012890889190870795,
+      "loss": 0.268,
+      "step": 2141
+    },
+    {
+      "epoch": 2.033942558746736,
+      "grad_norm": 0.0641472190618515,
+      "learning_rate": 0.00012885176513768637,
+      "loss": 0.2844,
+      "step": 2142
+    },
+    {
+      "epoch": 2.034892000949442,
+      "grad_norm": 0.06481951475143433,
+      "learning_rate": 0.00012879462809418325,
+      "loss": 0.2883,
+      "step": 2143
+    },
+    {
+      "epoch": 2.0358414431521483,
+      "grad_norm": 0.05889345332980156,
+      "learning_rate": 0.0001287374807985418,
+      "loss": 0.2688,
+      "step": 2144
+    },
+    {
+      "epoch": 2.036790885354854,
+      "grad_norm": 0.05446067079901695,
+      "learning_rate": 0.00012868032327110904,
+      "loss": 0.2699,
+      "step": 2145
+    },
+    {
+      "epoch": 2.03774032755756,
+      "grad_norm": 0.0558142326772213,
+      "learning_rate": 0.00012862315553223547,
+      "loss": 0.2662,
+      "step": 2146
+    },
+    {
+      "epoch": 2.038689769760266,
+      "grad_norm": 0.05485325679183006,
+      "learning_rate": 0.0001285659776022753,
+      "loss": 0.2684,
+      "step": 2147
+    },
+    {
+      "epoch": 2.039639211962972,
+      "grad_norm": 0.05541551858186722,
+      "learning_rate": 0.0001285087895015864,
+      "loss": 0.2674,
+      "step": 2148
+    },
+    {
+      "epoch": 2.0405886541656777,
+      "grad_norm": 0.10139881074428558,
+      "learning_rate": 0.0001284515912505301,
+      "loss": 0.2737,
+      "step": 2149
+    },
+    {
+      "epoch": 2.0415380963683836,
+      "grad_norm": 0.05179375782608986,
+      "learning_rate": 0.00012839438286947163,
+      "loss": 0.2647,
+      "step": 2150
+    },
+    {
+      "epoch": 2.0424875385710894,
+      "grad_norm": 0.0590873584151268,
+      "learning_rate": 0.0001283371643787795,
+      "loss": 0.2743,
+      "step": 2151
+    },
+    {
+      "epoch": 2.0434369807737953,
+      "grad_norm": 0.0546240359544754,
+      "learning_rate": 0.00012827993579882612,
+      "loss": 0.2798,
+      "step": 2152
+    },
+    {
+      "epoch": 2.044386422976501,
+      "grad_norm": 0.056896887719631195,
+      "learning_rate": 0.0001282226971499872,
+      "loss": 0.2717,
+      "step": 2153
+    },
+    {
+      "epoch": 2.045335865179207,
+      "grad_norm": 0.052284859120845795,
+      "learning_rate": 0.00012816544845264228,
+      "loss": 0.2719,
+      "step": 2154
+    },
+    {
+      "epoch": 2.046285307381913,
+      "grad_norm": 0.060961298644542694,
+      "learning_rate": 0.0001281081897271744,
+      "loss": 0.2764,
+      "step": 2155
+    },
+    {
+      "epoch": 2.047234749584619,
+      "grad_norm": 0.08830570429563522,
+      "learning_rate": 0.0001280509209939701,
+      "loss": 0.2946,
+      "step": 2156
+    },
+    {
+      "epoch": 2.048184191787325,
+      "grad_norm": 0.05548688769340515,
+      "learning_rate": 0.00012799364227341955,
+      "loss": 0.2647,
+      "step": 2157
+    },
+    {
+      "epoch": 2.049133633990031,
+      "grad_norm": 0.05134082958102226,
+      "learning_rate": 0.00012793635358591645,
+      "loss": 0.2724,
+      "step": 2158
+    },
+    {
+      "epoch": 2.050083076192737,
+      "grad_norm": 0.06974118202924728,
+      "learning_rate": 0.0001278790549518581,
+      "loss": 0.2873,
+      "step": 2159
+    },
+    {
+      "epoch": 2.0510325183954428,
+      "grad_norm": 0.06583964079618454,
+      "learning_rate": 0.0001278217463916453,
+      "loss": 0.2823,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0519819605981486,
+      "grad_norm": 0.05401783436536789,
+      "learning_rate": 0.00012776442792568232,
+      "loss": 0.2788,
+      "step": 2161
+    },
+    {
+      "epoch": 2.0529314028008545,
+      "grad_norm": 0.09343112260103226,
+      "learning_rate": 0.00012770709957437708,
+      "loss": 0.2824,
+      "step": 2162
+    },
+    {
+      "epoch": 2.0538808450035604,
+      "grad_norm": 0.11026190966367722,
+      "learning_rate": 0.00012764976135814094,
+      "loss": 0.2861,
+      "step": 2163
+    },
+    {
+      "epoch": 2.0548302872062663,
+      "grad_norm": 0.05160842835903168,
+      "learning_rate": 0.00012759241329738887,
+      "loss": 0.2615,
+      "step": 2164
+    },
+    {
+      "epoch": 2.055779729408972,
+      "grad_norm": 0.057216208428144455,
+      "learning_rate": 0.00012753505541253916,
+      "loss": 0.2757,
+      "step": 2165
+    },
+    {
+      "epoch": 2.056729171611678,
+      "grad_norm": 0.07923352718353271,
+      "learning_rate": 0.00012747768772401378,
+      "loss": 0.2659,
+      "step": 2166
+    },
+    {
+      "epoch": 2.057678613814384,
+      "grad_norm": 0.055502623319625854,
+      "learning_rate": 0.0001274203102522381,
+      "loss": 0.2757,
+      "step": 2167
+    },
+    {
+      "epoch": 2.0586280560170898,
+      "grad_norm": 0.10472196340560913,
+      "learning_rate": 0.00012736292301764098,
+      "loss": 0.2938,
+      "step": 2168
+    },
+    {
+      "epoch": 2.059577498219796,
+      "grad_norm": 0.1105305552482605,
+      "learning_rate": 0.00012730552604065475,
+      "loss": 0.2846,
+      "step": 2169
+    },
+    {
+      "epoch": 2.060526940422502,
+      "grad_norm": 0.06079312413930893,
+      "learning_rate": 0.0001272481193417153,
+      "loss": 0.2724,
+      "step": 2170
+    },
+    {
+      "epoch": 2.061476382625208,
+      "grad_norm": 0.06276509910821915,
+      "learning_rate": 0.00012719070294126182,
+      "loss": 0.2704,
+      "step": 2171
+    },
+    {
+      "epoch": 2.0624258248279137,
+      "grad_norm": 0.08746016025543213,
+      "learning_rate": 0.00012713327685973707,
+      "loss": 0.2834,
+      "step": 2172
+    },
+    {
+      "epoch": 2.0633752670306196,
+      "grad_norm": 0.053869761526584625,
+      "learning_rate": 0.0001270758411175873,
+      "loss": 0.2712,
+      "step": 2173
+    },
+    {
+      "epoch": 2.0643247092333254,
+      "grad_norm": 0.05118397995829582,
+      "learning_rate": 0.00012701839573526206,
+      "loss": 0.2737,
+      "step": 2174
+    },
+    {
+      "epoch": 2.0652741514360313,
+      "grad_norm": 0.05905655771493912,
+      "learning_rate": 0.0001269609407332144,
+      "loss": 0.2663,
+      "step": 2175
+    },
+    {
+      "epoch": 2.066223593638737,
+      "grad_norm": 0.049641139805316925,
+      "learning_rate": 0.00012690347613190082,
+      "loss": 0.263,
+      "step": 2176
+    },
+    {
+      "epoch": 2.067173035841443,
+      "grad_norm": 0.04823688417673111,
+      "learning_rate": 0.00012684600195178117,
+      "loss": 0.2667,
+      "step": 2177
+    },
+    {
+      "epoch": 2.068122478044149,
+      "grad_norm": 0.07979489117860794,
+      "learning_rate": 0.00012678851821331882,
+      "loss": 0.2854,
+      "step": 2178
+    },
+    {
+      "epoch": 2.069071920246855,
+      "grad_norm": 0.06123083457350731,
+      "learning_rate": 0.00012673102493698042,
+      "loss": 0.2832,
+      "step": 2179
+    },
+    {
+      "epoch": 2.0700213624495607,
+      "grad_norm": 0.07498030364513397,
+      "learning_rate": 0.00012667352214323614,
+      "loss": 0.3061,
+      "step": 2180
+    },
+    {
+      "epoch": 2.0709708046522666,
+      "grad_norm": 0.059050336480140686,
+      "learning_rate": 0.0001266160098525594,
+      "loss": 0.2623,
+      "step": 2181
+    },
+    {
+      "epoch": 2.071920246854973,
+      "grad_norm": 0.060739047825336456,
+      "learning_rate": 0.00012655848808542709,
+      "loss": 0.282,
+      "step": 2182
+    },
+    {
+      "epoch": 2.0728696890576788,
+      "grad_norm": 0.059133414179086685,
+      "learning_rate": 0.00012650095686231953,
+      "loss": 0.2637,
+      "step": 2183
+    },
+    {
+      "epoch": 2.0738191312603846,
+      "grad_norm": 0.05270388349890709,
+      "learning_rate": 0.00012644341620372023,
+      "loss": 0.2742,
+      "step": 2184
+    },
+    {
+      "epoch": 2.0747685734630905,
+      "grad_norm": 0.049184754490852356,
+      "learning_rate": 0.00012638586613011624,
+      "loss": 0.2582,
+      "step": 2185
+    },
+    {
+      "epoch": 2.0757180156657964,
+      "grad_norm": 0.05757623910903931,
+      "learning_rate": 0.0001263283066619978,
+      "loss": 0.2672,
+      "step": 2186
+    },
+    {
+      "epoch": 2.0766674578685023,
+      "grad_norm": 0.051976773887872696,
+      "learning_rate": 0.0001262707378198587,
+      "loss": 0.2769,
+      "step": 2187
+    },
+    {
+      "epoch": 2.077616900071208,
+      "grad_norm": 0.04786711558699608,
+      "learning_rate": 0.00012621315962419585,
+      "loss": 0.2661,
+      "step": 2188
+    },
+    {
+      "epoch": 2.078566342273914,
+      "grad_norm": 0.0624409057199955,
+      "learning_rate": 0.00012615557209550967,
+      "loss": 0.2867,
+      "step": 2189
+    },
+    {
+      "epoch": 2.07951578447662,
+      "grad_norm": 0.05563337355852127,
+      "learning_rate": 0.00012609797525430373,
+      "loss": 0.2778,
+      "step": 2190
+    },
+    {
+      "epoch": 2.0804652266793258,
+      "grad_norm": 0.04968985542654991,
+      "learning_rate": 0.00012604036912108505,
+      "loss": 0.2562,
+      "step": 2191
+    },
+    {
+      "epoch": 2.0814146688820316,
+      "grad_norm": 0.05211299657821655,
+      "learning_rate": 0.00012598275371636394,
+      "loss": 0.2746,
+      "step": 2192
+    },
+    {
+      "epoch": 2.0823641110847375,
+      "grad_norm": 0.0466628223657608,
+      "learning_rate": 0.00012592512906065397,
+      "loss": 0.2654,
+      "step": 2193
+    },
+    {
+      "epoch": 2.083313553287444,
+      "grad_norm": 0.056648485362529755,
+      "learning_rate": 0.000125867495174472,
+      "loss": 0.28,
+      "step": 2194
+    },
+    {
+      "epoch": 2.0842629954901497,
+      "grad_norm": 0.056760817766189575,
+      "learning_rate": 0.0001258098520783382,
+      "loss": 0.2732,
+      "step": 2195
+    },
+    {
+      "epoch": 2.0852124376928556,
+      "grad_norm": 0.05097498744726181,
+      "learning_rate": 0.00012575219979277602,
+      "loss": 0.261,
+      "step": 2196
+    },
+    {
+      "epoch": 2.0861618798955615,
+      "grad_norm": 0.05032607540488243,
+      "learning_rate": 0.00012569453833831222,
+      "loss": 0.2769,
+      "step": 2197
+    },
+    {
+      "epoch": 2.0871113220982673,
+      "grad_norm": 0.04438967630267143,
+      "learning_rate": 0.00012563686773547675,
+      "loss": 0.2561,
+      "step": 2198
+    },
+    {
+      "epoch": 2.088060764300973,
+      "grad_norm": 0.05397673696279526,
+      "learning_rate": 0.00012557918800480282,
+      "loss": 0.2712,
+      "step": 2199
+    },
+    {
+      "epoch": 2.089010206503679,
+      "grad_norm": 0.05158831924200058,
+      "learning_rate": 0.00012552149916682695,
+      "loss": 0.2685,
+      "step": 2200
+    },
+    {
+      "epoch": 2.089959648706385,
+      "grad_norm": 0.06279024481773376,
+      "learning_rate": 0.00012546380124208887,
+      "loss": 0.2722,
+      "step": 2201
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 0.04665720462799072,
+      "learning_rate": 0.00012540609425113156,
+      "loss": 0.2604,
+      "step": 2202
+    },
+    {
+      "epoch": 2.0918585331117967,
+      "grad_norm": 0.059546615928411484,
+      "learning_rate": 0.00012534837821450117,
+      "loss": 0.2721,
+      "step": 2203
+    },
+    {
+      "epoch": 2.0928079753145026,
+      "grad_norm": 0.0592176578938961,
+      "learning_rate": 0.0001252906531527472,
+      "loss": 0.2716,
+      "step": 2204
+    },
+    {
+      "epoch": 2.0937574175172085,
+      "grad_norm": 0.04968995600938797,
+      "learning_rate": 0.00012523291908642217,
+      "loss": 0.2474,
+      "step": 2205
+    },
+    {
+      "epoch": 2.0947068597199143,
+      "grad_norm": 0.052708033472299576,
+      "learning_rate": 0.00012517517603608203,
+      "loss": 0.2668,
+      "step": 2206
+    },
+    {
+      "epoch": 2.0956563019226206,
+      "grad_norm": 0.06978727877140045,
+      "learning_rate": 0.0001251174240222857,
+      "loss": 0.2729,
+      "step": 2207
+    },
+    {
+      "epoch": 2.0966057441253265,
+      "grad_norm": 0.061792004853487015,
+      "learning_rate": 0.0001250596630655955,
+      "loss": 0.2706,
+      "step": 2208
+    },
+    {
+      "epoch": 2.0975551863280324,
+      "grad_norm": 0.05177111551165581,
+      "learning_rate": 0.00012500189318657675,
+      "loss": 0.2759,
+      "step": 2209
+    },
+    {
+      "epoch": 2.0985046285307383,
+      "grad_norm": 0.05225459113717079,
+      "learning_rate": 0.00012494411440579814,
+      "loss": 0.2662,
+      "step": 2210
+    },
+    {
+      "epoch": 2.099454070733444,
+      "grad_norm": 0.046468012034893036,
+      "learning_rate": 0.00012488632674383134,
+      "loss": 0.2712,
+      "step": 2211
+    },
+    {
+      "epoch": 2.10040351293615,
+      "grad_norm": 0.044963154941797256,
+      "learning_rate": 0.00012482853022125132,
+      "loss": 0.2685,
+      "step": 2212
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5265,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 158,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.2839911292489564e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}