diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,78503 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9997324056729997,
+  "eval_steps": 500,
+  "global_step": 11210,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001783962180001784,
+      "grad_norm": 7.837475776672363,
+      "learning_rate": 0.0004999999901825538,
+      "loss": 6.5872,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003567924360003568,
+      "grad_norm": 32.637916564941406,
+      "learning_rate": 0.000499999960730216,
+      "loss": 7.833,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005351886540005352,
+      "grad_norm": 19.442291259765625,
+      "learning_rate": 0.0004999999116429887,
+      "loss": 6.3418,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007135848720007136,
+      "grad_norm": 7.7614946365356445,
+      "learning_rate": 0.0004999998429208761,
+      "loss": 5.3582,
+      "step": 4
+    },
+    {
+      "epoch": 0.000891981090000892,
+      "grad_norm": 5.833192348480225,
+      "learning_rate": 0.0004999997545638834,
+      "loss": 4.4165,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010703773080010704,
+      "grad_norm": 2.9772305488586426,
+      "learning_rate": 0.0004999996465720175,
+      "loss": 4.1561,
+      "step": 6
+    },
+    {
+      "epoch": 0.0012487735260012487,
+      "grad_norm": 3.942150354385376,
+      "learning_rate": 0.0004999995189452869,
+      "loss": 4.0736,
+      "step": 7
+    },
+    {
+      "epoch": 0.0014271697440014271,
+      "grad_norm": 2.2194981575012207,
+      "learning_rate": 0.0004999993716837017,
+      "loss": 3.9339,
+      "step": 8
+    },
+    {
+      "epoch": 0.0016055659620016055,
+      "grad_norm": 2.6826322078704834,
+      "learning_rate": 0.0004999992047872735,
+      "loss": 4.0066,
+      "step": 9
+    },
+    {
+      "epoch": 0.001783962180001784,
+      "grad_norm": 2.3299083709716797,
+      "learning_rate": 0.0004999990182560153,
+      "loss": 3.9545,
+      "step": 10
+    },
+    {
+      "epoch": 0.0019623583980019625,
+      "grad_norm": 2.3098955154418945,
+      "learning_rate": 0.0004999988120899418,
+      "loss": 3.8027,
+      "step": 11
+    },
+    {
+      "epoch": 0.0021407546160021407,
+      "grad_norm": 1.8366882801055908,
+      "learning_rate": 0.0004999985862890691,
+      "loss": 3.8252,
+      "step": 12
+    },
+    {
+      "epoch": 0.0023191508340023193,
+      "grad_norm": 1.689456820487976,
+      "learning_rate": 0.0004999983408534151,
+      "loss": 3.6765,
+      "step": 13
+    },
+    {
+      "epoch": 0.0024975470520024975,
+      "grad_norm": 2.1124043464660645,
+      "learning_rate": 0.000499998075782999,
+      "loss": 3.7322,
+      "step": 14
+    },
+    {
+      "epoch": 0.002675943270002676,
+      "grad_norm": 1.5068916082382202,
+      "learning_rate": 0.0004999977910778417,
+      "loss": 3.7449,
+      "step": 15
+    },
+    {
+      "epoch": 0.0028543394880028543,
+      "grad_norm": 1.2359281778335571,
+      "learning_rate": 0.0004999974867379652,
+      "loss": 3.6367,
+      "step": 16
+    },
+    {
+      "epoch": 0.003032735706003033,
+      "grad_norm": 1.5040236711502075,
+      "learning_rate": 0.0004999971627633939,
+      "loss": 3.5622,
+      "step": 17
+    },
+    {
+      "epoch": 0.003211131924003211,
+      "grad_norm": 1.7240302562713623,
+      "learning_rate": 0.000499996819154153,
+      "loss": 3.4124,
+      "step": 18
+    },
+    {
+      "epoch": 0.0033895281420033897,
+      "grad_norm": 1.6077510118484497,
+      "learning_rate": 0.0004999964559102693,
+      "loss": 3.6737,
+      "step": 19
+    },
+    {
+      "epoch": 0.003567924360003568,
+      "grad_norm": 1.3428990840911865,
+      "learning_rate": 0.0004999960730317718,
+      "loss": 3.4728,
+      "step": 20
+    },
+    {
+      "epoch": 0.0037463205780037465,
+      "grad_norm": 1.5660542249679565,
+      "learning_rate": 0.0004999956705186902,
+      "loss": 3.2715,
+      "step": 21
+    },
+    {
+      "epoch": 0.003924716796003925,
+      "grad_norm": 1.5012935400009155,
+      "learning_rate": 0.0004999952483710562,
+      "loss": 3.6257,
+      "step": 22
+    },
+    {
+      "epoch": 0.004103113014004103,
+      "grad_norm": 1.3174484968185425,
+      "learning_rate": 0.0004999948065889029,
+      "loss": 3.5645,
+      "step": 23
+    },
+    {
+      "epoch": 0.004281509232004281,
+      "grad_norm": 1.1299705505371094,
+      "learning_rate": 0.0004999943451722653,
+      "loss": 3.2416,
+      "step": 24
+    },
+    {
+      "epoch": 0.00445990545000446,
+      "grad_norm": 0.8905799984931946,
+      "learning_rate": 0.0004999938641211792,
+      "loss": 3.2397,
+      "step": 25
+    },
+    {
+      "epoch": 0.004638301668004639,
+      "grad_norm": 1.1655199527740479,
+      "learning_rate": 0.0004999933634356826,
+      "loss": 3.1717,
+      "step": 26
+    },
+    {
+      "epoch": 0.004816697886004816,
+      "grad_norm": 1.191267967224121,
+      "learning_rate": 0.0004999928431158149,
+      "loss": 3.3958,
+      "step": 27
+    },
+    {
+      "epoch": 0.004995094104004995,
+      "grad_norm": 1.0441665649414062,
+      "learning_rate": 0.0004999923031616169,
+      "loss": 3.3732,
+      "step": 28
+    },
+    {
+      "epoch": 0.005173490322005174,
+      "grad_norm": 1.2923051118850708,
+      "learning_rate": 0.0004999917435731309,
+      "loss": 3.3872,
+      "step": 29
+    },
+    {
+      "epoch": 0.005351886540005352,
+      "grad_norm": 1.1339762210845947,
+      "learning_rate": 0.000499991164350401,
+      "loss": 2.9439,
+      "step": 30
+    },
+    {
+      "epoch": 0.00553028275800553,
+      "grad_norm": 0.8818377256393433,
+      "learning_rate": 0.0004999905654934726,
+      "loss": 3.1922,
+      "step": 31
+    },
+    {
+      "epoch": 0.0057086789760057086,
+      "grad_norm": 0.7961766123771667,
+      "learning_rate": 0.0004999899470023929,
+      "loss": 3.2558,
+      "step": 32
+    },
+    {
+      "epoch": 0.005887075194005887,
+      "grad_norm": 1.180748701095581,
+      "learning_rate": 0.0004999893088772102,
+      "loss": 3.2286,
+      "step": 33
+    },
+    {
+      "epoch": 0.006065471412006066,
+      "grad_norm": 1.035274624824524,
+      "learning_rate": 0.0004999886511179748,
+      "loss": 3.4693,
+      "step": 34
+    },
+    {
+      "epoch": 0.0062438676300062435,
+      "grad_norm": 1.15408194065094,
+      "learning_rate": 0.0004999879737247382,
+      "loss": 3.2316,
+      "step": 35
+    },
+    {
+      "epoch": 0.006422263848006422,
+      "grad_norm": 0.8637914657592773,
+      "learning_rate": 0.0004999872766975539,
+      "loss": 2.9908,
+      "step": 36
+    },
+    {
+      "epoch": 0.006600660066006601,
+      "grad_norm": 1.0279821157455444,
+      "learning_rate": 0.0004999865600364764,
+      "loss": 3.1329,
+      "step": 37
+    },
+    {
+      "epoch": 0.006779056284006779,
+      "grad_norm": 0.9054232239723206,
+      "learning_rate": 0.000499985823741562,
+      "loss": 3.2613,
+      "step": 38
+    },
+    {
+      "epoch": 0.006957452502006957,
+      "grad_norm": 0.883553683757782,
+      "learning_rate": 0.0004999850678128687,
+      "loss": 3.0738,
+      "step": 39
+    },
+    {
+      "epoch": 0.007135848720007136,
+      "grad_norm": 0.9399454593658447,
+      "learning_rate": 0.0004999842922504556,
+      "loss": 3.2061,
+      "step": 40
+    },
+    {
+      "epoch": 0.007314244938007314,
+      "grad_norm": 1.0862022638320923,
+      "learning_rate": 0.0004999834970543839,
+      "loss": 3.0835,
+      "step": 41
+    },
+    {
+      "epoch": 0.007492641156007493,
+      "grad_norm": 0.9416713118553162,
+      "learning_rate": 0.0004999826822247159,
+      "loss": 3.0841,
+      "step": 42
+    },
+    {
+      "epoch": 0.007671037374007671,
+      "grad_norm": 1.0028291940689087,
+      "learning_rate": 0.0004999818477615155,
+      "loss": 3.0349,
+      "step": 43
+    },
+    {
+      "epoch": 0.00784943359200785,
+      "grad_norm": 0.8068011403083801,
+      "learning_rate": 0.0004999809936648484,
+      "loss": 3.0634,
+      "step": 44
+    },
+    {
+      "epoch": 0.008027829810008028,
+      "grad_norm": 2.0067076683044434,
+      "learning_rate": 0.0004999801199347817,
+      "loss": 3.2768,
+      "step": 45
+    },
+    {
+      "epoch": 0.008206226028008206,
+      "grad_norm": 0.8227775692939758,
+      "learning_rate": 0.000499979226571384,
+      "loss": 3.2637,
+      "step": 46
+    },
+    {
+      "epoch": 0.008384622246008385,
+      "grad_norm": 0.9798381924629211,
+      "learning_rate": 0.0004999783135747252,
+      "loss": 2.9619,
+      "step": 47
+    },
+    {
+      "epoch": 0.008563018464008563,
+      "grad_norm": 0.9678353071212769,
+      "learning_rate": 0.0004999773809448774,
+      "loss": 2.7488,
+      "step": 48
+    },
+    {
+      "epoch": 0.00874141468200874,
+      "grad_norm": 1.023210048675537,
+      "learning_rate": 0.0004999764286819137,
+      "loss": 3.1175,
+      "step": 49
+    },
+    {
+      "epoch": 0.00891981090000892,
+      "grad_norm": 0.892423689365387,
+      "learning_rate": 0.0004999754567859087,
+      "loss": 2.7036,
+      "step": 50
+    },
+    {
+      "epoch": 0.009098207118009098,
+      "grad_norm": 1.0287269353866577,
+      "learning_rate": 0.000499974465256939,
+      "loss": 2.892,
+      "step": 51
+    },
+    {
+      "epoch": 0.009276603336009277,
+      "grad_norm": 1.1491647958755493,
+      "learning_rate": 0.0004999734540950824,
+      "loss": 2.9976,
+      "step": 52
+    },
+    {
+      "epoch": 0.009454999554009455,
+      "grad_norm": 1.1946396827697754,
+      "learning_rate": 0.0004999724233004183,
+      "loss": 2.8391,
+      "step": 53
+    },
+    {
+      "epoch": 0.009633395772009633,
+      "grad_norm": 1.0518914461135864,
+      "learning_rate": 0.0004999713728730276,
+      "loss": 2.7981,
+      "step": 54
+    },
+    {
+      "epoch": 0.009811791990009812,
+      "grad_norm": 0.9127726554870605,
+      "learning_rate": 0.0004999703028129929,
+      "loss": 3.0045,
+      "step": 55
+    },
+    {
+      "epoch": 0.00999018820800999,
+      "grad_norm": 1.4292141199111938,
+      "learning_rate": 0.000499969213120398,
+      "loss": 2.8518,
+      "step": 56
+    },
+    {
+      "epoch": 0.010168584426010168,
+      "grad_norm": 1.256000280380249,
+      "learning_rate": 0.0004999681037953288,
+      "loss": 2.8525,
+      "step": 57
+    },
+    {
+      "epoch": 0.010346980644010347,
+      "grad_norm": 1.166292428970337,
+      "learning_rate": 0.0004999669748378723,
+      "loss": 2.9721,
+      "step": 58
+    },
+    {
+      "epoch": 0.010525376862010525,
+      "grad_norm": 1.197026014328003,
+      "learning_rate": 0.0004999658262481172,
+      "loss": 2.6609,
+      "step": 59
+    },
+    {
+      "epoch": 0.010703773080010704,
+      "grad_norm": 1.027642846107483,
+      "learning_rate": 0.0004999646580261537,
+      "loss": 2.7145,
+      "step": 60
+    },
+    {
+      "epoch": 0.010882169298010882,
+      "grad_norm": 1.0411521196365356,
+      "learning_rate": 0.0004999634701720734,
+      "loss": 2.895,
+      "step": 61
+    },
+    {
+      "epoch": 0.01106056551601106,
+      "grad_norm": 1.0868782997131348,
+      "learning_rate": 0.0004999622626859699,
+      "loss": 2.8614,
+      "step": 62
+    },
+    {
+      "epoch": 0.01123896173401124,
+      "grad_norm": 1.0062867403030396,
+      "learning_rate": 0.0004999610355679377,
+      "loss": 2.4979,
+      "step": 63
+    },
+    {
+      "epoch": 0.011417357952011417,
+      "grad_norm": 1.0268216133117676,
+      "learning_rate": 0.0004999597888180734,
+      "loss": 2.9,
+      "step": 64
+    },
+    {
+      "epoch": 0.011595754170011597,
+      "grad_norm": 0.9433121085166931,
+      "learning_rate": 0.0004999585224364748,
+      "loss": 3.0019,
+      "step": 65
+    },
+    {
+      "epoch": 0.011774150388011774,
+      "grad_norm": 0.9315336346626282,
+      "learning_rate": 0.0004999572364232414,
+      "loss": 2.7774,
+      "step": 66
+    },
+    {
+      "epoch": 0.011952546606011952,
+      "grad_norm": 0.8882467746734619,
+      "learning_rate": 0.0004999559307784743,
+      "loss": 2.7762,
+      "step": 67
+    },
+    {
+      "epoch": 0.012130942824012132,
+      "grad_norm": 0.9650992751121521,
+      "learning_rate": 0.000499954605502276,
+      "loss": 2.5736,
+      "step": 68
+    },
+    {
+      "epoch": 0.01230933904201231,
+      "grad_norm": 1.0080440044403076,
+      "learning_rate": 0.0004999532605947505,
+      "loss": 2.8355,
+      "step": 69
+    },
+    {
+      "epoch": 0.012487735260012487,
+      "grad_norm": 1.0923867225646973,
+      "learning_rate": 0.0004999518960560034,
+      "loss": 2.5415,
+      "step": 70
+    },
+    {
+      "epoch": 0.012666131478012667,
+      "grad_norm": 0.8062555193901062,
+      "learning_rate": 0.000499950511886142,
+      "loss": 2.1503,
+      "step": 71
+    },
+    {
+      "epoch": 0.012844527696012844,
+      "grad_norm": 1.1151552200317383,
+      "learning_rate": 0.000499949108085275,
+      "loss": 2.9247,
+      "step": 72
+    },
+    {
+      "epoch": 0.013022923914013024,
+      "grad_norm": 0.7492188811302185,
+      "learning_rate": 0.0004999476846535125,
+      "loss": 2.577,
+      "step": 73
+    },
+    {
+      "epoch": 0.013201320132013201,
+      "grad_norm": 1.0736455917358398,
+      "learning_rate": 0.0004999462415909664,
+      "loss": 2.7738,
+      "step": 74
+    },
+    {
+      "epoch": 0.01337971635001338,
+      "grad_norm": 0.9508681297302246,
+      "learning_rate": 0.0004999447788977502,
+      "loss": 2.4073,
+      "step": 75
+    },
+    {
+      "epoch": 0.013558112568013559,
+      "grad_norm": 0.8389919996261597,
+      "learning_rate": 0.0004999432965739786,
+      "loss": 2.7485,
+      "step": 76
+    },
+    {
+      "epoch": 0.013736508786013736,
+      "grad_norm": 1.055034875869751,
+      "learning_rate": 0.0004999417946197679,
+      "loss": 2.5765,
+      "step": 77
+    },
+    {
+      "epoch": 0.013914905004013914,
+      "grad_norm": 1.1897584199905396,
+      "learning_rate": 0.0004999402730352363,
+      "loss": 2.7678,
+      "step": 78
+    },
+    {
+      "epoch": 0.014093301222014094,
+      "grad_norm": 0.9991359710693359,
+      "learning_rate": 0.0004999387318205032,
+      "loss": 2.7551,
+      "step": 79
+    },
+    {
+      "epoch": 0.014271697440014271,
+      "grad_norm": 0.9848081469535828,
+      "learning_rate": 0.0004999371709756897,
+      "loss": 2.5171,
+      "step": 80
+    },
+    {
+      "epoch": 0.014450093658014451,
+      "grad_norm": 1.1168460845947266,
+      "learning_rate": 0.0004999355905009183,
+      "loss": 2.3899,
+      "step": 81
+    },
+    {
+      "epoch": 0.014628489876014629,
+      "grad_norm": 0.9936979413032532,
+      "learning_rate": 0.0004999339903963133,
+      "loss": 2.6265,
+      "step": 82
+    },
+    {
+      "epoch": 0.014806886094014806,
+      "grad_norm": 1.0733612775802612,
+      "learning_rate": 0.0004999323706620001,
+      "loss": 2.5981,
+      "step": 83
+    },
+    {
+      "epoch": 0.014985282312014986,
+      "grad_norm": 1.0748244524002075,
+      "learning_rate": 0.000499930731298106,
+      "loss": 2.5156,
+      "step": 84
+    },
+    {
+      "epoch": 0.015163678530015164,
+      "grad_norm": 0.8287347555160522,
+      "learning_rate": 0.00049992907230476,
+      "loss": 2.0732,
+      "step": 85
+    },
+    {
+      "epoch": 0.015342074748015341,
+      "grad_norm": 0.8893202543258667,
+      "learning_rate": 0.0004999273936820922,
+      "loss": 2.4538,
+      "step": 86
+    },
+    {
+      "epoch": 0.01552047096601552,
+      "grad_norm": 0.8320645689964294,
+      "learning_rate": 0.0004999256954302344,
+      "loss": 2.5106,
+      "step": 87
+    },
+    {
+      "epoch": 0.0156988671840157,
+      "grad_norm": 0.9351370930671692,
+      "learning_rate": 0.0004999239775493199,
+      "loss": 2.529,
+      "step": 88
+    },
+    {
+      "epoch": 0.015877263402015878,
+      "grad_norm": 0.8308570981025696,
+      "learning_rate": 0.0004999222400394839,
+      "loss": 2.234,
+      "step": 89
+    },
+    {
+      "epoch": 0.016055659620016056,
+      "grad_norm": 1.0600675344467163,
+      "learning_rate": 0.0004999204829008628,
+      "loss": 2.5008,
+      "step": 90
+    },
+    {
+      "epoch": 0.016234055838016234,
+      "grad_norm": 0.9695082902908325,
+      "learning_rate": 0.0004999187061335943,
+      "loss": 2.3691,
+      "step": 91
+    },
+    {
+      "epoch": 0.01641245205601641,
+      "grad_norm": 0.9149223566055298,
+      "learning_rate": 0.0004999169097378184,
+      "loss": 2.5968,
+      "step": 92
+    },
+    {
+      "epoch": 0.016590848274016592,
+      "grad_norm": 0.9797092080116272,
+      "learning_rate": 0.0004999150937136758,
+      "loss": 2.535,
+      "step": 93
+    },
+    {
+      "epoch": 0.01676924449201677,
+      "grad_norm": 0.7779546976089478,
+      "learning_rate": 0.0004999132580613094,
+      "loss": 2.0829,
+      "step": 94
+    },
+    {
+      "epoch": 0.016947640710016948,
+      "grad_norm": 0.8772634863853455,
+      "learning_rate": 0.0004999114027808632,
+      "loss": 2.2751,
+      "step": 95
+    },
+    {
+      "epoch": 0.017126036928017126,
+      "grad_norm": 0.828109860420227,
+      "learning_rate": 0.0004999095278724829,
+      "loss": 2.5915,
+      "step": 96
+    },
+    {
+      "epoch": 0.017304433146017303,
+      "grad_norm": 1.0454559326171875,
+      "learning_rate": 0.0004999076333363159,
+      "loss": 2.3756,
+      "step": 97
+    },
+    {
+      "epoch": 0.01748282936401748,
+      "grad_norm": 1.0165445804595947,
+      "learning_rate": 0.000499905719172511,
+      "loss": 2.3156,
+      "step": 98
+    },
+    {
+      "epoch": 0.017661225582017662,
+      "grad_norm": 0.9280771017074585,
+      "learning_rate": 0.0004999037853812183,
+      "loss": 2.4135,
+      "step": 99
+    },
+    {
+      "epoch": 0.01783962180001784,
+      "grad_norm": 0.9677728414535522,
+      "learning_rate": 0.0004999018319625898,
+      "loss": 2.2452,
+      "step": 100
+    },
+    {
+      "epoch": 0.018018018018018018,
+      "grad_norm": 0.9852823615074158,
+      "learning_rate": 0.0004998998589167791,
+      "loss": 2.1909,
+      "step": 101
+    },
+    {
+      "epoch": 0.018196414236018196,
+      "grad_norm": 1.07912278175354,
+      "learning_rate": 0.0004998978662439411,
+      "loss": 2.1278,
+      "step": 102
+    },
+    {
+      "epoch": 0.018374810454018373,
+      "grad_norm": 0.9431922435760498,
+      "learning_rate": 0.000499895853944232,
+      "loss": 2.2322,
+      "step": 103
+    },
+    {
+      "epoch": 0.018553206672018555,
+      "grad_norm": 0.8767147064208984,
+      "learning_rate": 0.0004998938220178102,
+      "loss": 2.1025,
+      "step": 104
+    },
+    {
+      "epoch": 0.018731602890018732,
+      "grad_norm": 1.079720377922058,
+      "learning_rate": 0.0004998917704648352,
+      "loss": 2.5964,
+      "step": 105
+    },
+    {
+      "epoch": 0.01890999910801891,
+      "grad_norm": 1.181444764137268,
+      "learning_rate": 0.000499889699285468,
+      "loss": 2.1149,
+      "step": 106
+    },
+    {
+      "epoch": 0.019088395326019088,
+      "grad_norm": 1.055641770362854,
+      "learning_rate": 0.0004998876084798714,
+      "loss": 2.3814,
+      "step": 107
+    },
+    {
+      "epoch": 0.019266791544019266,
+      "grad_norm": 0.9991637468338013,
+      "learning_rate": 0.0004998854980482095,
+      "loss": 2.4501,
+      "step": 108
+    },
+    {
+      "epoch": 0.019445187762019447,
+      "grad_norm": 0.878653883934021,
+      "learning_rate": 0.0004998833679906482,
+      "loss": 2.2617,
+      "step": 109
+    },
+    {
+      "epoch": 0.019623583980019624,
+      "grad_norm": 0.9334123134613037,
+      "learning_rate": 0.0004998812183073547,
+      "loss": 2.325,
+      "step": 110
+    },
+    {
+      "epoch": 0.019801980198019802,
+      "grad_norm": 0.7886718511581421,
+      "learning_rate": 0.0004998790489984978,
+      "loss": 2.1459,
+      "step": 111
+    },
+    {
+      "epoch": 0.01998037641601998,
+      "grad_norm": 0.8349006175994873,
+      "learning_rate": 0.0004998768600642479,
+      "loss": 2.3084,
+      "step": 112
+    },
+    {
+      "epoch": 0.020158772634020158,
+      "grad_norm": 0.9214372038841248,
+      "learning_rate": 0.000499874651504777,
+      "loss": 2.3816,
+      "step": 113
+    },
+    {
+      "epoch": 0.020337168852020335,
+      "grad_norm": 0.8497713804244995,
+      "learning_rate": 0.0004998724233202585,
+      "loss": 2.2302,
+      "step": 114
+    },
+    {
+      "epoch": 0.020515565070020517,
+      "grad_norm": 0.7643554210662842,
+      "learning_rate": 0.0004998701755108674,
+      "loss": 2.3578,
+      "step": 115
+    },
+    {
+      "epoch": 0.020693961288020694,
+      "grad_norm": 1.057411551475525,
+      "learning_rate": 0.0004998679080767802,
+      "loss": 2.1724,
+      "step": 116
+    },
+    {
+      "epoch": 0.020872357506020872,
+      "grad_norm": 1.0113954544067383,
+      "learning_rate": 0.000499865621018175,
+      "loss": 2.375,
+      "step": 117
+    },
+    {
+      "epoch": 0.02105075372402105,
+      "grad_norm": 0.8105971813201904,
+      "learning_rate": 0.0004998633143352315,
+      "loss": 2.0368,
+      "step": 118
+    },
+    {
+      "epoch": 0.021229149942021228,
+      "grad_norm": 1.2691274881362915,
+      "learning_rate": 0.0004998609880281309,
+      "loss": 2.2663,
+      "step": 119
+    },
+    {
+      "epoch": 0.02140754616002141,
+      "grad_norm": 0.8678051829338074,
+      "learning_rate": 0.0004998586420970557,
+      "loss": 2.0612,
+      "step": 120
+    },
+    {
+      "epoch": 0.021585942378021587,
+      "grad_norm": 0.7809403538703918,
+      "learning_rate": 0.0004998562765421903,
+      "loss": 2.0222,
+      "step": 121
+    },
+    {
+      "epoch": 0.021764338596021764,
+      "grad_norm": 0.8958665728569031,
+      "learning_rate": 0.0004998538913637205,
+      "loss": 2.2086,
+      "step": 122
+    },
+    {
+      "epoch": 0.021942734814021942,
+      "grad_norm": 0.8660983443260193,
+      "learning_rate": 0.0004998514865618335,
+      "loss": 2.0885,
+      "step": 123
+    },
+    {
+      "epoch": 0.02212113103202212,
+      "grad_norm": 0.8884083032608032,
+      "learning_rate": 0.0004998490621367184,
+      "loss": 2.0391,
+      "step": 124
+    },
+    {
+      "epoch": 0.0222995272500223,
+      "grad_norm": 0.9233642220497131,
+      "learning_rate": 0.0004998466180885653,
+      "loss": 2.3625,
+      "step": 125
+    },
+    {
+      "epoch": 0.02247792346802248,
+      "grad_norm": 0.9836124181747437,
+      "learning_rate": 0.0004998441544175666,
+      "loss": 2.2373,
+      "step": 126
+    },
+    {
+      "epoch": 0.022656319686022657,
+      "grad_norm": 0.8701616525650024,
+      "learning_rate": 0.0004998416711239153,
+      "loss": 2.2678,
+      "step": 127
+    },
+    {
+      "epoch": 0.022834715904022834,
+      "grad_norm": 0.8650346994400024,
+      "learning_rate": 0.0004998391682078067,
+      "loss": 2.1102,
+      "step": 128
+    },
+    {
+      "epoch": 0.023013112122023012,
+      "grad_norm": 1.1175918579101562,
+      "learning_rate": 0.0004998366456694374,
+      "loss": 2.1723,
+      "step": 129
+    },
+    {
+      "epoch": 0.023191508340023193,
+      "grad_norm": 0.8660110235214233,
+      "learning_rate": 0.0004998341035090055,
+      "loss": 2.0303,
+      "step": 130
+    },
+    {
+      "epoch": 0.02336990455802337,
+      "grad_norm": 0.7795286774635315,
+      "learning_rate": 0.0004998315417267105,
+      "loss": 2.2806,
+      "step": 131
+    },
+    {
+      "epoch": 0.02354830077602355,
+      "grad_norm": 0.8152791857719421,
+      "learning_rate": 0.0004998289603227538,
+      "loss": 2.0361,
+      "step": 132
+    },
+    {
+      "epoch": 0.023726696994023726,
+      "grad_norm": 0.8521414995193481,
+      "learning_rate": 0.0004998263592973381,
+      "loss": 2.3676,
+      "step": 133
+    },
+    {
+      "epoch": 0.023905093212023904,
+      "grad_norm": 0.8271788954734802,
+      "learning_rate": 0.0004998237386506676,
+      "loss": 2.2346,
+      "step": 134
+    },
+    {
+      "epoch": 0.024083489430024082,
+      "grad_norm": 0.8122360706329346,
+      "learning_rate": 0.0004998210983829482,
+      "loss": 2.2109,
+      "step": 135
+    },
+    {
+      "epoch": 0.024261885648024263,
+      "grad_norm": 0.8585101962089539,
+      "learning_rate": 0.0004998184384943871,
+      "loss": 2.2372,
+      "step": 136
+    },
+    {
+      "epoch": 0.02444028186602444,
+      "grad_norm": 1.338280200958252,
+      "learning_rate": 0.0004998157589851935,
+      "loss": 2.3498,
+      "step": 137
+    },
+    {
+      "epoch": 0.02461867808402462,
+      "grad_norm": 0.8073638677597046,
+      "learning_rate": 0.0004998130598555776,
+      "loss": 1.9356,
+      "step": 138
+    },
+    {
+      "epoch": 0.024797074302024796,
+      "grad_norm": 1.4015569686889648,
+      "learning_rate": 0.0004998103411057517,
+      "loss": 2.12,
+      "step": 139
+    },
+    {
+      "epoch": 0.024975470520024974,
+      "grad_norm": 3.684471845626831,
+      "learning_rate": 0.0004998076027359289,
+      "loss": 2.1707,
+      "step": 140
+    },
+    {
+      "epoch": 0.025153866738025155,
+      "grad_norm": 1.0060230493545532,
+      "learning_rate": 0.0004998048447463245,
+      "loss": 2.1753,
+      "step": 141
+    },
+    {
+      "epoch": 0.025332262956025333,
+      "grad_norm": 0.8520100116729736,
+      "learning_rate": 0.0004998020671371551,
+      "loss": 2.2941,
+      "step": 142
+    },
+    {
+      "epoch": 0.02551065917402551,
+      "grad_norm": 0.8406434059143066,
+      "learning_rate": 0.0004997992699086389,
+      "loss": 1.9893,
+      "step": 143
+    },
+    {
+      "epoch": 0.02568905539202569,
+      "grad_norm": 1.1668790578842163,
+      "learning_rate": 0.0004997964530609956,
+      "loss": 2.0301,
+      "step": 144
+    },
+    {
+      "epoch": 0.025867451610025866,
+      "grad_norm": 0.8116411566734314,
+      "learning_rate": 0.0004997936165944462,
+      "loss": 1.976,
+      "step": 145
+    },
+    {
+      "epoch": 0.026045847828026047,
+      "grad_norm": 1.0072658061981201,
+      "learning_rate": 0.0004997907605092138,
+      "loss": 2.0916,
+      "step": 146
+    },
+    {
+      "epoch": 0.026224244046026225,
+      "grad_norm": 1.1760748624801636,
+      "learning_rate": 0.0004997878848055225,
+      "loss": 1.9554,
+      "step": 147
+    },
+    {
+      "epoch": 0.026402640264026403,
+      "grad_norm": 0.9567267894744873,
+      "learning_rate": 0.0004997849894835982,
+      "loss": 2.1948,
+      "step": 148
+    },
+    {
+      "epoch": 0.02658103648202658,
+      "grad_norm": 1.109018325805664,
+      "learning_rate": 0.0004997820745436683,
+      "loss": 1.9971,
+      "step": 149
+    },
+    {
+      "epoch": 0.02675943270002676,
+      "grad_norm": 0.9176613092422485,
+      "learning_rate": 0.0004997791399859618,
+      "loss": 2.4487,
+      "step": 150
+    },
+    {
+      "epoch": 0.026937828918026936,
+      "grad_norm": 0.8620409369468689,
+      "learning_rate": 0.0004997761858107091,
+      "loss": 2.1291,
+      "step": 151
+    },
+    {
+      "epoch": 0.027116225136027117,
+      "grad_norm": 0.7816916108131409,
+      "learning_rate": 0.0004997732120181423,
+      "loss": 2.1157,
+      "step": 152
+    },
+    {
+      "epoch": 0.027294621354027295,
+      "grad_norm": 0.8684608340263367,
+      "learning_rate": 0.0004997702186084949,
+      "loss": 2.0322,
+      "step": 153
+    },
+    {
+      "epoch": 0.027473017572027473,
+      "grad_norm": 0.9221858382225037,
+      "learning_rate": 0.000499767205582002,
+      "loss": 1.9962,
+      "step": 154
+    },
+    {
+      "epoch": 0.02765141379002765,
+      "grad_norm": 0.9276618361473083,
+      "learning_rate": 0.0004997641729389002,
+      "loss": 2.2969,
+      "step": 155
+    },
+    {
+      "epoch": 0.02782981000802783,
+      "grad_norm": 0.9027115106582642,
+      "learning_rate": 0.0004997611206794278,
+      "loss": 2.1261,
+      "step": 156
+    },
+    {
+      "epoch": 0.02800820622602801,
+      "grad_norm": 0.7453658580780029,
+      "learning_rate": 0.0004997580488038245,
+      "loss": 1.9345,
+      "step": 157
+    },
+    {
+      "epoch": 0.028186602444028187,
+      "grad_norm": 0.8788852095603943,
+      "learning_rate": 0.0004997549573123314,
+      "loss": 2.096,
+      "step": 158
+    },
+    {
+      "epoch": 0.028364998662028365,
+      "grad_norm": 0.8055654764175415,
+      "learning_rate": 0.0004997518462051916,
+      "loss": 1.8397,
+      "step": 159
+    },
+    {
+      "epoch": 0.028543394880028543,
+      "grad_norm": 0.9946992993354797,
+      "learning_rate": 0.0004997487154826492,
+      "loss": 1.8003,
+      "step": 160
+    },
+    {
+      "epoch": 0.02872179109802872,
+      "grad_norm": 0.9577097296714783,
+      "learning_rate": 0.0004997455651449502,
+      "loss": 2.1386,
+      "step": 161
+    },
+    {
+      "epoch": 0.028900187316028902,
+      "grad_norm": 1.0292500257492065,
+      "learning_rate": 0.0004997423951923419,
+      "loss": 2.0328,
+      "step": 162
+    },
+    {
+      "epoch": 0.02907858353402908,
+      "grad_norm": 0.8872460126876831,
+      "learning_rate": 0.0004997392056250733,
+      "loss": 1.9567,
+      "step": 163
+    },
+    {
+      "epoch": 0.029256979752029257,
+      "grad_norm": 4.496621608734131,
+      "learning_rate": 0.0004997359964433952,
+      "loss": 2.2306,
+      "step": 164
+    },
+    {
+      "epoch": 0.029435375970029435,
+      "grad_norm": 1.3478816747665405,
+      "learning_rate": 0.0004997327676475593,
+      "loss": 1.9389,
+      "step": 165
+    },
+    {
+      "epoch": 0.029613772188029613,
+      "grad_norm": 1.18978750705719,
+      "learning_rate": 0.0004997295192378192,
+      "loss": 1.9183,
+      "step": 166
+    },
+    {
+      "epoch": 0.02979216840602979,
+      "grad_norm": 0.8998153805732727,
+      "learning_rate": 0.0004997262512144302,
+      "loss": 2.0974,
+      "step": 167
+    },
+    {
+      "epoch": 0.02997056462402997,
+      "grad_norm": 0.9464337229728699,
+      "learning_rate": 0.0004997229635776488,
+      "loss": 1.8474,
+      "step": 168
+    },
+    {
+      "epoch": 0.03014896084203015,
+      "grad_norm": 0.8955067992210388,
+      "learning_rate": 0.0004997196563277334,
+      "loss": 2.0262,
+      "step": 169
+    },
+    {
+      "epoch": 0.030327357060030327,
+      "grad_norm": 0.727933943271637,
+      "learning_rate": 0.0004997163294649437,
+      "loss": 1.7326,
+      "step": 170
+    },
+    {
+      "epoch": 0.030505753278030505,
+      "grad_norm": 1.0948885679244995,
+      "learning_rate": 0.0004997129829895409,
+      "loss": 2.1483,
+      "step": 171
+    },
+    {
+      "epoch": 0.030684149496030683,
+      "grad_norm": 0.9044740200042725,
+      "learning_rate": 0.0004997096169017879,
+      "loss": 1.8832,
+      "step": 172
+    },
+    {
+      "epoch": 0.030862545714030864,
+      "grad_norm": 1.0726819038391113,
+      "learning_rate": 0.0004997062312019489,
+      "loss": 1.9131,
+      "step": 173
+    },
+    {
+      "epoch": 0.03104094193203104,
+      "grad_norm": 0.9715486764907837,
+      "learning_rate": 0.0004997028258902902,
+      "loss": 2.0829,
+      "step": 174
+    },
+    {
+      "epoch": 0.03121933815003122,
+      "grad_norm": 0.8400108218193054,
+      "learning_rate": 0.0004996994009670788,
+      "loss": 1.9984,
+      "step": 175
+    },
+    {
+      "epoch": 0.0313977343680314,
+      "grad_norm": 0.9738950729370117,
+      "learning_rate": 0.000499695956432584,
+      "loss": 1.9813,
+      "step": 176
+    },
+    {
+      "epoch": 0.03157613058603158,
+      "grad_norm": 1.125439167022705,
+      "learning_rate": 0.0004996924922870762,
+      "loss": 1.8655,
+      "step": 177
+    },
+    {
+      "epoch": 0.031754526804031756,
+      "grad_norm": 1.0050427913665771,
+      "learning_rate": 0.0004996890085308275,
+      "loss": 2.1677,
+      "step": 178
+    },
+    {
+      "epoch": 0.031932923022031934,
+      "grad_norm": 0.7742742300033569,
+      "learning_rate": 0.0004996855051641116,
+      "loss": 2.0985,
+      "step": 179
+    },
+    {
+      "epoch": 0.03211131924003211,
+      "grad_norm": 1.3308488130569458,
+      "learning_rate": 0.0004996819821872035,
+      "loss": 1.7866,
+      "step": 180
+    },
+    {
+      "epoch": 0.03228971545803229,
+      "grad_norm": 0.776884913444519,
+      "learning_rate": 0.00049967843960038,
+      "loss": 1.8104,
+      "step": 181
+    },
+    {
+      "epoch": 0.03246811167603247,
+      "grad_norm": 0.8655677437782288,
+      "learning_rate": 0.0004996748774039192,
+      "loss": 2.0098,
+      "step": 182
+    },
+    {
+      "epoch": 0.032646507894032645,
+      "grad_norm": 0.7239806652069092,
+      "learning_rate": 0.000499671295598101,
+      "loss": 1.6073,
+      "step": 183
+    },
+    {
+      "epoch": 0.03282490411203282,
+      "grad_norm": 0.8207644820213318,
+      "learning_rate": 0.0004996676941832069,
+      "loss": 1.8347,
+      "step": 184
+    },
+    {
+      "epoch": 0.033003300330033,
+      "grad_norm": 0.7517510652542114,
+      "learning_rate": 0.0004996640731595194,
+      "loss": 2.0775,
+      "step": 185
+    },
+    {
+      "epoch": 0.033181696548033185,
+      "grad_norm": 0.7683122158050537,
+      "learning_rate": 0.000499660432527323,
+      "loss": 1.9587,
+      "step": 186
+    },
+    {
+      "epoch": 0.03336009276603336,
+      "grad_norm": 0.7496309280395508,
+      "learning_rate": 0.0004996567722869036,
+      "loss": 1.9149,
+      "step": 187
+    },
+    {
+      "epoch": 0.03353848898403354,
+      "grad_norm": 0.7267098426818848,
+      "learning_rate": 0.0004996530924385489,
+      "loss": 2.1301,
+      "step": 188
+    },
+    {
+      "epoch": 0.03371688520203372,
+      "grad_norm": 1.1069225072860718,
+      "learning_rate": 0.0004996493929825477,
+      "loss": 1.74,
+      "step": 189
+    },
+    {
+      "epoch": 0.033895281420033896,
+      "grad_norm": 0.795403778553009,
+      "learning_rate": 0.0004996456739191905,
+      "loss": 2.1685,
+      "step": 190
+    },
+    {
+      "epoch": 0.034073677638034074,
+      "grad_norm": 0.7450090050697327,
+      "learning_rate": 0.0004996419352487696,
+      "loss": 2.1011,
+      "step": 191
+    },
+    {
+      "epoch": 0.03425207385603425,
+      "grad_norm": 0.7762404084205627,
+      "learning_rate": 0.0004996381769715785,
+      "loss": 2.0786,
+      "step": 192
+    },
+    {
+      "epoch": 0.03443047007403443,
+      "grad_norm": 0.7745615839958191,
+      "learning_rate": 0.0004996343990879125,
+      "loss": 1.6322,
+      "step": 193
+    },
+    {
+      "epoch": 0.03460886629203461,
+      "grad_norm": 0.8149433732032776,
+      "learning_rate": 0.0004996306015980681,
+      "loss": 1.646,
+      "step": 194
+    },
+    {
+      "epoch": 0.034787262510034785,
+      "grad_norm": 0.8039779663085938,
+      "learning_rate": 0.0004996267845023437,
+      "loss": 1.6153,
+      "step": 195
+    },
+    {
+      "epoch": 0.03496565872803496,
+      "grad_norm": 0.7436607480049133,
+      "learning_rate": 0.0004996229478010392,
+      "loss": 1.9986,
+      "step": 196
+    },
+    {
+      "epoch": 0.03514405494603515,
+      "grad_norm": 1.0629464387893677,
+      "learning_rate": 0.0004996190914944556,
+      "loss": 1.7696,
+      "step": 197
+    },
+    {
+      "epoch": 0.035322451164035325,
+      "grad_norm": 0.7059940099716187,
+      "learning_rate": 0.0004996152155828961,
+      "loss": 1.75,
+      "step": 198
+    },
+    {
+      "epoch": 0.0355008473820355,
+      "grad_norm": 0.7760869860649109,
+      "learning_rate": 0.0004996113200666649,
+      "loss": 1.6804,
+      "step": 199
+    },
+    {
+      "epoch": 0.03567924360003568,
+      "grad_norm": 0.723596453666687,
+      "learning_rate": 0.000499607404946068,
+      "loss": 1.7709,
+      "step": 200
+    },
+    {
+      "epoch": 0.03585763981803586,
+      "grad_norm": 0.7342411279678345,
+      "learning_rate": 0.0004996034702214131,
+      "loss": 2.1074,
+      "step": 201
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 1.2572263479232788,
+      "learning_rate": 0.000499599515893009,
+      "loss": 1.8228,
+      "step": 202
+    },
+    {
+      "epoch": 0.03621443225403621,
+      "grad_norm": 0.9054358005523682,
+      "learning_rate": 0.0004995955419611663,
+      "loss": 2.0835,
+      "step": 203
+    },
+    {
+      "epoch": 0.03639282847203639,
+      "grad_norm": 0.8486645817756653,
+      "learning_rate": 0.0004995915484261971,
+      "loss": 2.1984,
+      "step": 204
+    },
+    {
+      "epoch": 0.03657122469003657,
+      "grad_norm": 0.7532051801681519,
+      "learning_rate": 0.0004995875352884152,
+      "loss": 1.9638,
+      "step": 205
+    },
+    {
+      "epoch": 0.03674962090803675,
+      "grad_norm": 0.9060899615287781,
+      "learning_rate": 0.0004995835025481357,
+      "loss": 1.8939,
+      "step": 206
+    },
+    {
+      "epoch": 0.03692801712603693,
+      "grad_norm": 0.7113338112831116,
+      "learning_rate": 0.0004995794502056751,
+      "loss": 1.6417,
+      "step": 207
+    },
+    {
+      "epoch": 0.03710641334403711,
+      "grad_norm": 3.6220555305480957,
+      "learning_rate": 0.0004995753782613521,
+      "loss": 1.7458,
+      "step": 208
+    },
+    {
+      "epoch": 0.03728480956203729,
+      "grad_norm": 1.1314635276794434,
+      "learning_rate": 0.0004995712867154863,
+      "loss": 1.9954,
+      "step": 209
+    },
+    {
+      "epoch": 0.037463205780037465,
+      "grad_norm": 0.8598477244377136,
+      "learning_rate": 0.0004995671755683989,
+      "loss": 1.7875,
+      "step": 210
+    },
+    {
+      "epoch": 0.03764160199803764,
+      "grad_norm": 0.725925087928772,
+      "learning_rate": 0.0004995630448204131,
+      "loss": 1.9205,
+      "step": 211
+    },
+    {
+      "epoch": 0.03781999821603782,
+      "grad_norm": 0.9679766297340393,
+      "learning_rate": 0.000499558894471853,
+      "loss": 2.1183,
+      "step": 212
+    },
+    {
+      "epoch": 0.037998394434038,
+      "grad_norm": 0.7008562684059143,
+      "learning_rate": 0.0004995547245230448,
+      "loss": 1.8477,
+      "step": 213
+    },
+    {
+      "epoch": 0.038176790652038176,
+      "grad_norm": 0.7993139624595642,
+      "learning_rate": 0.0004995505349743158,
+      "loss": 1.6427,
+      "step": 214
+    },
+    {
+      "epoch": 0.03835518687003835,
+      "grad_norm": 1.6259403228759766,
+      "learning_rate": 0.0004995463258259953,
+      "loss": 1.7742,
+      "step": 215
+    },
+    {
+      "epoch": 0.03853358308803853,
+      "grad_norm": 0.8765113353729248,
+      "learning_rate": 0.0004995420970784137,
+      "loss": 1.6492,
+      "step": 216
+    },
+    {
+      "epoch": 0.03871197930603871,
+      "grad_norm": 0.6928495168685913,
+      "learning_rate": 0.0004995378487319032,
+      "loss": 1.7392,
+      "step": 217
+    },
+    {
+      "epoch": 0.038890375524038893,
+      "grad_norm": 2.987717390060425,
+      "learning_rate": 0.0004995335807867975,
+      "loss": 1.8262,
+      "step": 218
+    },
+    {
+      "epoch": 0.03906877174203907,
+      "grad_norm": 1.99364173412323,
+      "learning_rate": 0.0004995292932434317,
+      "loss": 2.138,
+      "step": 219
+    },
+    {
+      "epoch": 0.03924716796003925,
+      "grad_norm": 1.4841055870056152,
+      "learning_rate": 0.0004995249861021425,
+      "loss": 1.7652,
+      "step": 220
+    },
+    {
+      "epoch": 0.03942556417803943,
+      "grad_norm": 0.9593233466148376,
+      "learning_rate": 0.0004995206593632685,
+      "loss": 2.2273,
+      "step": 221
+    },
+    {
+      "epoch": 0.039603960396039604,
+      "grad_norm": 0.7397097945213318,
+      "learning_rate": 0.0004995163130271491,
+      "loss": 1.6452,
+      "step": 222
+    },
+    {
+      "epoch": 0.03978235661403978,
+      "grad_norm": 0.7248395085334778,
+      "learning_rate": 0.0004995119470941259,
+      "loss": 1.7841,
+      "step": 223
+    },
+    {
+      "epoch": 0.03996075283203996,
+      "grad_norm": 0.8945801854133606,
+      "learning_rate": 0.0004995075615645418,
+      "loss": 1.8039,
+      "step": 224
+    },
+    {
+      "epoch": 0.04013914905004014,
+      "grad_norm": 1.0206605195999146,
+      "learning_rate": 0.0004995031564387411,
+      "loss": 1.8187,
+      "step": 225
+    },
+    {
+      "epoch": 0.040317545268040315,
+      "grad_norm": 0.802527129650116,
+      "learning_rate": 0.00049949873171707,
+      "loss": 1.7456,
+      "step": 226
+    },
+    {
+      "epoch": 0.04049594148604049,
+      "grad_norm": 0.8061395883560181,
+      "learning_rate": 0.0004994942873998757,
+      "loss": 1.7857,
+      "step": 227
+    },
+    {
+      "epoch": 0.04067433770404067,
+      "grad_norm": 0.8295297026634216,
+      "learning_rate": 0.0004994898234875074,
+      "loss": 1.6685,
+      "step": 228
+    },
+    {
+      "epoch": 0.040852733922040856,
+      "grad_norm": 0.9258911609649658,
+      "learning_rate": 0.000499485339980316,
+      "loss": 2.0146,
+      "step": 229
+    },
+    {
+      "epoch": 0.04103113014004103,
+      "grad_norm": 0.7650581002235413,
+      "learning_rate": 0.0004994808368786531,
+      "loss": 1.7862,
+      "step": 230
+    },
+    {
+      "epoch": 0.04120952635804121,
+      "grad_norm": 0.8616735339164734,
+      "learning_rate": 0.0004994763141828728,
+      "loss": 1.7706,
+      "step": 231
+    },
+    {
+      "epoch": 0.04138792257604139,
+      "grad_norm": 0.8802286386489868,
+      "learning_rate": 0.00049947177189333,
+      "loss": 1.7055,
+      "step": 232
+    },
+    {
+      "epoch": 0.04156631879404157,
+      "grad_norm": 0.6563844680786133,
+      "learning_rate": 0.0004994672100103818,
+      "loss": 1.7655,
+      "step": 233
+    },
+    {
+      "epoch": 0.041744715012041744,
+      "grad_norm": 0.8318186402320862,
+      "learning_rate": 0.0004994626285343861,
+      "loss": 1.7316,
+      "step": 234
+    },
+    {
+      "epoch": 0.04192311123004192,
+      "grad_norm": 0.8289990425109863,
+      "learning_rate": 0.0004994580274657029,
+      "loss": 1.9734,
+      "step": 235
+    },
+    {
+      "epoch": 0.0421015074480421,
+      "grad_norm": 0.8461620807647705,
+      "learning_rate": 0.0004994534068046936,
+      "loss": 1.8266,
+      "step": 236
+    },
+    {
+      "epoch": 0.04227990366604228,
+      "grad_norm": 0.8519775867462158,
+      "learning_rate": 0.0004994487665517212,
+      "loss": 1.8579,
+      "step": 237
+    },
+    {
+      "epoch": 0.042458299884042455,
+      "grad_norm": 0.707134485244751,
+      "learning_rate": 0.0004994441067071499,
+      "loss": 1.845,
+      "step": 238
+    },
+    {
+      "epoch": 0.04263669610204264,
+      "grad_norm": 1.2753993272781372,
+      "learning_rate": 0.0004994394272713459,
+      "loss": 1.8513,
+      "step": 239
+    },
+    {
+      "epoch": 0.04281509232004282,
+      "grad_norm": 0.9627553224563599,
+      "learning_rate": 0.0004994347282446765,
+      "loss": 2.2214,
+      "step": 240
+    },
+    {
+      "epoch": 0.042993488538042995,
+      "grad_norm": 0.8274680972099304,
+      "learning_rate": 0.0004994300096275108,
+      "loss": 1.8324,
+      "step": 241
+    },
+    {
+      "epoch": 0.04317188475604317,
+      "grad_norm": 1.0614216327667236,
+      "learning_rate": 0.0004994252714202198,
+      "loss": 1.8447,
+      "step": 242
+    },
+    {
+      "epoch": 0.04335028097404335,
+      "grad_norm": 0.853103518486023,
+      "learning_rate": 0.0004994205136231751,
+      "loss": 1.8614,
+      "step": 243
+    },
+    {
+      "epoch": 0.04352867719204353,
+      "grad_norm": 0.6934187412261963,
+      "learning_rate": 0.0004994157362367506,
+      "loss": 1.6686,
+      "step": 244
+    },
+    {
+      "epoch": 0.043707073410043706,
+      "grad_norm": 0.856378972530365,
+      "learning_rate": 0.0004994109392613215,
+      "loss": 1.7998,
+      "step": 245
+    },
+    {
+      "epoch": 0.043885469628043884,
+      "grad_norm": 0.8739482760429382,
+      "learning_rate": 0.0004994061226972647,
+      "loss": 1.7582,
+      "step": 246
+    },
+    {
+      "epoch": 0.04406386584604406,
+      "grad_norm": 0.904869794845581,
+      "learning_rate": 0.0004994012865449582,
+      "loss": 1.966,
+      "step": 247
+    },
+    {
+      "epoch": 0.04424226206404424,
+      "grad_norm": 0.8391656279563904,
+      "learning_rate": 0.0004993964308047821,
+      "loss": 1.7177,
+      "step": 248
+    },
+    {
+      "epoch": 0.04442065828204442,
+      "grad_norm": 0.9165405035018921,
+      "learning_rate": 0.0004993915554771175,
+      "loss": 1.8742,
+      "step": 249
+    },
+    {
+      "epoch": 0.0445990545000446,
+      "grad_norm": 0.7904878854751587,
+      "learning_rate": 0.0004993866605623475,
+      "loss": 1.8811,
+      "step": 250
+    },
+    {
+      "epoch": 0.04477745071804478,
+      "grad_norm": 0.7866482734680176,
+      "learning_rate": 0.0004993817460608566,
+      "loss": 1.6096,
+      "step": 251
+    },
+    {
+      "epoch": 0.04495584693604496,
+      "grad_norm": 0.8852795362472534,
+      "learning_rate": 0.0004993768119730306,
+      "loss": 1.8237,
+      "step": 252
+    },
+    {
+      "epoch": 0.045134243154045135,
+      "grad_norm": 0.9104618430137634,
+      "learning_rate": 0.0004993718582992572,
+      "loss": 1.8859,
+      "step": 253
+    },
+    {
+      "epoch": 0.04531263937204531,
+      "grad_norm": 0.7363934516906738,
+      "learning_rate": 0.0004993668850399252,
+      "loss": 1.3366,
+      "step": 254
+    },
+    {
+      "epoch": 0.04549103559004549,
+      "grad_norm": 0.7602558732032776,
+      "learning_rate": 0.0004993618921954254,
+      "loss": 2.0173,
+      "step": 255
+    },
+    {
+      "epoch": 0.04566943180804567,
+      "grad_norm": 0.8223975896835327,
+      "learning_rate": 0.00049935687976615,
+      "loss": 1.6692,
+      "step": 256
+    },
+    {
+      "epoch": 0.045847828026045846,
+      "grad_norm": 0.7281943559646606,
+      "learning_rate": 0.0004993518477524924,
+      "loss": 1.6527,
+      "step": 257
+    },
+    {
+      "epoch": 0.046026224244046024,
+      "grad_norm": 0.7140815854072571,
+      "learning_rate": 0.0004993467961548482,
+      "loss": 1.895,
+      "step": 258
+    },
+    {
+      "epoch": 0.0462046204620462,
+      "grad_norm": 1.0137954950332642,
+      "learning_rate": 0.0004993417249736138,
+      "loss": 2.0086,
+      "step": 259
+    },
+    {
+      "epoch": 0.046383016680046386,
+      "grad_norm": 0.7407935261726379,
+      "learning_rate": 0.0004993366342091876,
+      "loss": 1.6048,
+      "step": 260
+    },
+    {
+      "epoch": 0.046561412898046564,
+      "grad_norm": 0.7034161686897278,
+      "learning_rate": 0.0004993315238619695,
+      "loss": 1.8503,
+      "step": 261
+    },
+    {
+      "epoch": 0.04673980911604674,
+      "grad_norm": 0.6958416104316711,
+      "learning_rate": 0.0004993263939323608,
+      "loss": 1.6034,
+      "step": 262
+    },
+    {
+      "epoch": 0.04691820533404692,
+      "grad_norm": 0.7423539161682129,
+      "learning_rate": 0.0004993212444207644,
+      "loss": 1.598,
+      "step": 263
+    },
+    {
+      "epoch": 0.0470966015520471,
+      "grad_norm": 0.7731009125709534,
+      "learning_rate": 0.0004993160753275849,
+      "loss": 1.9353,
+      "step": 264
+    },
+    {
+      "epoch": 0.047274997770047275,
+      "grad_norm": 0.7424893379211426,
+      "learning_rate": 0.0004993108866532279,
+      "loss": 1.712,
+      "step": 265
+    },
+    {
+      "epoch": 0.04745339398804745,
+      "grad_norm": 0.8562301993370056,
+      "learning_rate": 0.0004993056783981013,
+      "loss": 1.8745,
+      "step": 266
+    },
+    {
+      "epoch": 0.04763179020604763,
+      "grad_norm": 1.6230698823928833,
+      "learning_rate": 0.000499300450562614,
+      "loss": 1.7503,
+      "step": 267
+    },
+    {
+      "epoch": 0.04781018642404781,
+      "grad_norm": 0.938933253288269,
+      "learning_rate": 0.0004992952031471765,
+      "loss": 1.683,
+      "step": 268
+    },
+    {
+      "epoch": 0.047988582642047986,
+      "grad_norm": 0.8198557496070862,
+      "learning_rate": 0.0004992899361522011,
+      "loss": 1.9077,
+      "step": 269
+    },
+    {
+      "epoch": 0.048166978860048164,
+      "grad_norm": 0.9028448462486267,
+      "learning_rate": 0.0004992846495781013,
+      "loss": 1.9227,
+      "step": 270
+    },
+    {
+      "epoch": 0.04834537507804835,
+      "grad_norm": 0.7884741425514221,
+      "learning_rate": 0.0004992793434252925,
+      "loss": 1.58,
+      "step": 271
+    },
+    {
+      "epoch": 0.048523771296048526,
+      "grad_norm": 1.0796072483062744,
+      "learning_rate": 0.0004992740176941912,
+      "loss": 1.6761,
+      "step": 272
+    },
+    {
+      "epoch": 0.048702167514048704,
+      "grad_norm": 0.9148755073547363,
+      "learning_rate": 0.0004992686723852161,
+      "loss": 1.7409,
+      "step": 273
+    },
+    {
+      "epoch": 0.04888056373204888,
+      "grad_norm": 0.7027503252029419,
+      "learning_rate": 0.0004992633074987864,
+      "loss": 1.4971,
+      "step": 274
+    },
+    {
+      "epoch": 0.04905895995004906,
+      "grad_norm": 0.8004087209701538,
+      "learning_rate": 0.000499257923035324,
+      "loss": 1.7677,
+      "step": 275
+    },
+    {
+      "epoch": 0.04923735616804924,
+      "grad_norm": 0.7200493216514587,
+      "learning_rate": 0.0004992525189952516,
+      "loss": 1.7596,
+      "step": 276
+    },
+    {
+      "epoch": 0.049415752386049415,
+      "grad_norm": 0.7689985036849976,
+      "learning_rate": 0.0004992470953789936,
+      "loss": 1.7032,
+      "step": 277
+    },
+    {
+      "epoch": 0.04959414860404959,
+      "grad_norm": 0.7748706936836243,
+      "learning_rate": 0.000499241652186976,
+      "loss": 1.972,
+      "step": 278
+    },
+    {
+      "epoch": 0.04977254482204977,
+      "grad_norm": 1.406790018081665,
+      "learning_rate": 0.0004992361894196263,
+      "loss": 1.5758,
+      "step": 279
+    },
+    {
+      "epoch": 0.04995094104004995,
+      "grad_norm": 0.6651303172111511,
+      "learning_rate": 0.0004992307070773734,
+      "loss": 1.6668,
+      "step": 280
+    },
+    {
+      "epoch": 0.050129337258050126,
+      "grad_norm": 0.7641019821166992,
+      "learning_rate": 0.0004992252051606481,
+      "loss": 1.8233,
+      "step": 281
+    },
+    {
+      "epoch": 0.05030773347605031,
+      "grad_norm": 0.6978675127029419,
+      "learning_rate": 0.0004992196836698825,
+      "loss": 1.6394,
+      "step": 282
+    },
+    {
+      "epoch": 0.05048612969405049,
+      "grad_norm": 0.7156147956848145,
+      "learning_rate": 0.0004992141426055101,
+      "loss": 1.6568,
+      "step": 283
+    },
+    {
+      "epoch": 0.050664525912050666,
+      "grad_norm": 0.791042149066925,
+      "learning_rate": 0.0004992085819679662,
+      "loss": 1.8138,
+      "step": 284
+    },
+    {
+      "epoch": 0.050842922130050844,
+      "grad_norm": 0.7295787334442139,
+      "learning_rate": 0.0004992030017576876,
+      "loss": 1.3886,
+      "step": 285
+    },
+    {
+      "epoch": 0.05102131834805102,
+      "grad_norm": 1.0952125787734985,
+      "learning_rate": 0.0004991974019751124,
+      "loss": 1.6662,
+      "step": 286
+    },
+    {
+      "epoch": 0.0511997145660512,
+      "grad_norm": 1.1296299695968628,
+      "learning_rate": 0.0004991917826206805,
+      "loss": 1.5014,
+      "step": 287
+    },
+    {
+      "epoch": 0.05137811078405138,
+      "grad_norm": 0.7258694767951965,
+      "learning_rate": 0.0004991861436948333,
+      "loss": 1.5864,
+      "step": 288
+    },
+    {
+      "epoch": 0.051556507002051555,
+      "grad_norm": 0.9456144571304321,
+      "learning_rate": 0.0004991804851980135,
+      "loss": 1.9251,
+      "step": 289
+    },
+    {
+      "epoch": 0.05173490322005173,
+      "grad_norm": 0.7830838561058044,
+      "learning_rate": 0.0004991748071306657,
+      "loss": 1.6641,
+      "step": 290
+    },
+    {
+      "epoch": 0.05191329943805191,
+      "grad_norm": 0.7046554684638977,
+      "learning_rate": 0.0004991691094932357,
+      "loss": 1.7749,
+      "step": 291
+    },
+    {
+      "epoch": 0.052091695656052095,
+      "grad_norm": 1.1261860132217407,
+      "learning_rate": 0.000499163392286171,
+      "loss": 1.7228,
+      "step": 292
+    },
+    {
+      "epoch": 0.05227009187405227,
+      "grad_norm": 0.7298058867454529,
+      "learning_rate": 0.0004991576555099208,
+      "loss": 1.666,
+      "step": 293
+    },
+    {
+      "epoch": 0.05244848809205245,
+      "grad_norm": 0.7975763082504272,
+      "learning_rate": 0.0004991518991649356,
+      "loss": 1.803,
+      "step": 294
+    },
+    {
+      "epoch": 0.05262688431005263,
+      "grad_norm": 1.0635559558868408,
+      "learning_rate": 0.0004991461232516675,
+      "loss": 1.4519,
+      "step": 295
+    },
+    {
+      "epoch": 0.052805280528052806,
+      "grad_norm": 0.9030710458755493,
+      "learning_rate": 0.00049914032777057,
+      "loss": 1.607,
+      "step": 296
+    },
+    {
+      "epoch": 0.052983676746052984,
+      "grad_norm": 1.0126208066940308,
+      "learning_rate": 0.0004991345127220983,
+      "loss": 1.6796,
+      "step": 297
+    },
+    {
+      "epoch": 0.05316207296405316,
+      "grad_norm": 0.8473824858665466,
+      "learning_rate": 0.0004991286781067094,
+      "loss": 1.6065,
+      "step": 298
+    },
+    {
+      "epoch": 0.05334046918205334,
+      "grad_norm": 0.7614424228668213,
+      "learning_rate": 0.0004991228239248611,
+      "loss": 1.4427,
+      "step": 299
+    },
+    {
+      "epoch": 0.05351886540005352,
+      "grad_norm": 0.8425862193107605,
+      "learning_rate": 0.0004991169501770135,
+      "loss": 1.6097,
+      "step": 300
+    },
+    {
+      "epoch": 0.053697261618053695,
+      "grad_norm": 0.7978277802467346,
+      "learning_rate": 0.0004991110568636278,
+      "loss": 1.6084,
+      "step": 301
+    },
+    {
+      "epoch": 0.05387565783605387,
+      "grad_norm": 0.9924404621124268,
+      "learning_rate": 0.000499105143985167,
+      "loss": 1.6886,
+      "step": 302
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 1.2324185371398926,
+      "learning_rate": 0.0004990992115420954,
+      "loss": 2.0878,
+      "step": 303
+    },
+    {
+      "epoch": 0.054232450272054235,
+      "grad_norm": 0.8193588256835938,
+      "learning_rate": 0.0004990932595348789,
+      "loss": 1.3275,
+      "step": 304
+    },
+    {
+      "epoch": 0.05441084649005441,
+      "grad_norm": 0.9794548153877258,
+      "learning_rate": 0.0004990872879639849,
+      "loss": 1.9818,
+      "step": 305
+    },
+    {
+      "epoch": 0.05458924270805459,
+      "grad_norm": 0.9556685090065002,
+      "learning_rate": 0.0004990812968298825,
+      "loss": 1.7266,
+      "step": 306
+    },
+    {
+      "epoch": 0.05476763892605477,
+      "grad_norm": 0.8109720945358276,
+      "learning_rate": 0.0004990752861330424,
+      "loss": 1.626,
+      "step": 307
+    },
+    {
+      "epoch": 0.054946035144054946,
+      "grad_norm": 0.7442929744720459,
+      "learning_rate": 0.0004990692558739363,
+      "loss": 1.6846,
+      "step": 308
+    },
+    {
+      "epoch": 0.055124431362055124,
+      "grad_norm": 0.8138058185577393,
+      "learning_rate": 0.0004990632060530381,
+      "loss": 1.4798,
+      "step": 309
+    },
+    {
+      "epoch": 0.0553028275800553,
+      "grad_norm": 1.426184892654419,
+      "learning_rate": 0.0004990571366708229,
+      "loss": 1.4324,
+      "step": 310
+    },
+    {
+      "epoch": 0.05548122379805548,
+      "grad_norm": 0.9690107703208923,
+      "learning_rate": 0.0004990510477277673,
+      "loss": 1.7228,
+      "step": 311
+    },
+    {
+      "epoch": 0.05565962001605566,
+      "grad_norm": 0.8152466416358948,
+      "learning_rate": 0.0004990449392243497,
+      "loss": 1.5993,
+      "step": 312
+    },
+    {
+      "epoch": 0.05583801623405584,
+      "grad_norm": 0.7555820345878601,
+      "learning_rate": 0.0004990388111610496,
+      "loss": 1.5512,
+      "step": 313
+    },
+    {
+      "epoch": 0.05601641245205602,
+      "grad_norm": 0.8185862302780151,
+      "learning_rate": 0.0004990326635383486,
+      "loss": 1.8048,
+      "step": 314
+    },
+    {
+      "epoch": 0.0561948086700562,
+      "grad_norm": 1.3000257015228271,
+      "learning_rate": 0.0004990264963567292,
+      "loss": 1.9454,
+      "step": 315
+    },
+    {
+      "epoch": 0.056373204888056375,
+      "grad_norm": 1.501250982284546,
+      "learning_rate": 0.0004990203096166761,
+      "loss": 1.7511,
+      "step": 316
+    },
+    {
+      "epoch": 0.05655160110605655,
+      "grad_norm": 0.7413790822029114,
+      "learning_rate": 0.000499014103318675,
+      "loss": 1.6659,
+      "step": 317
+    },
+    {
+      "epoch": 0.05672999732405673,
+      "grad_norm": 0.6846387982368469,
+      "learning_rate": 0.0004990078774632134,
+      "loss": 1.5829,
+      "step": 318
+    },
+    {
+      "epoch": 0.05690839354205691,
+      "grad_norm": 0.8253748416900635,
+      "learning_rate": 0.0004990016320507802,
+      "loss": 1.6716,
+      "step": 319
+    },
+    {
+      "epoch": 0.057086789760057086,
+      "grad_norm": 0.7216707468032837,
+      "learning_rate": 0.000498995367081866,
+      "loss": 1.6294,
+      "step": 320
+    },
+    {
+      "epoch": 0.05726518597805726,
+      "grad_norm": 0.8067865371704102,
+      "learning_rate": 0.0004989890825569628,
+      "loss": 1.7566,
+      "step": 321
+    },
+    {
+      "epoch": 0.05744358219605744,
+      "grad_norm": 0.867411732673645,
+      "learning_rate": 0.0004989827784765643,
+      "loss": 1.5055,
+      "step": 322
+    },
+    {
+      "epoch": 0.05762197841405762,
+      "grad_norm": 0.7605225443840027,
+      "learning_rate": 0.0004989764548411654,
+      "loss": 1.6925,
+      "step": 323
+    },
+    {
+      "epoch": 0.057800374632057804,
+      "grad_norm": 0.7990824580192566,
+      "learning_rate": 0.0004989701116512629,
+      "loss": 1.8807,
+      "step": 324
+    },
+    {
+      "epoch": 0.05797877085005798,
+      "grad_norm": 0.7364970445632935,
+      "learning_rate": 0.000498963748907355,
+      "loss": 1.6086,
+      "step": 325
+    },
+    {
+      "epoch": 0.05815716706805816,
+      "grad_norm": 0.7894652485847473,
+      "learning_rate": 0.0004989573666099415,
+      "loss": 1.9226,
+      "step": 326
+    },
+    {
+      "epoch": 0.05833556328605834,
+      "grad_norm": 0.8756299018859863,
+      "learning_rate": 0.0004989509647595234,
+      "loss": 1.7023,
+      "step": 327
+    },
+    {
+      "epoch": 0.058513959504058514,
+      "grad_norm": 1.0760093927383423,
+      "learning_rate": 0.0004989445433566037,
+      "loss": 1.4588,
+      "step": 328
+    },
+    {
+      "epoch": 0.05869235572205869,
+      "grad_norm": 0.8876066207885742,
+      "learning_rate": 0.0004989381024016867,
+      "loss": 1.7266,
+      "step": 329
+    },
+    {
+      "epoch": 0.05887075194005887,
+      "grad_norm": 4.222714900970459,
+      "learning_rate": 0.0004989316418952782,
+      "loss": 1.6749,
+      "step": 330
+    },
+    {
+      "epoch": 0.05904914815805905,
+      "grad_norm": 1.0172480344772339,
+      "learning_rate": 0.0004989251618378859,
+      "loss": 1.6138,
+      "step": 331
+    },
+    {
+      "epoch": 0.059227544376059225,
+      "grad_norm": 1.1439893245697021,
+      "learning_rate": 0.0004989186622300183,
+      "loss": 1.6511,
+      "step": 332
+    },
+    {
+      "epoch": 0.0594059405940594,
+      "grad_norm": 0.7901877760887146,
+      "learning_rate": 0.0004989121430721862,
+      "loss": 1.5239,
+      "step": 333
+    },
+    {
+      "epoch": 0.05958433681205958,
+      "grad_norm": 0.8215880393981934,
+      "learning_rate": 0.0004989056043649014,
+      "loss": 1.6242,
+      "step": 334
+    },
+    {
+      "epoch": 0.059762733030059766,
+      "grad_norm": 0.7913690805435181,
+      "learning_rate": 0.0004988990461086777,
+      "loss": 1.9562,
+      "step": 335
+    },
+    {
+      "epoch": 0.05994112924805994,
+      "grad_norm": 0.7676241397857666,
+      "learning_rate": 0.0004988924683040298,
+      "loss": 1.6554,
+      "step": 336
+    },
+    {
+      "epoch": 0.06011952546606012,
+      "grad_norm": 0.7700140476226807,
+      "learning_rate": 0.0004988858709514747,
+      "loss": 1.5174,
+      "step": 337
+    },
+    {
+      "epoch": 0.0602979216840603,
+      "grad_norm": 0.7194883227348328,
+      "learning_rate": 0.0004988792540515304,
+      "loss": 1.711,
+      "step": 338
+    },
+    {
+      "epoch": 0.06047631790206048,
+      "grad_norm": 0.737006425857544,
+      "learning_rate": 0.0004988726176047164,
+      "loss": 1.7161,
+      "step": 339
+    },
+    {
+      "epoch": 0.060654714120060654,
+      "grad_norm": 0.8978502154350281,
+      "learning_rate": 0.0004988659616115544,
+      "loss": 1.4597,
+      "step": 340
+    },
+    {
+      "epoch": 0.06083311033806083,
+      "grad_norm": 0.850236177444458,
+      "learning_rate": 0.0004988592860725667,
+      "loss": 1.7958,
+      "step": 341
+    },
+    {
+      "epoch": 0.06101150655606101,
+      "grad_norm": 0.6929709911346436,
+      "learning_rate": 0.0004988525909882779,
+      "loss": 1.4684,
+      "step": 342
+    },
+    {
+      "epoch": 0.06118990277406119,
+      "grad_norm": 0.8175438642501831,
+      "learning_rate": 0.0004988458763592135,
+      "loss": 1.5871,
+      "step": 343
+    },
+    {
+      "epoch": 0.061368298992061365,
+      "grad_norm": 1.5031802654266357,
+      "learning_rate": 0.0004988391421859011,
+      "loss": 1.8358,
+      "step": 344
+    },
+    {
+      "epoch": 0.06154669521006155,
+      "grad_norm": 0.8213221430778503,
+      "learning_rate": 0.0004988323884688696,
+      "loss": 1.6868,
+      "step": 345
+    },
+    {
+      "epoch": 0.06172509142806173,
+      "grad_norm": 0.855012059211731,
+      "learning_rate": 0.0004988256152086495,
+      "loss": 1.7024,
+      "step": 346
+    },
+    {
+      "epoch": 0.061903487646061905,
+      "grad_norm": 0.9920960664749146,
+      "learning_rate": 0.0004988188224057724,
+      "loss": 1.641,
+      "step": 347
+    },
+    {
+      "epoch": 0.06208188386406208,
+      "grad_norm": 0.7211305499076843,
+      "learning_rate": 0.0004988120100607723,
+      "loss": 1.3817,
+      "step": 348
+    },
+    {
+      "epoch": 0.06226028008206226,
+      "grad_norm": 0.8306823968887329,
+      "learning_rate": 0.0004988051781741839,
+      "loss": 1.6479,
+      "step": 349
+    },
+    {
+      "epoch": 0.06243867630006244,
+      "grad_norm": 0.6879592537879944,
+      "learning_rate": 0.0004987983267465439,
+      "loss": 1.7006,
+      "step": 350
+    },
+    {
+      "epoch": 0.06261707251806262,
+      "grad_norm": 0.7928709387779236,
+      "learning_rate": 0.0004987914557783905,
+      "loss": 1.7222,
+      "step": 351
+    },
+    {
+      "epoch": 0.0627954687360628,
+      "grad_norm": 0.8470203876495361,
+      "learning_rate": 0.000498784565270263,
+      "loss": 1.7225,
+      "step": 352
+    },
+    {
+      "epoch": 0.06297386495406297,
+      "grad_norm": 0.7351035475730896,
+      "learning_rate": 0.0004987776552227029,
+      "loss": 1.574,
+      "step": 353
+    },
+    {
+      "epoch": 0.06315226117206316,
+      "grad_norm": 0.8006918430328369,
+      "learning_rate": 0.0004987707256362529,
+      "loss": 1.9217,
+      "step": 354
+    },
+    {
+      "epoch": 0.06333065739006333,
+      "grad_norm": 0.7959868907928467,
+      "learning_rate": 0.0004987637765114571,
+      "loss": 1.6191,
+      "step": 355
+    },
+    {
+      "epoch": 0.06350905360806351,
+      "grad_norm": 0.6624523401260376,
+      "learning_rate": 0.0004987568078488613,
+      "loss": 1.4849,
+      "step": 356
+    },
+    {
+      "epoch": 0.06368744982606368,
+      "grad_norm": 0.665393054485321,
+      "learning_rate": 0.0004987498196490129,
+      "loss": 1.67,
+      "step": 357
+    },
+    {
+      "epoch": 0.06386584604406387,
+      "grad_norm": 1.2895721197128296,
+      "learning_rate": 0.0004987428119124607,
+      "loss": 1.6514,
+      "step": 358
+    },
+    {
+      "epoch": 0.06404424226206404,
+      "grad_norm": 0.7853053212165833,
+      "learning_rate": 0.0004987357846397551,
+      "loss": 1.528,
+      "step": 359
+    },
+    {
+      "epoch": 0.06422263848006422,
+      "grad_norm": 0.7855213284492493,
+      "learning_rate": 0.000498728737831448,
+      "loss": 1.3609,
+      "step": 360
+    },
+    {
+      "epoch": 0.06440103469806441,
+      "grad_norm": 0.9577522277832031,
+      "learning_rate": 0.0004987216714880929,
+      "loss": 1.6681,
+      "step": 361
+    },
+    {
+      "epoch": 0.06457943091606458,
+      "grad_norm": 0.7544964551925659,
+      "learning_rate": 0.0004987145856102448,
+      "loss": 1.5707,
+      "step": 362
+    },
+    {
+      "epoch": 0.06475782713406476,
+      "grad_norm": 0.9541155695915222,
+      "learning_rate": 0.00049870748019846,
+      "loss": 1.5559,
+      "step": 363
+    },
+    {
+      "epoch": 0.06493622335206493,
+      "grad_norm": 0.8656521439552307,
+      "learning_rate": 0.0004987003552532969,
+      "loss": 1.5096,
+      "step": 364
+    },
+    {
+      "epoch": 0.06511461957006512,
+      "grad_norm": 0.9872167706489563,
+      "learning_rate": 0.0004986932107753148,
+      "loss": 1.575,
+      "step": 365
+    },
+    {
+      "epoch": 0.06529301578806529,
+      "grad_norm": 0.7449053525924683,
+      "learning_rate": 0.000498686046765075,
+      "loss": 1.3222,
+      "step": 366
+    },
+    {
+      "epoch": 0.06547141200606547,
+      "grad_norm": 0.7896831631660461,
+      "learning_rate": 0.0004986788632231401,
+      "loss": 1.6182,
+      "step": 367
+    },
+    {
+      "epoch": 0.06564980822406564,
+      "grad_norm": 0.8574457168579102,
+      "learning_rate": 0.0004986716601500744,
+      "loss": 1.6993,
+      "step": 368
+    },
+    {
+      "epoch": 0.06582820444206583,
+      "grad_norm": 0.8115587830543518,
+      "learning_rate": 0.0004986644375464434,
+      "loss": 1.6287,
+      "step": 369
+    },
+    {
+      "epoch": 0.066006600660066,
+      "grad_norm": 1.8400505781173706,
+      "learning_rate": 0.0004986571954128145,
+      "loss": 1.6949,
+      "step": 370
+    },
+    {
+      "epoch": 0.06618499687806619,
+      "grad_norm": 0.8843457698822021,
+      "learning_rate": 0.0004986499337497565,
+      "loss": 1.3939,
+      "step": 371
+    },
+    {
+      "epoch": 0.06636339309606637,
+      "grad_norm": 0.9792248010635376,
+      "learning_rate": 0.0004986426525578398,
+      "loss": 1.6735,
+      "step": 372
+    },
+    {
+      "epoch": 0.06654178931406654,
+      "grad_norm": 0.8991268277168274,
+      "learning_rate": 0.000498635351837636,
+      "loss": 1.6638,
+      "step": 373
+    },
+    {
+      "epoch": 0.06672018553206673,
+      "grad_norm": 1.0878491401672363,
+      "learning_rate": 0.0004986280315897188,
+      "loss": 1.3675,
+      "step": 374
+    },
+    {
+      "epoch": 0.0668985817500669,
+      "grad_norm": 0.9121967554092407,
+      "learning_rate": 0.0004986206918146629,
+      "loss": 1.4673,
+      "step": 375
+    },
+    {
+      "epoch": 0.06707697796806708,
+      "grad_norm": 0.9130116105079651,
+      "learning_rate": 0.0004986133325130448,
+      "loss": 1.8798,
+      "step": 376
+    },
+    {
+      "epoch": 0.06725537418606725,
+      "grad_norm": 2.4044368267059326,
+      "learning_rate": 0.0004986059536854427,
+      "loss": 1.396,
+      "step": 377
+    },
+    {
+      "epoch": 0.06743377040406744,
+      "grad_norm": 1.2057360410690308,
+      "learning_rate": 0.0004985985553324359,
+      "loss": 1.6874,
+      "step": 378
+    },
+    {
+      "epoch": 0.06761216662206761,
+      "grad_norm": 0.9697250723838806,
+      "learning_rate": 0.0004985911374546056,
+      "loss": 1.6581,
+      "step": 379
+    },
+    {
+      "epoch": 0.06779056284006779,
+      "grad_norm": 8.421307563781738,
+      "learning_rate": 0.0004985837000525343,
+      "loss": 1.4384,
+      "step": 380
+    },
+    {
+      "epoch": 0.06796895905806796,
+      "grad_norm": 1.0877493619918823,
+      "learning_rate": 0.0004985762431268062,
+      "loss": 1.6454,
+      "step": 381
+    },
+    {
+      "epoch": 0.06814735527606815,
+      "grad_norm": 0.8904185891151428,
+      "learning_rate": 0.0004985687666780069,
+      "loss": 1.611,
+      "step": 382
+    },
+    {
+      "epoch": 0.06832575149406833,
+      "grad_norm": 0.9544228911399841,
+      "learning_rate": 0.0004985612707067237,
+      "loss": 1.3702,
+      "step": 383
+    },
+    {
+      "epoch": 0.0685041477120685,
+      "grad_norm": 0.9072893261909485,
+      "learning_rate": 0.0004985537552135451,
+      "loss": 1.6441,
+      "step": 384
+    },
+    {
+      "epoch": 0.06868254393006869,
+      "grad_norm": 0.8297262191772461,
+      "learning_rate": 0.0004985462201990617,
+      "loss": 1.7376,
+      "step": 385
+    },
+    {
+      "epoch": 0.06886094014806886,
+      "grad_norm": 1.0085870027542114,
+      "learning_rate": 0.000498538665663865,
+      "loss": 1.5545,
+      "step": 386
+    },
+    {
+      "epoch": 0.06903933636606904,
+      "grad_norm": 0.7521477341651917,
+      "learning_rate": 0.0004985310916085485,
+      "loss": 1.77,
+      "step": 387
+    },
+    {
+      "epoch": 0.06921773258406921,
+      "grad_norm": 0.8971685767173767,
+      "learning_rate": 0.000498523498033707,
+      "loss": 1.6057,
+      "step": 388
+    },
+    {
+      "epoch": 0.0693961288020694,
+      "grad_norm": 0.7471177577972412,
+      "learning_rate": 0.000498515884939937,
+      "loss": 1.6315,
+      "step": 389
+    },
+    {
+      "epoch": 0.06957452502006957,
+      "grad_norm": 1.5577987432479858,
+      "learning_rate": 0.0004985082523278363,
+      "loss": 1.7456,
+      "step": 390
+    },
+    {
+      "epoch": 0.06975292123806975,
+      "grad_norm": 0.7326628565788269,
+      "learning_rate": 0.0004985006001980044,
+      "loss": 1.8009,
+      "step": 391
+    },
+    {
+      "epoch": 0.06993131745606992,
+      "grad_norm": 0.6636003851890564,
+      "learning_rate": 0.0004984929285510423,
+      "loss": 1.6494,
+      "step": 392
+    },
+    {
+      "epoch": 0.07010971367407011,
+      "grad_norm": 0.7074577808380127,
+      "learning_rate": 0.0004984852373875524,
+      "loss": 1.5621,
+      "step": 393
+    },
+    {
+      "epoch": 0.0702881098920703,
+      "grad_norm": 0.682092547416687,
+      "learning_rate": 0.000498477526708139,
+      "loss": 1.3208,
+      "step": 394
+    },
+    {
+      "epoch": 0.07046650611007046,
+      "grad_norm": 0.692871630191803,
+      "learning_rate": 0.0004984697965134076,
+      "loss": 1.2526,
+      "step": 395
+    },
+    {
+      "epoch": 0.07064490232807065,
+      "grad_norm": 0.6794890761375427,
+      "learning_rate": 0.0004984620468039653,
+      "loss": 1.772,
+      "step": 396
+    },
+    {
+      "epoch": 0.07082329854607082,
+      "grad_norm": 0.9123995304107666,
+      "learning_rate": 0.0004984542775804207,
+      "loss": 1.6829,
+      "step": 397
+    },
+    {
+      "epoch": 0.071001694764071,
+      "grad_norm": 0.9195268154144287,
+      "learning_rate": 0.0004984464888433842,
+      "loss": 1.5221,
+      "step": 398
+    },
+    {
+      "epoch": 0.07118009098207118,
+      "grad_norm": 0.6694832444190979,
+      "learning_rate": 0.0004984386805934672,
+      "loss": 1.5026,
+      "step": 399
+    },
+    {
+      "epoch": 0.07135848720007136,
+      "grad_norm": 0.7524397969245911,
+      "learning_rate": 0.0004984308528312833,
+      "loss": 1.6038,
+      "step": 400
+    },
+    {
+      "epoch": 0.07153688341807153,
+      "grad_norm": 1.0407260656356812,
+      "learning_rate": 0.000498423005557447,
+      "loss": 1.9059,
+      "step": 401
+    },
+    {
+      "epoch": 0.07171527963607172,
+      "grad_norm": 0.9182829260826111,
+      "learning_rate": 0.0004984151387725748,
+      "loss": 1.8141,
+      "step": 402
+    },
+    {
+      "epoch": 0.07189367585407189,
+      "grad_norm": 0.67998206615448,
+      "learning_rate": 0.0004984072524772845,
+      "loss": 1.607,
+      "step": 403
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 0.6888518333435059,
+      "learning_rate": 0.0004983993466721955,
+      "loss": 1.8157,
+      "step": 404
+    },
+    {
+      "epoch": 0.07225046829007226,
+      "grad_norm": 0.9161537289619446,
+      "learning_rate": 0.0004983914213579287,
+      "loss": 1.28,
+      "step": 405
+    },
+    {
+      "epoch": 0.07242886450807243,
+      "grad_norm": 0.8046262860298157,
+      "learning_rate": 0.0004983834765351065,
+      "loss": 1.7245,
+      "step": 406
+    },
+    {
+      "epoch": 0.07260726072607261,
+      "grad_norm": 0.7060288786888123,
+      "learning_rate": 0.000498375512204353,
+      "loss": 1.3604,
+      "step": 407
+    },
+    {
+      "epoch": 0.07278565694407278,
+      "grad_norm": 0.6991462707519531,
+      "learning_rate": 0.0004983675283662936,
+      "loss": 1.3504,
+      "step": 408
+    },
+    {
+      "epoch": 0.07296405316207297,
+      "grad_norm": 0.7093139886856079,
+      "learning_rate": 0.0004983595250215556,
+      "loss": 1.3739,
+      "step": 409
+    },
+    {
+      "epoch": 0.07314244938007314,
+      "grad_norm": 0.7139406800270081,
+      "learning_rate": 0.0004983515021707672,
+      "loss": 1.6803,
+      "step": 410
+    },
+    {
+      "epoch": 0.07332084559807332,
+      "grad_norm": 0.6523078680038452,
+      "learning_rate": 0.0004983434598145587,
+      "loss": 1.5356,
+      "step": 411
+    },
+    {
+      "epoch": 0.0734992418160735,
+      "grad_norm": 0.7209059596061707,
+      "learning_rate": 0.0004983353979535617,
+      "loss": 1.5146,
+      "step": 412
+    },
+    {
+      "epoch": 0.07367763803407368,
+      "grad_norm": 0.8449965715408325,
+      "learning_rate": 0.0004983273165884096,
+      "loss": 1.5411,
+      "step": 413
+    },
+    {
+      "epoch": 0.07385603425207386,
+      "grad_norm": 0.8013091683387756,
+      "learning_rate": 0.0004983192157197368,
+      "loss": 1.6204,
+      "step": 414
+    },
+    {
+      "epoch": 0.07403443047007403,
+      "grad_norm": 0.7177372574806213,
+      "learning_rate": 0.0004983110953481796,
+      "loss": 1.3559,
+      "step": 415
+    },
+    {
+      "epoch": 0.07421282668807422,
+      "grad_norm": 1.297348976135254,
+      "learning_rate": 0.0004983029554743759,
+      "loss": 1.3917,
+      "step": 416
+    },
+    {
+      "epoch": 0.07439122290607439,
+      "grad_norm": 0.7415265440940857,
+      "learning_rate": 0.0004982947960989649,
+      "loss": 1.4611,
+      "step": 417
+    },
+    {
+      "epoch": 0.07456961912407457,
+      "grad_norm": 0.7333908677101135,
+      "learning_rate": 0.0004982866172225875,
+      "loss": 1.5049,
+      "step": 418
+    },
+    {
+      "epoch": 0.07474801534207474,
+      "grad_norm": 0.6998400092124939,
+      "learning_rate": 0.000498278418845886,
+      "loss": 1.5281,
+      "step": 419
+    },
+    {
+      "epoch": 0.07492641156007493,
+      "grad_norm": 0.8220930099487305,
+      "learning_rate": 0.0004982702009695044,
+      "loss": 1.6659,
+      "step": 420
+    },
+    {
+      "epoch": 0.0751048077780751,
+      "grad_norm": 0.6637287139892578,
+      "learning_rate": 0.0004982619635940879,
+      "loss": 1.5122,
+      "step": 421
+    },
+    {
+      "epoch": 0.07528320399607528,
+      "grad_norm": 0.6980449557304382,
+      "learning_rate": 0.0004982537067202837,
+      "loss": 1.4508,
+      "step": 422
+    },
+    {
+      "epoch": 0.07546160021407546,
+      "grad_norm": 0.7847539782524109,
+      "learning_rate": 0.0004982454303487403,
+      "loss": 1.5511,
+      "step": 423
+    },
+    {
+      "epoch": 0.07563999643207564,
+      "grad_norm": 0.7317553758621216,
+      "learning_rate": 0.0004982371344801074,
+      "loss": 1.6737,
+      "step": 424
+    },
+    {
+      "epoch": 0.07581839265007582,
+      "grad_norm": 0.816692054271698,
+      "learning_rate": 0.000498228819115037,
+      "loss": 1.6792,
+      "step": 425
+    },
+    {
+      "epoch": 0.075996788868076,
+      "grad_norm": 0.6910188794136047,
+      "learning_rate": 0.0004982204842541818,
+      "loss": 1.6721,
+      "step": 426
+    },
+    {
+      "epoch": 0.07617518508607618,
+      "grad_norm": 0.6814787983894348,
+      "learning_rate": 0.0004982121298981967,
+      "loss": 1.3176,
+      "step": 427
+    },
+    {
+      "epoch": 0.07635358130407635,
+      "grad_norm": 1.1283695697784424,
+      "learning_rate": 0.0004982037560477377,
+      "loss": 1.6094,
+      "step": 428
+    },
+    {
+      "epoch": 0.07653197752207654,
+      "grad_norm": 0.7835364937782288,
+      "learning_rate": 0.0004981953627034625,
+      "loss": 1.6079,
+      "step": 429
+    },
+    {
+      "epoch": 0.0767103737400767,
+      "grad_norm": 0.9054422974586487,
+      "learning_rate": 0.0004981869498660304,
+      "loss": 1.5844,
+      "step": 430
+    },
+    {
+      "epoch": 0.07688876995807689,
+      "grad_norm": 1.4797576665878296,
+      "learning_rate": 0.000498178517536102,
+      "loss": 1.6547,
+      "step": 431
+    },
+    {
+      "epoch": 0.07706716617607706,
+      "grad_norm": 0.9180254340171814,
+      "learning_rate": 0.0004981700657143396,
+      "loss": 1.6352,
+      "step": 432
+    },
+    {
+      "epoch": 0.07724556239407725,
+      "grad_norm": 0.703322172164917,
+      "learning_rate": 0.0004981615944014071,
+      "loss": 1.3254,
+      "step": 433
+    },
+    {
+      "epoch": 0.07742395861207742,
+      "grad_norm": 0.794771134853363,
+      "learning_rate": 0.0004981531035979697,
+      "loss": 1.6643,
+      "step": 434
+    },
+    {
+      "epoch": 0.0776023548300776,
+      "grad_norm": 0.7554542422294617,
+      "learning_rate": 0.0004981445933046944,
+      "loss": 1.393,
+      "step": 435
+    },
+    {
+      "epoch": 0.07778075104807779,
+      "grad_norm": 0.8355424404144287,
+      "learning_rate": 0.0004981360635222495,
+      "loss": 1.7901,
+      "step": 436
+    },
+    {
+      "epoch": 0.07795914726607796,
+      "grad_norm": 0.6845025420188904,
+      "learning_rate": 0.0004981275142513049,
+      "loss": 1.6268,
+      "step": 437
+    },
+    {
+      "epoch": 0.07813754348407814,
+      "grad_norm": 0.7474300265312195,
+      "learning_rate": 0.0004981189454925322,
+      "loss": 1.2496,
+      "step": 438
+    },
+    {
+      "epoch": 0.07831593970207831,
+      "grad_norm": 0.6945734620094299,
+      "learning_rate": 0.0004981103572466042,
+      "loss": 1.7841,
+      "step": 439
+    },
+    {
+      "epoch": 0.0784943359200785,
+      "grad_norm": 0.6879515051841736,
+      "learning_rate": 0.0004981017495141955,
+      "loss": 1.4973,
+      "step": 440
+    },
+    {
+      "epoch": 0.07867273213807867,
+      "grad_norm": 0.9907708168029785,
+      "learning_rate": 0.0004980931222959823,
+      "loss": 1.9778,
+      "step": 441
+    },
+    {
+      "epoch": 0.07885112835607885,
+      "grad_norm": 1.6716070175170898,
+      "learning_rate": 0.000498084475592642,
+      "loss": 1.8797,
+      "step": 442
+    },
+    {
+      "epoch": 0.07902952457407902,
+      "grad_norm": 0.8019148707389832,
+      "learning_rate": 0.0004980758094048536,
+      "loss": 1.4991,
+      "step": 443
+    },
+    {
+      "epoch": 0.07920792079207921,
+      "grad_norm": 0.7047806978225708,
+      "learning_rate": 0.000498067123733298,
+      "loss": 1.4349,
+      "step": 444
+    },
+    {
+      "epoch": 0.07938631701007938,
+      "grad_norm": 0.7296415567398071,
+      "learning_rate": 0.0004980584185786573,
+      "loss": 1.504,
+      "step": 445
+    },
+    {
+      "epoch": 0.07956471322807956,
+      "grad_norm": 0.8090648651123047,
+      "learning_rate": 0.0004980496939416151,
+      "loss": 1.5402,
+      "step": 446
+    },
+    {
+      "epoch": 0.07974310944607975,
+      "grad_norm": 0.7694994807243347,
+      "learning_rate": 0.0004980409498228566,
+      "loss": 1.6298,
+      "step": 447
+    },
+    {
+      "epoch": 0.07992150566407992,
+      "grad_norm": 1.529492735862732,
+      "learning_rate": 0.0004980321862230688,
+      "loss": 1.4919,
+      "step": 448
+    },
+    {
+      "epoch": 0.0800999018820801,
+      "grad_norm": 0.7055628299713135,
+      "learning_rate": 0.0004980234031429397,
+      "loss": 1.3756,
+      "step": 449
+    },
+    {
+      "epoch": 0.08027829810008028,
+      "grad_norm": 0.7243545055389404,
+      "learning_rate": 0.0004980146005831592,
+      "loss": 1.4947,
+      "step": 450
+    },
+    {
+      "epoch": 0.08045669431808046,
+      "grad_norm": 0.7006012797355652,
+      "learning_rate": 0.0004980057785444189,
+      "loss": 1.6987,
+      "step": 451
+    },
+    {
+      "epoch": 0.08063509053608063,
+      "grad_norm": 0.7193490862846375,
+      "learning_rate": 0.0004979969370274113,
+      "loss": 1.6076,
+      "step": 452
+    },
+    {
+      "epoch": 0.08081348675408082,
+      "grad_norm": 0.7039264440536499,
+      "learning_rate": 0.0004979880760328312,
+      "loss": 1.708,
+      "step": 453
+    },
+    {
+      "epoch": 0.08099188297208099,
+      "grad_norm": 0.7600789666175842,
+      "learning_rate": 0.0004979791955613741,
+      "loss": 1.9332,
+      "step": 454
+    },
+    {
+      "epoch": 0.08117027919008117,
+      "grad_norm": 4.260866165161133,
+      "learning_rate": 0.0004979702956137378,
+      "loss": 1.5614,
+      "step": 455
+    },
+    {
+      "epoch": 0.08134867540808134,
+      "grad_norm": 0.7909694314002991,
+      "learning_rate": 0.0004979613761906212,
+      "loss": 1.5609,
+      "step": 456
+    },
+    {
+      "epoch": 0.08152707162608153,
+      "grad_norm": 1.7039103507995605,
+      "learning_rate": 0.0004979524372927248,
+      "loss": 1.6149,
+      "step": 457
+    },
+    {
+      "epoch": 0.08170546784408171,
+      "grad_norm": 0.7937911152839661,
+      "learning_rate": 0.0004979434789207506,
+      "loss": 1.4368,
+      "step": 458
+    },
+    {
+      "epoch": 0.08188386406208188,
+      "grad_norm": 1.0153762102127075,
+      "learning_rate": 0.0004979345010754024,
+      "loss": 1.4232,
+      "step": 459
+    },
+    {
+      "epoch": 0.08206226028008207,
+      "grad_norm": 0.7721362709999084,
+      "learning_rate": 0.0004979255037573851,
+      "loss": 1.6384,
+      "step": 460
+    },
+    {
+      "epoch": 0.08224065649808224,
+      "grad_norm": 0.6385205388069153,
+      "learning_rate": 0.0004979164869674055,
+      "loss": 1.2598,
+      "step": 461
+    },
+    {
+      "epoch": 0.08241905271608242,
+      "grad_norm": 0.7460764646530151,
+      "learning_rate": 0.0004979074507061716,
+      "loss": 1.4009,
+      "step": 462
+    },
+    {
+      "epoch": 0.08259744893408259,
+      "grad_norm": 0.6863163113594055,
+      "learning_rate": 0.0004978983949743932,
+      "loss": 1.156,
+      "step": 463
+    },
+    {
+      "epoch": 0.08277584515208278,
+      "grad_norm": 0.9043185710906982,
+      "learning_rate": 0.0004978893197727817,
+      "loss": 1.4024,
+      "step": 464
+    },
+    {
+      "epoch": 0.08295424137008295,
+      "grad_norm": 0.8068817853927612,
+      "learning_rate": 0.0004978802251020494,
+      "loss": 1.5392,
+      "step": 465
+    },
+    {
+      "epoch": 0.08313263758808313,
+      "grad_norm": 0.8445210456848145,
+      "learning_rate": 0.0004978711109629112,
+      "loss": 1.3355,
+      "step": 466
+    },
+    {
+      "epoch": 0.08331103380608332,
+      "grad_norm": 0.9372073411941528,
+      "learning_rate": 0.0004978619773560824,
+      "loss": 1.5414,
+      "step": 467
+    },
+    {
+      "epoch": 0.08348943002408349,
+      "grad_norm": 0.7277926206588745,
+      "learning_rate": 0.0004978528242822807,
+      "loss": 1.3025,
+      "step": 468
+    },
+    {
+      "epoch": 0.08366782624208367,
+      "grad_norm": 0.8729844093322754,
+      "learning_rate": 0.0004978436517422248,
+      "loss": 1.5157,
+      "step": 469
+    },
+    {
+      "epoch": 0.08384622246008384,
+      "grad_norm": 0.6946988105773926,
+      "learning_rate": 0.0004978344597366351,
+      "loss": 1.5723,
+      "step": 470
+    },
+    {
+      "epoch": 0.08402461867808403,
+      "grad_norm": 0.6131045818328857,
+      "learning_rate": 0.0004978252482662337,
+      "loss": 1.3419,
+      "step": 471
+    },
+    {
+      "epoch": 0.0842030148960842,
+      "grad_norm": 0.8643115758895874,
+      "learning_rate": 0.0004978160173317438,
+      "loss": 1.609,
+      "step": 472
+    },
+    {
+      "epoch": 0.08438141111408438,
+      "grad_norm": 0.8250529170036316,
+      "learning_rate": 0.0004978067669338906,
+      "loss": 1.7085,
+      "step": 473
+    },
+    {
+      "epoch": 0.08455980733208456,
+      "grad_norm": 0.8330903649330139,
+      "learning_rate": 0.0004977974970734006,
+      "loss": 1.5074,
+      "step": 474
+    },
+    {
+      "epoch": 0.08473820355008474,
+      "grad_norm": 0.7635675072669983,
+      "learning_rate": 0.0004977882077510018,
+      "loss": 1.4812,
+      "step": 475
+    },
+    {
+      "epoch": 0.08491659976808491,
+      "grad_norm": 1.1247259378433228,
+      "learning_rate": 0.0004977788989674238,
+      "loss": 1.4147,
+      "step": 476
+    },
+    {
+      "epoch": 0.0850949959860851,
+      "grad_norm": 0.6841705441474915,
+      "learning_rate": 0.0004977695707233977,
+      "loss": 1.3934,
+      "step": 477
+    },
+    {
+      "epoch": 0.08527339220408528,
+      "grad_norm": 0.7214035987854004,
+      "learning_rate": 0.0004977602230196561,
+      "loss": 1.3389,
+      "step": 478
+    },
+    {
+      "epoch": 0.08545178842208545,
+      "grad_norm": 0.7651268243789673,
+      "learning_rate": 0.0004977508558569332,
+      "loss": 1.5814,
+      "step": 479
+    },
+    {
+      "epoch": 0.08563018464008564,
+      "grad_norm": 0.7980949282646179,
+      "learning_rate": 0.0004977414692359648,
+      "loss": 1.5766,
+      "step": 480
+    },
+    {
+      "epoch": 0.0858085808580858,
+      "grad_norm": 2.5435357093811035,
+      "learning_rate": 0.0004977320631574879,
+      "loss": 1.4437,
+      "step": 481
+    },
+    {
+      "epoch": 0.08598697707608599,
+      "grad_norm": 0.8370219469070435,
+      "learning_rate": 0.0004977226376222415,
+      "loss": 1.3928,
+      "step": 482
+    },
+    {
+      "epoch": 0.08616537329408616,
+      "grad_norm": 0.6488362550735474,
+      "learning_rate": 0.0004977131926309656,
+      "loss": 1.6902,
+      "step": 483
+    },
+    {
+      "epoch": 0.08634376951208635,
+      "grad_norm": 0.7219479084014893,
+      "learning_rate": 0.0004977037281844023,
+      "loss": 1.5692,
+      "step": 484
+    },
+    {
+      "epoch": 0.08652216573008652,
+      "grad_norm": 1.9494524002075195,
+      "learning_rate": 0.0004976942442832946,
+      "loss": 1.5357,
+      "step": 485
+    },
+    {
+      "epoch": 0.0867005619480867,
+      "grad_norm": 0.8252092003822327,
+      "learning_rate": 0.0004976847409283876,
+      "loss": 1.566,
+      "step": 486
+    },
+    {
+      "epoch": 0.08687895816608687,
+      "grad_norm": 0.624910295009613,
+      "learning_rate": 0.0004976752181204277,
+      "loss": 1.5799,
+      "step": 487
+    },
+    {
+      "epoch": 0.08705735438408706,
+      "grad_norm": 0.8229494094848633,
+      "learning_rate": 0.0004976656758601628,
+      "loss": 1.3698,
+      "step": 488
+    },
+    {
+      "epoch": 0.08723575060208724,
+      "grad_norm": 0.8363699316978455,
+      "learning_rate": 0.0004976561141483421,
+      "loss": 1.3137,
+      "step": 489
+    },
+    {
+      "epoch": 0.08741414682008741,
+      "grad_norm": 0.665203869342804,
+      "learning_rate": 0.0004976465329857169,
+      "loss": 1.579,
+      "step": 490
+    },
+    {
+      "epoch": 0.0875925430380876,
+      "grad_norm": 0.7317224144935608,
+      "learning_rate": 0.0004976369323730396,
+      "loss": 1.6404,
+      "step": 491
+    },
+    {
+      "epoch": 0.08777093925608777,
+      "grad_norm": 0.7230332493782043,
+      "learning_rate": 0.0004976273123110642,
+      "loss": 1.5054,
+      "step": 492
+    },
+    {
+      "epoch": 0.08794933547408795,
+      "grad_norm": 0.9114980697631836,
+      "learning_rate": 0.0004976176728005462,
+      "loss": 1.5787,
+      "step": 493
+    },
+    {
+      "epoch": 0.08812773169208812,
+      "grad_norm": 0.7472654581069946,
+      "learning_rate": 0.0004976080138422428,
+      "loss": 1.6348,
+      "step": 494
+    },
+    {
+      "epoch": 0.08830612791008831,
+      "grad_norm": 0.7824138402938843,
+      "learning_rate": 0.0004975983354369124,
+      "loss": 1.5032,
+      "step": 495
+    },
+    {
+      "epoch": 0.08848452412808848,
+      "grad_norm": 0.6334506869316101,
+      "learning_rate": 0.0004975886375853155,
+      "loss": 1.3295,
+      "step": 496
+    },
+    {
+      "epoch": 0.08866292034608866,
+      "grad_norm": 1.3684911727905273,
+      "learning_rate": 0.0004975789202882133,
+      "loss": 1.6436,
+      "step": 497
+    },
+    {
+      "epoch": 0.08884131656408883,
+      "grad_norm": 0.7262062430381775,
+      "learning_rate": 0.0004975691835463694,
+      "loss": 1.4747,
+      "step": 498
+    },
+    {
+      "epoch": 0.08901971278208902,
+      "grad_norm": 0.6823199987411499,
+      "learning_rate": 0.0004975594273605484,
+      "loss": 1.4783,
+      "step": 499
+    },
+    {
+      "epoch": 0.0891981090000892,
+      "grad_norm": 0.5289890766143799,
+      "learning_rate": 0.0004975496517315164,
+      "loss": 1.1087,
+      "step": 500
+    },
+    {
+      "epoch": 0.08937650521808937,
+      "grad_norm": 0.8217448592185974,
+      "learning_rate": 0.0004975398566600413,
+      "loss": 1.9442,
+      "step": 501
+    },
+    {
+      "epoch": 0.08955490143608956,
+      "grad_norm": 0.6704925894737244,
+      "learning_rate": 0.0004975300421468925,
+      "loss": 1.6343,
+      "step": 502
+    },
+    {
+      "epoch": 0.08973329765408973,
+      "grad_norm": 0.8133907914161682,
+      "learning_rate": 0.0004975202081928405,
+      "loss": 1.7413,
+      "step": 503
+    },
+    {
+      "epoch": 0.08991169387208992,
+      "grad_norm": 0.6570431590080261,
+      "learning_rate": 0.0004975103547986581,
+      "loss": 1.4866,
+      "step": 504
+    },
+    {
+      "epoch": 0.09009009009009009,
+      "grad_norm": 0.7188069820404053,
+      "learning_rate": 0.0004975004819651188,
+      "loss": 1.6916,
+      "step": 505
+    },
+    {
+      "epoch": 0.09026848630809027,
+      "grad_norm": 1.0071742534637451,
+      "learning_rate": 0.0004974905896929981,
+      "loss": 1.458,
+      "step": 506
+    },
+    {
+      "epoch": 0.09044688252609044,
+      "grad_norm": 0.7110443711280823,
+      "learning_rate": 0.0004974806779830731,
+      "loss": 1.4134,
+      "step": 507
+    },
+    {
+      "epoch": 0.09062527874409063,
+      "grad_norm": 0.6911363005638123,
+      "learning_rate": 0.0004974707468361221,
+      "loss": 1.5558,
+      "step": 508
+    },
+    {
+      "epoch": 0.0908036749620908,
+      "grad_norm": 0.632495641708374,
+      "learning_rate": 0.0004974607962529252,
+      "loss": 1.2942,
+      "step": 509
+    },
+    {
+      "epoch": 0.09098207118009098,
+      "grad_norm": 0.7343394160270691,
+      "learning_rate": 0.0004974508262342638,
+      "loss": 1.5239,
+      "step": 510
+    },
+    {
+      "epoch": 0.09116046739809117,
+      "grad_norm": 1.4106143712997437,
+      "learning_rate": 0.000497440836780921,
+      "loss": 1.3772,
+      "step": 511
+    },
+    {
+      "epoch": 0.09133886361609134,
+      "grad_norm": 0.7864501476287842,
+      "learning_rate": 0.0004974308278936813,
+      "loss": 1.672,
+      "step": 512
+    },
+    {
+      "epoch": 0.09151725983409152,
+      "grad_norm": 0.7116146087646484,
+      "learning_rate": 0.000497420799573331,
+      "loss": 1.3471,
+      "step": 513
+    },
+    {
+      "epoch": 0.09169565605209169,
+      "grad_norm": 0.6454455256462097,
+      "learning_rate": 0.0004974107518206575,
+      "loss": 1.5374,
+      "step": 514
+    },
+    {
+      "epoch": 0.09187405227009188,
+      "grad_norm": 1.5066065788269043,
+      "learning_rate": 0.00049740068463645,
+      "loss": 1.1575,
+      "step": 515
+    },
+    {
+      "epoch": 0.09205244848809205,
+      "grad_norm": 0.8750366568565369,
+      "learning_rate": 0.0004973905980214992,
+      "loss": 1.342,
+      "step": 516
+    },
+    {
+      "epoch": 0.09223084470609223,
+      "grad_norm": 1.1781761646270752,
+      "learning_rate": 0.0004973804919765973,
+      "loss": 1.5306,
+      "step": 517
+    },
+    {
+      "epoch": 0.0924092409240924,
+      "grad_norm": 0.7154638767242432,
+      "learning_rate": 0.0004973703665025381,
+      "loss": 1.3626,
+      "step": 518
+    },
+    {
+      "epoch": 0.09258763714209259,
+      "grad_norm": 0.8861932158470154,
+      "learning_rate": 0.0004973602216001166,
+      "loss": 1.4045,
+      "step": 519
+    },
+    {
+      "epoch": 0.09276603336009277,
+      "grad_norm": 0.8264048099517822,
+      "learning_rate": 0.0004973500572701299,
+      "loss": 1.5786,
+      "step": 520
+    },
+    {
+      "epoch": 0.09294442957809294,
+      "grad_norm": 0.6503618955612183,
+      "learning_rate": 0.0004973398735133761,
+      "loss": 1.2953,
+      "step": 521
+    },
+    {
+      "epoch": 0.09312282579609313,
+      "grad_norm": 1.5494425296783447,
+      "learning_rate": 0.000497329670330655,
+      "loss": 1.7067,
+      "step": 522
+    },
+    {
+      "epoch": 0.0933012220140933,
+      "grad_norm": 0.7740722298622131,
+      "learning_rate": 0.0004973194477227681,
+      "loss": 1.7424,
+      "step": 523
+    },
+    {
+      "epoch": 0.09347961823209348,
+      "grad_norm": 0.6886745095252991,
+      "learning_rate": 0.0004973092056905181,
+      "loss": 1.4517,
+      "step": 524
+    },
+    {
+      "epoch": 0.09365801445009365,
+      "grad_norm": 0.9685157537460327,
+      "learning_rate": 0.0004972989442347097,
+      "loss": 1.3515,
+      "step": 525
+    },
+    {
+      "epoch": 0.09383641066809384,
+      "grad_norm": 0.6641685962677002,
+      "learning_rate": 0.0004972886633561486,
+      "loss": 1.4647,
+      "step": 526
+    },
+    {
+      "epoch": 0.09401480688609401,
+      "grad_norm": 0.705767035484314,
+      "learning_rate": 0.0004972783630556421,
+      "loss": 1.6306,
+      "step": 527
+    },
+    {
+      "epoch": 0.0941932031040942,
+      "grad_norm": 0.6863012909889221,
+      "learning_rate": 0.0004972680433339995,
+      "loss": 1.5197,
+      "step": 528
+    },
+    {
+      "epoch": 0.09437159932209437,
+      "grad_norm": 3.830077886581421,
+      "learning_rate": 0.0004972577041920312,
+      "loss": 1.2183,
+      "step": 529
+    },
+    {
+      "epoch": 0.09454999554009455,
+      "grad_norm": 0.8609939217567444,
+      "learning_rate": 0.0004972473456305493,
+      "loss": 1.5419,
+      "step": 530
+    },
+    {
+      "epoch": 0.09472839175809473,
+      "grad_norm": 1.5111016035079956,
+      "learning_rate": 0.0004972369676503671,
+      "loss": 1.4507,
+      "step": 531
+    },
+    {
+      "epoch": 0.0949067879760949,
+      "grad_norm": 0.8439610600471497,
+      "learning_rate": 0.0004972265702523,
+      "loss": 1.5155,
+      "step": 532
+    },
+    {
+      "epoch": 0.09508518419409509,
+      "grad_norm": 0.9149817228317261,
+      "learning_rate": 0.0004972161534371643,
+      "loss": 1.727,
+      "step": 533
+    },
+    {
+      "epoch": 0.09526358041209526,
+      "grad_norm": 1.0200440883636475,
+      "learning_rate": 0.0004972057172057784,
+      "loss": 1.6166,
+      "step": 534
+    },
+    {
+      "epoch": 0.09544197663009545,
+      "grad_norm": 0.9794648885726929,
+      "learning_rate": 0.0004971952615589619,
+      "loss": 1.5486,
+      "step": 535
+    },
+    {
+      "epoch": 0.09562037284809562,
+      "grad_norm": 1.2553775310516357,
+      "learning_rate": 0.0004971847864975358,
+      "loss": 1.5297,
+      "step": 536
+    },
+    {
+      "epoch": 0.0957987690660958,
+      "grad_norm": 0.9116058349609375,
+      "learning_rate": 0.0004971742920223231,
+      "loss": 1.7309,
+      "step": 537
+    },
+    {
+      "epoch": 0.09597716528409597,
+      "grad_norm": 1.0907950401306152,
+      "learning_rate": 0.0004971637781341478,
+      "loss": 1.4598,
+      "step": 538
+    },
+    {
+      "epoch": 0.09615556150209616,
+      "grad_norm": 1.32576322555542,
+      "learning_rate": 0.0004971532448338357,
+      "loss": 1.3408,
+      "step": 539
+    },
+    {
+      "epoch": 0.09633395772009633,
+      "grad_norm": 0.7169795036315918,
+      "learning_rate": 0.0004971426921222141,
+      "loss": 1.3837,
+      "step": 540
+    },
+    {
+      "epoch": 0.09651235393809651,
+      "grad_norm": 1.4852778911590576,
+      "learning_rate": 0.0004971321200001119,
+      "loss": 1.3068,
+      "step": 541
+    },
+    {
+      "epoch": 0.0966907501560967,
+      "grad_norm": 5.810755729675293,
+      "learning_rate": 0.0004971215284683592,
+      "loss": 1.3056,
+      "step": 542
+    },
+    {
+      "epoch": 0.09686914637409687,
+      "grad_norm": 1.1541168689727783,
+      "learning_rate": 0.0004971109175277882,
+      "loss": 1.3768,
+      "step": 543
+    },
+    {
+      "epoch": 0.09704754259209705,
+      "grad_norm": 0.8170933127403259,
+      "learning_rate": 0.000497100287179232,
+      "loss": 1.6132,
+      "step": 544
+    },
+    {
+      "epoch": 0.09722593881009722,
+      "grad_norm": 0.8202213644981384,
+      "learning_rate": 0.0004970896374235255,
+      "loss": 1.7538,
+      "step": 545
+    },
+    {
+      "epoch": 0.09740433502809741,
+      "grad_norm": 0.7768821120262146,
+      "learning_rate": 0.0004970789682615052,
+      "loss": 1.5705,
+      "step": 546
+    },
+    {
+      "epoch": 0.09758273124609758,
+      "grad_norm": 1.4286799430847168,
+      "learning_rate": 0.0004970682796940091,
+      "loss": 1.5751,
+      "step": 547
+    },
+    {
+      "epoch": 0.09776112746409776,
+      "grad_norm": 2.075110912322998,
+      "learning_rate": 0.0004970575717218767,
+      "loss": 1.5745,
+      "step": 548
+    },
+    {
+      "epoch": 0.09793952368209793,
+      "grad_norm": 0.8903990387916565,
+      "learning_rate": 0.0004970468443459489,
+      "loss": 1.4809,
+      "step": 549
+    },
+    {
+      "epoch": 0.09811791990009812,
+      "grad_norm": 0.8855850696563721,
+      "learning_rate": 0.0004970360975670682,
+      "loss": 1.7035,
+      "step": 550
+    },
+    {
+      "epoch": 0.09829631611809829,
+      "grad_norm": 0.8159774541854858,
+      "learning_rate": 0.0004970253313860787,
+      "loss": 1.2233,
+      "step": 551
+    },
+    {
+      "epoch": 0.09847471233609847,
+      "grad_norm": 2.709984302520752,
+      "learning_rate": 0.0004970145458038261,
+      "loss": 1.497,
+      "step": 552
+    },
+    {
+      "epoch": 0.09865310855409866,
+      "grad_norm": 1.0269032716751099,
+      "learning_rate": 0.0004970037408211573,
+      "loss": 1.5952,
+      "step": 553
+    },
+    {
+      "epoch": 0.09883150477209883,
+      "grad_norm": 1.1201132535934448,
+      "learning_rate": 0.000496992916438921,
+      "loss": 1.3275,
+      "step": 554
+    },
+    {
+      "epoch": 0.09900990099009901,
+      "grad_norm": 0.8947777152061462,
+      "learning_rate": 0.0004969820726579673,
+      "loss": 1.4586,
+      "step": 555
+    },
+    {
+      "epoch": 0.09918829720809919,
+      "grad_norm": 0.9649630784988403,
+      "learning_rate": 0.0004969712094791479,
+      "loss": 1.4422,
+      "step": 556
+    },
+    {
+      "epoch": 0.09936669342609937,
+      "grad_norm": 0.9738620519638062,
+      "learning_rate": 0.0004969603269033159,
+      "loss": 1.4171,
+      "step": 557
+    },
+    {
+      "epoch": 0.09954508964409954,
+      "grad_norm": 0.7529584169387817,
+      "learning_rate": 0.0004969494249313262,
+      "loss": 1.2664,
+      "step": 558
+    },
+    {
+      "epoch": 0.09972348586209973,
+      "grad_norm": 0.8182567358016968,
+      "learning_rate": 0.000496938503564035,
+      "loss": 1.5923,
+      "step": 559
+    },
+    {
+      "epoch": 0.0999018820800999,
+      "grad_norm": 0.7630960941314697,
+      "learning_rate": 0.0004969275628023,
+      "loss": 1.3102,
+      "step": 560
+    },
+    {
+      "epoch": 0.10008027829810008,
+      "grad_norm": 0.7095026969909668,
+      "learning_rate": 0.0004969166026469803,
+      "loss": 1.3446,
+      "step": 561
+    },
+    {
+      "epoch": 0.10025867451610025,
+      "grad_norm": 0.7493692636489868,
+      "learning_rate": 0.0004969056230989371,
+      "loss": 1.5735,
+      "step": 562
+    },
+    {
+      "epoch": 0.10043707073410044,
+      "grad_norm": 0.6589444279670715,
+      "learning_rate": 0.0004968946241590324,
+      "loss": 1.2374,
+      "step": 563
+    },
+    {
+      "epoch": 0.10061546695210062,
+      "grad_norm": 1.1761233806610107,
+      "learning_rate": 0.0004968836058281301,
+      "loss": 1.4974,
+      "step": 564
+    },
+    {
+      "epoch": 0.10079386317010079,
+      "grad_norm": 0.6165998578071594,
+      "learning_rate": 0.0004968725681070957,
+      "loss": 1.1654,
+      "step": 565
+    },
+    {
+      "epoch": 0.10097225938810098,
+      "grad_norm": 0.7280291318893433,
+      "learning_rate": 0.0004968615109967961,
+      "loss": 1.4225,
+      "step": 566
+    },
+    {
+      "epoch": 0.10115065560610115,
+      "grad_norm": 1.2140183448791504,
+      "learning_rate": 0.0004968504344980997,
+      "loss": 1.6689,
+      "step": 567
+    },
+    {
+      "epoch": 0.10132905182410133,
+      "grad_norm": 0.7183876037597656,
+      "learning_rate": 0.0004968393386118763,
+      "loss": 1.5669,
+      "step": 568
+    },
+    {
+      "epoch": 0.1015074480421015,
+      "grad_norm": 0.6638524532318115,
+      "learning_rate": 0.0004968282233389974,
+      "loss": 1.1266,
+      "step": 569
+    },
+    {
+      "epoch": 0.10168584426010169,
+      "grad_norm": 0.7190263271331787,
+      "learning_rate": 0.0004968170886803362,
+      "loss": 1.363,
+      "step": 570
+    },
+    {
+      "epoch": 0.10186424047810186,
+      "grad_norm": 0.8314970135688782,
+      "learning_rate": 0.000496805934636767,
+      "loss": 1.6896,
+      "step": 571
+    },
+    {
+      "epoch": 0.10204263669610204,
+      "grad_norm": 0.6682543158531189,
+      "learning_rate": 0.0004967947612091659,
+      "loss": 1.4487,
+      "step": 572
+    },
+    {
+      "epoch": 0.10222103291410223,
+      "grad_norm": 0.806235134601593,
+      "learning_rate": 0.0004967835683984105,
+      "loss": 1.3844,
+      "step": 573
+    },
+    {
+      "epoch": 0.1023994291321024,
+      "grad_norm": 0.7934510707855225,
+      "learning_rate": 0.0004967723562053798,
+      "loss": 1.5275,
+      "step": 574
+    },
+    {
+      "epoch": 0.10257782535010258,
+      "grad_norm": 0.6725155115127563,
+      "learning_rate": 0.0004967611246309544,
+      "loss": 1.3457,
+      "step": 575
+    },
+    {
+      "epoch": 0.10275622156810275,
+      "grad_norm": 1.1390475034713745,
+      "learning_rate": 0.0004967498736760165,
+      "loss": 1.5195,
+      "step": 576
+    },
+    {
+      "epoch": 0.10293461778610294,
+      "grad_norm": 0.7463436126708984,
+      "learning_rate": 0.0004967386033414498,
+      "loss": 1.4778,
+      "step": 577
+    },
+    {
+      "epoch": 0.10311301400410311,
+      "grad_norm": 0.6401159763336182,
+      "learning_rate": 0.0004967273136281392,
+      "loss": 1.0325,
+      "step": 578
+    },
+    {
+      "epoch": 0.1032914102221033,
+      "grad_norm": 0.7498372793197632,
+      "learning_rate": 0.0004967160045369716,
+      "loss": 1.6773,
+      "step": 579
+    },
+    {
+      "epoch": 0.10346980644010347,
+      "grad_norm": 0.6850476861000061,
+      "learning_rate": 0.0004967046760688353,
+      "loss": 1.4975,
+      "step": 580
+    },
+    {
+      "epoch": 0.10364820265810365,
+      "grad_norm": 0.6645762324333191,
+      "learning_rate": 0.0004966933282246198,
+      "loss": 1.5224,
+      "step": 581
+    },
+    {
+      "epoch": 0.10382659887610382,
+      "grad_norm": 0.9928523898124695,
+      "learning_rate": 0.0004966819610052165,
+      "loss": 1.2436,
+      "step": 582
+    },
+    {
+      "epoch": 0.104004995094104,
+      "grad_norm": 2.335838794708252,
+      "learning_rate": 0.0004966705744115182,
+      "loss": 1.4003,
+      "step": 583
+    },
+    {
+      "epoch": 0.10418339131210419,
+      "grad_norm": 0.7920882701873779,
+      "learning_rate": 0.0004966591684444191,
+      "loss": 1.3903,
+      "step": 584
+    },
+    {
+      "epoch": 0.10436178753010436,
+      "grad_norm": 1.7378654479980469,
+      "learning_rate": 0.000496647743104815,
+      "loss": 1.5465,
+      "step": 585
+    },
+    {
+      "epoch": 0.10454018374810455,
+      "grad_norm": 1.0265392065048218,
+      "learning_rate": 0.0004966362983936033,
+      "loss": 1.639,
+      "step": 586
+    },
+    {
+      "epoch": 0.10471857996610472,
+      "grad_norm": 25.1031436920166,
+      "learning_rate": 0.0004966248343116828,
+      "loss": 1.8437,
+      "step": 587
+    },
+    {
+      "epoch": 0.1048969761841049,
+      "grad_norm": 2.059586524963379,
+      "learning_rate": 0.0004966133508599541,
+      "loss": 1.4263,
+      "step": 588
+    },
+    {
+      "epoch": 0.10507537240210507,
+      "grad_norm": 1.6019502878189087,
+      "learning_rate": 0.0004966018480393188,
+      "loss": 1.2302,
+      "step": 589
+    },
+    {
+      "epoch": 0.10525376862010526,
+      "grad_norm": 7.920602798461914,
+      "learning_rate": 0.0004965903258506806,
+      "loss": 1.5145,
+      "step": 590
+    },
+    {
+      "epoch": 0.10543216483810543,
+      "grad_norm": 8.575254440307617,
+      "learning_rate": 0.0004965787842949443,
+      "loss": 1.138,
+      "step": 591
+    },
+    {
+      "epoch": 0.10561056105610561,
+      "grad_norm": 9.764083862304688,
+      "learning_rate": 0.0004965672233730163,
+      "loss": 1.8386,
+      "step": 592
+    },
+    {
+      "epoch": 0.10578895727410578,
+      "grad_norm": 2.423666000366211,
+      "learning_rate": 0.0004965556430858049,
+      "loss": 1.2881,
+      "step": 593
+    },
+    {
+      "epoch": 0.10596735349210597,
+      "grad_norm": 1.2473442554473877,
+      "learning_rate": 0.0004965440434342191,
+      "loss": 1.1922,
+      "step": 594
+    },
+    {
+      "epoch": 0.10614574971010615,
+      "grad_norm": 0.9045082330703735,
+      "learning_rate": 0.0004965324244191704,
+      "loss": 1.3325,
+      "step": 595
+    },
+    {
+      "epoch": 0.10632414592810632,
+      "grad_norm": 1.219525694847107,
+      "learning_rate": 0.0004965207860415711,
+      "loss": 1.6335,
+      "step": 596
+    },
+    {
+      "epoch": 0.10650254214610651,
+      "grad_norm": 5.38252592086792,
+      "learning_rate": 0.0004965091283023355,
+      "loss": 1.425,
+      "step": 597
+    },
+    {
+      "epoch": 0.10668093836410668,
+      "grad_norm": 2.2056992053985596,
+      "learning_rate": 0.0004964974512023789,
+      "loss": 1.5319,
+      "step": 598
+    },
+    {
+      "epoch": 0.10685933458210686,
+      "grad_norm": 0.9872840642929077,
+      "learning_rate": 0.0004964857547426186,
+      "loss": 1.1613,
+      "step": 599
+    },
+    {
+      "epoch": 0.10703773080010703,
+      "grad_norm": 1.287277102470398,
+      "learning_rate": 0.0004964740389239732,
+      "loss": 1.6958,
+      "step": 600
+    },
+    {
+      "epoch": 0.10721612701810722,
+      "grad_norm": 1.0453811883926392,
+      "learning_rate": 0.0004964623037473628,
+      "loss": 1.7414,
+      "step": 601
+    },
+    {
+      "epoch": 0.10739452323610739,
+      "grad_norm": 1.040404200553894,
+      "learning_rate": 0.0004964505492137092,
+      "loss": 1.482,
+      "step": 602
+    },
+    {
+      "epoch": 0.10757291945410757,
+      "grad_norm": 0.5612449645996094,
+      "learning_rate": 0.0004964387753239355,
+      "loss": 1.3086,
+      "step": 603
+    },
+    {
+      "epoch": 0.10775131567210774,
+      "grad_norm": 42.32016372680664,
+      "learning_rate": 0.0004964269820789664,
+      "loss": 1.3425,
+      "step": 604
+    },
+    {
+      "epoch": 0.10792971189010793,
+      "grad_norm": 51.25515365600586,
+      "learning_rate": 0.0004964151694797282,
+      "loss": 1.6645,
+      "step": 605
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 1.5014013051986694,
+      "learning_rate": 0.0004964033375271486,
+      "loss": 1.5083,
+      "step": 606
+    },
+    {
+      "epoch": 0.10828650432610828,
+      "grad_norm": 2.11055588722229,
+      "learning_rate": 0.0004963914862221569,
+      "loss": 1.2624,
+      "step": 607
+    },
+    {
+      "epoch": 0.10846490054410847,
+      "grad_norm": 1.610378623008728,
+      "learning_rate": 0.000496379615565684,
+      "loss": 1.5693,
+      "step": 608
+    },
+    {
+      "epoch": 0.10864329676210864,
+      "grad_norm": 0.88837069272995,
+      "learning_rate": 0.000496367725558662,
+      "loss": 1.5379,
+      "step": 609
+    },
+    {
+      "epoch": 0.10882169298010883,
+      "grad_norm": 1.831912875175476,
+      "learning_rate": 0.000496355816202025,
+      "loss": 1.3983,
+      "step": 610
+    },
+    {
+      "epoch": 0.109000089198109,
+      "grad_norm": 0.7277733087539673,
+      "learning_rate": 0.0004963438874967081,
+      "loss": 1.5717,
+      "step": 611
+    },
+    {
+      "epoch": 0.10917848541610918,
+      "grad_norm": 0.7317838072776794,
+      "learning_rate": 0.0004963319394436483,
+      "loss": 1.2432,
+      "step": 612
+    },
+    {
+      "epoch": 0.10935688163410935,
+      "grad_norm": 5.225924491882324,
+      "learning_rate": 0.000496319972043784,
+      "loss": 1.3503,
+      "step": 613
+    },
+    {
+      "epoch": 0.10953527785210954,
+      "grad_norm": 1.8268285989761353,
+      "learning_rate": 0.0004963079852980551,
+      "loss": 1.5332,
+      "step": 614
+    },
+    {
+      "epoch": 0.1097136740701097,
+      "grad_norm": 0.9872994422912598,
+      "learning_rate": 0.000496295979207403,
+      "loss": 1.5261,
+      "step": 615
+    },
+    {
+      "epoch": 0.10989207028810989,
+      "grad_norm": 8.904617309570312,
+      "learning_rate": 0.0004962839537727706,
+      "loss": 1.4517,
+      "step": 616
+    },
+    {
+      "epoch": 0.11007046650611008,
+      "grad_norm": 1.2641127109527588,
+      "learning_rate": 0.0004962719089951027,
+      "loss": 1.6554,
+      "step": 617
+    },
+    {
+      "epoch": 0.11024886272411025,
+      "grad_norm": 0.7095901370048523,
+      "learning_rate": 0.0004962598448753448,
+      "loss": 1.5814,
+      "step": 618
+    },
+    {
+      "epoch": 0.11042725894211043,
+      "grad_norm": 0.8702186346054077,
+      "learning_rate": 0.0004962477614144448,
+      "loss": 1.2384,
+      "step": 619
+    },
+    {
+      "epoch": 0.1106056551601106,
+      "grad_norm": 0.9688485860824585,
+      "learning_rate": 0.0004962356586133515,
+      "loss": 1.6793,
+      "step": 620
+    },
+    {
+      "epoch": 0.11078405137811079,
+      "grad_norm": 0.9615238308906555,
+      "learning_rate": 0.0004962235364730157,
+      "loss": 1.4395,
+      "step": 621
+    },
+    {
+      "epoch": 0.11096244759611096,
+      "grad_norm": 1.0574663877487183,
+      "learning_rate": 0.0004962113949943891,
+      "loss": 1.6388,
+      "step": 622
+    },
+    {
+      "epoch": 0.11114084381411114,
+      "grad_norm": 0.8646270036697388,
+      "learning_rate": 0.0004961992341784256,
+      "loss": 1.4026,
+      "step": 623
+    },
+    {
+      "epoch": 0.11131924003211131,
+      "grad_norm": 0.6639731526374817,
+      "learning_rate": 0.0004961870540260801,
+      "loss": 1.2448,
+      "step": 624
+    },
+    {
+      "epoch": 0.1114976362501115,
+      "grad_norm": 0.8900001049041748,
+      "learning_rate": 0.0004961748545383094,
+      "loss": 1.66,
+      "step": 625
+    },
+    {
+      "epoch": 0.11167603246811168,
+      "grad_norm": 1.1087528467178345,
+      "learning_rate": 0.0004961626357160716,
+      "loss": 1.3272,
+      "step": 626
+    },
+    {
+      "epoch": 0.11185442868611185,
+      "grad_norm": 0.6758600473403931,
+      "learning_rate": 0.0004961503975603262,
+      "loss": 1.1703,
+      "step": 627
+    },
+    {
+      "epoch": 0.11203282490411204,
+      "grad_norm": 0.7017180323600769,
+      "learning_rate": 0.0004961381400720346,
+      "loss": 1.6468,
+      "step": 628
+    },
+    {
+      "epoch": 0.11221122112211221,
+      "grad_norm": 1.1828702688217163,
+      "learning_rate": 0.0004961258632521595,
+      "loss": 1.2218,
+      "step": 629
+    },
+    {
+      "epoch": 0.1123896173401124,
+      "grad_norm": 0.7270997762680054,
+      "learning_rate": 0.0004961135671016647,
+      "loss": 1.435,
+      "step": 630
+    },
+    {
+      "epoch": 0.11256801355811256,
+      "grad_norm": 0.6998748183250427,
+      "learning_rate": 0.0004961012516215166,
+      "loss": 1.4037,
+      "step": 631
+    },
+    {
+      "epoch": 0.11274640977611275,
+      "grad_norm": 0.6797101497650146,
+      "learning_rate": 0.0004960889168126819,
+      "loss": 1.3327,
+      "step": 632
+    },
+    {
+      "epoch": 0.11292480599411292,
+      "grad_norm": 0.7535882592201233,
+      "learning_rate": 0.0004960765626761296,
+      "loss": 1.2727,
+      "step": 633
+    },
+    {
+      "epoch": 0.1131032022121131,
+      "grad_norm": 0.7744285464286804,
+      "learning_rate": 0.00049606418921283,
+      "loss": 1.3627,
+      "step": 634
+    },
+    {
+      "epoch": 0.11328159843011328,
+      "grad_norm": 1.608994483947754,
+      "learning_rate": 0.0004960517964237548,
+      "loss": 1.4621,
+      "step": 635
+    },
+    {
+      "epoch": 0.11345999464811346,
+      "grad_norm": 5.228842258453369,
+      "learning_rate": 0.0004960393843098775,
+      "loss": 1.2311,
+      "step": 636
+    },
+    {
+      "epoch": 0.11363839086611364,
+      "grad_norm": 0.7488652467727661,
+      "learning_rate": 0.0004960269528721728,
+      "loss": 1.3578,
+      "step": 637
+    },
+    {
+      "epoch": 0.11381678708411382,
+      "grad_norm": 0.6869979500770569,
+      "learning_rate": 0.0004960145021116171,
+      "loss": 1.4778,
+      "step": 638
+    },
+    {
+      "epoch": 0.113995183302114,
+      "grad_norm": 0.8673625588417053,
+      "learning_rate": 0.0004960020320291882,
+      "loss": 1.5803,
+      "step": 639
+    },
+    {
+      "epoch": 0.11417357952011417,
+      "grad_norm": 1.1181416511535645,
+      "learning_rate": 0.0004959895426258656,
+      "loss": 1.2287,
+      "step": 640
+    },
+    {
+      "epoch": 0.11435197573811436,
+      "grad_norm": 1.0992931127548218,
+      "learning_rate": 0.0004959770339026301,
+      "loss": 1.5307,
+      "step": 641
+    },
+    {
+      "epoch": 0.11453037195611453,
+      "grad_norm": 0.8505348563194275,
+      "learning_rate": 0.0004959645058604644,
+      "loss": 1.2554,
+      "step": 642
+    },
+    {
+      "epoch": 0.11470876817411471,
+      "grad_norm": 1.2507725954055786,
+      "learning_rate": 0.0004959519585003521,
+      "loss": 1.3439,
+      "step": 643
+    },
+    {
+      "epoch": 0.11488716439211488,
+      "grad_norm": 0.9388852119445801,
+      "learning_rate": 0.0004959393918232789,
+      "loss": 1.6679,
+      "step": 644
+    },
+    {
+      "epoch": 0.11506556061011507,
+      "grad_norm": 0.9033030271530151,
+      "learning_rate": 0.0004959268058302318,
+      "loss": 1.5723,
+      "step": 645
+    },
+    {
+      "epoch": 0.11524395682811524,
+      "grad_norm": 0.7030625939369202,
+      "learning_rate": 0.0004959142005221991,
+      "loss": 1.4426,
+      "step": 646
+    },
+    {
+      "epoch": 0.11542235304611542,
+      "grad_norm": 0.7073930501937866,
+      "learning_rate": 0.0004959015759001708,
+      "loss": 1.3419,
+      "step": 647
+    },
+    {
+      "epoch": 0.11560074926411561,
+      "grad_norm": 0.7212799191474915,
+      "learning_rate": 0.0004958889319651386,
+      "loss": 1.4213,
+      "step": 648
+    },
+    {
+      "epoch": 0.11577914548211578,
+      "grad_norm": 0.7988331317901611,
+      "learning_rate": 0.0004958762687180956,
+      "loss": 1.5477,
+      "step": 649
+    },
+    {
+      "epoch": 0.11595754170011596,
+      "grad_norm": 0.6533039212226868,
+      "learning_rate": 0.0004958635861600362,
+      "loss": 1.3518,
+      "step": 650
+    },
+    {
+      "epoch": 0.11613593791811613,
+      "grad_norm": 1.167569637298584,
+      "learning_rate": 0.0004958508842919565,
+      "loss": 1.5068,
+      "step": 651
+    },
+    {
+      "epoch": 0.11631433413611632,
+      "grad_norm": 0.8685398101806641,
+      "learning_rate": 0.0004958381631148543,
+      "loss": 1.5344,
+      "step": 652
+    },
+    {
+      "epoch": 0.11649273035411649,
+      "grad_norm": 0.6582388281822205,
+      "learning_rate": 0.0004958254226297284,
+      "loss": 1.4228,
+      "step": 653
+    },
+    {
+      "epoch": 0.11667112657211667,
+      "grad_norm": 0.6874244213104248,
+      "learning_rate": 0.0004958126628375797,
+      "loss": 1.3626,
+      "step": 654
+    },
+    {
+      "epoch": 0.11684952279011684,
+      "grad_norm": 0.6385990977287292,
+      "learning_rate": 0.0004957998837394102,
+      "loss": 1.4008,
+      "step": 655
+    },
+    {
+      "epoch": 0.11702791900811703,
+      "grad_norm": 0.7000042796134949,
+      "learning_rate": 0.0004957870853362237,
+      "loss": 1.4344,
+      "step": 656
+    },
+    {
+      "epoch": 0.1172063152261172,
+      "grad_norm": 0.6392539143562317,
+      "learning_rate": 0.0004957742676290251,
+      "loss": 1.2199,
+      "step": 657
+    },
+    {
+      "epoch": 0.11738471144411738,
+      "grad_norm": 0.6661501526832581,
+      "learning_rate": 0.0004957614306188214,
+      "loss": 1.5327,
+      "step": 658
+    },
+    {
+      "epoch": 0.11756310766211757,
+      "grad_norm": 0.6941922903060913,
+      "learning_rate": 0.0004957485743066207,
+      "loss": 1.157,
+      "step": 659
+    },
+    {
+      "epoch": 0.11774150388011774,
+      "grad_norm": 0.8714244365692139,
+      "learning_rate": 0.0004957356986934326,
+      "loss": 1.4848,
+      "step": 660
+    },
+    {
+      "epoch": 0.11791990009811792,
+      "grad_norm": 0.5838067531585693,
+      "learning_rate": 0.0004957228037802687,
+      "loss": 1.3737,
+      "step": 661
+    },
+    {
+      "epoch": 0.1180982963161181,
+      "grad_norm": 0.9917201399803162,
+      "learning_rate": 0.0004957098895681414,
+      "loss": 1.5201,
+      "step": 662
+    },
+    {
+      "epoch": 0.11827669253411828,
+      "grad_norm": 5.74738883972168,
+      "learning_rate": 0.0004956969560580651,
+      "loss": 1.3064,
+      "step": 663
+    },
+    {
+      "epoch": 0.11845508875211845,
+      "grad_norm": 0.6317707896232605,
+      "learning_rate": 0.0004956840032510556,
+      "loss": 1.1942,
+      "step": 664
+    },
+    {
+      "epoch": 0.11863348497011864,
+      "grad_norm": 0.7513093948364258,
+      "learning_rate": 0.0004956710311481302,
+      "loss": 1.4782,
+      "step": 665
+    },
+    {
+      "epoch": 0.1188118811881188,
+      "grad_norm": 0.8310014605522156,
+      "learning_rate": 0.0004956580397503078,
+      "loss": 1.5899,
+      "step": 666
+    },
+    {
+      "epoch": 0.11899027740611899,
+      "grad_norm": 1.0249019861221313,
+      "learning_rate": 0.0004956450290586087,
+      "loss": 1.1852,
+      "step": 667
+    },
+    {
+      "epoch": 0.11916867362411916,
+      "grad_norm": 1.868385910987854,
+      "learning_rate": 0.0004956319990740547,
+      "loss": 1.4115,
+      "step": 668
+    },
+    {
+      "epoch": 0.11934706984211935,
+      "grad_norm": 0.7085257768630981,
+      "learning_rate": 0.0004956189497976691,
+      "loss": 1.578,
+      "step": 669
+    },
+    {
+      "epoch": 0.11952546606011953,
+      "grad_norm": 0.607820451259613,
+      "learning_rate": 0.0004956058812304769,
+      "loss": 1.4083,
+      "step": 670
+    },
+    {
+      "epoch": 0.1197038622781197,
+      "grad_norm": 0.6980863213539124,
+      "learning_rate": 0.0004955927933735046,
+      "loss": 1.2609,
+      "step": 671
+    },
+    {
+      "epoch": 0.11988225849611989,
+      "grad_norm": 0.5971283316612244,
+      "learning_rate": 0.0004955796862277799,
+      "loss": 1.2083,
+      "step": 672
+    },
+    {
+      "epoch": 0.12006065471412006,
+      "grad_norm": 0.641476571559906,
+      "learning_rate": 0.0004955665597943323,
+      "loss": 1.2241,
+      "step": 673
+    },
+    {
+      "epoch": 0.12023905093212024,
+      "grad_norm": 0.7836436629295349,
+      "learning_rate": 0.0004955534140741928,
+      "loss": 1.511,
+      "step": 674
+    },
+    {
+      "epoch": 0.12041744715012041,
+      "grad_norm": 0.716119110584259,
+      "learning_rate": 0.0004955402490683939,
+      "loss": 1.7792,
+      "step": 675
+    },
+    {
+      "epoch": 0.1205958433681206,
+      "grad_norm": 0.7199398279190063,
+      "learning_rate": 0.0004955270647779695,
+      "loss": 1.3504,
+      "step": 676
+    },
+    {
+      "epoch": 0.12077423958612077,
+      "grad_norm": 1.583276629447937,
+      "learning_rate": 0.000495513861203955,
+      "loss": 1.4377,
+      "step": 677
+    },
+    {
+      "epoch": 0.12095263580412095,
+      "grad_norm": 0.5903739333152771,
+      "learning_rate": 0.0004955006383473876,
+      "loss": 1.2851,
+      "step": 678
+    },
+    {
+      "epoch": 0.12113103202212114,
+      "grad_norm": 0.85763019323349,
+      "learning_rate": 0.0004954873962093056,
+      "loss": 1.6958,
+      "step": 679
+    },
+    {
+      "epoch": 0.12130942824012131,
+      "grad_norm": 0.5933085083961487,
+      "learning_rate": 0.0004954741347907492,
+      "loss": 1.4475,
+      "step": 680
+    },
+    {
+      "epoch": 0.1214878244581215,
+      "grad_norm": 0.6299702525138855,
+      "learning_rate": 0.0004954608540927599,
+      "loss": 1.5114,
+      "step": 681
+    },
+    {
+      "epoch": 0.12166622067612166,
+      "grad_norm": 0.588097333908081,
+      "learning_rate": 0.0004954475541163807,
+      "loss": 1.2504,
+      "step": 682
+    },
+    {
+      "epoch": 0.12184461689412185,
+      "grad_norm": 0.7014090418815613,
+      "learning_rate": 0.0004954342348626562,
+      "loss": 1.7175,
+      "step": 683
+    },
+    {
+      "epoch": 0.12202301311212202,
+      "grad_norm": 0.6995732188224792,
+      "learning_rate": 0.0004954208963326327,
+      "loss": 1.4363,
+      "step": 684
+    },
+    {
+      "epoch": 0.1222014093301222,
+      "grad_norm": 0.5883218050003052,
+      "learning_rate": 0.0004954075385273574,
+      "loss": 1.385,
+      "step": 685
+    },
+    {
+      "epoch": 0.12237980554812238,
+      "grad_norm": 0.6858941912651062,
+      "learning_rate": 0.0004953941614478797,
+      "loss": 1.5209,
+      "step": 686
+    },
+    {
+      "epoch": 0.12255820176612256,
+      "grad_norm": 2.024874687194824,
+      "learning_rate": 0.0004953807650952502,
+      "loss": 1.4826,
+      "step": 687
+    },
+    {
+      "epoch": 0.12273659798412273,
+      "grad_norm": 0.6671762466430664,
+      "learning_rate": 0.000495367349470521,
+      "loss": 1.3409,
+      "step": 688
+    },
+    {
+      "epoch": 0.12291499420212292,
+      "grad_norm": 0.6432123184204102,
+      "learning_rate": 0.0004953539145747457,
+      "loss": 1.3691,
+      "step": 689
+    },
+    {
+      "epoch": 0.1230933904201231,
+      "grad_norm": 0.6529083251953125,
+      "learning_rate": 0.0004953404604089796,
+      "loss": 1.4347,
+      "step": 690
+    },
+    {
+      "epoch": 0.12327178663812327,
+      "grad_norm": 0.6190119981765747,
+      "learning_rate": 0.0004953269869742792,
+      "loss": 1.2487,
+      "step": 691
+    },
+    {
+      "epoch": 0.12345018285612346,
+      "grad_norm": 0.6484703421592712,
+      "learning_rate": 0.000495313494271703,
+      "loss": 1.5288,
+      "step": 692
+    },
+    {
+      "epoch": 0.12362857907412363,
+      "grad_norm": 0.6164755821228027,
+      "learning_rate": 0.0004952999823023104,
+      "loss": 1.2645,
+      "step": 693
+    },
+    {
+      "epoch": 0.12380697529212381,
+      "grad_norm": 0.6002447009086609,
+      "learning_rate": 0.0004952864510671628,
+      "loss": 1.2468,
+      "step": 694
+    },
+    {
+      "epoch": 0.12398537151012398,
+      "grad_norm": 0.6057468056678772,
+      "learning_rate": 0.0004952729005673229,
+      "loss": 1.5349,
+      "step": 695
+    },
+    {
+      "epoch": 0.12416376772812417,
+      "grad_norm": 0.5792039036750793,
+      "learning_rate": 0.0004952593308038549,
+      "loss": 1.1507,
+      "step": 696
+    },
+    {
+      "epoch": 0.12434216394612434,
+      "grad_norm": 0.7286936044692993,
+      "learning_rate": 0.0004952457417778247,
+      "loss": 1.4709,
+      "step": 697
+    },
+    {
+      "epoch": 0.12452056016412452,
+      "grad_norm": 0.6991254687309265,
+      "learning_rate": 0.0004952321334902993,
+      "loss": 1.4715,
+      "step": 698
+    },
+    {
+      "epoch": 0.12469895638212469,
+      "grad_norm": 2.0999631881713867,
+      "learning_rate": 0.0004952185059423478,
+      "loss": 1.4724,
+      "step": 699
+    },
+    {
+      "epoch": 0.12487735260012488,
+      "grad_norm": 0.7571597099304199,
+      "learning_rate": 0.0004952048591350403,
+      "loss": 1.4842,
+      "step": 700
+    },
+    {
+      "epoch": 0.12505574881812506,
+      "grad_norm": 0.7109989523887634,
+      "learning_rate": 0.0004951911930694487,
+      "loss": 1.2872,
+      "step": 701
+    },
+    {
+      "epoch": 0.12523414503612523,
+      "grad_norm": 3.125427484512329,
+      "learning_rate": 0.0004951775077466463,
+      "loss": 1.1666,
+      "step": 702
+    },
+    {
+      "epoch": 0.1254125412541254,
+      "grad_norm": 0.9019857048988342,
+      "learning_rate": 0.0004951638031677081,
+      "loss": 1.3661,
+      "step": 703
+    },
+    {
+      "epoch": 0.1255909374721256,
+      "grad_norm": 0.7541314363479614,
+      "learning_rate": 0.00049515007933371,
+      "loss": 1.2676,
+      "step": 704
+    },
+    {
+      "epoch": 0.12576933369012577,
+      "grad_norm": 0.6455008387565613,
+      "learning_rate": 0.0004951363362457304,
+      "loss": 1.2095,
+      "step": 705
+    },
+    {
+      "epoch": 0.12594772990812594,
+      "grad_norm": 0.7301307916641235,
+      "learning_rate": 0.0004951225739048484,
+      "loss": 1.5632,
+      "step": 706
+    },
+    {
+      "epoch": 0.12612612612612611,
+      "grad_norm": 0.6640974879264832,
+      "learning_rate": 0.0004951087923121449,
+      "loss": 1.4562,
+      "step": 707
+    },
+    {
+      "epoch": 0.1263045223441263,
+      "grad_norm": 0.6812661290168762,
+      "learning_rate": 0.0004950949914687023,
+      "loss": 1.2051,
+      "step": 708
+    },
+    {
+      "epoch": 0.12648291856212648,
+      "grad_norm": 1.516214370727539,
+      "learning_rate": 0.0004950811713756047,
+      "loss": 1.3933,
+      "step": 709
+    },
+    {
+      "epoch": 0.12666131478012665,
+      "grad_norm": 0.8335700631141663,
+      "learning_rate": 0.0004950673320339372,
+      "loss": 1.6518,
+      "step": 710
+    },
+    {
+      "epoch": 0.12683971099812685,
+      "grad_norm": 0.6542342901229858,
+      "learning_rate": 0.0004950534734447869,
+      "loss": 1.3769,
+      "step": 711
+    },
+    {
+      "epoch": 0.12701810721612702,
+      "grad_norm": 0.8510260581970215,
+      "learning_rate": 0.0004950395956092423,
+      "loss": 1.3511,
+      "step": 712
+    },
+    {
+      "epoch": 0.1271965034341272,
+      "grad_norm": 2.36539626121521,
+      "learning_rate": 0.0004950256985283934,
+      "loss": 1.35,
+      "step": 713
+    },
+    {
+      "epoch": 0.12737489965212737,
+      "grad_norm": 0.6594715118408203,
+      "learning_rate": 0.0004950117822033315,
+      "loss": 1.2426,
+      "step": 714
+    },
+    {
+      "epoch": 0.12755329587012756,
+      "grad_norm": 0.619835615158081,
+      "learning_rate": 0.0004949978466351495,
+      "loss": 1.256,
+      "step": 715
+    },
+    {
+      "epoch": 0.12773169208812774,
+      "grad_norm": 0.792833685874939,
+      "learning_rate": 0.0004949838918249423,
+      "loss": 1.4526,
+      "step": 716
+    },
+    {
+      "epoch": 0.1279100883061279,
+      "grad_norm": 0.7402137517929077,
+      "learning_rate": 0.0004949699177738056,
+      "loss": 1.4895,
+      "step": 717
+    },
+    {
+      "epoch": 0.12808848452412808,
+      "grad_norm": 0.7301040291786194,
+      "learning_rate": 0.0004949559244828369,
+      "loss": 1.435,
+      "step": 718
+    },
+    {
+      "epoch": 0.12826688074212828,
+      "grad_norm": 0.6888275742530823,
+      "learning_rate": 0.0004949419119531354,
+      "loss": 1.4007,
+      "step": 719
+    },
+    {
+      "epoch": 0.12844527696012845,
+      "grad_norm": 13.487852096557617,
+      "learning_rate": 0.0004949278801858015,
+      "loss": 1.2581,
+      "step": 720
+    },
+    {
+      "epoch": 0.12862367317812862,
+      "grad_norm": 0.8741459846496582,
+      "learning_rate": 0.0004949138291819372,
+      "loss": 1.3751,
+      "step": 721
+    },
+    {
+      "epoch": 0.12880206939612882,
+      "grad_norm": 14.084300994873047,
+      "learning_rate": 0.0004948997589426463,
+      "loss": 1.6692,
+      "step": 722
+    },
+    {
+      "epoch": 0.128980465614129,
+      "grad_norm": 1.240329623222351,
+      "learning_rate": 0.0004948856694690337,
+      "loss": 1.6749,
+      "step": 723
+    },
+    {
+      "epoch": 0.12915886183212916,
+      "grad_norm": 0.7830163240432739,
+      "learning_rate": 0.000494871560762206,
+      "loss": 1.5147,
+      "step": 724
+    },
+    {
+      "epoch": 0.12933725805012933,
+      "grad_norm": 0.8917336463928223,
+      "learning_rate": 0.0004948574328232713,
+      "loss": 1.4268,
+      "step": 725
+    },
+    {
+      "epoch": 0.12951565426812953,
+      "grad_norm": 0.8025479912757874,
+      "learning_rate": 0.000494843285653339,
+      "loss": 1.4808,
+      "step": 726
+    },
+    {
+      "epoch": 0.1296940504861297,
+      "grad_norm": 0.8113172650337219,
+      "learning_rate": 0.0004948291192535206,
+      "loss": 1.6472,
+      "step": 727
+    },
+    {
+      "epoch": 0.12987244670412987,
+      "grad_norm": 0.6620337963104248,
+      "learning_rate": 0.0004948149336249286,
+      "loss": 1.4085,
+      "step": 728
+    },
+    {
+      "epoch": 0.13005084292213004,
+      "grad_norm": 1.0298174619674683,
+      "learning_rate": 0.0004948007287686769,
+      "loss": 1.2783,
+      "step": 729
+    },
+    {
+      "epoch": 0.13022923914013024,
+      "grad_norm": 0.6767082810401917,
+      "learning_rate": 0.0004947865046858814,
+      "loss": 1.2708,
+      "step": 730
+    },
+    {
+      "epoch": 0.1304076353581304,
+      "grad_norm": 0.7776159644126892,
+      "learning_rate": 0.0004947722613776591,
+      "loss": 1.1688,
+      "step": 731
+    },
+    {
+      "epoch": 0.13058603157613058,
+      "grad_norm": 0.7061336636543274,
+      "learning_rate": 0.0004947579988451288,
+      "loss": 1.4495,
+      "step": 732
+    },
+    {
+      "epoch": 0.13076442779413078,
+      "grad_norm": 0.5810830593109131,
+      "learning_rate": 0.0004947437170894105,
+      "loss": 1.0836,
+      "step": 733
+    },
+    {
+      "epoch": 0.13094282401213095,
+      "grad_norm": 0.6294886469841003,
+      "learning_rate": 0.000494729416111626,
+      "loss": 1.3136,
+      "step": 734
+    },
+    {
+      "epoch": 0.13112122023013112,
+      "grad_norm": 0.6610152721405029,
+      "learning_rate": 0.0004947150959128986,
+      "loss": 1.4622,
+      "step": 735
+    },
+    {
+      "epoch": 0.1312996164481313,
+      "grad_norm": 0.6016314625740051,
+      "learning_rate": 0.0004947007564943527,
+      "loss": 1.1905,
+      "step": 736
+    },
+    {
+      "epoch": 0.1314780126661315,
+      "grad_norm": 0.6377681493759155,
+      "learning_rate": 0.0004946863978571148,
+      "loss": 1.2777,
+      "step": 737
+    },
+    {
+      "epoch": 0.13165640888413166,
+      "grad_norm": 0.6914573907852173,
+      "learning_rate": 0.0004946720200023125,
+      "loss": 1.4642,
+      "step": 738
+    },
+    {
+      "epoch": 0.13183480510213183,
+      "grad_norm": 0.6614610552787781,
+      "learning_rate": 0.000494657622931075,
+      "loss": 1.4415,
+      "step": 739
+    },
+    {
+      "epoch": 0.132013201320132,
+      "grad_norm": 1.029931902885437,
+      "learning_rate": 0.0004946432066445331,
+      "loss": 1.2224,
+      "step": 740
+    },
+    {
+      "epoch": 0.1321915975381322,
+      "grad_norm": 0.7482788562774658,
+      "learning_rate": 0.000494628771143819,
+      "loss": 1.4412,
+      "step": 741
+    },
+    {
+      "epoch": 0.13236999375613237,
+      "grad_norm": 0.9311326146125793,
+      "learning_rate": 0.0004946143164300665,
+      "loss": 1.5545,
+      "step": 742
+    },
+    {
+      "epoch": 0.13254838997413254,
+      "grad_norm": 0.6425170302391052,
+      "learning_rate": 0.0004945998425044109,
+      "loss": 1.3134,
+      "step": 743
+    },
+    {
+      "epoch": 0.13272678619213274,
+      "grad_norm": 0.674411952495575,
+      "learning_rate": 0.0004945853493679889,
+      "loss": 1.4689,
+      "step": 744
+    },
+    {
+      "epoch": 0.1329051824101329,
+      "grad_norm": 0.6528111696243286,
+      "learning_rate": 0.0004945708370219388,
+      "loss": 1.3335,
+      "step": 745
+    },
+    {
+      "epoch": 0.13308357862813308,
+      "grad_norm": 0.663161039352417,
+      "learning_rate": 0.0004945563054674005,
+      "loss": 1.6369,
+      "step": 746
+    },
+    {
+      "epoch": 0.13326197484613325,
+      "grad_norm": 1.808029294013977,
+      "learning_rate": 0.0004945417547055151,
+      "loss": 1.2105,
+      "step": 747
+    },
+    {
+      "epoch": 0.13344037106413345,
+      "grad_norm": 0.683954656124115,
+      "learning_rate": 0.0004945271847374255,
+      "loss": 1.4248,
+      "step": 748
+    },
+    {
+      "epoch": 0.13361876728213362,
+      "grad_norm": 0.5650470852851868,
+      "learning_rate": 0.0004945125955642761,
+      "loss": 1.2412,
+      "step": 749
+    },
+    {
+      "epoch": 0.1337971635001338,
+      "grad_norm": 0.7181941866874695,
+      "learning_rate": 0.0004944979871872126,
+      "loss": 1.5134,
+      "step": 750
+    },
+    {
+      "epoch": 0.13397555971813396,
+      "grad_norm": 0.6759237051010132,
+      "learning_rate": 0.0004944833596073825,
+      "loss": 1.4109,
+      "step": 751
+    },
+    {
+      "epoch": 0.13415395593613416,
+      "grad_norm": 1.112088918685913,
+      "learning_rate": 0.0004944687128259345,
+      "loss": 1.4521,
+      "step": 752
+    },
+    {
+      "epoch": 0.13433235215413433,
+      "grad_norm": 0.8726287484169006,
+      "learning_rate": 0.000494454046844019,
+      "loss": 1.4233,
+      "step": 753
+    },
+    {
+      "epoch": 0.1345107483721345,
+      "grad_norm": 0.61822509765625,
+      "learning_rate": 0.0004944393616627879,
+      "loss": 1.1642,
+      "step": 754
+    },
+    {
+      "epoch": 0.1346891445901347,
+      "grad_norm": 0.7049110531806946,
+      "learning_rate": 0.0004944246572833945,
+      "loss": 1.3176,
+      "step": 755
+    },
+    {
+      "epoch": 0.13486754080813487,
+      "grad_norm": 0.7006151676177979,
+      "learning_rate": 0.0004944099337069936,
+      "loss": 1.4621,
+      "step": 756
+    },
+    {
+      "epoch": 0.13504593702613504,
+      "grad_norm": 0.907748818397522,
+      "learning_rate": 0.0004943951909347419,
+      "loss": 1.4472,
+      "step": 757
+    },
+    {
+      "epoch": 0.13522433324413521,
+      "grad_norm": 1.7227107286453247,
+      "learning_rate": 0.0004943804289677969,
+      "loss": 1.381,
+      "step": 758
+    },
+    {
+      "epoch": 0.1354027294621354,
+      "grad_norm": 0.8136081695556641,
+      "learning_rate": 0.0004943656478073182,
+      "loss": 1.2206,
+      "step": 759
+    },
+    {
+      "epoch": 0.13558112568013558,
+      "grad_norm": 1.1220283508300781,
+      "learning_rate": 0.0004943508474544667,
+      "loss": 1.5002,
+      "step": 760
+    },
+    {
+      "epoch": 0.13575952189813575,
+      "grad_norm": 2.818793296813965,
+      "learning_rate": 0.0004943360279104047,
+      "loss": 1.4188,
+      "step": 761
+    },
+    {
+      "epoch": 0.13593791811613593,
+      "grad_norm": 7.106777191162109,
+      "learning_rate": 0.0004943211891762964,
+      "loss": 1.6358,
+      "step": 762
+    },
+    {
+      "epoch": 0.13611631433413612,
+      "grad_norm": 1.419669270515442,
+      "learning_rate": 0.0004943063312533069,
+      "loss": 1.4433,
+      "step": 763
+    },
+    {
+      "epoch": 0.1362947105521363,
+      "grad_norm": 1.2436542510986328,
+      "learning_rate": 0.0004942914541426033,
+      "loss": 1.5007,
+      "step": 764
+    },
+    {
+      "epoch": 0.13647310677013647,
+      "grad_norm": 1.268811821937561,
+      "learning_rate": 0.0004942765578453541,
+      "loss": 1.6141,
+      "step": 765
+    },
+    {
+      "epoch": 0.13665150298813666,
+      "grad_norm": 1.0248541831970215,
+      "learning_rate": 0.000494261642362729,
+      "loss": 1.2407,
+      "step": 766
+    },
+    {
+      "epoch": 0.13682989920613683,
+      "grad_norm": 0.8062700629234314,
+      "learning_rate": 0.0004942467076958999,
+      "loss": 1.2817,
+      "step": 767
+    },
+    {
+      "epoch": 0.137008295424137,
+      "grad_norm": 1.0193575620651245,
+      "learning_rate": 0.0004942317538460392,
+      "loss": 1.2323,
+      "step": 768
+    },
+    {
+      "epoch": 0.13718669164213718,
+      "grad_norm": 1.2625796794891357,
+      "learning_rate": 0.0004942167808143218,
+      "loss": 1.6804,
+      "step": 769
+    },
+    {
+      "epoch": 0.13736508786013737,
+      "grad_norm": 0.7931640148162842,
+      "learning_rate": 0.0004942017886019236,
+      "loss": 1.2993,
+      "step": 770
+    },
+    {
+      "epoch": 0.13754348407813755,
+      "grad_norm": 1.35048508644104,
+      "learning_rate": 0.0004941867772100218,
+      "loss": 1.4314,
+      "step": 771
+    },
+    {
+      "epoch": 0.13772188029613772,
+      "grad_norm": 1.2043832540512085,
+      "learning_rate": 0.0004941717466397957,
+      "loss": 1.4401,
+      "step": 772
+    },
+    {
+      "epoch": 0.1379002765141379,
+      "grad_norm": 0.8232206106185913,
+      "learning_rate": 0.0004941566968924258,
+      "loss": 1.4218,
+      "step": 773
+    },
+    {
+      "epoch": 0.13807867273213809,
+      "grad_norm": 0.7783534526824951,
+      "learning_rate": 0.0004941416279690939,
+      "loss": 1.5267,
+      "step": 774
+    },
+    {
+      "epoch": 0.13825706895013826,
+      "grad_norm": 0.6729874014854431,
+      "learning_rate": 0.0004941265398709835,
+      "loss": 1.368,
+      "step": 775
+    },
+    {
+      "epoch": 0.13843546516813843,
+      "grad_norm": 0.7547962069511414,
+      "learning_rate": 0.0004941114325992798,
+      "loss": 1.3377,
+      "step": 776
+    },
+    {
+      "epoch": 0.13861386138613863,
+      "grad_norm": 0.7145437002182007,
+      "learning_rate": 0.0004940963061551693,
+      "loss": 1.3902,
+      "step": 777
+    },
+    {
+      "epoch": 0.1387922576041388,
+      "grad_norm": 0.6485962271690369,
+      "learning_rate": 0.0004940811605398399,
+      "loss": 1.0085,
+      "step": 778
+    },
+    {
+      "epoch": 0.13897065382213897,
+      "grad_norm": 0.630803644657135,
+      "learning_rate": 0.0004940659957544813,
+      "loss": 1.4345,
+      "step": 779
+    },
+    {
+      "epoch": 0.13914905004013914,
+      "grad_norm": 1.3144874572753906,
+      "learning_rate": 0.0004940508118002842,
+      "loss": 1.2102,
+      "step": 780
+    },
+    {
+      "epoch": 0.13932744625813934,
+      "grad_norm": 0.6318569779396057,
+      "learning_rate": 0.0004940356086784415,
+      "loss": 1.1991,
+      "step": 781
+    },
+    {
+      "epoch": 0.1395058424761395,
+      "grad_norm": 0.9692159295082092,
+      "learning_rate": 0.0004940203863901472,
+      "loss": 1.1838,
+      "step": 782
+    },
+    {
+      "epoch": 0.13968423869413968,
+      "grad_norm": 0.6397761106491089,
+      "learning_rate": 0.0004940051449365966,
+      "loss": 1.301,
+      "step": 783
+    },
+    {
+      "epoch": 0.13986263491213985,
+      "grad_norm": 0.7027068734169006,
+      "learning_rate": 0.000493989884318987,
+      "loss": 1.1906,
+      "step": 784
+    },
+    {
+      "epoch": 0.14004103113014005,
+      "grad_norm": 0.5374085307121277,
+      "learning_rate": 0.0004939746045385168,
+      "loss": 1.2157,
+      "step": 785
+    },
+    {
+      "epoch": 0.14021942734814022,
+      "grad_norm": 0.7548506855964661,
+      "learning_rate": 0.0004939593055963863,
+      "loss": 1.2354,
+      "step": 786
+    },
+    {
+      "epoch": 0.1403978235661404,
+      "grad_norm": 0.8988766670227051,
+      "learning_rate": 0.0004939439874937967,
+      "loss": 1.2367,
+      "step": 787
+    },
+    {
+      "epoch": 0.1405762197841406,
+      "grad_norm": 0.6512690186500549,
+      "learning_rate": 0.0004939286502319515,
+      "loss": 1.1598,
+      "step": 788
+    },
+    {
+      "epoch": 0.14075461600214076,
+      "grad_norm": 0.664986789226532,
+      "learning_rate": 0.0004939132938120551,
+      "loss": 1.4823,
+      "step": 789
+    },
+    {
+      "epoch": 0.14093301222014093,
+      "grad_norm": 0.6778481006622314,
+      "learning_rate": 0.0004938979182353134,
+      "loss": 1.2677,
+      "step": 790
+    },
+    {
+      "epoch": 0.1411114084381411,
+      "grad_norm": 0.570472240447998,
+      "learning_rate": 0.0004938825235029343,
+      "loss": 0.8828,
+      "step": 791
+    },
+    {
+      "epoch": 0.1412898046561413,
+      "grad_norm": 0.9883807897567749,
+      "learning_rate": 0.0004938671096161267,
+      "loss": 1.2965,
+      "step": 792
+    },
+    {
+      "epoch": 0.14146820087414147,
+      "grad_norm": 0.7180954217910767,
+      "learning_rate": 0.0004938516765761011,
+      "loss": 1.4603,
+      "step": 793
+    },
+    {
+      "epoch": 0.14164659709214164,
+      "grad_norm": 0.6821011900901794,
+      "learning_rate": 0.00049383622438407,
+      "loss": 1.6537,
+      "step": 794
+    },
+    {
+      "epoch": 0.1418249933101418,
+      "grad_norm": 0.6514670848846436,
+      "learning_rate": 0.0004938207530412467,
+      "loss": 1.1712,
+      "step": 795
+    },
+    {
+      "epoch": 0.142003389528142,
+      "grad_norm": 0.5831206440925598,
+      "learning_rate": 0.0004938052625488464,
+      "loss": 1.2547,
+      "step": 796
+    },
+    {
+      "epoch": 0.14218178574614218,
+      "grad_norm": 0.6372058987617493,
+      "learning_rate": 0.0004937897529080856,
+      "loss": 1.3056,
+      "step": 797
+    },
+    {
+      "epoch": 0.14236018196414235,
+      "grad_norm": 0.6786195635795593,
+      "learning_rate": 0.0004937742241201826,
+      "loss": 1.5176,
+      "step": 798
+    },
+    {
+      "epoch": 0.14253857818214255,
+      "grad_norm": 0.6339353322982788,
+      "learning_rate": 0.000493758676186357,
+      "loss": 1.2932,
+      "step": 799
+    },
+    {
+      "epoch": 0.14271697440014272,
+      "grad_norm": 0.6407662630081177,
+      "learning_rate": 0.0004937431091078297,
+      "loss": 1.2913,
+      "step": 800
+    },
+    {
+      "epoch": 0.1428953706181429,
+      "grad_norm": 0.7081295251846313,
+      "learning_rate": 0.0004937275228858235,
+      "loss": 1.2895,
+      "step": 801
+    },
+    {
+      "epoch": 0.14307376683614306,
+      "grad_norm": 0.5862438678741455,
+      "learning_rate": 0.0004937119175215627,
+      "loss": 1.2476,
+      "step": 802
+    },
+    {
+      "epoch": 0.14325216305414326,
+      "grad_norm": 0.6276576519012451,
+      "learning_rate": 0.0004936962930162728,
+      "loss": 1.2236,
+      "step": 803
+    },
+    {
+      "epoch": 0.14343055927214343,
+      "grad_norm": 0.6015689969062805,
+      "learning_rate": 0.0004936806493711808,
+      "loss": 1.2797,
+      "step": 804
+    },
+    {
+      "epoch": 0.1436089554901436,
+      "grad_norm": 1.0078672170639038,
+      "learning_rate": 0.0004936649865875155,
+      "loss": 1.5178,
+      "step": 805
+    },
+    {
+      "epoch": 0.14378735170814377,
+      "grad_norm": 0.6235846281051636,
+      "learning_rate": 0.000493649304666507,
+      "loss": 1.2727,
+      "step": 806
+    },
+    {
+      "epoch": 0.14396574792614397,
+      "grad_norm": 0.6958335638046265,
+      "learning_rate": 0.0004936336036093869,
+      "loss": 1.4622,
+      "step": 807
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 0.7218196392059326,
+      "learning_rate": 0.0004936178834173884,
+      "loss": 1.3002,
+      "step": 808
+    },
+    {
+      "epoch": 0.1443225403621443,
+      "grad_norm": 0.5912010669708252,
+      "learning_rate": 0.0004936021440917462,
+      "loss": 1.406,
+      "step": 809
+    },
+    {
+      "epoch": 0.1445009365801445,
+      "grad_norm": 1.7334743738174438,
+      "learning_rate": 0.0004935863856336965,
+      "loss": 1.2606,
+      "step": 810
+    },
+    {
+      "epoch": 0.14467933279814468,
+      "grad_norm": 0.7477133274078369,
+      "learning_rate": 0.000493570608044477,
+      "loss": 1.3254,
+      "step": 811
+    },
+    {
+      "epoch": 0.14485772901614485,
+      "grad_norm": 0.6125087141990662,
+      "learning_rate": 0.0004935548113253266,
+      "loss": 1.448,
+      "step": 812
+    },
+    {
+      "epoch": 0.14503612523414502,
+      "grad_norm": 0.6391841173171997,
+      "learning_rate": 0.0004935389954774861,
+      "loss": 1.0645,
+      "step": 813
+    },
+    {
+      "epoch": 0.14521452145214522,
+      "grad_norm": 0.6377097368240356,
+      "learning_rate": 0.0004935231605021977,
+      "loss": 1.2738,
+      "step": 814
+    },
+    {
+      "epoch": 0.1453929176701454,
+      "grad_norm": 0.6253454685211182,
+      "learning_rate": 0.0004935073064007052,
+      "loss": 1.6608,
+      "step": 815
+    },
+    {
+      "epoch": 0.14557131388814556,
+      "grad_norm": 0.7210360169410706,
+      "learning_rate": 0.0004934914331742536,
+      "loss": 1.5487,
+      "step": 816
+    },
+    {
+      "epoch": 0.14574971010614576,
+      "grad_norm": 0.589590311050415,
+      "learning_rate": 0.0004934755408240896,
+      "loss": 1.1513,
+      "step": 817
+    },
+    {
+      "epoch": 0.14592810632414593,
+      "grad_norm": 0.6044402718544006,
+      "learning_rate": 0.0004934596293514614,
+      "loss": 1.2131,
+      "step": 818
+    },
+    {
+      "epoch": 0.1461065025421461,
+      "grad_norm": 0.6587312817573547,
+      "learning_rate": 0.0004934436987576186,
+      "loss": 1.5939,
+      "step": 819
+    },
+    {
+      "epoch": 0.14628489876014628,
+      "grad_norm": 0.5897753238677979,
+      "learning_rate": 0.0004934277490438126,
+      "loss": 1.2628,
+      "step": 820
+    },
+    {
+      "epoch": 0.14646329497814647,
+      "grad_norm": 0.5983254909515381,
+      "learning_rate": 0.0004934117802112959,
+      "loss": 1.1709,
+      "step": 821
+    },
+    {
+      "epoch": 0.14664169119614665,
+      "grad_norm": 0.6504422426223755,
+      "learning_rate": 0.0004933957922613227,
+      "loss": 1.4531,
+      "step": 822
+    },
+    {
+      "epoch": 0.14682008741414682,
+      "grad_norm": 0.5873280763626099,
+      "learning_rate": 0.0004933797851951487,
+      "loss": 1.1531,
+      "step": 823
+    },
+    {
+      "epoch": 0.146998483632147,
+      "grad_norm": 2.0532479286193848,
+      "learning_rate": 0.0004933637590140311,
+      "loss": 1.4533,
+      "step": 824
+    },
+    {
+      "epoch": 0.14717687985014719,
+      "grad_norm": 0.6756530404090881,
+      "learning_rate": 0.0004933477137192287,
+      "loss": 1.4504,
+      "step": 825
+    },
+    {
+      "epoch": 0.14735527606814736,
+      "grad_norm": 0.5908591747283936,
+      "learning_rate": 0.0004933316493120015,
+      "loss": 1.1947,
+      "step": 826
+    },
+    {
+      "epoch": 0.14753367228614753,
+      "grad_norm": 0.5556735396385193,
+      "learning_rate": 0.0004933155657936112,
+      "loss": 1.0962,
+      "step": 827
+    },
+    {
+      "epoch": 0.14771206850414773,
+      "grad_norm": 0.5515411496162415,
+      "learning_rate": 0.0004932994631653212,
+      "loss": 1.1774,
+      "step": 828
+    },
+    {
+      "epoch": 0.1478904647221479,
+      "grad_norm": 0.8183857202529907,
+      "learning_rate": 0.000493283341428396,
+      "loss": 1.3823,
+      "step": 829
+    },
+    {
+      "epoch": 0.14806886094014807,
+      "grad_norm": 0.6181221008300781,
+      "learning_rate": 0.0004932672005841019,
+      "loss": 1.4347,
+      "step": 830
+    },
+    {
+      "epoch": 0.14824725715814824,
+      "grad_norm": 0.632533848285675,
+      "learning_rate": 0.0004932510406337065,
+      "loss": 1.6746,
+      "step": 831
+    },
+    {
+      "epoch": 0.14842565337614844,
+      "grad_norm": 0.679986834526062,
+      "learning_rate": 0.0004932348615784791,
+      "loss": 1.2888,
+      "step": 832
+    },
+    {
+      "epoch": 0.1486040495941486,
+      "grad_norm": 0.6554014086723328,
+      "learning_rate": 0.0004932186634196903,
+      "loss": 1.4259,
+      "step": 833
+    },
+    {
+      "epoch": 0.14878244581214878,
+      "grad_norm": 0.6169226169586182,
+      "learning_rate": 0.0004932024461586124,
+      "loss": 1.206,
+      "step": 834
+    },
+    {
+      "epoch": 0.14896084203014895,
+      "grad_norm": 0.6929764747619629,
+      "learning_rate": 0.000493186209796519,
+      "loss": 1.2828,
+      "step": 835
+    },
+    {
+      "epoch": 0.14913923824814915,
+      "grad_norm": 0.5675246119499207,
+      "learning_rate": 0.0004931699543346854,
+      "loss": 1.4443,
+      "step": 836
+    },
+    {
+      "epoch": 0.14931763446614932,
+      "grad_norm": 0.6860451698303223,
+      "learning_rate": 0.0004931536797743881,
+      "loss": 1.3001,
+      "step": 837
+    },
+    {
+      "epoch": 0.1494960306841495,
+      "grad_norm": 0.6203287243843079,
+      "learning_rate": 0.0004931373861169055,
+      "loss": 1.6086,
+      "step": 838
+    },
+    {
+      "epoch": 0.1496744269021497,
+      "grad_norm": 0.5558333992958069,
+      "learning_rate": 0.0004931210733635172,
+      "loss": 1.2781,
+      "step": 839
+    },
+    {
+      "epoch": 0.14985282312014986,
+      "grad_norm": 0.6401587128639221,
+      "learning_rate": 0.0004931047415155044,
+      "loss": 1.278,
+      "step": 840
+    },
+    {
+      "epoch": 0.15003121933815003,
+      "grad_norm": 0.6404081583023071,
+      "learning_rate": 0.00049308839057415,
+      "loss": 1.2592,
+      "step": 841
+    },
+    {
+      "epoch": 0.1502096155561502,
+      "grad_norm": 0.646500289440155,
+      "learning_rate": 0.0004930720205407378,
+      "loss": 1.4263,
+      "step": 842
+    },
+    {
+      "epoch": 0.1503880117741504,
+      "grad_norm": 0.6502192616462708,
+      "learning_rate": 0.0004930556314165538,
+      "loss": 1.3815,
+      "step": 843
+    },
+    {
+      "epoch": 0.15056640799215057,
+      "grad_norm": 0.7043633460998535,
+      "learning_rate": 0.0004930392232028851,
+      "loss": 1.3289,
+      "step": 844
+    },
+    {
+      "epoch": 0.15074480421015074,
+      "grad_norm": 0.566808819770813,
+      "learning_rate": 0.0004930227959010204,
+      "loss": 1.226,
+      "step": 845
+    },
+    {
+      "epoch": 0.1509232004281509,
+      "grad_norm": 0.5885351896286011,
+      "learning_rate": 0.0004930063495122498,
+      "loss": 1.0517,
+      "step": 846
+    },
+    {
+      "epoch": 0.1511015966461511,
+      "grad_norm": 0.696162760257721,
+      "learning_rate": 0.0004929898840378651,
+      "loss": 1.4365,
+      "step": 847
+    },
+    {
+      "epoch": 0.15127999286415128,
+      "grad_norm": 0.676964282989502,
+      "learning_rate": 0.0004929733994791597,
+      "loss": 1.5412,
+      "step": 848
+    },
+    {
+      "epoch": 0.15145838908215145,
+      "grad_norm": 0.5642474293708801,
+      "learning_rate": 0.0004929568958374278,
+      "loss": 1.155,
+      "step": 849
+    },
+    {
+      "epoch": 0.15163678530015165,
+      "grad_norm": 0.9656538367271423,
+      "learning_rate": 0.0004929403731139659,
+      "loss": 1.7266,
+      "step": 850
+    },
+    {
+      "epoch": 0.15181518151815182,
+      "grad_norm": 0.6227328777313232,
+      "learning_rate": 0.0004929238313100717,
+      "loss": 1.4751,
+      "step": 851
+    },
+    {
+      "epoch": 0.151993577736152,
+      "grad_norm": 0.6424437761306763,
+      "learning_rate": 0.0004929072704270444,
+      "loss": 1.4661,
+      "step": 852
+    },
+    {
+      "epoch": 0.15217197395415216,
+      "grad_norm": 0.5338221192359924,
+      "learning_rate": 0.0004928906904661845,
+      "loss": 1.2984,
+      "step": 853
+    },
+    {
+      "epoch": 0.15235037017215236,
+      "grad_norm": 0.594628632068634,
+      "learning_rate": 0.0004928740914287944,
+      "loss": 1.1771,
+      "step": 854
+    },
+    {
+      "epoch": 0.15252876639015253,
+      "grad_norm": 1.8328267335891724,
+      "learning_rate": 0.0004928574733161775,
+      "loss": 1.241,
+      "step": 855
+    },
+    {
+      "epoch": 0.1527071626081527,
+      "grad_norm": 4.153714656829834,
+      "learning_rate": 0.0004928408361296393,
+      "loss": 1.2908,
+      "step": 856
+    },
+    {
+      "epoch": 0.15288555882615287,
+      "grad_norm": 0.6561400294303894,
+      "learning_rate": 0.0004928241798704862,
+      "loss": 1.1795,
+      "step": 857
+    },
+    {
+      "epoch": 0.15306395504415307,
+      "grad_norm": 0.9632255434989929,
+      "learning_rate": 0.0004928075045400267,
+      "loss": 1.3548,
+      "step": 858
+    },
+    {
+      "epoch": 0.15324235126215324,
+      "grad_norm": 0.6676768064498901,
+      "learning_rate": 0.0004927908101395701,
+      "loss": 1.3005,
+      "step": 859
+    },
+    {
+      "epoch": 0.1534207474801534,
+      "grad_norm": 0.6931350827217102,
+      "learning_rate": 0.0004927740966704278,
+      "loss": 1.4234,
+      "step": 860
+    },
+    {
+      "epoch": 0.1535991436981536,
+      "grad_norm": 0.582984983921051,
+      "learning_rate": 0.0004927573641339125,
+      "loss": 1.286,
+      "step": 861
+    },
+    {
+      "epoch": 0.15377753991615378,
+      "grad_norm": 1.9343520402908325,
+      "learning_rate": 0.0004927406125313382,
+      "loss": 1.1929,
+      "step": 862
+    },
+    {
+      "epoch": 0.15395593613415395,
+      "grad_norm": 0.8297525644302368,
+      "learning_rate": 0.0004927238418640208,
+      "loss": 1.3183,
+      "step": 863
+    },
+    {
+      "epoch": 0.15413433235215412,
+      "grad_norm": 0.6805528998374939,
+      "learning_rate": 0.0004927070521332772,
+      "loss": 1.3282,
+      "step": 864
+    },
+    {
+      "epoch": 0.15431272857015432,
+      "grad_norm": 0.8657769560813904,
+      "learning_rate": 0.0004926902433404261,
+      "loss": 1.253,
+      "step": 865
+    },
+    {
+      "epoch": 0.1544911247881545,
+      "grad_norm": 0.7054359912872314,
+      "learning_rate": 0.0004926734154867878,
+      "loss": 1.2187,
+      "step": 866
+    },
+    {
+      "epoch": 0.15466952100615466,
+      "grad_norm": 0.6124011874198914,
+      "learning_rate": 0.0004926565685736839,
+      "loss": 1.1525,
+      "step": 867
+    },
+    {
+      "epoch": 0.15484791722415484,
+      "grad_norm": 0.6714728474617004,
+      "learning_rate": 0.0004926397026024375,
+      "loss": 1.1008,
+      "step": 868
+    },
+    {
+      "epoch": 0.15502631344215503,
+      "grad_norm": 0.7281408905982971,
+      "learning_rate": 0.0004926228175743733,
+      "loss": 1.4181,
+      "step": 869
+    },
+    {
+      "epoch": 0.1552047096601552,
+      "grad_norm": 0.6333009004592896,
+      "learning_rate": 0.0004926059134908173,
+      "loss": 1.3329,
+      "step": 870
+    },
+    {
+      "epoch": 0.15538310587815538,
+      "grad_norm": 0.6089525818824768,
+      "learning_rate": 0.0004925889903530973,
+      "loss": 1.257,
+      "step": 871
+    },
+    {
+      "epoch": 0.15556150209615557,
+      "grad_norm": 0.6068894267082214,
+      "learning_rate": 0.0004925720481625426,
+      "loss": 1.126,
+      "step": 872
+    },
+    {
+      "epoch": 0.15573989831415574,
+      "grad_norm": 0.594664990901947,
+      "learning_rate": 0.0004925550869204835,
+      "loss": 1.3666,
+      "step": 873
+    },
+    {
+      "epoch": 0.15591829453215592,
+      "grad_norm": 0.6410947442054749,
+      "learning_rate": 0.0004925381066282522,
+      "loss": 1.6397,
+      "step": 874
+    },
+    {
+      "epoch": 0.1560966907501561,
+      "grad_norm": 0.6146419644355774,
+      "learning_rate": 0.0004925211072871824,
+      "loss": 1.4024,
+      "step": 875
+    },
+    {
+      "epoch": 0.15627508696815628,
+      "grad_norm": 0.6187763810157776,
+      "learning_rate": 0.0004925040888986091,
+      "loss": 1.4961,
+      "step": 876
+    },
+    {
+      "epoch": 0.15645348318615646,
+      "grad_norm": 0.6881486773490906,
+      "learning_rate": 0.0004924870514638691,
+      "loss": 1.5297,
+      "step": 877
+    },
+    {
+      "epoch": 0.15663187940415663,
+      "grad_norm": 0.69078129529953,
+      "learning_rate": 0.0004924699949843004,
+      "loss": 1.2651,
+      "step": 878
+    },
+    {
+      "epoch": 0.1568102756221568,
+      "grad_norm": 0.5774492621421814,
+      "learning_rate": 0.0004924529194612428,
+      "loss": 1.4284,
+      "step": 879
+    },
+    {
+      "epoch": 0.156988671840157,
+      "grad_norm": 0.5534959435462952,
+      "learning_rate": 0.0004924358248960372,
+      "loss": 1.4157,
+      "step": 880
+    },
+    {
+      "epoch": 0.15716706805815717,
+      "grad_norm": 0.5735335350036621,
+      "learning_rate": 0.000492418711290026,
+      "loss": 1.1534,
+      "step": 881
+    },
+    {
+      "epoch": 0.15734546427615734,
+      "grad_norm": 0.5932490229606628,
+      "learning_rate": 0.0004924015786445537,
+      "loss": 1.3835,
+      "step": 882
+    },
+    {
+      "epoch": 0.15752386049415754,
+      "grad_norm": 0.6592042446136475,
+      "learning_rate": 0.0004923844269609657,
+      "loss": 1.2475,
+      "step": 883
+    },
+    {
+      "epoch": 0.1577022567121577,
+      "grad_norm": 0.6756531000137329,
+      "learning_rate": 0.0004923672562406092,
+      "loss": 1.5197,
+      "step": 884
+    },
+    {
+      "epoch": 0.15788065293015788,
+      "grad_norm": 0.6795757412910461,
+      "learning_rate": 0.0004923500664848326,
+      "loss": 1.466,
+      "step": 885
+    },
+    {
+      "epoch": 0.15805904914815805,
+      "grad_norm": 0.6750994920730591,
+      "learning_rate": 0.0004923328576949862,
+      "loss": 1.494,
+      "step": 886
+    },
+    {
+      "epoch": 0.15823744536615825,
+      "grad_norm": 0.5709784030914307,
+      "learning_rate": 0.0004923156298724213,
+      "loss": 1.2837,
+      "step": 887
+    },
+    {
+      "epoch": 0.15841584158415842,
+      "grad_norm": 0.6380211114883423,
+      "learning_rate": 0.000492298383018491,
+      "loss": 1.4147,
+      "step": 888
+    },
+    {
+      "epoch": 0.1585942378021586,
+      "grad_norm": 0.6981920599937439,
+      "learning_rate": 0.0004922811171345502,
+      "loss": 1.297,
+      "step": 889
+    },
+    {
+      "epoch": 0.15877263402015876,
+      "grad_norm": 0.6991307139396667,
+      "learning_rate": 0.0004922638322219546,
+      "loss": 1.445,
+      "step": 890
+    },
+    {
+      "epoch": 0.15895103023815896,
+      "grad_norm": 0.6050926446914673,
+      "learning_rate": 0.000492246528282062,
+      "loss": 1.3701,
+      "step": 891
+    },
+    {
+      "epoch": 0.15912942645615913,
+      "grad_norm": 0.6252057552337646,
+      "learning_rate": 0.0004922292053162312,
+      "loss": 1.1218,
+      "step": 892
+    },
+    {
+      "epoch": 0.1593078226741593,
+      "grad_norm": 0.6167547702789307,
+      "learning_rate": 0.0004922118633258228,
+      "loss": 1.2547,
+      "step": 893
+    },
+    {
+      "epoch": 0.1594862188921595,
+      "grad_norm": 0.5581344366073608,
+      "learning_rate": 0.0004921945023121989,
+      "loss": 1.2666,
+      "step": 894
+    },
+    {
+      "epoch": 0.15966461511015967,
+      "grad_norm": 0.6345050930976868,
+      "learning_rate": 0.0004921771222767231,
+      "loss": 1.3042,
+      "step": 895
+    },
+    {
+      "epoch": 0.15984301132815984,
+      "grad_norm": 0.6487632393836975,
+      "learning_rate": 0.0004921597232207604,
+      "loss": 1.3426,
+      "step": 896
+    },
+    {
+      "epoch": 0.16002140754616,
+      "grad_norm": 0.6063089966773987,
+      "learning_rate": 0.0004921423051456772,
+      "loss": 1.2359,
+      "step": 897
+    },
+    {
+      "epoch": 0.1601998037641602,
+      "grad_norm": 0.6432583928108215,
+      "learning_rate": 0.0004921248680528414,
+      "loss": 1.2015,
+      "step": 898
+    },
+    {
+      "epoch": 0.16037819998216038,
+      "grad_norm": 0.6074740290641785,
+      "learning_rate": 0.0004921074119436229,
+      "loss": 1.0048,
+      "step": 899
+    },
+    {
+      "epoch": 0.16055659620016055,
+      "grad_norm": 0.6494312882423401,
+      "learning_rate": 0.0004920899368193923,
+      "loss": 1.2296,
+      "step": 900
+    },
+    {
+      "epoch": 0.16073499241816072,
+      "grad_norm": 0.6664918661117554,
+      "learning_rate": 0.0004920724426815222,
+      "loss": 1.4211,
+      "step": 901
+    },
+    {
+      "epoch": 0.16091338863616092,
+      "grad_norm": 0.6054812073707581,
+      "learning_rate": 0.0004920549295313867,
+      "loss": 1.1764,
+      "step": 902
+    },
+    {
+      "epoch": 0.1610917848541611,
+      "grad_norm": 0.5707181096076965,
+      "learning_rate": 0.0004920373973703612,
+      "loss": 1.3884,
+      "step": 903
+    },
+    {
+      "epoch": 0.16127018107216126,
+      "grad_norm": 0.5909110307693481,
+      "learning_rate": 0.0004920198461998227,
+      "loss": 1.2596,
+      "step": 904
+    },
+    {
+      "epoch": 0.16144857729016146,
+      "grad_norm": 0.592135488986969,
+      "learning_rate": 0.0004920022760211496,
+      "loss": 1.2927,
+      "step": 905
+    },
+    {
+      "epoch": 0.16162697350816163,
+      "grad_norm": 0.6415541172027588,
+      "learning_rate": 0.0004919846868357218,
+      "loss": 1.3682,
+      "step": 906
+    },
+    {
+      "epoch": 0.1618053697261618,
+      "grad_norm": 0.5409737229347229,
+      "learning_rate": 0.0004919670786449208,
+      "loss": 1.2196,
+      "step": 907
+    },
+    {
+      "epoch": 0.16198376594416197,
+      "grad_norm": 25.429161071777344,
+      "learning_rate": 0.0004919494514501298,
+      "loss": 1.6991,
+      "step": 908
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 4.827995777130127,
+      "learning_rate": 0.0004919318052527328,
+      "loss": 1.4752,
+      "step": 909
+    },
+    {
+      "epoch": 0.16234055838016234,
+      "grad_norm": 0.8540498614311218,
+      "learning_rate": 0.0004919141400541161,
+      "loss": 1.6279,
+      "step": 910
+    },
+    {
+      "epoch": 0.1625189545981625,
+      "grad_norm": 0.6418678760528564,
+      "learning_rate": 0.0004918964558556668,
+      "loss": 1.4733,
+      "step": 911
+    },
+    {
+      "epoch": 0.16269735081616268,
+      "grad_norm": 0.7732519507408142,
+      "learning_rate": 0.0004918787526587739,
+      "loss": 1.2503,
+      "step": 912
+    },
+    {
+      "epoch": 0.16287574703416288,
+      "grad_norm": 0.678031325340271,
+      "learning_rate": 0.000491861030464828,
+      "loss": 1.208,
+      "step": 913
+    },
+    {
+      "epoch": 0.16305414325216305,
+      "grad_norm": 0.6401492357254028,
+      "learning_rate": 0.0004918432892752208,
+      "loss": 1.0655,
+      "step": 914
+    },
+    {
+      "epoch": 0.16323253947016322,
+      "grad_norm": 0.904754102230072,
+      "learning_rate": 0.0004918255290913457,
+      "loss": 1.3065,
+      "step": 915
+    },
+    {
+      "epoch": 0.16341093568816342,
+      "grad_norm": 0.6582340002059937,
+      "learning_rate": 0.0004918077499145977,
+      "loss": 1.3543,
+      "step": 916
+    },
+    {
+      "epoch": 0.1635893319061636,
+      "grad_norm": 1.050347089767456,
+      "learning_rate": 0.000491789951746373,
+      "loss": 1.3799,
+      "step": 917
+    },
+    {
+      "epoch": 0.16376772812416376,
+      "grad_norm": 6.393340587615967,
+      "learning_rate": 0.0004917721345880696,
+      "loss": 1.2498,
+      "step": 918
+    },
+    {
+      "epoch": 0.16394612434216393,
+      "grad_norm": 4.43874454498291,
+      "learning_rate": 0.0004917542984410867,
+      "loss": 1.3987,
+      "step": 919
+    },
+    {
+      "epoch": 0.16412452056016413,
+      "grad_norm": 2.1143059730529785,
+      "learning_rate": 0.0004917364433068253,
+      "loss": 1.4354,
+      "step": 920
+    },
+    {
+      "epoch": 0.1643029167781643,
+      "grad_norm": 2.551389455795288,
+      "learning_rate": 0.0004917185691866876,
+      "loss": 1.4108,
+      "step": 921
+    },
+    {
+      "epoch": 0.16448131299616447,
+      "grad_norm": 4.1894426345825195,
+      "learning_rate": 0.0004917006760820776,
+      "loss": 1.4468,
+      "step": 922
+    },
+    {
+      "epoch": 0.16465970921416467,
+      "grad_norm": 0.6970653533935547,
+      "learning_rate": 0.0004916827639944005,
+      "loss": 1.3746,
+      "step": 923
+    },
+    {
+      "epoch": 0.16483810543216484,
+      "grad_norm": 0.559752345085144,
+      "learning_rate": 0.0004916648329250631,
+      "loss": 1.1099,
+      "step": 924
+    },
+    {
+      "epoch": 0.16501650165016502,
+      "grad_norm": 0.6053521037101746,
+      "learning_rate": 0.0004916468828754737,
+      "loss": 1.3007,
+      "step": 925
+    },
+    {
+      "epoch": 0.16519489786816519,
+      "grad_norm": 0.5920356512069702,
+      "learning_rate": 0.000491628913847042,
+      "loss": 1.5636,
+      "step": 926
+    },
+    {
+      "epoch": 0.16537329408616538,
+      "grad_norm": 0.5610060095787048,
+      "learning_rate": 0.0004916109258411795,
+      "loss": 1.0346,
+      "step": 927
+    },
+    {
+      "epoch": 0.16555169030416556,
+      "grad_norm": 0.6758180260658264,
+      "learning_rate": 0.0004915929188592989,
+      "loss": 1.0975,
+      "step": 928
+    },
+    {
+      "epoch": 0.16573008652216573,
+      "grad_norm": 0.6995829343795776,
+      "learning_rate": 0.0004915748929028145,
+      "loss": 1.574,
+      "step": 929
+    },
+    {
+      "epoch": 0.1659084827401659,
+      "grad_norm": 0.6252689957618713,
+      "learning_rate": 0.0004915568479731417,
+      "loss": 1.3219,
+      "step": 930
+    },
+    {
+      "epoch": 0.1660868789581661,
+      "grad_norm": 0.7008538842201233,
+      "learning_rate": 0.0004915387840716982,
+      "loss": 1.3542,
+      "step": 931
+    },
+    {
+      "epoch": 0.16626527517616627,
+      "grad_norm": 0.6104601621627808,
+      "learning_rate": 0.0004915207011999025,
+      "loss": 1.2171,
+      "step": 932
+    },
+    {
+      "epoch": 0.16644367139416644,
+      "grad_norm": 0.5563907623291016,
+      "learning_rate": 0.0004915025993591748,
+      "loss": 1.2237,
+      "step": 933
+    },
+    {
+      "epoch": 0.16662206761216664,
+      "grad_norm": 0.7092035412788391,
+      "learning_rate": 0.000491484478550937,
+      "loss": 1.5328,
+      "step": 934
+    },
+    {
+      "epoch": 0.1668004638301668,
+      "grad_norm": 0.5745419263839722,
+      "learning_rate": 0.0004914663387766121,
+      "loss": 1.2546,
+      "step": 935
+    },
+    {
+      "epoch": 0.16697886004816698,
+      "grad_norm": 0.6714827418327332,
+      "learning_rate": 0.0004914481800376248,
+      "loss": 1.2575,
+      "step": 936
+    },
+    {
+      "epoch": 0.16715725626616715,
+      "grad_norm": 0.6761769652366638,
+      "learning_rate": 0.0004914300023354015,
+      "loss": 1.4926,
+      "step": 937
+    },
+    {
+      "epoch": 0.16733565248416735,
+      "grad_norm": 0.5255780220031738,
+      "learning_rate": 0.0004914118056713695,
+      "loss": 1.0914,
+      "step": 938
+    },
+    {
+      "epoch": 0.16751404870216752,
+      "grad_norm": 0.5681695938110352,
+      "learning_rate": 0.0004913935900469584,
+      "loss": 1.1869,
+      "step": 939
+    },
+    {
+      "epoch": 0.1676924449201677,
+      "grad_norm": 0.6561183929443359,
+      "learning_rate": 0.0004913753554635985,
+      "loss": 1.6089,
+      "step": 940
+    },
+    {
+      "epoch": 0.16787084113816786,
+      "grad_norm": 0.6278382539749146,
+      "learning_rate": 0.0004913571019227221,
+      "loss": 1.3546,
+      "step": 941
+    },
+    {
+      "epoch": 0.16804923735616806,
+      "grad_norm": 0.6552897095680237,
+      "learning_rate": 0.0004913388294257627,
+      "loss": 1.1408,
+      "step": 942
+    },
+    {
+      "epoch": 0.16822763357416823,
+      "grad_norm": 0.7079983353614807,
+      "learning_rate": 0.0004913205379741555,
+      "loss": 1.704,
+      "step": 943
+    },
+    {
+      "epoch": 0.1684060297921684,
+      "grad_norm": 0.8556866645812988,
+      "learning_rate": 0.0004913022275693372,
+      "loss": 1.3536,
+      "step": 944
+    },
+    {
+      "epoch": 0.1685844260101686,
+      "grad_norm": 0.6923747062683105,
+      "learning_rate": 0.0004912838982127456,
+      "loss": 1.2432,
+      "step": 945
+    },
+    {
+      "epoch": 0.16876282222816877,
+      "grad_norm": 0.6427940726280212,
+      "learning_rate": 0.0004912655499058207,
+      "loss": 1.0336,
+      "step": 946
+    },
+    {
+      "epoch": 0.16894121844616894,
+      "grad_norm": 0.7308516502380371,
+      "learning_rate": 0.0004912471826500032,
+      "loss": 1.5323,
+      "step": 947
+    },
+    {
+      "epoch": 0.1691196146641691,
+      "grad_norm": 0.6658211946487427,
+      "learning_rate": 0.0004912287964467358,
+      "loss": 1.3749,
+      "step": 948
+    },
+    {
+      "epoch": 0.1692980108821693,
+      "grad_norm": 0.5936264395713806,
+      "learning_rate": 0.0004912103912974626,
+      "loss": 1.1564,
+      "step": 949
+    },
+    {
+      "epoch": 0.16947640710016948,
+      "grad_norm": 0.6166514754295349,
+      "learning_rate": 0.000491191967203629,
+      "loss": 1.5516,
+      "step": 950
+    },
+    {
+      "epoch": 0.16965480331816965,
+      "grad_norm": 0.6150327920913696,
+      "learning_rate": 0.0004911735241666821,
+      "loss": 1.4399,
+      "step": 951
+    },
+    {
+      "epoch": 0.16983319953616982,
+      "grad_norm": 0.6488068699836731,
+      "learning_rate": 0.0004911550621880704,
+      "loss": 1.4561,
+      "step": 952
+    },
+    {
+      "epoch": 0.17001159575417002,
+      "grad_norm": 0.565139889717102,
+      "learning_rate": 0.0004911365812692439,
+      "loss": 1.3605,
+      "step": 953
+    },
+    {
+      "epoch": 0.1701899919721702,
+      "grad_norm": 0.5767154693603516,
+      "learning_rate": 0.0004911180814116541,
+      "loss": 1.2484,
+      "step": 954
+    },
+    {
+      "epoch": 0.17036838819017036,
+      "grad_norm": 0.6060677766799927,
+      "learning_rate": 0.0004910995626167539,
+      "loss": 1.1449,
+      "step": 955
+    },
+    {
+      "epoch": 0.17054678440817056,
+      "grad_norm": 0.5569210648536682,
+      "learning_rate": 0.0004910810248859979,
+      "loss": 1.2146,
+      "step": 956
+    },
+    {
+      "epoch": 0.17072518062617073,
+      "grad_norm": 1.7405076026916504,
+      "learning_rate": 0.0004910624682208418,
+      "loss": 1.1172,
+      "step": 957
+    },
+    {
+      "epoch": 0.1709035768441709,
+      "grad_norm": 0.6181856393814087,
+      "learning_rate": 0.0004910438926227433,
+      "loss": 1.3699,
+      "step": 958
+    },
+    {
+      "epoch": 0.17108197306217107,
+      "grad_norm": 0.6462327837944031,
+      "learning_rate": 0.0004910252980931611,
+      "loss": 1.3112,
+      "step": 959
+    },
+    {
+      "epoch": 0.17126036928017127,
+      "grad_norm": 0.5451391935348511,
+      "learning_rate": 0.0004910066846335558,
+      "loss": 1.1995,
+      "step": 960
+    },
+    {
+      "epoch": 0.17143876549817144,
+      "grad_norm": 0.5842812061309814,
+      "learning_rate": 0.0004909880522453891,
+      "loss": 1.2324,
+      "step": 961
+    },
+    {
+      "epoch": 0.1716171617161716,
+      "grad_norm": 0.56785649061203,
+      "learning_rate": 0.0004909694009301247,
+      "loss": 1.0771,
+      "step": 962
+    },
+    {
+      "epoch": 0.17179555793417178,
+      "grad_norm": 0.6656644344329834,
+      "learning_rate": 0.000490950730689227,
+      "loss": 1.3911,
+      "step": 963
+    },
+    {
+      "epoch": 0.17197395415217198,
+      "grad_norm": 0.6386826038360596,
+      "learning_rate": 0.0004909320415241627,
+      "loss": 1.1475,
+      "step": 964
+    },
+    {
+      "epoch": 0.17215235037017215,
+      "grad_norm": 0.6195803284645081,
+      "learning_rate": 0.0004909133334363996,
+      "loss": 1.3983,
+      "step": 965
+    },
+    {
+      "epoch": 0.17233074658817232,
+      "grad_norm": 0.6595250368118286,
+      "learning_rate": 0.000490894606427407,
+      "loss": 1.1298,
+      "step": 966
+    },
+    {
+      "epoch": 0.17250914280617252,
+      "grad_norm": 0.5828666090965271,
+      "learning_rate": 0.0004908758604986555,
+      "loss": 1.0375,
+      "step": 967
+    },
+    {
+      "epoch": 0.1726875390241727,
+      "grad_norm": 0.5903612375259399,
+      "learning_rate": 0.0004908570956516176,
+      "loss": 1.3681,
+      "step": 968
+    },
+    {
+      "epoch": 0.17286593524217286,
+      "grad_norm": 0.587887167930603,
+      "learning_rate": 0.0004908383118877672,
+      "loss": 1.1854,
+      "step": 969
+    },
+    {
+      "epoch": 0.17304433146017303,
+      "grad_norm": 0.5458610653877258,
+      "learning_rate": 0.0004908195092085794,
+      "loss": 1.3502,
+      "step": 970
+    },
+    {
+      "epoch": 0.17322272767817323,
+      "grad_norm": 0.5689892768859863,
+      "learning_rate": 0.0004908006876155309,
+      "loss": 1.072,
+      "step": 971
+    },
+    {
+      "epoch": 0.1734011238961734,
+      "grad_norm": 0.5631921887397766,
+      "learning_rate": 0.0004907818471101,
+      "loss": 1.1866,
+      "step": 972
+    },
+    {
+      "epoch": 0.17357952011417357,
+      "grad_norm": 2.34973406791687,
+      "learning_rate": 0.0004907629876937665,
+      "loss": 1.3553,
+      "step": 973
+    },
+    {
+      "epoch": 0.17375791633217375,
+      "grad_norm": 0.5664592385292053,
+      "learning_rate": 0.0004907441093680115,
+      "loss": 1.2058,
+      "step": 974
+    },
+    {
+      "epoch": 0.17393631255017394,
+      "grad_norm": 0.6143642663955688,
+      "learning_rate": 0.0004907252121343178,
+      "loss": 1.4795,
+      "step": 975
+    },
+    {
+      "epoch": 0.17411470876817411,
+      "grad_norm": 0.5856114625930786,
+      "learning_rate": 0.0004907062959941695,
+      "loss": 1.2277,
+      "step": 976
+    },
+    {
+      "epoch": 0.17429310498617429,
+      "grad_norm": 0.5994637608528137,
+      "learning_rate": 0.0004906873609490523,
+      "loss": 1.3448,
+      "step": 977
+    },
+    {
+      "epoch": 0.17447150120417448,
+      "grad_norm": 0.5852928161621094,
+      "learning_rate": 0.0004906684070004534,
+      "loss": 1.1422,
+      "step": 978
+    },
+    {
+      "epoch": 0.17464989742217465,
+      "grad_norm": 1.4769107103347778,
+      "learning_rate": 0.0004906494341498614,
+      "loss": 1.351,
+      "step": 979
+    },
+    {
+      "epoch": 0.17482829364017483,
+      "grad_norm": 0.6843876838684082,
+      "learning_rate": 0.0004906304423987663,
+      "loss": 1.1986,
+      "step": 980
+    },
+    {
+      "epoch": 0.175006689858175,
+      "grad_norm": 0.7055404782295227,
+      "learning_rate": 0.0004906114317486599,
+      "loss": 1.3418,
+      "step": 981
+    },
+    {
+      "epoch": 0.1751850860761752,
+      "grad_norm": 0.6515609622001648,
+      "learning_rate": 0.000490592402201035,
+      "loss": 1.011,
+      "step": 982
+    },
+    {
+      "epoch": 0.17536348229417537,
+      "grad_norm": 0.6670236587524414,
+      "learning_rate": 0.0004905733537573867,
+      "loss": 1.2945,
+      "step": 983
+    },
+    {
+      "epoch": 0.17554187851217554,
+      "grad_norm": 0.6488009691238403,
+      "learning_rate": 0.0004905542864192104,
+      "loss": 1.3415,
+      "step": 984
+    },
+    {
+      "epoch": 0.1757202747301757,
+      "grad_norm": 1.1373000144958496,
+      "learning_rate": 0.0004905352001880041,
+      "loss": 1.562,
+      "step": 985
+    },
+    {
+      "epoch": 0.1758986709481759,
+      "grad_norm": 0.5701956748962402,
+      "learning_rate": 0.0004905160950652667,
+      "loss": 1.0377,
+      "step": 986
+    },
+    {
+      "epoch": 0.17607706716617608,
+      "grad_norm": 2.7552855014801025,
+      "learning_rate": 0.0004904969710524987,
+      "loss": 1.2074,
+      "step": 987
+    },
+    {
+      "epoch": 0.17625546338417625,
+      "grad_norm": 0.6393254995346069,
+      "learning_rate": 0.0004904778281512021,
+      "loss": 1.1729,
+      "step": 988
+    },
+    {
+      "epoch": 0.17643385960217645,
+      "grad_norm": 0.6372584104537964,
+      "learning_rate": 0.0004904586663628804,
+      "loss": 1.2313,
+      "step": 989
+    },
+    {
+      "epoch": 0.17661225582017662,
+      "grad_norm": 0.6509175896644592,
+      "learning_rate": 0.0004904394856890384,
+      "loss": 1.3779,
+      "step": 990
+    },
+    {
+      "epoch": 0.1767906520381768,
+      "grad_norm": 0.668499231338501,
+      "learning_rate": 0.0004904202861311827,
+      "loss": 1.3286,
+      "step": 991
+    },
+    {
+      "epoch": 0.17696904825617696,
+      "grad_norm": 0.6670610308647156,
+      "learning_rate": 0.0004904010676908213,
+      "loss": 1.3467,
+      "step": 992
+    },
+    {
+      "epoch": 0.17714744447417716,
+      "grad_norm": 0.6278952956199646,
+      "learning_rate": 0.0004903818303694633,
+      "loss": 1.2545,
+      "step": 993
+    },
+    {
+      "epoch": 0.17732584069217733,
+      "grad_norm": 0.6193589568138123,
+      "learning_rate": 0.0004903625741686199,
+      "loss": 1.2678,
+      "step": 994
+    },
+    {
+      "epoch": 0.1775042369101775,
+      "grad_norm": 0.6413539052009583,
+      "learning_rate": 0.0004903432990898033,
+      "loss": 1.5118,
+      "step": 995
+    },
+    {
+      "epoch": 0.17768263312817767,
+      "grad_norm": 1.0503791570663452,
+      "learning_rate": 0.0004903240051345276,
+      "loss": 1.3905,
+      "step": 996
+    },
+    {
+      "epoch": 0.17786102934617787,
+      "grad_norm": 0.5560838580131531,
+      "learning_rate": 0.0004903046923043077,
+      "loss": 0.9987,
+      "step": 997
+    },
+    {
+      "epoch": 0.17803942556417804,
+      "grad_norm": 0.7070960402488708,
+      "learning_rate": 0.0004902853606006609,
+      "loss": 1.3521,
+      "step": 998
+    },
+    {
+      "epoch": 0.1782178217821782,
+      "grad_norm": 0.7161464691162109,
+      "learning_rate": 0.0004902660100251051,
+      "loss": 1.1022,
+      "step": 999
+    },
+    {
+      "epoch": 0.1783962180001784,
+      "grad_norm": 0.6511627435684204,
+      "learning_rate": 0.0004902466405791604,
+      "loss": 1.284,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17857461421817858,
+      "grad_norm": 0.6245794892311096,
+      "learning_rate": 0.0004902272522643478,
+      "loss": 1.1996,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17875301043617875,
+      "grad_norm": 0.6342217922210693,
+      "learning_rate": 0.0004902078450821904,
+      "loss": 1.2702,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17893140665417892,
+      "grad_norm": 0.5756664872169495,
+      "learning_rate": 0.0004901884190342121,
+      "loss": 1.2978,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17910980287217912,
+      "grad_norm": 0.653351902961731,
+      "learning_rate": 0.0004901689741219386,
+      "loss": 1.4452,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1792881990901793,
+      "grad_norm": 0.6309232115745544,
+      "learning_rate": 0.0004901495103468974,
+      "loss": 1.4723,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17946659530817946,
+      "grad_norm": 0.617276668548584,
+      "learning_rate": 0.000490130027710617,
+      "loss": 1.2532,
+      "step": 1006
+    },
+    {
+      "epoch": 0.17964499152617963,
+      "grad_norm": 0.599210798740387,
+      "learning_rate": 0.0004901105262146275,
+      "loss": 1.2724,
+      "step": 1007
+    },
+    {
+      "epoch": 0.17982338774417983,
+      "grad_norm": 0.9026014804840088,
+      "learning_rate": 0.0004900910058604606,
+      "loss": 1.2691,
+      "step": 1008
+    },
+    {
+      "epoch": 0.18000178396218,
+      "grad_norm": 0.7611701488494873,
+      "learning_rate": 0.0004900714666496494,
+      "loss": 1.2221,
+      "step": 1009
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 0.6050606966018677,
+      "learning_rate": 0.0004900519085837285,
+      "loss": 1.2423,
+      "step": 1010
+    },
+    {
+      "epoch": 0.18035857639818037,
+      "grad_norm": 0.6327321529388428,
+      "learning_rate": 0.0004900323316642341,
+      "loss": 1.5898,
+      "step": 1011
+    },
+    {
+      "epoch": 0.18053697261618054,
+      "grad_norm": 3.784043788909912,
+      "learning_rate": 0.0004900127358927036,
+      "loss": 1.3516,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1807153688341807,
+      "grad_norm": 1.5190318822860718,
+      "learning_rate": 0.0004899931212706761,
+      "loss": 1.2567,
+      "step": 1013
+    },
+    {
+      "epoch": 0.18089376505218088,
+      "grad_norm": 0.7018091082572937,
+      "learning_rate": 0.0004899734877996922,
+      "loss": 1.3094,
+      "step": 1014
+    },
+    {
+      "epoch": 0.18107216127018108,
+      "grad_norm": 1.1731371879577637,
+      "learning_rate": 0.0004899538354812937,
+      "loss": 1.3155,
+      "step": 1015
+    },
+    {
+      "epoch": 0.18125055748818125,
+      "grad_norm": 0.6494358777999878,
+      "learning_rate": 0.0004899341643170243,
+      "loss": 1.2901,
+      "step": 1016
+    },
+    {
+      "epoch": 0.18142895370618142,
+      "grad_norm": 0.6760775446891785,
+      "learning_rate": 0.0004899144743084289,
+      "loss": 1.3969,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1816073499241816,
+      "grad_norm": 0.6156429648399353,
+      "learning_rate": 0.000489894765457054,
+      "loss": 1.1901,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1817857461421818,
+      "grad_norm": 0.9414156079292297,
+      "learning_rate": 0.0004898750377644474,
+      "loss": 1.1387,
+      "step": 1019
+    },
+    {
+      "epoch": 0.18196414236018196,
+      "grad_norm": 0.6034333109855652,
+      "learning_rate": 0.0004898552912321586,
+      "loss": 1.3314,
+      "step": 1020
+    },
+    {
+      "epoch": 0.18214253857818213,
+      "grad_norm": 0.5967845916748047,
+      "learning_rate": 0.0004898355258617383,
+      "loss": 1.3435,
+      "step": 1021
+    },
+    {
+      "epoch": 0.18232093479618233,
+      "grad_norm": 0.6214697957038879,
+      "learning_rate": 0.000489815741654739,
+      "loss": 1.1333,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1824993310141825,
+      "grad_norm": 0.7541596293449402,
+      "learning_rate": 0.0004897959386127148,
+      "loss": 1.1722,
+      "step": 1023
+    },
+    {
+      "epoch": 0.18267772723218267,
+      "grad_norm": 0.6246985793113708,
+      "learning_rate": 0.0004897761167372205,
+      "loss": 1.6439,
+      "step": 1024
+    },
+    {
+      "epoch": 0.18285612345018284,
+      "grad_norm": 2.4204161167144775,
+      "learning_rate": 0.0004897562760298134,
+      "loss": 1.442,
+      "step": 1025
+    },
+    {
+      "epoch": 0.18303451966818304,
+      "grad_norm": 8.560677528381348,
+      "learning_rate": 0.0004897364164920514,
+      "loss": 1.6211,
+      "step": 1026
+    },
+    {
+      "epoch": 0.18321291588618321,
+      "grad_norm": 0.8490382432937622,
+      "learning_rate": 0.0004897165381254945,
+      "loss": 1.2842,
+      "step": 1027
+    },
+    {
+      "epoch": 0.18339131210418338,
+      "grad_norm": 0.6359208226203918,
+      "learning_rate": 0.0004896966409317038,
+      "loss": 1.1782,
+      "step": 1028
+    },
+    {
+      "epoch": 0.18356970832218358,
+      "grad_norm": 0.6399941444396973,
+      "learning_rate": 0.0004896767249122421,
+      "loss": 1.1316,
+      "step": 1029
+    },
+    {
+      "epoch": 0.18374810454018375,
+      "grad_norm": 0.680789589881897,
+      "learning_rate": 0.0004896567900686736,
+      "loss": 1.2123,
+      "step": 1030
+    },
+    {
+      "epoch": 0.18392650075818393,
+      "grad_norm": 0.7316074371337891,
+      "learning_rate": 0.0004896368364025639,
+      "loss": 1.4407,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1841048969761841,
+      "grad_norm": 1.80794095993042,
+      "learning_rate": 0.0004896168639154802,
+      "loss": 1.3116,
+      "step": 1032
+    },
+    {
+      "epoch": 0.1842832931941843,
+      "grad_norm": 1.869382619857788,
+      "learning_rate": 0.000489596872608991,
+      "loss": 1.3759,
+      "step": 1033
+    },
+    {
+      "epoch": 0.18446168941218447,
+      "grad_norm": 0.9936087727546692,
+      "learning_rate": 0.0004895768624846667,
+      "loss": 1.4019,
+      "step": 1034
+    },
+    {
+      "epoch": 0.18464008563018464,
+      "grad_norm": 1.2543047666549683,
+      "learning_rate": 0.0004895568335440786,
+      "loss": 1.2554,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1848184818481848,
+      "grad_norm": 1.1524609327316284,
+      "learning_rate": 0.0004895367857888,
+      "loss": 1.1582,
+      "step": 1036
+    },
+    {
+      "epoch": 0.184996878066185,
+      "grad_norm": 0.8680905103683472,
+      "learning_rate": 0.0004895167192204053,
+      "loss": 1.4434,
+      "step": 1037
+    },
+    {
+      "epoch": 0.18517527428418518,
+      "grad_norm": 0.7361879348754883,
+      "learning_rate": 0.0004894966338404705,
+      "loss": 1.3415,
+      "step": 1038
+    },
+    {
+      "epoch": 0.18535367050218535,
+      "grad_norm": 0.6387485861778259,
+      "learning_rate": 0.0004894765296505732,
+      "loss": 1.2213,
+      "step": 1039
+    },
+    {
+      "epoch": 0.18553206672018555,
+      "grad_norm": 2.19881534576416,
+      "learning_rate": 0.0004894564066522923,
+      "loss": 1.1374,
+      "step": 1040
+    },
+    {
+      "epoch": 0.18571046293818572,
+      "grad_norm": 0.7040894627571106,
+      "learning_rate": 0.0004894362648472082,
+      "loss": 1.3711,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1858888591561859,
+      "grad_norm": 0.6495254039764404,
+      "learning_rate": 0.000489416104236903,
+      "loss": 1.106,
+      "step": 1042
+    },
+    {
+      "epoch": 0.18606725537418606,
+      "grad_norm": 0.832444429397583,
+      "learning_rate": 0.00048939592482296,
+      "loss": 1.3915,
+      "step": 1043
+    },
+    {
+      "epoch": 0.18624565159218626,
+      "grad_norm": 0.5508606433868408,
+      "learning_rate": 0.000489375726606964,
+      "loss": 0.9048,
+      "step": 1044
+    },
+    {
+      "epoch": 0.18642404781018643,
+      "grad_norm": 0.6425449848175049,
+      "learning_rate": 0.0004893555095905013,
+      "loss": 1.3206,
+      "step": 1045
+    },
+    {
+      "epoch": 0.1866024440281866,
+      "grad_norm": 0.5754274129867554,
+      "learning_rate": 0.0004893352737751601,
+      "loss": 1.2648,
+      "step": 1046
+    },
+    {
+      "epoch": 0.18678084024618677,
+      "grad_norm": 0.5647794604301453,
+      "learning_rate": 0.0004893150191625295,
+      "loss": 1.171,
+      "step": 1047
+    },
+    {
+      "epoch": 0.18695923646418697,
+      "grad_norm": 0.6045464277267456,
+      "learning_rate": 0.0004892947457542002,
+      "loss": 0.9138,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18713763268218714,
+      "grad_norm": 0.7438944578170776,
+      "learning_rate": 0.0004892744535517646,
+      "loss": 1.4838,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1873160289001873,
+      "grad_norm": 0.5550808310508728,
+      "learning_rate": 0.0004892541425568163,
+      "loss": 1.2546,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1874944251181875,
+      "grad_norm": 0.574200451374054,
+      "learning_rate": 0.0004892338127709507,
+      "loss": 1.2506,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18767282133618768,
+      "grad_norm": 1.666972041130066,
+      "learning_rate": 0.0004892134641957642,
+      "loss": 1.389,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18785121755418785,
+      "grad_norm": 0.6545609831809998,
+      "learning_rate": 0.0004891930968328554,
+      "loss": 1.2801,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18802961377218802,
+      "grad_norm": 1.3686062097549438,
+      "learning_rate": 0.0004891727106838236,
+      "loss": 1.2763,
+      "step": 1054
+    },
+    {
+      "epoch": 0.18820800999018822,
+      "grad_norm": 1.248139500617981,
+      "learning_rate": 0.0004891523057502701,
+      "loss": 1.4209,
+      "step": 1055
+    },
+    {
+      "epoch": 0.1883864062081884,
+      "grad_norm": 0.7310791611671448,
+      "learning_rate": 0.0004891318820337974,
+      "loss": 1.4907,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18856480242618856,
+      "grad_norm": 0.6431208848953247,
+      "learning_rate": 0.0004891114395360096,
+      "loss": 1.5441,
+      "step": 1057
+    },
+    {
+      "epoch": 0.18874319864418873,
+      "grad_norm": 2.24438738822937,
+      "learning_rate": 0.0004890909782585121,
+      "loss": 1.2141,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18892159486218893,
+      "grad_norm": 4.204579830169678,
+      "learning_rate": 0.0004890704982029122,
+      "loss": 1.4053,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1890999910801891,
+      "grad_norm": 0.8106665015220642,
+      "learning_rate": 0.0004890499993708182,
+      "loss": 1.1461,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18927838729818927,
+      "grad_norm": 0.7268997430801392,
+      "learning_rate": 0.0004890294817638401,
+      "loss": 1.1508,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18945678351618947,
+      "grad_norm": 0.9513046741485596,
+      "learning_rate": 0.0004890089453835894,
+      "loss": 1.2753,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18963517973418964,
+      "grad_norm": 0.7538788318634033,
+      "learning_rate": 0.000488988390231679,
+      "loss": 1.088,
+      "step": 1063
+    },
+    {
+      "epoch": 0.1898135759521898,
+      "grad_norm": 0.6763918399810791,
+      "learning_rate": 0.0004889678163097233,
+      "loss": 1.1825,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18999197217018998,
+      "grad_norm": 0.7881626486778259,
+      "learning_rate": 0.0004889472236193381,
+      "loss": 1.3298,
+      "step": 1065
+    },
+    {
+      "epoch": 0.19017036838819018,
+      "grad_norm": 0.6589952111244202,
+      "learning_rate": 0.0004889266121621407,
+      "loss": 1.2218,
+      "step": 1066
+    },
+    {
+      "epoch": 0.19034876460619035,
+      "grad_norm": 0.7166324853897095,
+      "learning_rate": 0.0004889059819397501,
+      "loss": 1.5962,
+      "step": 1067
+    },
+    {
+      "epoch": 0.19052716082419052,
+      "grad_norm": 0.6386669874191284,
+      "learning_rate": 0.0004888853329537865,
+      "loss": 1.2342,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1907055570421907,
+      "grad_norm": 0.7325186729431152,
+      "learning_rate": 0.0004888646652058716,
+      "loss": 1.2617,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1908839532601909,
+      "grad_norm": 0.5728015899658203,
+      "learning_rate": 0.0004888439786976287,
+      "loss": 1.3414,
+      "step": 1070
+    },
+    {
+      "epoch": 0.19106234947819106,
+      "grad_norm": 0.6659836769104004,
+      "learning_rate": 0.0004888232734306825,
+      "loss": 1.2854,
+      "step": 1071
+    },
+    {
+      "epoch": 0.19124074569619123,
+      "grad_norm": 0.7242840528488159,
+      "learning_rate": 0.0004888025494066592,
+      "loss": 1.4503,
+      "step": 1072
+    },
+    {
+      "epoch": 0.19141914191419143,
+      "grad_norm": 0.5824533700942993,
+      "learning_rate": 0.0004887818066271864,
+      "loss": 1.3502,
+      "step": 1073
+    },
+    {
+      "epoch": 0.1915975381321916,
+      "grad_norm": 0.6332762241363525,
+      "learning_rate": 0.0004887610450938932,
+      "loss": 1.0815,
+      "step": 1074
+    },
+    {
+      "epoch": 0.19177593435019177,
+      "grad_norm": 1.1359984874725342,
+      "learning_rate": 0.0004887402648084104,
+      "loss": 1.2975,
+      "step": 1075
+    },
+    {
+      "epoch": 0.19195433056819194,
+      "grad_norm": 0.8233804106712341,
+      "learning_rate": 0.0004887194657723699,
+      "loss": 1.1015,
+      "step": 1076
+    },
+    {
+      "epoch": 0.19213272678619214,
+      "grad_norm": 0.6143079996109009,
+      "learning_rate": 0.0004886986479874052,
+      "loss": 1.2998,
+      "step": 1077
+    },
+    {
+      "epoch": 0.1923111230041923,
+      "grad_norm": 1.4513376951217651,
+      "learning_rate": 0.0004886778114551515,
+      "loss": 1.3469,
+      "step": 1078
+    },
+    {
+      "epoch": 0.19248951922219248,
+      "grad_norm": 0.6782541275024414,
+      "learning_rate": 0.0004886569561772452,
+      "loss": 1.1655,
+      "step": 1079
+    },
+    {
+      "epoch": 0.19266791544019266,
+      "grad_norm": 0.6781834959983826,
+      "learning_rate": 0.0004886360821553242,
+      "loss": 1.4877,
+      "step": 1080
+    },
+    {
+      "epoch": 0.19284631165819285,
+      "grad_norm": 0.6629863977432251,
+      "learning_rate": 0.000488615189391028,
+      "loss": 1.5273,
+      "step": 1081
+    },
+    {
+      "epoch": 0.19302470787619302,
+      "grad_norm": 0.5616167783737183,
+      "learning_rate": 0.0004885942778859976,
+      "loss": 1.1904,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1932031040941932,
+      "grad_norm": 0.5528833270072937,
+      "learning_rate": 0.0004885733476418752,
+      "loss": 0.9634,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1933815003121934,
+      "grad_norm": 0.7818634510040283,
+      "learning_rate": 0.0004885523986603048,
+      "loss": 1.3033,
+      "step": 1084
+    },
+    {
+      "epoch": 0.19355989653019356,
+      "grad_norm": 0.6581936478614807,
+      "learning_rate": 0.0004885314309429316,
+      "loss": 1.4896,
+      "step": 1085
+    },
+    {
+      "epoch": 0.19373829274819374,
+      "grad_norm": 0.652714192867279,
+      "learning_rate": 0.0004885104444914026,
+      "loss": 1.2453,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1939166889661939,
+      "grad_norm": 3.307530641555786,
+      "learning_rate": 0.0004884894393073658,
+      "loss": 1.173,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1940950851841941,
+      "grad_norm": 0.685991644859314,
+      "learning_rate": 0.0004884684153924711,
+      "loss": 0.9765,
+      "step": 1088
+    },
+    {
+      "epoch": 0.19427348140219428,
+      "grad_norm": 0.678576648235321,
+      "learning_rate": 0.0004884473727483697,
+      "loss": 1.1766,
+      "step": 1089
+    },
+    {
+      "epoch": 0.19445187762019445,
+      "grad_norm": 0.7492129802703857,
+      "learning_rate": 0.0004884263113767143,
+      "loss": 1.4711,
+      "step": 1090
+    },
+    {
+      "epoch": 0.19463027383819462,
+      "grad_norm": 0.6048595309257507,
+      "learning_rate": 0.0004884052312791588,
+      "loss": 1.0355,
+      "step": 1091
+    },
+    {
+      "epoch": 0.19480867005619482,
+      "grad_norm": 0.6212369203567505,
+      "learning_rate": 0.0004883841324573592,
+      "loss": 1.2895,
+      "step": 1092
+    },
+    {
+      "epoch": 0.194987066274195,
+      "grad_norm": 0.5634220838546753,
+      "learning_rate": 0.0004883630149129725,
+      "loss": 0.9151,
+      "step": 1093
+    },
+    {
+      "epoch": 0.19516546249219516,
+      "grad_norm": 0.6526708602905273,
+      "learning_rate": 0.000488341878647657,
+      "loss": 1.3422,
+      "step": 1094
+    },
+    {
+      "epoch": 0.19534385871019536,
+      "grad_norm": 0.5777380466461182,
+      "learning_rate": 0.000488320723663073,
+      "loss": 1.3257,
+      "step": 1095
+    },
+    {
+      "epoch": 0.19552225492819553,
+      "grad_norm": 0.9771785736083984,
+      "learning_rate": 0.0004882995499608819,
+      "loss": 1.4458,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1957006511461957,
+      "grad_norm": 0.942666232585907,
+      "learning_rate": 0.00048827835754274674,
+      "loss": 1.3359,
+      "step": 1097
+    },
+    {
+      "epoch": 0.19587904736419587,
+      "grad_norm": 0.6071109175682068,
+      "learning_rate": 0.00048825714641033185,
+      "loss": 1.2979,
+      "step": 1098
+    },
+    {
+      "epoch": 0.19605744358219607,
+      "grad_norm": 0.9285879135131836,
+      "learning_rate": 0.0004882359165653033,
+      "loss": 1.2042,
+      "step": 1099
+    },
+    {
+      "epoch": 0.19623583980019624,
+      "grad_norm": 0.6125981211662292,
+      "learning_rate": 0.00048821466800932835,
+      "loss": 1.2767,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1964142360181964,
+      "grad_norm": 0.549712061882019,
+      "learning_rate": 0.00048819340074407583,
+      "loss": 1.0559,
+      "step": 1101
+    },
+    {
+      "epoch": 0.19659263223619658,
+      "grad_norm": 0.6521591544151306,
+      "learning_rate": 0.00048817211477121617,
+      "loss": 1.0833,
+      "step": 1102
+    },
+    {
+      "epoch": 0.19677102845419678,
+      "grad_norm": 0.6663747429847717,
+      "learning_rate": 0.0004881508100924211,
+      "loss": 1.4768,
+      "step": 1103
+    },
+    {
+      "epoch": 0.19694942467219695,
+      "grad_norm": 1.3679718971252441,
+      "learning_rate": 0.00048812948670936385,
+      "loss": 1.43,
+      "step": 1104
+    },
+    {
+      "epoch": 0.19712782089019712,
+      "grad_norm": 0.7124596238136292,
+      "learning_rate": 0.0004881081446237192,
+      "loss": 1.0491,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19730621710819732,
+      "grad_norm": 0.6204227209091187,
+      "learning_rate": 0.0004880867838371634,
+      "loss": 1.2674,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1974846133261975,
+      "grad_norm": 1.057226300239563,
+      "learning_rate": 0.00048806540435137404,
+      "loss": 1.2598,
+      "step": 1107
+    },
+    {
+      "epoch": 0.19766300954419766,
+      "grad_norm": 0.7484321594238281,
+      "learning_rate": 0.00048804400616803026,
+      "loss": 1.2401,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19784140576219783,
+      "grad_norm": 12.229165077209473,
+      "learning_rate": 0.0004880225892888126,
+      "loss": 1.2016,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19801980198019803,
+      "grad_norm": 0.7973271608352661,
+      "learning_rate": 0.0004880011537154033,
+      "loss": 1.4108,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1981981981981982,
+      "grad_norm": 1.3274434804916382,
+      "learning_rate": 0.0004879796994494858,
+      "loss": 1.013,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19837659441619837,
+      "grad_norm": 0.8325060606002808,
+      "learning_rate": 0.00048795822649274506,
+      "loss": 1.3261,
+      "step": 1112
+    },
+    {
+      "epoch": 0.19855499063419854,
+      "grad_norm": 0.9182140231132507,
+      "learning_rate": 0.0004879367348468676,
+      "loss": 1.1414,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19873338685219874,
+      "grad_norm": 0.7945131063461304,
+      "learning_rate": 0.0004879152245135415,
+      "loss": 1.3566,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1989117830701989,
+      "grad_norm": 0.7313281297683716,
+      "learning_rate": 0.00048789369549445596,
+      "loss": 1.2533,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19909017928819908,
+      "grad_norm": 0.7198520302772522,
+      "learning_rate": 0.00048787214779130196,
+      "loss": 1.4449,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19926857550619928,
+      "grad_norm": 0.6107659935951233,
+      "learning_rate": 0.00048785058140577185,
+      "loss": 1.1473,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19944697172419945,
+      "grad_norm": 0.6234009861946106,
+      "learning_rate": 0.0004878289963395594,
+      "loss": 1.0506,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19962536794219962,
+      "grad_norm": 0.6401305198669434,
+      "learning_rate": 0.0004878073925943599,
+      "loss": 1.2247,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1998037641601998,
+      "grad_norm": 0.6568447947502136,
+      "learning_rate": 0.0004877857701718702,
+      "loss": 1.2032,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1999821603782,
+      "grad_norm": 0.6193853616714478,
+      "learning_rate": 0.0004877641290737884,
+      "loss": 1.2807,
+      "step": 1121
+    },
+    {
+      "epoch": 0.20016055659620016,
+      "grad_norm": 0.605709969997406,
+      "learning_rate": 0.0004877424693018142,
+      "loss": 1.3687,
+      "step": 1122
+    },
+    {
+      "epoch": 0.20033895281420033,
+      "grad_norm": 0.5979405045509338,
+      "learning_rate": 0.0004877207908576488,
+      "loss": 1.1466,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2005173490322005,
+      "grad_norm": 0.5929616093635559,
+      "learning_rate": 0.00048769909374299483,
+      "loss": 1.3955,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2006957452502007,
+      "grad_norm": 0.6005091071128845,
+      "learning_rate": 0.00048767737795955623,
+      "loss": 1.468,
+      "step": 1125
+    },
+    {
+      "epoch": 0.20087414146820087,
+      "grad_norm": 0.5641143918037415,
+      "learning_rate": 0.0004876556435090387,
+      "loss": 1.495,
+      "step": 1126
+    },
+    {
+      "epoch": 0.20105253768620104,
+      "grad_norm": 0.6153658628463745,
+      "learning_rate": 0.0004876338903931492,
+      "loss": 1.1535,
+      "step": 1127
+    },
+    {
+      "epoch": 0.20123093390420124,
+      "grad_norm": 0.5413753390312195,
+      "learning_rate": 0.0004876121186135962,
+      "loss": 1.002,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2014093301222014,
+      "grad_norm": 0.573294460773468,
+      "learning_rate": 0.00048759032817208964,
+      "loss": 1.3631,
+      "step": 1129
+    },
+    {
+      "epoch": 0.20158772634020158,
+      "grad_norm": 0.569232702255249,
+      "learning_rate": 0.00048756851907034093,
+      "loss": 1.105,
+      "step": 1130
+    },
+    {
+      "epoch": 0.20176612255820175,
+      "grad_norm": 0.6847321391105652,
+      "learning_rate": 0.000487546691310063,
+      "loss": 1.1282,
+      "step": 1131
+    },
+    {
+      "epoch": 0.20194451877620195,
+      "grad_norm": 0.7410063743591309,
+      "learning_rate": 0.00048752484489297023,
+      "loss": 1.2903,
+      "step": 1132
+    },
+    {
+      "epoch": 0.20212291499420212,
+      "grad_norm": 0.6502756476402283,
+      "learning_rate": 0.0004875029798207783,
+      "loss": 1.4029,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2023013112122023,
+      "grad_norm": 0.569385826587677,
+      "learning_rate": 0.0004874810960952045,
+      "loss": 1.2896,
+      "step": 1134
+    },
+    {
+      "epoch": 0.2024797074302025,
+      "grad_norm": 0.5342479348182678,
+      "learning_rate": 0.00048745919371796765,
+      "loss": 1.1572,
+      "step": 1135
+    },
+    {
+      "epoch": 0.20265810364820266,
+      "grad_norm": 0.750543475151062,
+      "learning_rate": 0.0004874372726907879,
+      "loss": 1.3087,
+      "step": 1136
+    },
+    {
+      "epoch": 0.20283649986620284,
+      "grad_norm": 0.6126813292503357,
+      "learning_rate": 0.00048741533301538685,
+      "loss": 1.4071,
+      "step": 1137
+    },
+    {
+      "epoch": 0.203014896084203,
+      "grad_norm": 0.648357093334198,
+      "learning_rate": 0.00048739337469348785,
+      "loss": 1.3212,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2031932923022032,
+      "grad_norm": 0.6305087208747864,
+      "learning_rate": 0.00048737139772681525,
+      "loss": 1.1886,
+      "step": 1139
+    },
+    {
+      "epoch": 0.20337168852020338,
+      "grad_norm": 0.5908694863319397,
+      "learning_rate": 0.0004873494021170953,
+      "loss": 1.1184,
+      "step": 1140
+    },
+    {
+      "epoch": 0.20355008473820355,
+      "grad_norm": 0.6326526999473572,
+      "learning_rate": 0.0004873273878660555,
+      "loss": 1.358,
+      "step": 1141
+    },
+    {
+      "epoch": 0.20372848095620372,
+      "grad_norm": 0.9167863130569458,
+      "learning_rate": 0.00048730535497542465,
+      "loss": 1.3994,
+      "step": 1142
+    },
+    {
+      "epoch": 0.20390687717420392,
+      "grad_norm": 0.5850123167037964,
+      "learning_rate": 0.0004872833034469334,
+      "loss": 1.4304,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2040852733922041,
+      "grad_norm": 0.5937606692314148,
+      "learning_rate": 0.00048726123328231367,
+      "loss": 1.1403,
+      "step": 1144
+    },
+    {
+      "epoch": 0.20426366961020426,
+      "grad_norm": 0.6057407855987549,
+      "learning_rate": 0.00048723914448329863,
+      "loss": 1.5015,
+      "step": 1145
+    },
+    {
+      "epoch": 0.20444206582820446,
+      "grad_norm": 0.5308231711387634,
+      "learning_rate": 0.0004872170370516234,
+      "loss": 1.2105,
+      "step": 1146
+    },
+    {
+      "epoch": 0.20462046204620463,
+      "grad_norm": 1.8873573541641235,
+      "learning_rate": 0.0004871949109890241,
+      "loss": 1.3811,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2047988582642048,
+      "grad_norm": 0.6975631713867188,
+      "learning_rate": 0.0004871727662972386,
+      "loss": 1.331,
+      "step": 1148
+    },
+    {
+      "epoch": 0.20497725448220497,
+      "grad_norm": 4.755294322967529,
+      "learning_rate": 0.00048715060297800606,
+      "loss": 1.3582,
+      "step": 1149
+    },
+    {
+      "epoch": 0.20515565070020517,
+      "grad_norm": 5.158052921295166,
+      "learning_rate": 0.0004871284210330673,
+      "loss": 1.445,
+      "step": 1150
+    },
+    {
+      "epoch": 0.20533404691820534,
+      "grad_norm": 9.641525268554688,
+      "learning_rate": 0.0004871062204641643,
+      "loss": 1.25,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2055124431362055,
+      "grad_norm": 1.7820740938186646,
+      "learning_rate": 0.00048708400127304085,
+      "loss": 1.4857,
+      "step": 1152
+    },
+    {
+      "epoch": 0.20569083935420568,
+      "grad_norm": 1.1601697206497192,
+      "learning_rate": 0.0004870617634614419,
+      "loss": 1.3956,
+      "step": 1153
+    },
+    {
+      "epoch": 0.20586923557220588,
+      "grad_norm": 0.9471563100814819,
+      "learning_rate": 0.0004870395070311141,
+      "loss": 1.2032,
+      "step": 1154
+    },
+    {
+      "epoch": 0.20604763179020605,
+      "grad_norm": 1.270116925239563,
+      "learning_rate": 0.00048701723198380545,
+      "loss": 1.5996,
+      "step": 1155
+    },
+    {
+      "epoch": 0.20622602800820622,
+      "grad_norm": 1.2946140766143799,
+      "learning_rate": 0.00048699493832126534,
+      "loss": 1.2046,
+      "step": 1156
+    },
+    {
+      "epoch": 0.20640442422620642,
+      "grad_norm": 0.6727069020271301,
+      "learning_rate": 0.00048697262604524485,
+      "loss": 1.1865,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2065828204442066,
+      "grad_norm": 0.7931843400001526,
+      "learning_rate": 0.00048695029515749615,
+      "loss": 1.2591,
+      "step": 1158
+    },
+    {
+      "epoch": 0.20676121666220676,
+      "grad_norm": 0.7511464953422546,
+      "learning_rate": 0.0004869279456597733,
+      "loss": 1.2946,
+      "step": 1159
+    },
+    {
+      "epoch": 0.20693961288020693,
+      "grad_norm": 0.6185610294342041,
+      "learning_rate": 0.00048690557755383157,
+      "loss": 1.0467,
+      "step": 1160
+    },
+    {
+      "epoch": 0.20711800909820713,
+      "grad_norm": 1.2515671253204346,
+      "learning_rate": 0.00048688319084142775,
+      "loss": 1.3274,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2072964053162073,
+      "grad_norm": 66.89749908447266,
+      "learning_rate": 0.00048686078552432,
+      "loss": 1.4026,
+      "step": 1162
+    },
+    {
+      "epoch": 0.20747480153420747,
+      "grad_norm": 1.6944903135299683,
+      "learning_rate": 0.0004868383616042682,
+      "loss": 1.2216,
+      "step": 1163
+    },
+    {
+      "epoch": 0.20765319775220764,
+      "grad_norm": 0.8577490448951721,
+      "learning_rate": 0.0004868159190830333,
+      "loss": 1.1645,
+      "step": 1164
+    },
+    {
+      "epoch": 0.20783159397020784,
+      "grad_norm": 0.8082415461540222,
+      "learning_rate": 0.00048679345796237805,
+      "loss": 1.3516,
+      "step": 1165
+    },
+    {
+      "epoch": 0.208009990188208,
+      "grad_norm": 0.6959142088890076,
+      "learning_rate": 0.00048677097824406646,
+      "loss": 1.4433,
+      "step": 1166
+    },
+    {
+      "epoch": 0.20818838640620818,
+      "grad_norm": 0.6914120316505432,
+      "learning_rate": 0.0004867484799298642,
+      "loss": 1.2154,
+      "step": 1167
+    },
+    {
+      "epoch": 0.20836678262420838,
+      "grad_norm": 0.6086413860321045,
+      "learning_rate": 0.00048672596302153814,
+      "loss": 1.3274,
+      "step": 1168
+    },
+    {
+      "epoch": 0.20854517884220855,
+      "grad_norm": 0.6582692861557007,
+      "learning_rate": 0.0004867034275208569,
+      "loss": 1.0341,
+      "step": 1169
+    },
+    {
+      "epoch": 0.20872357506020872,
+      "grad_norm": 0.9762216210365295,
+      "learning_rate": 0.0004866808734295903,
+      "loss": 1.4215,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2089019712782089,
+      "grad_norm": 45.90314483642578,
+      "learning_rate": 0.00048665830074950966,
+      "loss": 1.5744,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2090803674962091,
+      "grad_norm": 2.054628372192383,
+      "learning_rate": 0.00048663570948238806,
+      "loss": 1.3446,
+      "step": 1172
+    },
+    {
+      "epoch": 0.20925876371420926,
+      "grad_norm": 0.6909850239753723,
+      "learning_rate": 0.00048661309962999956,
+      "loss": 1.378,
+      "step": 1173
+    },
+    {
+      "epoch": 0.20943715993220943,
+      "grad_norm": 0.8046495318412781,
+      "learning_rate": 0.00048659047119412014,
+      "loss": 1.2083,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2096155561502096,
+      "grad_norm": 0.6526099443435669,
+      "learning_rate": 0.00048656782417652686,
+      "loss": 1.3512,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2097939523682098,
+      "grad_norm": 0.6259192824363708,
+      "learning_rate": 0.0004865451585789985,
+      "loss": 1.0709,
+      "step": 1176
+    },
+    {
+      "epoch": 0.20997234858620997,
+      "grad_norm": 0.5453342795372009,
+      "learning_rate": 0.0004865224744033151,
+      "loss": 1.0345,
+      "step": 1177
+    },
+    {
+      "epoch": 0.21015074480421014,
+      "grad_norm": 0.6736587285995483,
+      "learning_rate": 0.0004864997716512584,
+      "loss": 1.1594,
+      "step": 1178
+    },
+    {
+      "epoch": 0.21032914102221034,
+      "grad_norm": 0.6084941625595093,
+      "learning_rate": 0.0004864770503246114,
+      "loss": 1.1488,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2105075372402105,
+      "grad_norm": 0.6036438941955566,
+      "learning_rate": 0.00048645431042515866,
+      "loss": 1.2402,
+      "step": 1180
+    },
+    {
+      "epoch": 0.21068593345821068,
+      "grad_norm": 0.6231617331504822,
+      "learning_rate": 0.0004864315519546861,
+      "loss": 0.8676,
+      "step": 1181
+    },
+    {
+      "epoch": 0.21086432967621085,
+      "grad_norm": 0.6669137477874756,
+      "learning_rate": 0.00048640877491498127,
+      "loss": 1.3004,
+      "step": 1182
+    },
+    {
+      "epoch": 0.21104272589421105,
+      "grad_norm": 0.6060863137245178,
+      "learning_rate": 0.0004863859793078329,
+      "loss": 1.2152,
+      "step": 1183
+    },
+    {
+      "epoch": 0.21122112211221122,
+      "grad_norm": 0.5332959294319153,
+      "learning_rate": 0.0004863631651350315,
+      "loss": 1.0145,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2113995183302114,
+      "grad_norm": 0.5904478430747986,
+      "learning_rate": 0.0004863403323983688,
+      "loss": 1.2652,
+      "step": 1185
+    },
+    {
+      "epoch": 0.21157791454821157,
+      "grad_norm": 0.6089318990707397,
+      "learning_rate": 0.0004863174810996381,
+      "loss": 1.2047,
+      "step": 1186
+    },
+    {
+      "epoch": 0.21175631076621176,
+      "grad_norm": 0.5423917770385742,
+      "learning_rate": 0.00048629461124063413,
+      "loss": 1.0264,
+      "step": 1187
+    },
+    {
+      "epoch": 0.21193470698421193,
+      "grad_norm": 0.5629234910011292,
+      "learning_rate": 0.00048627172282315304,
+      "loss": 1.1713,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2121131032022121,
+      "grad_norm": 0.7865850925445557,
+      "learning_rate": 0.00048624881584899253,
+      "loss": 1.3266,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2122914994202123,
+      "grad_norm": 0.5741007328033447,
+      "learning_rate": 0.00048622589031995173,
+      "loss": 1.0702,
+      "step": 1190
+    },
+    {
+      "epoch": 0.21246989563821247,
+      "grad_norm": 0.7023731470108032,
+      "learning_rate": 0.0004862029462378311,
+      "loss": 1.2425,
+      "step": 1191
+    },
+    {
+      "epoch": 0.21264829185621265,
+      "grad_norm": 0.6102539300918579,
+      "learning_rate": 0.0004861799836044328,
+      "loss": 1.3116,
+      "step": 1192
+    },
+    {
+      "epoch": 0.21282668807421282,
+      "grad_norm": 0.6883063316345215,
+      "learning_rate": 0.0004861570024215602,
+      "loss": 1.4091,
+      "step": 1193
+    },
+    {
+      "epoch": 0.21300508429221301,
+      "grad_norm": 0.5917825698852539,
+      "learning_rate": 0.00048613400269101824,
+      "loss": 1.3487,
+      "step": 1194
+    },
+    {
+      "epoch": 0.21318348051021319,
+      "grad_norm": 0.5880262851715088,
+      "learning_rate": 0.00048611098441461335,
+      "loss": 1.2805,
+      "step": 1195
+    },
+    {
+      "epoch": 0.21336187672821336,
+      "grad_norm": 0.633368730545044,
+      "learning_rate": 0.00048608794759415333,
+      "loss": 1.0161,
+      "step": 1196
+    },
+    {
+      "epoch": 0.21354027294621353,
+      "grad_norm": 0.659824550151825,
+      "learning_rate": 0.00048606489223144744,
+      "loss": 1.2684,
+      "step": 1197
+    },
+    {
+      "epoch": 0.21371866916421373,
+      "grad_norm": 5.748763561248779,
+      "learning_rate": 0.0004860418183283066,
+      "loss": 1.2505,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2138970653822139,
+      "grad_norm": 0.6809664368629456,
+      "learning_rate": 0.00048601872588654283,
+      "loss": 1.2108,
+      "step": 1199
+    },
+    {
+      "epoch": 0.21407546160021407,
+      "grad_norm": 0.6129047870635986,
+      "learning_rate": 0.00048599561490796995,
+      "loss": 1.0522,
+      "step": 1200
+    },
+    {
+      "epoch": 0.21425385781821427,
+      "grad_norm": 0.6308250427246094,
+      "learning_rate": 0.0004859724853944031,
+      "loss": 1.2155,
+      "step": 1201
+    },
+    {
+      "epoch": 0.21443225403621444,
+      "grad_norm": 0.5974701642990112,
+      "learning_rate": 0.00048594933734765866,
+      "loss": 1.2953,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2146106502542146,
+      "grad_norm": 0.6131543517112732,
+      "learning_rate": 0.00048592617076955493,
+      "loss": 1.3646,
+      "step": 1203
+    },
+    {
+      "epoch": 0.21478904647221478,
+      "grad_norm": 0.6104893684387207,
+      "learning_rate": 0.00048590298566191116,
+      "loss": 1.629,
+      "step": 1204
+    },
+    {
+      "epoch": 0.21496744269021498,
+      "grad_norm": 0.8043147921562195,
+      "learning_rate": 0.0004858797820265485,
+      "loss": 1.2002,
+      "step": 1205
+    },
+    {
+      "epoch": 0.21514583890821515,
+      "grad_norm": 0.61087965965271,
+      "learning_rate": 0.0004858565598652892,
+      "loss": 1.3669,
+      "step": 1206
+    },
+    {
+      "epoch": 0.21532423512621532,
+      "grad_norm": 0.5664741396903992,
+      "learning_rate": 0.0004858333191799572,
+      "loss": 1.5041,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2155026313442155,
+      "grad_norm": 0.6045664548873901,
+      "learning_rate": 0.0004858100599723778,
+      "loss": 1.3528,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2156810275622157,
+      "grad_norm": 0.8556034564971924,
+      "learning_rate": 0.00048578678224437777,
+      "loss": 1.3253,
+      "step": 1209
+    },
+    {
+      "epoch": 0.21585942378021586,
+      "grad_norm": 0.7069472670555115,
+      "learning_rate": 0.0004857634859977854,
+      "loss": 1.3242,
+      "step": 1210
+    },
+    {
+      "epoch": 0.21603781999821603,
+      "grad_norm": 0.5635353922843933,
+      "learning_rate": 0.00048574017123443025,
+      "loss": 1.1976,
+      "step": 1211
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.6107478141784668,
+      "learning_rate": 0.00048571683795614346,
+      "loss": 1.2678,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2163946124342164,
+      "grad_norm": 0.6950213313102722,
+      "learning_rate": 0.0004856934861647577,
+      "loss": 1.1423,
+      "step": 1213
+    },
+    {
+      "epoch": 0.21657300865221657,
+      "grad_norm": 0.7915967106819153,
+      "learning_rate": 0.00048567011586210697,
+      "loss": 1.5536,
+      "step": 1214
+    },
+    {
+      "epoch": 0.21675140487021674,
+      "grad_norm": 0.5662855505943298,
+      "learning_rate": 0.00048564672705002663,
+      "loss": 1.081,
+      "step": 1215
+    },
+    {
+      "epoch": 0.21692980108821694,
+      "grad_norm": 0.7298009991645813,
+      "learning_rate": 0.0004856233197303539,
+      "loss": 1.328,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2171081973062171,
+      "grad_norm": 0.5519408583641052,
+      "learning_rate": 0.000485599893904927,
+      "loss": 1.2547,
+      "step": 1217
+    },
+    {
+      "epoch": 0.21728659352421728,
+      "grad_norm": 0.6197732090950012,
+      "learning_rate": 0.0004855764495755858,
+      "loss": 1.2905,
+      "step": 1218
+    },
+    {
+      "epoch": 0.21746498974221745,
+      "grad_norm": 0.6033538579940796,
+      "learning_rate": 0.00048555298674417166,
+      "loss": 1.1542,
+      "step": 1219
+    },
+    {
+      "epoch": 0.21764338596021765,
+      "grad_norm": 0.7289498448371887,
+      "learning_rate": 0.00048552950541252727,
+      "loss": 1.4689,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21782178217821782,
+      "grad_norm": 0.5700170397758484,
+      "learning_rate": 0.0004855060055824969,
+      "loss": 0.9664,
+      "step": 1221
+    },
+    {
+      "epoch": 0.218000178396218,
+      "grad_norm": 0.6593589782714844,
+      "learning_rate": 0.00048548248725592617,
+      "loss": 1.2289,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2181785746142182,
+      "grad_norm": 0.7251504063606262,
+      "learning_rate": 0.00048545895043466226,
+      "loss": 1.4661,
+      "step": 1223
+    },
+    {
+      "epoch": 0.21835697083221836,
+      "grad_norm": 0.6131605505943298,
+      "learning_rate": 0.00048543539512055367,
+      "loss": 1.182,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21853536705021853,
+      "grad_norm": 0.6947142481803894,
+      "learning_rate": 0.00048541182131545054,
+      "loss": 1.5384,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2187137632682187,
+      "grad_norm": 0.6361053586006165,
+      "learning_rate": 0.00048538822902120424,
+      "loss": 1.4648,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2188921594862189,
+      "grad_norm": 0.5528272390365601,
+      "learning_rate": 0.0004853646182396677,
+      "loss": 1.1138,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21907055570421907,
+      "grad_norm": 0.6252143979072571,
+      "learning_rate": 0.0004853409889726953,
+      "loss": 1.402,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21924895192221924,
+      "grad_norm": 0.6612663865089417,
+      "learning_rate": 0.000485317341222143,
+      "loss": 1.3349,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2194273481402194,
+      "grad_norm": 0.5366940498352051,
+      "learning_rate": 0.00048529367498986785,
+      "loss": 1.0664,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2196057443582196,
+      "grad_norm": 1.6709811687469482,
+      "learning_rate": 0.00048526999027772883,
+      "loss": 1.0385,
+      "step": 1231
+    },
+    {
+      "epoch": 0.21978414057621978,
+      "grad_norm": 0.737723708152771,
+      "learning_rate": 0.000485246287087586,
+      "loss": 1.3627,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21996253679421995,
+      "grad_norm": 0.6255607008934021,
+      "learning_rate": 0.0004852225654213009,
+      "loss": 1.1994,
+      "step": 1233
+    },
+    {
+      "epoch": 0.22014093301222015,
+      "grad_norm": 0.6627714037895203,
+      "learning_rate": 0.00048519882528073687,
+      "loss": 1.3014,
+      "step": 1234
+    },
+    {
+      "epoch": 0.22031932923022032,
+      "grad_norm": 0.5637165307998657,
+      "learning_rate": 0.00048517506666775835,
+      "loss": 1.1434,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2204977254482205,
+      "grad_norm": 0.8759331703186035,
+      "learning_rate": 0.00048515128958423123,
+      "loss": 1.3106,
+      "step": 1236
+    },
+    {
+      "epoch": 0.22067612166622066,
+      "grad_norm": 0.5403386950492859,
+      "learning_rate": 0.000485127494032023,
+      "loss": 1.0078,
+      "step": 1237
+    },
+    {
+      "epoch": 0.22085451788422086,
+      "grad_norm": 0.628709614276886,
+      "learning_rate": 0.0004851036800130026,
+      "loss": 1.1152,
+      "step": 1238
+    },
+    {
+      "epoch": 0.22103291410222103,
+      "grad_norm": 0.7338658571243286,
+      "learning_rate": 0.00048507984752904035,
+      "loss": 1.4054,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2212113103202212,
+      "grad_norm": 0.6589036583900452,
+      "learning_rate": 0.000485055996582008,
+      "loss": 1.252,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2213897065382214,
+      "grad_norm": 0.672472357749939,
+      "learning_rate": 0.0004850321271737789,
+      "loss": 1.4335,
+      "step": 1241
+    },
+    {
+      "epoch": 0.22156810275622157,
+      "grad_norm": 0.6786160469055176,
+      "learning_rate": 0.0004850082393062276,
+      "loss": 1.3884,
+      "step": 1242
+    },
+    {
+      "epoch": 0.22174649897422175,
+      "grad_norm": 1.999408483505249,
+      "learning_rate": 0.00048498433298123036,
+      "loss": 1.2055,
+      "step": 1243
+    },
+    {
+      "epoch": 0.22192489519222192,
+      "grad_norm": 0.7847235202789307,
+      "learning_rate": 0.00048496040820066467,
+      "loss": 1.1009,
+      "step": 1244
+    },
+    {
+      "epoch": 0.22210329141022211,
+      "grad_norm": 0.6475226879119873,
+      "learning_rate": 0.0004849364649664097,
+      "loss": 1.3426,
+      "step": 1245
+    },
+    {
+      "epoch": 0.22228168762822229,
+      "grad_norm": 0.56676185131073,
+      "learning_rate": 0.0004849125032803459,
+      "loss": 1.0316,
+      "step": 1246
+    },
+    {
+      "epoch": 0.22246008384622246,
+      "grad_norm": 0.6990684270858765,
+      "learning_rate": 0.00048488852314435503,
+      "loss": 1.2661,
+      "step": 1247
+    },
+    {
+      "epoch": 0.22263848006422263,
+      "grad_norm": 0.6938690543174744,
+      "learning_rate": 0.0004848645245603208,
+      "loss": 1.3882,
+      "step": 1248
+    },
+    {
+      "epoch": 0.22281687628222283,
+      "grad_norm": 2.238471508026123,
+      "learning_rate": 0.00048484050753012784,
+      "loss": 1.3287,
+      "step": 1249
+    },
+    {
+      "epoch": 0.222995272500223,
+      "grad_norm": 0.8358233571052551,
+      "learning_rate": 0.0004848164720556624,
+      "loss": 1.1643,
+      "step": 1250
+    },
+    {
+      "epoch": 0.22317366871822317,
+      "grad_norm": 2.564124822616577,
+      "learning_rate": 0.00048479241813881237,
+      "loss": 1.2204,
+      "step": 1251
+    },
+    {
+      "epoch": 0.22335206493622337,
+      "grad_norm": 1.5827924013137817,
+      "learning_rate": 0.00048476834578146683,
+      "loss": 1.4587,
+      "step": 1252
+    },
+    {
+      "epoch": 0.22353046115422354,
+      "grad_norm": 1.2166929244995117,
+      "learning_rate": 0.00048474425498551643,
+      "loss": 1.6144,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2237088573722237,
+      "grad_norm": 8.74429702758789,
+      "learning_rate": 0.00048472014575285326,
+      "loss": 2.0011,
+      "step": 1254
+    },
+    {
+      "epoch": 0.22388725359022388,
+      "grad_norm": 11.300551414489746,
+      "learning_rate": 0.0004846960180853709,
+      "loss": 2.7429,
+      "step": 1255
+    },
+    {
+      "epoch": 0.22406564980822408,
+      "grad_norm": 1.6926195621490479,
+      "learning_rate": 0.00048467187198496426,
+      "loss": 1.1161,
+      "step": 1256
+    },
+    {
+      "epoch": 0.22424404602622425,
+      "grad_norm": 0.9503287672996521,
+      "learning_rate": 0.00048464770745352984,
+      "loss": 1.4126,
+      "step": 1257
+    },
+    {
+      "epoch": 0.22442244224422442,
+      "grad_norm": 1.7149012088775635,
+      "learning_rate": 0.00048462352449296533,
+      "loss": 1.318,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2246008384622246,
+      "grad_norm": 9.479455947875977,
+      "learning_rate": 0.00048459932310517017,
+      "loss": 1.2574,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2247792346802248,
+      "grad_norm": 12.350288391113281,
+      "learning_rate": 0.0004845751032920452,
+      "loss": 1.4817,
+      "step": 1260
+    },
+    {
+      "epoch": 0.22495763089822496,
+      "grad_norm": 16.48604393005371,
+      "learning_rate": 0.0004845508650554926,
+      "loss": 1.3518,
+      "step": 1261
+    },
+    {
+      "epoch": 0.22513602711622513,
+      "grad_norm": 6.238397598266602,
+      "learning_rate": 0.00048452660839741593,
+      "loss": 0.9875,
+      "step": 1262
+    },
+    {
+      "epoch": 0.22531442333422533,
+      "grad_norm": 1.4720149040222168,
+      "learning_rate": 0.0004845023333197204,
+      "loss": 1.3612,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2254928195522255,
+      "grad_norm": 2.780336856842041,
+      "learning_rate": 0.0004844780398243125,
+      "loss": 1.3331,
+      "step": 1264
+    },
+    {
+      "epoch": 0.22567121577022567,
+      "grad_norm": 1.0210349559783936,
+      "learning_rate": 0.0004844537279131002,
+      "loss": 1.5075,
+      "step": 1265
+    },
+    {
+      "epoch": 0.22584961198822584,
+      "grad_norm": 1.9614717960357666,
+      "learning_rate": 0.0004844293975879931,
+      "loss": 1.1595,
+      "step": 1266
+    },
+    {
+      "epoch": 0.22602800820622604,
+      "grad_norm": 0.7749541997909546,
+      "learning_rate": 0.000484405048850902,
+      "loss": 1.3533,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2262064044242262,
+      "grad_norm": 1.7619514465332031,
+      "learning_rate": 0.00048438068170373916,
+      "loss": 1.0186,
+      "step": 1268
+    },
+    {
+      "epoch": 0.22638480064222638,
+      "grad_norm": 0.9682341814041138,
+      "learning_rate": 0.0004843562961484185,
+      "loss": 1.18,
+      "step": 1269
+    },
+    {
+      "epoch": 0.22656319686022655,
+      "grad_norm": 0.7163513898849487,
+      "learning_rate": 0.00048433189218685516,
+      "loss": 1.2096,
+      "step": 1270
+    },
+    {
+      "epoch": 0.22674159307822675,
+      "grad_norm": 0.5259701013565063,
+      "learning_rate": 0.0004843074698209658,
+      "loss": 1.1834,
+      "step": 1271
+    },
+    {
+      "epoch": 0.22691998929622692,
+      "grad_norm": 0.8285301923751831,
+      "learning_rate": 0.00048428302905266865,
+      "loss": 1.1552,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2270983855142271,
+      "grad_norm": 0.6478269100189209,
+      "learning_rate": 0.0004842585698838832,
+      "loss": 1.2377,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2272767817322273,
+      "grad_norm": 0.7141563296318054,
+      "learning_rate": 0.0004842340923165305,
+      "loss": 1.0266,
+      "step": 1274
+    },
+    {
+      "epoch": 0.22745517795022746,
+      "grad_norm": 1.3157190084457397,
+      "learning_rate": 0.000484209596352533,
+      "loss": 1.139,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22763357416822763,
+      "grad_norm": 1.4093172550201416,
+      "learning_rate": 0.0004841850819938145,
+      "loss": 1.0584,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2278119703862278,
+      "grad_norm": 0.805065929889679,
+      "learning_rate": 0.0004841605492423006,
+      "loss": 1.3242,
+      "step": 1277
+    },
+    {
+      "epoch": 0.227990366604228,
+      "grad_norm": 0.5310962796211243,
+      "learning_rate": 0.00048413599809991783,
+      "loss": 1.0605,
+      "step": 1278
+    },
+    {
+      "epoch": 0.22816876282222817,
+      "grad_norm": 0.5967015027999878,
+      "learning_rate": 0.00048411142856859447,
+      "loss": 1.1277,
+      "step": 1279
+    },
+    {
+      "epoch": 0.22834715904022834,
+      "grad_norm": 0.6595317721366882,
+      "learning_rate": 0.00048408684065026034,
+      "loss": 1.3209,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2285255552582285,
+      "grad_norm": 0.583308219909668,
+      "learning_rate": 0.0004840622343468465,
+      "loss": 1.1515,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2287039514762287,
+      "grad_norm": 0.9470168948173523,
+      "learning_rate": 0.0004840376096602854,
+      "loss": 1.1452,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22888234769422888,
+      "grad_norm": 0.6106230616569519,
+      "learning_rate": 0.00048401296659251125,
+      "loss": 1.5261,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22906074391222905,
+      "grad_norm": 0.656694233417511,
+      "learning_rate": 0.00048398830514545933,
+      "loss": 1.4598,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22923914013022925,
+      "grad_norm": 0.7199791669845581,
+      "learning_rate": 0.0004839636253210667,
+      "loss": 1.0677,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22941753634822942,
+      "grad_norm": 0.748662531375885,
+      "learning_rate": 0.0004839389271212715,
+      "loss": 1.1233,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2295959325662296,
+      "grad_norm": 0.8158852458000183,
+      "learning_rate": 0.00048391421054801376,
+      "loss": 1.2534,
+      "step": 1287
+    },
+    {
+      "epoch": 0.22977432878422976,
+      "grad_norm": 1.9527506828308105,
+      "learning_rate": 0.0004838894756032345,
+      "loss": 1.2561,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22995272500222996,
+      "grad_norm": 1.1054450273513794,
+      "learning_rate": 0.0004838647222888766,
+      "loss": 1.0608,
+      "step": 1289
+    },
+    {
+      "epoch": 0.23013112122023013,
+      "grad_norm": 0.5444337725639343,
+      "learning_rate": 0.0004838399506068839,
+      "loss": 1.193,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2303095174382303,
+      "grad_norm": 0.6436626315116882,
+      "learning_rate": 0.0004838151605592022,
+      "loss": 1.4263,
+      "step": 1291
+    },
+    {
+      "epoch": 0.23048791365623048,
+      "grad_norm": 0.7222276926040649,
+      "learning_rate": 0.00048379035214777833,
+      "loss": 1.482,
+      "step": 1292
+    },
+    {
+      "epoch": 0.23066630987423067,
+      "grad_norm": 0.7238490581512451,
+      "learning_rate": 0.00048376552537456085,
+      "loss": 1.1443,
+      "step": 1293
+    },
+    {
+      "epoch": 0.23084470609223084,
+      "grad_norm": 0.9711647033691406,
+      "learning_rate": 0.00048374068024149966,
+      "loss": 1.2471,
+      "step": 1294
+    },
+    {
+      "epoch": 0.23102310231023102,
+      "grad_norm": 0.54137122631073,
+      "learning_rate": 0.000483715816750546,
+      "loss": 1.1579,
+      "step": 1295
+    },
+    {
+      "epoch": 0.23120149852823121,
+      "grad_norm": 0.8173871636390686,
+      "learning_rate": 0.0004836909349036527,
+      "loss": 1.4205,
+      "step": 1296
+    },
+    {
+      "epoch": 0.23137989474623138,
+      "grad_norm": 8.518444061279297,
+      "learning_rate": 0.0004836660347027738,
+      "loss": 1.0929,
+      "step": 1297
+    },
+    {
+      "epoch": 0.23155829096423156,
+      "grad_norm": 0.603255569934845,
+      "learning_rate": 0.0004836411161498652,
+      "loss": 1.2047,
+      "step": 1298
+    },
+    {
+      "epoch": 0.23173668718223173,
+      "grad_norm": 0.6280937790870667,
+      "learning_rate": 0.0004836161792468839,
+      "loss": 1.2314,
+      "step": 1299
+    },
+    {
+      "epoch": 0.23191508340023193,
+      "grad_norm": 0.5769376158714294,
+      "learning_rate": 0.00048359122399578835,
+      "loss": 1.077,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2320934796182321,
+      "grad_norm": 0.6672999858856201,
+      "learning_rate": 0.00048356625039853865,
+      "loss": 1.1322,
+      "step": 1301
+    },
+    {
+      "epoch": 0.23227187583623227,
+      "grad_norm": 1.1087464094161987,
+      "learning_rate": 0.00048354125845709604,
+      "loss": 1.5064,
+      "step": 1302
+    },
+    {
+      "epoch": 0.23245027205423244,
+      "grad_norm": 0.6089938879013062,
+      "learning_rate": 0.0004835162481734237,
+      "loss": 0.9383,
+      "step": 1303
+    },
+    {
+      "epoch": 0.23262866827223264,
+      "grad_norm": 0.6506130695343018,
+      "learning_rate": 0.0004834912195494855,
+      "loss": 1.2445,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2328070644902328,
+      "grad_norm": 0.6899453401565552,
+      "learning_rate": 0.0004834661725872475,
+      "loss": 1.2801,
+      "step": 1305
+    },
+    {
+      "epoch": 0.23298546070823298,
+      "grad_norm": 0.9193609952926636,
+      "learning_rate": 0.00048344110728867673,
+      "loss": 1.3748,
+      "step": 1306
+    },
+    {
+      "epoch": 0.23316385692623318,
+      "grad_norm": 0.905149519443512,
+      "learning_rate": 0.0004834160236557419,
+      "loss": 1.2297,
+      "step": 1307
+    },
+    {
+      "epoch": 0.23334225314423335,
+      "grad_norm": 0.5802024602890015,
+      "learning_rate": 0.00048339092169041306,
+      "loss": 1.0281,
+      "step": 1308
+    },
+    {
+      "epoch": 0.23352064936223352,
+      "grad_norm": 0.644010603427887,
+      "learning_rate": 0.0004833658013946616,
+      "loss": 1.4227,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2336990455802337,
+      "grad_norm": 0.5992226600646973,
+      "learning_rate": 0.0004833406627704605,
+      "loss": 1.229,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2338774417982339,
+      "grad_norm": 0.5435013771057129,
+      "learning_rate": 0.00048331550581978423,
+      "loss": 1.2623,
+      "step": 1311
+    },
+    {
+      "epoch": 0.23405583801623406,
+      "grad_norm": 0.5842602252960205,
+      "learning_rate": 0.0004832903305446085,
+      "loss": 1.3458,
+      "step": 1312
+    },
+    {
+      "epoch": 0.23423423423423423,
+      "grad_norm": 0.6166499257087708,
+      "learning_rate": 0.00048326513694691055,
+      "loss": 1.1492,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2344126304522344,
+      "grad_norm": 0.586531400680542,
+      "learning_rate": 0.0004832399250286692,
+      "loss": 1.179,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2345910266702346,
+      "grad_norm": 1.1618958711624146,
+      "learning_rate": 0.0004832146947918645,
+      "loss": 1.0842,
+      "step": 1315
+    },
+    {
+      "epoch": 0.23476942288823477,
+      "grad_norm": 0.5545247197151184,
+      "learning_rate": 0.000483189446238478,
+      "loss": 1.0436,
+      "step": 1316
+    },
+    {
+      "epoch": 0.23494781910623494,
+      "grad_norm": 0.5850834846496582,
+      "learning_rate": 0.00048316417937049275,
+      "loss": 1.1775,
+      "step": 1317
+    },
+    {
+      "epoch": 0.23512621532423514,
+      "grad_norm": 0.654733419418335,
+      "learning_rate": 0.0004831388941898932,
+      "loss": 1.2601,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2353046115422353,
+      "grad_norm": 0.5766509175300598,
+      "learning_rate": 0.00048311359069866524,
+      "loss": 1.2645,
+      "step": 1319
+    },
+    {
+      "epoch": 0.23548300776023548,
+      "grad_norm": 0.577642560005188,
+      "learning_rate": 0.00048308826889879614,
+      "loss": 1.1026,
+      "step": 1320
+    },
+    {
+      "epoch": 0.23566140397823565,
+      "grad_norm": 0.6444463729858398,
+      "learning_rate": 0.0004830629287922748,
+      "loss": 1.2969,
+      "step": 1321
+    },
+    {
+      "epoch": 0.23583980019623585,
+      "grad_norm": 0.7246226072311401,
+      "learning_rate": 0.00048303757038109117,
+      "loss": 1.1957,
+      "step": 1322
+    },
+    {
+      "epoch": 0.23601819641423602,
+      "grad_norm": 0.7728081345558167,
+      "learning_rate": 0.00048301219366723714,
+      "loss": 1.1193,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2361965926322362,
+      "grad_norm": 0.6015149354934692,
+      "learning_rate": 0.00048298679865270574,
+      "loss": 1.3462,
+      "step": 1324
+    },
+    {
+      "epoch": 0.23637498885023636,
+      "grad_norm": 0.5546750426292419,
+      "learning_rate": 0.0004829613853394914,
+      "loss": 1.2633,
+      "step": 1325
+    },
+    {
+      "epoch": 0.23655338506823656,
+      "grad_norm": 0.7235816717147827,
+      "learning_rate": 0.0004829359537295901,
+      "loss": 1.5118,
+      "step": 1326
+    },
+    {
+      "epoch": 0.23673178128623673,
+      "grad_norm": 0.6748101711273193,
+      "learning_rate": 0.0004829105038249992,
+      "loss": 1.1664,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2369101775042369,
+      "grad_norm": 0.7603464722633362,
+      "learning_rate": 0.0004828850356277176,
+      "loss": 1.2698,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2370885737222371,
+      "grad_norm": 0.6377245187759399,
+      "learning_rate": 0.0004828595491397455,
+      "loss": 0.9762,
+      "step": 1329
+    },
+    {
+      "epoch": 0.23726696994023727,
+      "grad_norm": 0.6112083792686462,
+      "learning_rate": 0.0004828340443630846,
+      "loss": 0.9882,
+      "step": 1330
+    },
+    {
+      "epoch": 0.23744536615823744,
+      "grad_norm": 0.5406150817871094,
+      "learning_rate": 0.00048280852129973807,
+      "loss": 0.9828,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2376237623762376,
+      "grad_norm": 0.592891275882721,
+      "learning_rate": 0.0004827829799517105,
+      "loss": 1.0252,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2378021585942378,
+      "grad_norm": 0.564150869846344,
+      "learning_rate": 0.0004827574203210078,
+      "loss": 1.2145,
+      "step": 1333
+    },
+    {
+      "epoch": 0.23798055481223798,
+      "grad_norm": 0.8308655619621277,
+      "learning_rate": 0.0004827318424096375,
+      "loss": 0.9988,
+      "step": 1334
+    },
+    {
+      "epoch": 0.23815895103023815,
+      "grad_norm": 0.6468952894210815,
+      "learning_rate": 0.00048270624621960846,
+      "loss": 1.0637,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23833734724823832,
+      "grad_norm": 0.5933250784873962,
+      "learning_rate": 0.000482680631752931,
+      "loss": 1.2847,
+      "step": 1336
+    },
+    {
+      "epoch": 0.23851574346623852,
+      "grad_norm": 0.8203790187835693,
+      "learning_rate": 0.0004826549990116168,
+      "loss": 1.3192,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2386941396842387,
+      "grad_norm": 0.8273621201515198,
+      "learning_rate": 0.0004826293479976791,
+      "loss": 1.2088,
+      "step": 1338
+    },
+    {
+      "epoch": 0.23887253590223886,
+      "grad_norm": 0.6572198271751404,
+      "learning_rate": 0.0004826036787131326,
+      "loss": 1.3699,
+      "step": 1339
+    },
+    {
+      "epoch": 0.23905093212023906,
+      "grad_norm": 0.6077407002449036,
+      "learning_rate": 0.0004825779911599932,
+      "loss": 1.2602,
+      "step": 1340
+    },
+    {
+      "epoch": 0.23922932833823923,
+      "grad_norm": 0.58214271068573,
+      "learning_rate": 0.00048255228534027845,
+      "loss": 0.9227,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2394077245562394,
+      "grad_norm": 0.5967569947242737,
+      "learning_rate": 0.0004825265612560073,
+      "loss": 1.1136,
+      "step": 1342
+    },
+    {
+      "epoch": 0.23958612077423957,
+      "grad_norm": 0.5940139889717102,
+      "learning_rate": 0.0004825008189092001,
+      "loss": 1.1137,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23976451699223977,
+      "grad_norm": 0.5328962802886963,
+      "learning_rate": 0.00048247505830187863,
+      "loss": 1.0354,
+      "step": 1344
+    },
+    {
+      "epoch": 0.23994291321023994,
+      "grad_norm": 0.548477828502655,
+      "learning_rate": 0.00048244927943606617,
+      "loss": 0.9724,
+      "step": 1345
+    },
+    {
+      "epoch": 0.24012130942824012,
+      "grad_norm": 0.6590238809585571,
+      "learning_rate": 0.0004824234823137873,
+      "loss": 1.2702,
+      "step": 1346
+    },
+    {
+      "epoch": 0.24029970564624029,
+      "grad_norm": 0.5709405541419983,
+      "learning_rate": 0.0004823976669370681,
+      "loss": 1.3085,
+      "step": 1347
+    },
+    {
+      "epoch": 0.24047810186424048,
+      "grad_norm": 0.6574723720550537,
+      "learning_rate": 0.0004823718333079362,
+      "loss": 1.5324,
+      "step": 1348
+    },
+    {
+      "epoch": 0.24065649808224066,
+      "grad_norm": 0.5990821123123169,
+      "learning_rate": 0.00048234598142842043,
+      "loss": 1.1941,
+      "step": 1349
+    },
+    {
+      "epoch": 0.24083489430024083,
+      "grad_norm": 0.6324416995048523,
+      "learning_rate": 0.0004823201113005514,
+      "loss": 1.3861,
+      "step": 1350
+    },
+    {
+      "epoch": 0.24101329051824102,
+      "grad_norm": 0.5532766580581665,
+      "learning_rate": 0.0004822942229263607,
+      "loss": 1.1808,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2411916867362412,
+      "grad_norm": 0.5371944904327393,
+      "learning_rate": 0.00048226831630788174,
+      "loss": 0.9721,
+      "step": 1352
+    },
+    {
+      "epoch": 0.24137008295424137,
+      "grad_norm": 0.6007786393165588,
+      "learning_rate": 0.0004822423914471492,
+      "loss": 1.2506,
+      "step": 1353
+    },
+    {
+      "epoch": 0.24154847917224154,
+      "grad_norm": 0.5308194756507874,
+      "learning_rate": 0.0004822164483461991,
+      "loss": 1.1275,
+      "step": 1354
+    },
+    {
+      "epoch": 0.24172687539024174,
+      "grad_norm": 0.5771823525428772,
+      "learning_rate": 0.00048219048700706913,
+      "loss": 1.138,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2419052716082419,
+      "grad_norm": 0.6079913377761841,
+      "learning_rate": 0.00048216450743179817,
+      "loss": 1.4897,
+      "step": 1356
+    },
+    {
+      "epoch": 0.24208366782624208,
+      "grad_norm": 0.5161705613136292,
+      "learning_rate": 0.0004821385096224268,
+      "loss": 1.0735,
+      "step": 1357
+    },
+    {
+      "epoch": 0.24226206404424228,
+      "grad_norm": 0.6866333484649658,
+      "learning_rate": 0.00048211249358099675,
+      "loss": 1.316,
+      "step": 1358
+    },
+    {
+      "epoch": 0.24244046026224245,
+      "grad_norm": 0.5263490080833435,
+      "learning_rate": 0.0004820864593095513,
+      "loss": 1.1486,
+      "step": 1359
+    },
+    {
+      "epoch": 0.24261885648024262,
+      "grad_norm": 0.5457320213317871,
+      "learning_rate": 0.0004820604068101352,
+      "loss": 1.076,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2427972526982428,
+      "grad_norm": 0.5485474467277527,
+      "learning_rate": 0.00048203433608479465,
+      "loss": 1.4718,
+      "step": 1361
+    },
+    {
+      "epoch": 0.242975648916243,
+      "grad_norm": 0.5592158436775208,
+      "learning_rate": 0.0004820082471355772,
+      "loss": 1.123,
+      "step": 1362
+    },
+    {
+      "epoch": 0.24315404513424316,
+      "grad_norm": 0.6344873905181885,
+      "learning_rate": 0.0004819821399645319,
+      "loss": 1.2592,
+      "step": 1363
+    },
+    {
+      "epoch": 0.24333244135224333,
+      "grad_norm": 0.5310791730880737,
+      "learning_rate": 0.00048195601457370907,
+      "loss": 1.1477,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2435108375702435,
+      "grad_norm": 0.5795305371284485,
+      "learning_rate": 0.0004819298709651607,
+      "loss": 1.3579,
+      "step": 1365
+    },
+    {
+      "epoch": 0.2436892337882437,
+      "grad_norm": 0.6433924436569214,
+      "learning_rate": 0.0004819037091409401,
+      "loss": 1.3058,
+      "step": 1366
+    },
+    {
+      "epoch": 0.24386763000624387,
+      "grad_norm": 0.6262935996055603,
+      "learning_rate": 0.00048187752910310196,
+      "loss": 1.4104,
+      "step": 1367
+    },
+    {
+      "epoch": 0.24404602622424404,
+      "grad_norm": 0.6437758207321167,
+      "learning_rate": 0.0004818513308537025,
+      "loss": 1.4279,
+      "step": 1368
+    },
+    {
+      "epoch": 0.24422442244224424,
+      "grad_norm": 0.5018840432167053,
+      "learning_rate": 0.00048182511439479926,
+      "loss": 1.0561,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2444028186602444,
+      "grad_norm": 0.5912278890609741,
+      "learning_rate": 0.00048179887972845124,
+      "loss": 1.2844,
+      "step": 1370
+    },
+    {
+      "epoch": 0.24458121487824458,
+      "grad_norm": 0.5428308844566345,
+      "learning_rate": 0.0004817726268567191,
+      "loss": 1.1105,
+      "step": 1371
+    },
+    {
+      "epoch": 0.24475961109624475,
+      "grad_norm": 0.5821102857589722,
+      "learning_rate": 0.0004817463557816644,
+      "loss": 1.3421,
+      "step": 1372
+    },
+    {
+      "epoch": 0.24493800731424495,
+      "grad_norm": 0.5811814665794373,
+      "learning_rate": 0.0004817200665053508,
+      "loss": 1.1854,
+      "step": 1373
+    },
+    {
+      "epoch": 0.24511640353224512,
+      "grad_norm": 0.5271221399307251,
+      "learning_rate": 0.00048169375902984283,
+      "loss": 1.1517,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2452947997502453,
+      "grad_norm": 0.592476487159729,
+      "learning_rate": 0.00048166743335720675,
+      "loss": 1.459,
+      "step": 1375
+    },
+    {
+      "epoch": 0.24547319596824546,
+      "grad_norm": 0.5430518984794617,
+      "learning_rate": 0.00048164108948951014,
+      "loss": 1.0736,
+      "step": 1376
+    },
+    {
+      "epoch": 0.24565159218624566,
+      "grad_norm": 0.601703405380249,
+      "learning_rate": 0.00048161472742882204,
+      "loss": 1.358,
+      "step": 1377
+    },
+    {
+      "epoch": 0.24582998840424583,
+      "grad_norm": 0.7517264485359192,
+      "learning_rate": 0.0004815883471772129,
+      "loss": 1.1916,
+      "step": 1378
+    },
+    {
+      "epoch": 0.246008384622246,
+      "grad_norm": 0.6537923216819763,
+      "learning_rate": 0.00048156194873675466,
+      "loss": 1.3555,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2461867808402462,
+      "grad_norm": 0.5317115187644958,
+      "learning_rate": 0.0004815355321095206,
+      "loss": 1.1173,
+      "step": 1380
+    },
+    {
+      "epoch": 0.24636517705824637,
+      "grad_norm": 0.5854525566101074,
+      "learning_rate": 0.00048150909729758554,
+      "loss": 1.2466,
+      "step": 1381
+    },
+    {
+      "epoch": 0.24654357327624654,
+      "grad_norm": 0.5315707325935364,
+      "learning_rate": 0.0004814826443030256,
+      "loss": 1.0387,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2467219694942467,
+      "grad_norm": 0.5110334157943726,
+      "learning_rate": 0.00048145617312791837,
+      "loss": 1.0514,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2469003657122469,
+      "grad_norm": 0.5571600198745728,
+      "learning_rate": 0.0004814296837743428,
+      "loss": 1.1521,
+      "step": 1384
+    },
+    {
+      "epoch": 0.24707876193024708,
+      "grad_norm": 0.5861849188804626,
+      "learning_rate": 0.0004814031762443796,
+      "loss": 1.26,
+      "step": 1385
+    },
+    {
+      "epoch": 0.24725715814824725,
+      "grad_norm": 0.5479332208633423,
+      "learning_rate": 0.00048137665054011044,
+      "loss": 1.1034,
+      "step": 1386
+    },
+    {
+      "epoch": 0.24743555436624742,
+      "grad_norm": 0.5551263689994812,
+      "learning_rate": 0.0004813501066636187,
+      "loss": 1.0253,
+      "step": 1387
+    },
+    {
+      "epoch": 0.24761395058424762,
+      "grad_norm": 0.6580809354782104,
+      "learning_rate": 0.00048132354461698923,
+      "loss": 1.2498,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2477923468022478,
+      "grad_norm": 0.5350015759468079,
+      "learning_rate": 0.0004812969644023081,
+      "loss": 1.1355,
+      "step": 1389
+    },
+    {
+      "epoch": 0.24797074302024796,
+      "grad_norm": 0.5560418367385864,
+      "learning_rate": 0.00048127036602166285,
+      "loss": 1.0944,
+      "step": 1390
+    },
+    {
+      "epoch": 0.24814913923824816,
+      "grad_norm": 0.5966114401817322,
+      "learning_rate": 0.00048124374947714263,
+      "loss": 1.0267,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24832753545624833,
+      "grad_norm": 0.5361310243606567,
+      "learning_rate": 0.0004812171147708378,
+      "loss": 1.1974,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2485059316742485,
+      "grad_norm": 0.5971125960350037,
+      "learning_rate": 0.00048119046190484027,
+      "loss": 1.2728,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24868432789224867,
+      "grad_norm": 0.5017391443252563,
+      "learning_rate": 0.0004811637908812434,
+      "loss": 1.0279,
+      "step": 1394
+    },
+    {
+      "epoch": 0.24886272411024887,
+      "grad_norm": 0.5720884203910828,
+      "learning_rate": 0.00048113710170214185,
+      "loss": 1.3989,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24904112032824904,
+      "grad_norm": 0.5782856345176697,
+      "learning_rate": 0.00048111039436963177,
+      "loss": 1.2528,
+      "step": 1396
+    },
+    {
+      "epoch": 0.24921951654624921,
+      "grad_norm": 0.5723220109939575,
+      "learning_rate": 0.00048108366888581077,
+      "loss": 1.0068,
+      "step": 1397
+    },
+    {
+      "epoch": 0.24939791276424939,
+      "grad_norm": 0.5289225578308105,
+      "learning_rate": 0.00048105692525277793,
+      "loss": 0.9357,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24957630898224958,
+      "grad_norm": 0.5223532319068909,
+      "learning_rate": 0.00048103016347263356,
+      "loss": 1.0046,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24975470520024975,
+      "grad_norm": 0.6456426978111267,
+      "learning_rate": 0.0004810033835474796,
+      "loss": 1.5626,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24993310141824993,
+      "grad_norm": 0.5884763598442078,
+      "learning_rate": 0.00048097658547941927,
+      "loss": 1.4618,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2501114976362501,
+      "grad_norm": 0.5833611488342285,
+      "learning_rate": 0.00048094976927055735,
+      "loss": 1.396,
+      "step": 1402
+    },
+    {
+      "epoch": 0.25028989385425027,
+      "grad_norm": 0.5198600888252258,
+      "learning_rate": 0.00048092293492299986,
+      "loss": 1.1263,
+      "step": 1403
+    },
+    {
+      "epoch": 0.25046829007225047,
+      "grad_norm": 0.5795050859451294,
+      "learning_rate": 0.00048089608243885454,
+      "loss": 1.2508,
+      "step": 1404
+    },
+    {
+      "epoch": 0.25064668629025066,
+      "grad_norm": 0.7242421507835388,
+      "learning_rate": 0.0004808692118202302,
+      "loss": 1.2361,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2508250825082508,
+      "grad_norm": 0.6138631105422974,
+      "learning_rate": 0.00048084232306923737,
+      "loss": 1.3318,
+      "step": 1406
+    },
+    {
+      "epoch": 0.251003478726251,
+      "grad_norm": 0.5212423801422119,
+      "learning_rate": 0.0004808154161879877,
+      "loss": 1.2418,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2511818749442512,
+      "grad_norm": 0.5970634818077087,
+      "learning_rate": 0.0004807884911785947,
+      "loss": 1.1968,
+      "step": 1408
+    },
+    {
+      "epoch": 0.25136027116225135,
+      "grad_norm": 0.6800307035446167,
+      "learning_rate": 0.0004807615480431729,
+      "loss": 1.2915,
+      "step": 1409
+    },
+    {
+      "epoch": 0.25153866738025155,
+      "grad_norm": 0.5408998727798462,
+      "learning_rate": 0.0004807345867838384,
+      "loss": 0.9737,
+      "step": 1410
+    },
+    {
+      "epoch": 0.25171706359825174,
+      "grad_norm": 0.6598271727561951,
+      "learning_rate": 0.00048070760740270873,
+      "loss": 1.2425,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2518954598162519,
+      "grad_norm": 0.6781688332557678,
+      "learning_rate": 0.0004806806099019029,
+      "loss": 1.0545,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2520738560342521,
+      "grad_norm": 0.6171199679374695,
+      "learning_rate": 0.00048065359428354115,
+      "loss": 1.1862,
+      "step": 1413
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 0.5618746876716614,
+      "learning_rate": 0.00048062656054974546,
+      "loss": 1.3709,
+      "step": 1414
+    },
+    {
+      "epoch": 0.25243064847025243,
+      "grad_norm": 2.3955955505371094,
+      "learning_rate": 0.0004805995087026389,
+      "loss": 1.2006,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2526090446882526,
+      "grad_norm": 0.6166795492172241,
+      "learning_rate": 0.0004805724387443462,
+      "loss": 1.3305,
+      "step": 1416
+    },
+    {
+      "epoch": 0.25278744090625277,
+      "grad_norm": 0.5519673228263855,
+      "learning_rate": 0.00048054535067699333,
+      "loss": 1.1071,
+      "step": 1417
+    },
+    {
+      "epoch": 0.25296583712425297,
+      "grad_norm": 0.6297189593315125,
+      "learning_rate": 0.0004805182445027079,
+      "loss": 1.4462,
+      "step": 1418
+    },
+    {
+      "epoch": 0.25314423334225317,
+      "grad_norm": 0.5049257278442383,
+      "learning_rate": 0.0004804911202236187,
+      "loss": 0.9542,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2533226295602533,
+      "grad_norm": 0.5491045713424683,
+      "learning_rate": 0.0004804639778418561,
+      "loss": 1.1949,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2535010257782535,
+      "grad_norm": 0.6792704463005066,
+      "learning_rate": 0.00048043681735955183,
+      "loss": 1.4088,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2536794219962537,
+      "grad_norm": 0.543935239315033,
+      "learning_rate": 0.0004804096387788391,
+      "loss": 1.3297,
+      "step": 1422
+    },
+    {
+      "epoch": 0.25385781821425385,
+      "grad_norm": 0.6691182851791382,
+      "learning_rate": 0.00048038244210185253,
+      "loss": 1.1259,
+      "step": 1423
+    },
+    {
+      "epoch": 0.25403621443225405,
+      "grad_norm": 0.6011500954627991,
+      "learning_rate": 0.0004803552273307281,
+      "loss": 1.2238,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2542146106502542,
+      "grad_norm": 0.667797863483429,
+      "learning_rate": 0.0004803279944676032,
+      "loss": 1.206,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2543930068682544,
+      "grad_norm": 0.676861047744751,
+      "learning_rate": 0.0004803007435146168,
+      "loss": 1.0519,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2545714030862546,
+      "grad_norm": 0.5236627459526062,
+      "learning_rate": 0.00048027347447390914,
+      "loss": 1.0689,
+      "step": 1427
+    },
+    {
+      "epoch": 0.25474979930425473,
+      "grad_norm": 0.6431361436843872,
+      "learning_rate": 0.00048024618734762183,
+      "loss": 1.0586,
+      "step": 1428
+    },
+    {
+      "epoch": 0.25492819552225493,
+      "grad_norm": 0.6550641655921936,
+      "learning_rate": 0.00048021888213789797,
+      "loss": 1.2463,
+      "step": 1429
+    },
+    {
+      "epoch": 0.25510659174025513,
+      "grad_norm": 0.5982055068016052,
+      "learning_rate": 0.0004801915588468823,
+      "loss": 1.1063,
+      "step": 1430
+    },
+    {
+      "epoch": 0.25528498795825527,
+      "grad_norm": 0.5872926115989685,
+      "learning_rate": 0.00048016421747672054,
+      "loss": 1.2272,
+      "step": 1431
+    },
+    {
+      "epoch": 0.25546338417625547,
+      "grad_norm": 0.6466538310050964,
+      "learning_rate": 0.0004801368580295603,
+      "loss": 1.1145,
+      "step": 1432
+    },
+    {
+      "epoch": 0.25564178039425567,
+      "grad_norm": 0.5384798645973206,
+      "learning_rate": 0.0004801094805075502,
+      "loss": 1.1214,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2558201766122558,
+      "grad_norm": 0.5813401937484741,
+      "learning_rate": 0.00048008208491284054,
+      "loss": 1.2142,
+      "step": 1434
+    },
+    {
+      "epoch": 0.255998572830256,
+      "grad_norm": 0.5265066623687744,
+      "learning_rate": 0.00048005467124758296,
+      "loss": 1.0926,
+      "step": 1435
+    },
+    {
+      "epoch": 0.25617696904825615,
+      "grad_norm": 0.5532288551330566,
+      "learning_rate": 0.0004800272395139305,
+      "loss": 1.2191,
+      "step": 1436
+    },
+    {
+      "epoch": 0.25635536526625635,
+      "grad_norm": 0.5451021194458008,
+      "learning_rate": 0.00047999978971403765,
+      "loss": 1.1107,
+      "step": 1437
+    },
+    {
+      "epoch": 0.25653376148425655,
+      "grad_norm": 0.5395956039428711,
+      "learning_rate": 0.0004799723218500602,
+      "loss": 1.2165,
+      "step": 1438
+    },
+    {
+      "epoch": 0.2567121577022567,
+      "grad_norm": 0.5500022172927856,
+      "learning_rate": 0.00047994483592415574,
+      "loss": 1.3835,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2568905539202569,
+      "grad_norm": 0.5290274024009705,
+      "learning_rate": 0.0004799173319384826,
+      "loss": 0.966,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2570689501382571,
+      "grad_norm": 0.6000217795372009,
+      "learning_rate": 0.0004798898098952013,
+      "loss": 1.3913,
+      "step": 1441
+    },
+    {
+      "epoch": 0.25724734635625723,
+      "grad_norm": 0.5701634883880615,
+      "learning_rate": 0.0004798622697964732,
+      "loss": 1.0587,
+      "step": 1442
+    },
+    {
+      "epoch": 0.25742574257425743,
+      "grad_norm": 0.5983652472496033,
+      "learning_rate": 0.00047983471164446135,
+      "loss": 1.2844,
+      "step": 1443
+    },
+    {
+      "epoch": 0.25760413879225763,
+      "grad_norm": 0.5445163249969482,
+      "learning_rate": 0.0004798071354413302,
+      "loss": 1.1172,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2577825350102578,
+      "grad_norm": 0.5581110715866089,
+      "learning_rate": 0.0004797795411892455,
+      "loss": 1.1949,
+      "step": 1445
+    },
+    {
+      "epoch": 0.257960931228258,
+      "grad_norm": 0.585059642791748,
+      "learning_rate": 0.0004797519288903745,
+      "loss": 1.187,
+      "step": 1446
+    },
+    {
+      "epoch": 0.2581393274462581,
+      "grad_norm": 0.6527815461158752,
+      "learning_rate": 0.00047972429854688595,
+      "loss": 1.1534,
+      "step": 1447
+    },
+    {
+      "epoch": 0.2583177236642583,
+      "grad_norm": 0.7034025192260742,
+      "learning_rate": 0.00047969665016094976,
+      "loss": 1.0801,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2584961198822585,
+      "grad_norm": 1.5441768169403076,
+      "learning_rate": 0.00047966898373473754,
+      "loss": 1.3137,
+      "step": 1449
+    },
+    {
+      "epoch": 0.25867451610025866,
+      "grad_norm": 0.5990326404571533,
+      "learning_rate": 0.0004796412992704221,
+      "loss": 1.3426,
+      "step": 1450
+    },
+    {
+      "epoch": 0.25885291231825885,
+      "grad_norm": 0.5593619346618652,
+      "learning_rate": 0.0004796135967701779,
+      "loss": 1.2633,
+      "step": 1451
+    },
+    {
+      "epoch": 0.25903130853625905,
+      "grad_norm": 0.5932374000549316,
+      "learning_rate": 0.00047958587623618066,
+      "loss": 1.111,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2592097047542592,
+      "grad_norm": 0.9718675017356873,
+      "learning_rate": 0.0004795581376706075,
+      "loss": 0.9386,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2593881009722594,
+      "grad_norm": 0.5909976363182068,
+      "learning_rate": 0.00047953038107563696,
+      "loss": 1.0972,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2595664971902596,
+      "grad_norm": 0.550532877445221,
+      "learning_rate": 0.000479502606453449,
+      "loss": 0.9393,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25974489340825974,
+      "grad_norm": 0.5949730277061462,
+      "learning_rate": 0.00047947481380622523,
+      "loss": 1.1803,
+      "step": 1456
+    },
+    {
+      "epoch": 0.25992328962625993,
+      "grad_norm": 4.032388210296631,
+      "learning_rate": 0.0004794470031361482,
+      "loss": 1.5666,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2601016858442601,
+      "grad_norm": 0.5784013271331787,
+      "learning_rate": 0.0004794191744454024,
+      "loss": 1.3467,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2602800820622603,
+      "grad_norm": 0.6242582201957703,
+      "learning_rate": 0.00047939132773617334,
+      "loss": 1.1208,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2604584782802605,
+      "grad_norm": 0.7698078155517578,
+      "learning_rate": 0.00047936346301064806,
+      "loss": 1.0162,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2606368744982606,
+      "grad_norm": 0.8937914967536926,
+      "learning_rate": 0.0004793355802710151,
+      "loss": 1.3314,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2608152707162608,
+      "grad_norm": 0.8126354217529297,
+      "learning_rate": 0.0004793076795194644,
+      "loss": 1.1186,
+      "step": 1462
+    },
+    {
+      "epoch": 0.260993666934261,
+      "grad_norm": 0.6322470307350159,
+      "learning_rate": 0.0004792797607581872,
+      "loss": 1.1986,
+      "step": 1463
+    },
+    {
+      "epoch": 0.26117206315226116,
+      "grad_norm": 0.6327349543571472,
+      "learning_rate": 0.0004792518239893763,
+      "loss": 1.1737,
+      "step": 1464
+    },
+    {
+      "epoch": 0.26135045937026136,
+      "grad_norm": 0.5458834767341614,
+      "learning_rate": 0.00047922386921522576,
+      "loss": 1.0218,
+      "step": 1465
+    },
+    {
+      "epoch": 0.26152885558826156,
+      "grad_norm": 0.685711681842804,
+      "learning_rate": 0.0004791958964379312,
+      "loss": 1.2088,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2617072518062617,
+      "grad_norm": 0.7403172850608826,
+      "learning_rate": 0.0004791679056596895,
+      "loss": 1.4212,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2618856480242619,
+      "grad_norm": 0.6377992033958435,
+      "learning_rate": 0.0004791398968826991,
+      "loss": 1.2633,
+      "step": 1468
+    },
+    {
+      "epoch": 0.26206404424226204,
+      "grad_norm": 0.6038281321525574,
+      "learning_rate": 0.0004791118701091599,
+      "loss": 1.2117,
+      "step": 1469
+    },
+    {
+      "epoch": 0.26224244046026224,
+      "grad_norm": 2.225592613220215,
+      "learning_rate": 0.0004790838253412729,
+      "loss": 1.2892,
+      "step": 1470
+    },
+    {
+      "epoch": 0.26242083667826244,
+      "grad_norm": 0.5641833543777466,
+      "learning_rate": 0.0004790557625812409,
+      "loss": 1.1663,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2625992328962626,
+      "grad_norm": 0.5233538746833801,
+      "learning_rate": 0.00047902768183126797,
+      "loss": 1.1302,
+      "step": 1472
+    },
+    {
+      "epoch": 0.2627776291142628,
+      "grad_norm": 0.5378060340881348,
+      "learning_rate": 0.0004789995830935594,
+      "loss": 1.0252,
+      "step": 1473
+    },
+    {
+      "epoch": 0.262956025332263,
+      "grad_norm": 0.5588990449905396,
+      "learning_rate": 0.0004789714663703221,
+      "loss": 1.2761,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2631344215502631,
+      "grad_norm": 0.5507195591926575,
+      "learning_rate": 0.00047894333166376434,
+      "loss": 1.2709,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2633128177682633,
+      "grad_norm": 0.6437628865242004,
+      "learning_rate": 0.00047891517897609594,
+      "loss": 1.1084,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2634912139862635,
+      "grad_norm": 0.5252813100814819,
+      "learning_rate": 0.0004788870083095278,
+      "loss": 1.1753,
+      "step": 1477
+    },
+    {
+      "epoch": 0.26366961020426366,
+      "grad_norm": 0.721181333065033,
+      "learning_rate": 0.00047885881966627255,
+      "loss": 1.3439,
+      "step": 1478
+    },
+    {
+      "epoch": 0.26384800642226386,
+      "grad_norm": 0.6929256916046143,
+      "learning_rate": 0.000478830613048544,
+      "loss": 1.15,
+      "step": 1479
+    },
+    {
+      "epoch": 0.264026402640264,
+      "grad_norm": 0.5794677734375,
+      "learning_rate": 0.0004788023884585577,
+      "loss": 0.991,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2642047988582642,
+      "grad_norm": 0.5833317041397095,
+      "learning_rate": 0.00047877414589853024,
+      "loss": 1.23,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2643831950762644,
+      "grad_norm": 0.5699701309204102,
+      "learning_rate": 0.0004787458853706798,
+      "loss": 1.0603,
+      "step": 1482
+    },
+    {
+      "epoch": 0.26456159129426454,
+      "grad_norm": 0.6602557897567749,
+      "learning_rate": 0.00047871760687722597,
+      "loss": 1.2557,
+      "step": 1483
+    },
+    {
+      "epoch": 0.26473998751226474,
+      "grad_norm": 0.5460817813873291,
+      "learning_rate": 0.0004786893104203897,
+      "loss": 1.0461,
+      "step": 1484
+    },
+    {
+      "epoch": 0.26491838373026494,
+      "grad_norm": 0.5744298100471497,
+      "learning_rate": 0.0004786609960023934,
+      "loss": 1.3121,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2650967799482651,
+      "grad_norm": 0.6171600222587585,
+      "learning_rate": 0.00047863266362546095,
+      "loss": 1.264,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2652751761662653,
+      "grad_norm": 0.5980626344680786,
+      "learning_rate": 0.00047860431329181744,
+      "loss": 1.4093,
+      "step": 1487
+    },
+    {
+      "epoch": 0.2654535723842655,
+      "grad_norm": 0.674322247505188,
+      "learning_rate": 0.0004785759450036895,
+      "loss": 1.2626,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2656319686022656,
+      "grad_norm": 0.7270940542221069,
+      "learning_rate": 0.0004785475587633052,
+      "loss": 1.0291,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2658103648202658,
+      "grad_norm": 0.5281042456626892,
+      "learning_rate": 0.00047851915457289404,
+      "loss": 1.0533,
+      "step": 1490
+    },
+    {
+      "epoch": 0.26598876103826596,
+      "grad_norm": 0.498197466135025,
+      "learning_rate": 0.0004784907324346868,
+      "loss": 0.8067,
+      "step": 1491
+    },
+    {
+      "epoch": 0.26616715725626616,
+      "grad_norm": 0.583159863948822,
+      "learning_rate": 0.00047846229235091575,
+      "loss": 0.9819,
+      "step": 1492
+    },
+    {
+      "epoch": 0.26634555347426636,
+      "grad_norm": 0.5671996474266052,
+      "learning_rate": 0.0004784338343238146,
+      "loss": 1.2338,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2665239496922665,
+      "grad_norm": 0.5304687023162842,
+      "learning_rate": 0.0004784053583556184,
+      "loss": 1.2323,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2667023459102667,
+      "grad_norm": 0.8755894303321838,
+      "learning_rate": 0.0004783768644485636,
+      "loss": 1.0927,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2668807421282669,
+      "grad_norm": 0.5539624691009521,
+      "learning_rate": 0.0004783483526048882,
+      "loss": 1.1342,
+      "step": 1496
+    },
+    {
+      "epoch": 0.26705913834626704,
+      "grad_norm": 0.5056388974189758,
+      "learning_rate": 0.0004783198228268314,
+      "loss": 1.0296,
+      "step": 1497
+    },
+    {
+      "epoch": 0.26723753456426724,
+      "grad_norm": 0.5661635994911194,
+      "learning_rate": 0.00047829127511663395,
+      "loss": 1.0945,
+      "step": 1498
+    },
+    {
+      "epoch": 0.26741593078226744,
+      "grad_norm": 0.5565497875213623,
+      "learning_rate": 0.00047826270947653803,
+      "loss": 1.0299,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2675943270002676,
+      "grad_norm": 0.5293715596199036,
+      "learning_rate": 0.0004782341259087872,
+      "loss": 1.1213,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2677727232182678,
+      "grad_norm": 0.6879062056541443,
+      "learning_rate": 0.00047820552441562626,
+      "loss": 1.1906,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2679511194362679,
+      "grad_norm": 0.6181793212890625,
+      "learning_rate": 0.00047817690499930165,
+      "loss": 1.2696,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2681295156542681,
+      "grad_norm": 0.8244830369949341,
+      "learning_rate": 0.00047814826766206115,
+      "loss": 1.1163,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2683079118722683,
+      "grad_norm": 0.6630935072898865,
+      "learning_rate": 0.0004781196124061539,
+      "loss": 1.2271,
+      "step": 1504
+    },
+    {
+      "epoch": 0.26848630809026847,
+      "grad_norm": 0.6512210369110107,
+      "learning_rate": 0.0004780909392338304,
+      "loss": 1.3788,
+      "step": 1505
+    },
+    {
+      "epoch": 0.26866470430826866,
+      "grad_norm": 0.900384247303009,
+      "learning_rate": 0.00047806224814734275,
+      "loss": 1.1045,
+      "step": 1506
+    },
+    {
+      "epoch": 0.26884310052626886,
+      "grad_norm": 1.1172149181365967,
+      "learning_rate": 0.0004780335391489442,
+      "loss": 1.116,
+      "step": 1507
+    },
+    {
+      "epoch": 0.269021496744269,
+      "grad_norm": 0.6222420930862427,
+      "learning_rate": 0.00047800481224088965,
+      "loss": 1.0094,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2691998929622692,
+      "grad_norm": 0.7224119901657104,
+      "learning_rate": 0.00047797606742543526,
+      "loss": 1.0397,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2693782891802694,
+      "grad_norm": 0.5364158749580383,
+      "learning_rate": 0.00047794730470483867,
+      "loss": 1.0348,
+      "step": 1510
+    },
+    {
+      "epoch": 0.26955668539826955,
+      "grad_norm": 0.5481920838356018,
+      "learning_rate": 0.00047791852408135885,
+      "loss": 0.8935,
+      "step": 1511
+    },
+    {
+      "epoch": 0.26973508161626975,
+      "grad_norm": 4.288261890411377,
+      "learning_rate": 0.0004778897255572562,
+      "loss": 0.963,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2699134778342699,
+      "grad_norm": 0.6022629141807556,
+      "learning_rate": 0.00047786090913479255,
+      "loss": 1.0822,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2700918740522701,
+      "grad_norm": 0.609437882900238,
+      "learning_rate": 0.00047783207481623126,
+      "loss": 0.8158,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 0.5600314736366272,
+      "learning_rate": 0.00047780322260383674,
+      "loss": 1.1993,
+      "step": 1515
+    },
+    {
+      "epoch": 0.27044866648827043,
+      "grad_norm": 0.6165355443954468,
+      "learning_rate": 0.00047777435249987525,
+      "loss": 1.1175,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2706270627062706,
+      "grad_norm": 0.5497164726257324,
+      "learning_rate": 0.00047774546450661407,
+      "loss": 0.9579,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2708054589242708,
+      "grad_norm": 0.5620499849319458,
+      "learning_rate": 0.0004777165586263221,
+      "loss": 1.162,
+      "step": 1518
+    },
+    {
+      "epoch": 0.27098385514227097,
+      "grad_norm": 0.6000719666481018,
+      "learning_rate": 0.00047768763486126964,
+      "loss": 1.0962,
+      "step": 1519
+    },
+    {
+      "epoch": 0.27116225136027117,
+      "grad_norm": 0.6456661224365234,
+      "learning_rate": 0.00047765869321372834,
+      "loss": 1.2405,
+      "step": 1520
+    },
+    {
+      "epoch": 0.27134064757827137,
+      "grad_norm": 0.6640433073043823,
+      "learning_rate": 0.00047762973368597117,
+      "loss": 1.2917,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2715190437962715,
+      "grad_norm": 0.5571081042289734,
+      "learning_rate": 0.0004776007562802728,
+      "loss": 1.1319,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2716974400142717,
+      "grad_norm": 0.6709672212600708,
+      "learning_rate": 0.0004775717609989089,
+      "loss": 1.1581,
+      "step": 1523
+    },
+    {
+      "epoch": 0.27187583623227185,
+      "grad_norm": 0.6140976548194885,
+      "learning_rate": 0.00047754274784415673,
+      "loss": 1.0881,
+      "step": 1524
+    },
+    {
+      "epoch": 0.27205423245027205,
+      "grad_norm": 0.5703726410865784,
+      "learning_rate": 0.0004775137168182952,
+      "loss": 1.1018,
+      "step": 1525
+    },
+    {
+      "epoch": 0.27223262866827225,
+      "grad_norm": 1.8433053493499756,
+      "learning_rate": 0.00047748466792360425,
+      "loss": 1.0196,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2724110248862724,
+      "grad_norm": 0.5568527579307556,
+      "learning_rate": 0.0004774556011623653,
+      "loss": 0.9479,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2725894211042726,
+      "grad_norm": 0.6627964377403259,
+      "learning_rate": 0.00047742651653686133,
+      "loss": 0.8791,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2727678173222728,
+      "grad_norm": 0.655723512172699,
+      "learning_rate": 0.00047739741404937666,
+      "loss": 1.5567,
+      "step": 1529
+    },
+    {
+      "epoch": 0.27294621354027293,
+      "grad_norm": 1.177575707435608,
+      "learning_rate": 0.00047736829370219694,
+      "loss": 1.1893,
+      "step": 1530
+    },
+    {
+      "epoch": 0.27312460975827313,
+      "grad_norm": 0.5675112009048462,
+      "learning_rate": 0.0004773391554976093,
+      "loss": 1.2078,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2733030059762733,
+      "grad_norm": 0.6315929889678955,
+      "learning_rate": 0.00047730999943790216,
+      "loss": 1.0456,
+      "step": 1532
+    },
+    {
+      "epoch": 0.27348140219427347,
+      "grad_norm": 1.2165286540985107,
+      "learning_rate": 0.0004772808255253655,
+      "loss": 1.3296,
+      "step": 1533
+    },
+    {
+      "epoch": 0.27365979841227367,
+      "grad_norm": 0.6034713983535767,
+      "learning_rate": 0.00047725163376229063,
+      "loss": 1.2784,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2738381946302738,
+      "grad_norm": 0.5982022285461426,
+      "learning_rate": 0.0004772224241509702,
+      "loss": 1.2338,
+      "step": 1535
+    },
+    {
+      "epoch": 0.274016590848274,
+      "grad_norm": 0.797183096408844,
+      "learning_rate": 0.00047719319669369843,
+      "loss": 1.2911,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2741949870662742,
+      "grad_norm": 0.5509144067764282,
+      "learning_rate": 0.0004771639513927707,
+      "loss": 1.3141,
+      "step": 1537
+    },
+    {
+      "epoch": 0.27437338328427435,
+      "grad_norm": 0.6552438735961914,
+      "learning_rate": 0.000477134688250484,
+      "loss": 1.3726,
+      "step": 1538
+    },
+    {
+      "epoch": 0.27455177950227455,
+      "grad_norm": 0.5664661526679993,
+      "learning_rate": 0.0004771054072691367,
+      "loss": 1.484,
+      "step": 1539
+    },
+    {
+      "epoch": 0.27473017572027475,
+      "grad_norm": 0.657057523727417,
+      "learning_rate": 0.0004770761084510283,
+      "loss": 1.5254,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2749085719382749,
+      "grad_norm": 0.5660455822944641,
+      "learning_rate": 0.00047704679179846014,
+      "loss": 1.1427,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2750869681562751,
+      "grad_norm": 0.5524396300315857,
+      "learning_rate": 0.00047701745731373467,
+      "loss": 1.1335,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2752653643742753,
+      "grad_norm": 0.5262104272842407,
+      "learning_rate": 0.00047698810499915577,
+      "loss": 1.0167,
+      "step": 1543
+    },
+    {
+      "epoch": 0.27544376059227543,
+      "grad_norm": 0.5536545515060425,
+      "learning_rate": 0.0004769587348570288,
+      "loss": 1.0548,
+      "step": 1544
+    },
+    {
+      "epoch": 0.27562215681027563,
+      "grad_norm": 0.5155540108680725,
+      "learning_rate": 0.0004769293468896605,
+      "loss": 1.1564,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2758005530282758,
+      "grad_norm": 0.5515694618225098,
+      "learning_rate": 0.00047689994109935884,
+      "loss": 1.1428,
+      "step": 1546
+    },
+    {
+      "epoch": 0.275978949246276,
+      "grad_norm": 0.5768205523490906,
+      "learning_rate": 0.00047687051748843357,
+      "loss": 1.371,
+      "step": 1547
+    },
+    {
+      "epoch": 0.27615734546427617,
+      "grad_norm": 0.6669476628303528,
+      "learning_rate": 0.0004768410760591955,
+      "loss": 1.3972,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2763357416822763,
+      "grad_norm": 1.1227506399154663,
+      "learning_rate": 0.0004768116168139568,
+      "loss": 0.9764,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2765141379002765,
+      "grad_norm": 0.5911063551902771,
+      "learning_rate": 0.00047678213975503136,
+      "loss": 1.2668,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2766925341182767,
+      "grad_norm": 0.47411879897117615,
+      "learning_rate": 0.00047675264488473436,
+      "loss": 0.9734,
+      "step": 1551
+    },
+    {
+      "epoch": 0.27687093033627685,
+      "grad_norm": 0.5374860763549805,
+      "learning_rate": 0.0004767231322053821,
+      "loss": 1.1893,
+      "step": 1552
+    },
+    {
+      "epoch": 0.27704932655427705,
+      "grad_norm": 0.5706403851509094,
+      "learning_rate": 0.00047669360171929265,
+      "loss": 1.1484,
+      "step": 1553
+    },
+    {
+      "epoch": 0.27722772277227725,
+      "grad_norm": 0.6070848703384399,
+      "learning_rate": 0.0004766640534287853,
+      "loss": 1.3872,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2774061189902774,
+      "grad_norm": 0.5827165246009827,
+      "learning_rate": 0.00047663448733618066,
+      "loss": 1.3347,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2775845152082776,
+      "grad_norm": 3.5079944133758545,
+      "learning_rate": 0.00047660490344380094,
+      "loss": 1.1723,
+      "step": 1556
+    },
+    {
+      "epoch": 0.27776291142627774,
+      "grad_norm": 0.6186500787734985,
+      "learning_rate": 0.00047657530175396955,
+      "loss": 1.5012,
+      "step": 1557
+    },
+    {
+      "epoch": 0.27794130764427794,
+      "grad_norm": 0.6020243763923645,
+      "learning_rate": 0.0004765456822690116,
+      "loss": 1.0751,
+      "step": 1558
+    },
+    {
+      "epoch": 0.27811970386227813,
+      "grad_norm": 0.5379880666732788,
+      "learning_rate": 0.00047651604499125325,
+      "loss": 0.9193,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2782981000802783,
+      "grad_norm": 0.6437864899635315,
+      "learning_rate": 0.0004764863899230221,
+      "loss": 1.2344,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2784764962982785,
+      "grad_norm": 0.5693362951278687,
+      "learning_rate": 0.00047645671706664737,
+      "loss": 1.1764,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2786548925162787,
+      "grad_norm": 0.5577165484428406,
+      "learning_rate": 0.00047642702642445954,
+      "loss": 1.2728,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2788332887342788,
+      "grad_norm": 0.5637865662574768,
+      "learning_rate": 0.0004763973179987906,
+      "loss": 1.3064,
+      "step": 1563
+    },
+    {
+      "epoch": 0.279011684952279,
+      "grad_norm": 0.5422176122665405,
+      "learning_rate": 0.00047636759179197366,
+      "loss": 0.9754,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2791900811702792,
+      "grad_norm": 0.5627367496490479,
+      "learning_rate": 0.00047633784780634343,
+      "loss": 1.0802,
+      "step": 1565
+    },
+    {
+      "epoch": 0.27936847738827936,
+      "grad_norm": 0.6842982769012451,
+      "learning_rate": 0.0004763080860442361,
+      "loss": 1.0098,
+      "step": 1566
+    },
+    {
+      "epoch": 0.27954687360627956,
+      "grad_norm": 0.6131782531738281,
+      "learning_rate": 0.00047627830650798903,
+      "loss": 1.4362,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2797252698242797,
+      "grad_norm": 0.6176859736442566,
+      "learning_rate": 0.00047624850919994113,
+      "loss": 1.3148,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2799036660422799,
+      "grad_norm": 0.6015002727508545,
+      "learning_rate": 0.00047621869412243275,
+      "loss": 1.1703,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2800820622602801,
+      "grad_norm": 0.750379204750061,
+      "learning_rate": 0.0004761888612778054,
+      "loss": 1.2353,
+      "step": 1570
+    },
+    {
+      "epoch": 0.28026045847828024,
+      "grad_norm": 0.567866325378418,
+      "learning_rate": 0.0004761590106684023,
+      "loss": 1.2543,
+      "step": 1571
+    },
+    {
+      "epoch": 0.28043885469628044,
+      "grad_norm": 0.5942226052284241,
+      "learning_rate": 0.0004761291422965678,
+      "loss": 1.2128,
+      "step": 1572
+    },
+    {
+      "epoch": 0.28061725091428064,
+      "grad_norm": 0.4971591830253601,
+      "learning_rate": 0.00047609925616464777,
+      "loss": 1.0342,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2807956471322808,
+      "grad_norm": 0.5671881437301636,
+      "learning_rate": 0.0004760693522749894,
+      "loss": 1.0644,
+      "step": 1574
+    },
+    {
+      "epoch": 0.280974043350281,
+      "grad_norm": 0.887310266494751,
+      "learning_rate": 0.00047603943062994147,
+      "loss": 1.0307,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2811524395682812,
+      "grad_norm": 0.5171691179275513,
+      "learning_rate": 0.00047600949123185386,
+      "loss": 0.9699,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2813308357862813,
+      "grad_norm": 0.799001932144165,
+      "learning_rate": 0.00047597953408307813,
+      "loss": 1.2391,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2815092320042815,
+      "grad_norm": 0.5518134832382202,
+      "learning_rate": 0.00047594955918596704,
+      "loss": 0.9825,
+      "step": 1578
+    },
+    {
+      "epoch": 0.28168762822228166,
+      "grad_norm": 0.512997567653656,
+      "learning_rate": 0.00047591956654287484,
+      "loss": 1.1596,
+      "step": 1579
+    },
+    {
+      "epoch": 0.28186602444028186,
+      "grad_norm": 0.5603839755058289,
+      "learning_rate": 0.00047588955615615705,
+      "loss": 1.1735,
+      "step": 1580
+    },
+    {
+      "epoch": 0.28204442065828206,
+      "grad_norm": 0.5397359728813171,
+      "learning_rate": 0.0004758595280281707,
+      "loss": 1.1707,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2822228168762822,
+      "grad_norm": 0.5633308291435242,
+      "learning_rate": 0.0004758294821612742,
+      "loss": 1.1632,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2824012130942824,
+      "grad_norm": 1.6372519731521606,
+      "learning_rate": 0.00047579941855782745,
+      "loss": 0.9778,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2825796093122826,
+      "grad_norm": 0.790706217288971,
+      "learning_rate": 0.00047576933722019146,
+      "loss": 1.2109,
+      "step": 1584
+    },
+    {
+      "epoch": 0.28275800553028274,
+      "grad_norm": 0.5796968340873718,
+      "learning_rate": 0.0004757392381507289,
+      "loss": 1.3131,
+      "step": 1585
+    },
+    {
+      "epoch": 0.28293640174828294,
+      "grad_norm": 0.7055728435516357,
+      "learning_rate": 0.0004757091213518037,
+      "loss": 1.1037,
+      "step": 1586
+    },
+    {
+      "epoch": 0.28311479796628314,
+      "grad_norm": 1.4809973239898682,
+      "learning_rate": 0.00047567898682578124,
+      "loss": 1.1257,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2832931941842833,
+      "grad_norm": 0.6054453253746033,
+      "learning_rate": 0.00047564883457502835,
+      "loss": 1.0544,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2834715904022835,
+      "grad_norm": 6.443731307983398,
+      "learning_rate": 0.0004756186646019131,
+      "loss": 0.8428,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2836499866202836,
+      "grad_norm": 9.057873725891113,
+      "learning_rate": 0.0004755884769088049,
+      "loss": 1.5444,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2838283828382838,
+      "grad_norm": 0.7593368291854858,
+      "learning_rate": 0.00047555827149807484,
+      "loss": 1.1721,
+      "step": 1591
+    },
+    {
+      "epoch": 0.284006779056284,
+      "grad_norm": 0.6235383749008179,
+      "learning_rate": 0.00047552804837209525,
+      "loss": 1.3072,
+      "step": 1592
+    },
+    {
+      "epoch": 0.28418517527428416,
+      "grad_norm": 0.9655506610870361,
+      "learning_rate": 0.0004754978075332398,
+      "loss": 1.1129,
+      "step": 1593
+    },
+    {
+      "epoch": 0.28436357149228436,
+      "grad_norm": 0.7398672699928284,
+      "learning_rate": 0.0004754675489838835,
+      "loss": 0.8643,
+      "step": 1594
+    },
+    {
+      "epoch": 0.28454196771028456,
+      "grad_norm": 0.6276271343231201,
+      "learning_rate": 0.000475437272726403,
+      "loss": 1.1541,
+      "step": 1595
+    },
+    {
+      "epoch": 0.2847203639282847,
+      "grad_norm": 0.6548689603805542,
+      "learning_rate": 0.0004754069787631761,
+      "loss": 1.024,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2848987601462849,
+      "grad_norm": 0.7249738574028015,
+      "learning_rate": 0.0004753766670965821,
+      "loss": 1.0359,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2850771563642851,
+      "grad_norm": 0.6566236615180969,
+      "learning_rate": 0.0004753463377290016,
+      "loss": 0.8981,
+      "step": 1598
+    },
+    {
+      "epoch": 0.28525555258228524,
+      "grad_norm": 0.7142555117607117,
+      "learning_rate": 0.0004753159906628167,
+      "loss": 1.2243,
+      "step": 1599
+    },
+    {
+      "epoch": 0.28543394880028544,
+      "grad_norm": 0.5189533233642578,
+      "learning_rate": 0.00047528562590041097,
+      "loss": 1.1284,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2856123450182856,
+      "grad_norm": 0.619817316532135,
+      "learning_rate": 0.000475255243444169,
+      "loss": 1.0063,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2857907412362858,
+      "grad_norm": 3.838261127471924,
+      "learning_rate": 0.00047522484329647725,
+      "loss": 1.0014,
+      "step": 1602
+    },
+    {
+      "epoch": 0.285969137454286,
+      "grad_norm": 0.6709682941436768,
+      "learning_rate": 0.0004751944254597232,
+      "loss": 1.1519,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2861475336722861,
+      "grad_norm": 0.6216676235198975,
+      "learning_rate": 0.0004751639899362958,
+      "loss": 1.17,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2863259298902863,
+      "grad_norm": 2.6110217571258545,
+      "learning_rate": 0.00047513353672858565,
+      "loss": 1.3296,
+      "step": 1605
+    },
+    {
+      "epoch": 0.2865043261082865,
+      "grad_norm": 0.6154805421829224,
+      "learning_rate": 0.0004751030658389843,
+      "loss": 1.4798,
+      "step": 1606
+    },
+    {
+      "epoch": 0.28668272232628667,
+      "grad_norm": 0.6929340958595276,
+      "learning_rate": 0.00047507257726988515,
+      "loss": 0.8626,
+      "step": 1607
+    },
+    {
+      "epoch": 0.28686111854428686,
+      "grad_norm": 0.5480323433876038,
+      "learning_rate": 0.00047504207102368255,
+      "loss": 1.0958,
+      "step": 1608
+    },
+    {
+      "epoch": 0.28703951476228706,
+      "grad_norm": 0.6278098821640015,
+      "learning_rate": 0.00047501154710277255,
+      "loss": 1.4003,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2872179109802872,
+      "grad_norm": 0.5721479058265686,
+      "learning_rate": 0.0004749810055095525,
+      "loss": 0.9552,
+      "step": 1610
+    },
+    {
+      "epoch": 0.2873963071982874,
+      "grad_norm": 0.5661513209342957,
+      "learning_rate": 0.000474950446246421,
+      "loss": 1.3078,
+      "step": 1611
+    },
+    {
+      "epoch": 0.28757470341628755,
+      "grad_norm": 0.508573055267334,
+      "learning_rate": 0.00047491986931577835,
+      "loss": 1.1201,
+      "step": 1612
+    },
+    {
+      "epoch": 0.28775309963428775,
+      "grad_norm": 0.5927449464797974,
+      "learning_rate": 0.00047488927472002596,
+      "loss": 1.26,
+      "step": 1613
+    },
+    {
+      "epoch": 0.28793149585228794,
+      "grad_norm": 0.5334576964378357,
+      "learning_rate": 0.00047485866246156665,
+      "loss": 1.313,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2881098920702881,
+      "grad_norm": 0.5566635131835938,
+      "learning_rate": 0.0004748280325428048,
+      "loss": 1.4478,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 0.47653868794441223,
+      "learning_rate": 0.000474797384966146,
+      "loss": 1.0392,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2884666845062885,
+      "grad_norm": 0.6128924489021301,
+      "learning_rate": 0.0004747667197339974,
+      "loss": 1.2961,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2886450807242886,
+      "grad_norm": 0.4769141376018524,
+      "learning_rate": 0.0004747360368487672,
+      "loss": 0.9468,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2888234769422888,
+      "grad_norm": 0.5372063517570496,
+      "learning_rate": 0.0004747053363128655,
+      "loss": 1.1281,
+      "step": 1619
+    },
+    {
+      "epoch": 0.289001873160289,
+      "grad_norm": 0.499072402715683,
+      "learning_rate": 0.0004746746181287034,
+      "loss": 1.094,
+      "step": 1620
+    },
+    {
+      "epoch": 0.28918026937828917,
+      "grad_norm": 0.6395072937011719,
+      "learning_rate": 0.0004746438822986934,
+      "loss": 1.3721,
+      "step": 1621
+    },
+    {
+      "epoch": 0.28935866559628937,
+      "grad_norm": 0.5611110925674438,
+      "learning_rate": 0.00047461312882524954,
+      "loss": 1.1692,
+      "step": 1622
+    },
+    {
+      "epoch": 0.28953706181428956,
+      "grad_norm": 0.6090511679649353,
+      "learning_rate": 0.0004745823577107873,
+      "loss": 1.175,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2897154580322897,
+      "grad_norm": 0.5791161060333252,
+      "learning_rate": 0.0004745515689577233,
+      "loss": 1.2759,
+      "step": 1624
+    },
+    {
+      "epoch": 0.2898938542502899,
+      "grad_norm": 0.5539647936820984,
+      "learning_rate": 0.0004745207625684756,
+      "loss": 1.1852,
+      "step": 1625
+    },
+    {
+      "epoch": 0.29007225046829005,
+      "grad_norm": 1.770546317100525,
+      "learning_rate": 0.000474489938545464,
+      "loss": 1.0354,
+      "step": 1626
+    },
+    {
+      "epoch": 0.29025064668629025,
+      "grad_norm": 0.640678882598877,
+      "learning_rate": 0.0004744590968911091,
+      "loss": 1.2128,
+      "step": 1627
+    },
+    {
+      "epoch": 0.29042904290429045,
+      "grad_norm": 0.6019294261932373,
+      "learning_rate": 0.00047442823760783336,
+      "loss": 1.3319,
+      "step": 1628
+    },
+    {
+      "epoch": 0.2906074391222906,
+      "grad_norm": 0.6891437768936157,
+      "learning_rate": 0.0004743973606980604,
+      "loss": 0.8761,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2907858353402908,
+      "grad_norm": 0.5156970620155334,
+      "learning_rate": 0.0004743664661642153,
+      "loss": 1.139,
+      "step": 1630
+    },
+    {
+      "epoch": 0.290964231558291,
+      "grad_norm": 0.5733403563499451,
+      "learning_rate": 0.0004743355540087245,
+      "loss": 1.3032,
+      "step": 1631
+    },
+    {
+      "epoch": 0.29114262777629113,
+      "grad_norm": 0.7720729112625122,
+      "learning_rate": 0.00047430462423401587,
+      "loss": 1.1981,
+      "step": 1632
+    },
+    {
+      "epoch": 0.29132102399429133,
+      "grad_norm": 0.4814129173755646,
+      "learning_rate": 0.00047427367684251855,
+      "loss": 0.7758,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2914994202122915,
+      "grad_norm": 0.6132440567016602,
+      "learning_rate": 0.0004742427118366632,
+      "loss": 1.4416,
+      "step": 1634
+    },
+    {
+      "epoch": 0.29167781643029167,
+      "grad_norm": 0.5137991309165955,
+      "learning_rate": 0.0004742117292188817,
+      "loss": 0.913,
+      "step": 1635
+    },
+    {
+      "epoch": 0.29185621264829187,
+      "grad_norm": 0.561093270778656,
+      "learning_rate": 0.0004741807289916075,
+      "loss": 1.0205,
+      "step": 1636
+    },
+    {
+      "epoch": 0.292034608866292,
+      "grad_norm": 0.5749049782752991,
+      "learning_rate": 0.00047414971115727536,
+      "loss": 1.1973,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2922130050842922,
+      "grad_norm": 0.5146419405937195,
+      "learning_rate": 0.00047411867571832135,
+      "loss": 1.0004,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2923914013022924,
+      "grad_norm": 0.6099346280097961,
+      "learning_rate": 0.00047408762267718297,
+      "loss": 1.1748,
+      "step": 1639
+    },
+    {
+      "epoch": 0.29256979752029255,
+      "grad_norm": 0.5745267271995544,
+      "learning_rate": 0.0004740565520362991,
+      "loss": 1.1932,
+      "step": 1640
+    },
+    {
+      "epoch": 0.29274819373829275,
+      "grad_norm": 0.5441005229949951,
+      "learning_rate": 0.0004740254637981101,
+      "loss": 1.2205,
+      "step": 1641
+    },
+    {
+      "epoch": 0.29292658995629295,
+      "grad_norm": 0.5015900135040283,
+      "learning_rate": 0.00047399435796505754,
+      "loss": 0.8114,
+      "step": 1642
+    },
+    {
+      "epoch": 0.2931049861742931,
+      "grad_norm": 0.6179179549217224,
+      "learning_rate": 0.0004739632345395846,
+      "loss": 1.1337,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2932833823922933,
+      "grad_norm": 0.5103684663772583,
+      "learning_rate": 0.0004739320935241355,
+      "loss": 0.8642,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2934617786102935,
+      "grad_norm": 0.68010014295578,
+      "learning_rate": 0.0004739009349211561,
+      "loss": 1.284,
+      "step": 1645
+    },
+    {
+      "epoch": 0.29364017482829363,
+      "grad_norm": 0.5158190131187439,
+      "learning_rate": 0.0004738697587330937,
+      "loss": 1.1043,
+      "step": 1646
+    },
+    {
+      "epoch": 0.29381857104629383,
+      "grad_norm": 0.5020380616188049,
+      "learning_rate": 0.00047383856496239677,
+      "loss": 0.9218,
+      "step": 1647
+    },
+    {
+      "epoch": 0.293996967264294,
+      "grad_norm": 0.5497924089431763,
+      "learning_rate": 0.00047380735361151526,
+      "loss": 1.3251,
+      "step": 1648
+    },
+    {
+      "epoch": 0.29417536348229417,
+      "grad_norm": 0.5309639573097229,
+      "learning_rate": 0.00047377612468290053,
+      "loss": 1.0265,
+      "step": 1649
+    },
+    {
+      "epoch": 0.29435375970029437,
+      "grad_norm": 0.528611958026886,
+      "learning_rate": 0.0004737448781790052,
+      "loss": 1.1084,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2945321559182945,
+      "grad_norm": 0.7254324555397034,
+      "learning_rate": 0.0004737136141022836,
+      "loss": 1.2578,
+      "step": 1651
+    },
+    {
+      "epoch": 0.2947105521362947,
+      "grad_norm": 0.6030469536781311,
+      "learning_rate": 0.0004736823324551909,
+      "loss": 1.0314,
+      "step": 1652
+    },
+    {
+      "epoch": 0.2948889483542949,
+      "grad_norm": 0.5361880660057068,
+      "learning_rate": 0.00047365103324018405,
+      "loss": 1.1275,
+      "step": 1653
+    },
+    {
+      "epoch": 0.29506734457229505,
+      "grad_norm": 0.5910381078720093,
+      "learning_rate": 0.00047361971645972135,
+      "loss": 0.9726,
+      "step": 1654
+    },
+    {
+      "epoch": 0.29524574079029525,
+      "grad_norm": 0.6346861124038696,
+      "learning_rate": 0.00047358838211626234,
+      "loss": 1.0678,
+      "step": 1655
+    },
+    {
+      "epoch": 0.29542413700829545,
+      "grad_norm": 2.6803693771362305,
+      "learning_rate": 0.0004735570302122679,
+      "loss": 1.2372,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2956025332262956,
+      "grad_norm": 0.571451723575592,
+      "learning_rate": 0.0004735256607502006,
+      "loss": 1.1094,
+      "step": 1657
+    },
+    {
+      "epoch": 0.2957809294442958,
+      "grad_norm": 0.5796542167663574,
+      "learning_rate": 0.0004734942737325242,
+      "loss": 1.05,
+      "step": 1658
+    },
+    {
+      "epoch": 0.29595932566229594,
+      "grad_norm": 0.6353934407234192,
+      "learning_rate": 0.00047346286916170356,
+      "loss": 1.1633,
+      "step": 1659
+    },
+    {
+      "epoch": 0.29613772188029613,
+      "grad_norm": 0.7311161160469055,
+      "learning_rate": 0.00047343144704020543,
+      "loss": 1.1351,
+      "step": 1660
+    },
+    {
+      "epoch": 0.29631611809829633,
+      "grad_norm": 0.5946211218833923,
+      "learning_rate": 0.00047340000737049756,
+      "loss": 1.0721,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2964945143162965,
+      "grad_norm": 0.6070974469184875,
+      "learning_rate": 0.00047336855015504923,
+      "loss": 1.0058,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2966729105342967,
+      "grad_norm": 0.5165389180183411,
+      "learning_rate": 0.0004733370753963311,
+      "loss": 0.9556,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2968513067522969,
+      "grad_norm": 0.7445652484893799,
+      "learning_rate": 0.0004733055830968152,
+      "loss": 0.9277,
+      "step": 1664
+    },
+    {
+      "epoch": 0.297029702970297,
+      "grad_norm": 0.520530641078949,
+      "learning_rate": 0.0004732740732589749,
+      "loss": 1.0856,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2972080991882972,
+      "grad_norm": 0.6134181618690491,
+      "learning_rate": 0.00047324254588528497,
+      "loss": 1.3027,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2973864954062974,
+      "grad_norm": 0.5679133534431458,
+      "learning_rate": 0.00047321100097822154,
+      "loss": 1.1749,
+      "step": 1667
+    },
+    {
+      "epoch": 0.29756489162429756,
+      "grad_norm": 0.5943863391876221,
+      "learning_rate": 0.0004731794385402621,
+      "loss": 1.2488,
+      "step": 1668
+    },
+    {
+      "epoch": 0.29774328784229775,
+      "grad_norm": 0.5449469089508057,
+      "learning_rate": 0.00047314785857388575,
+      "loss": 1.1212,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2979216840602979,
+      "grad_norm": 0.882116973400116,
+      "learning_rate": 0.0004731162610815725,
+      "loss": 1.2176,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2981000802782981,
+      "grad_norm": 0.6479673385620117,
+      "learning_rate": 0.0004730846460658041,
+      "loss": 1.3039,
+      "step": 1671
+    },
+    {
+      "epoch": 0.2982784764962983,
+      "grad_norm": 0.5808786749839783,
+      "learning_rate": 0.0004730530135290637,
+      "loss": 1.0677,
+      "step": 1672
+    },
+    {
+      "epoch": 0.29845687271429844,
+      "grad_norm": 0.527851402759552,
+      "learning_rate": 0.0004730213634738355,
+      "loss": 1.0218,
+      "step": 1673
+    },
+    {
+      "epoch": 0.29863526893229864,
+      "grad_norm": 0.8989919424057007,
+      "learning_rate": 0.00047298969590260545,
+      "loss": 1.0192,
+      "step": 1674
+    },
+    {
+      "epoch": 0.29881366515029884,
+      "grad_norm": 0.5305653214454651,
+      "learning_rate": 0.0004729580108178606,
+      "loss": 1.0954,
+      "step": 1675
+    },
+    {
+      "epoch": 0.298992061368299,
+      "grad_norm": 0.6758263111114502,
+      "learning_rate": 0.0004729263082220896,
+      "loss": 1.1321,
+      "step": 1676
+    },
+    {
+      "epoch": 0.2991704575862992,
+      "grad_norm": 0.7814352512359619,
+      "learning_rate": 0.0004728945881177822,
+      "loss": 0.8791,
+      "step": 1677
+    },
+    {
+      "epoch": 0.2993488538042994,
+      "grad_norm": 1.4971033334732056,
+      "learning_rate": 0.00047286285050742984,
+      "loss": 1.2261,
+      "step": 1678
+    },
+    {
+      "epoch": 0.2995272500222995,
+      "grad_norm": 0.6250385046005249,
+      "learning_rate": 0.0004728310953935251,
+      "loss": 1.0965,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2997056462402997,
+      "grad_norm": 0.5605043172836304,
+      "learning_rate": 0.00047279932277856195,
+      "loss": 1.2065,
+      "step": 1680
+    },
+    {
+      "epoch": 0.29988404245829986,
+      "grad_norm": 0.5820797085762024,
+      "learning_rate": 0.0004727675326650359,
+      "loss": 0.892,
+      "step": 1681
+    },
+    {
+      "epoch": 0.30006243867630006,
+      "grad_norm": 0.626677930355072,
+      "learning_rate": 0.0004727357250554437,
+      "loss": 1.2638,
+      "step": 1682
+    },
+    {
+      "epoch": 0.30024083489430026,
+      "grad_norm": 0.5350278615951538,
+      "learning_rate": 0.00047270389995228353,
+      "loss": 1.2105,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3004192311123004,
+      "grad_norm": 0.6105818748474121,
+      "learning_rate": 0.0004726720573580549,
+      "loss": 1.0251,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3005976273303006,
+      "grad_norm": 0.6602348685264587,
+      "learning_rate": 0.00047264019727525866,
+      "loss": 1.3229,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3007760235483008,
+      "grad_norm": 0.5458388924598694,
+      "learning_rate": 0.00047260831970639716,
+      "loss": 1.2786,
+      "step": 1686
+    },
+    {
+      "epoch": 0.30095441976630094,
+      "grad_norm": 0.6178166270256042,
+      "learning_rate": 0.000472576424653974,
+      "loss": 1.1324,
+      "step": 1687
+    },
+    {
+      "epoch": 0.30113281598430114,
+      "grad_norm": 0.6096773147583008,
+      "learning_rate": 0.0004725445121204943,
+      "loss": 1.0911,
+      "step": 1688
+    },
+    {
+      "epoch": 0.30131121220230134,
+      "grad_norm": 0.6458967924118042,
+      "learning_rate": 0.0004725125821084643,
+      "loss": 1.0113,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3014896084203015,
+      "grad_norm": 0.6352248191833496,
+      "learning_rate": 0.00047248063462039194,
+      "loss": 1.206,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3016680046383017,
+      "grad_norm": 0.7942283749580383,
+      "learning_rate": 0.0004724486696587862,
+      "loss": 1.3854,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3018464008563018,
+      "grad_norm": 0.6200905442237854,
+      "learning_rate": 0.00047241668722615773,
+      "loss": 1.0581,
+      "step": 1692
+    },
+    {
+      "epoch": 0.302024797074302,
+      "grad_norm": 0.6159539222717285,
+      "learning_rate": 0.0004723846873250183,
+      "loss": 1.15,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3022031932923022,
+      "grad_norm": 0.5503310561180115,
+      "learning_rate": 0.00047235266995788127,
+      "loss": 1.0242,
+      "step": 1694
+    },
+    {
+      "epoch": 0.30238158951030236,
+      "grad_norm": 0.8836873173713684,
+      "learning_rate": 0.00047232063512726125,
+      "loss": 1.2961,
+      "step": 1695
+    },
+    {
+      "epoch": 0.30255998572830256,
+      "grad_norm": 0.6245954036712646,
+      "learning_rate": 0.0004722885828356742,
+      "loss": 1.2354,
+      "step": 1696
+    },
+    {
+      "epoch": 0.30273838194630276,
+      "grad_norm": 1.8814623355865479,
+      "learning_rate": 0.0004722565130856375,
+      "loss": 0.9584,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3029167781643029,
+      "grad_norm": 0.6291859149932861,
+      "learning_rate": 0.0004722244258796699,
+      "loss": 1.3523,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3030951743823031,
+      "grad_norm": 0.52773517370224,
+      "learning_rate": 0.00047219232122029154,
+      "loss": 1.1431,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3032735706003033,
+      "grad_norm": 0.6015790104866028,
+      "learning_rate": 0.0004721601991100239,
+      "loss": 1.2411,
+      "step": 1700
+    },
+    {
+      "epoch": 0.30345196681830344,
+      "grad_norm": 0.5504578351974487,
+      "learning_rate": 0.0004721280595513898,
+      "loss": 1.1042,
+      "step": 1701
+    },
+    {
+      "epoch": 0.30363036303630364,
+      "grad_norm": 0.5806344747543335,
+      "learning_rate": 0.00047209590254691347,
+      "loss": 1.4875,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3038087592543038,
+      "grad_norm": 0.6146537065505981,
+      "learning_rate": 0.0004720637280991206,
+      "loss": 1.3344,
+      "step": 1703
+    },
+    {
+      "epoch": 0.303987155472304,
+      "grad_norm": 0.5845167636871338,
+      "learning_rate": 0.00047203153621053803,
+      "loss": 1.3061,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3041655516903042,
+      "grad_norm": 0.555408775806427,
+      "learning_rate": 0.0004719993268836942,
+      "loss": 0.9783,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3043439479083043,
+      "grad_norm": 0.7523461580276489,
+      "learning_rate": 0.00047196710012111865,
+      "loss": 1.0337,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3045223441263045,
+      "grad_norm": 0.5478354096412659,
+      "learning_rate": 0.00047193485592534267,
+      "loss": 1.1427,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3047007403443047,
+      "grad_norm": 0.5235744118690491,
+      "learning_rate": 0.0004719025942988986,
+      "loss": 1.1299,
+      "step": 1708
+    },
+    {
+      "epoch": 0.30487913656230486,
+      "grad_norm": 0.5737703442573547,
+      "learning_rate": 0.00047187031524432033,
+      "loss": 1.259,
+      "step": 1709
+    },
+    {
+      "epoch": 0.30505753278030506,
+      "grad_norm": 0.515281617641449,
+      "learning_rate": 0.0004718380187641429,
+      "loss": 1.0624,
+      "step": 1710
+    },
+    {
+      "epoch": 0.30523592899830526,
+      "grad_norm": 0.4825386703014374,
+      "learning_rate": 0.000471805704860903,
+      "loss": 0.9467,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3054143252163054,
+      "grad_norm": 0.73399418592453,
+      "learning_rate": 0.00047177337353713843,
+      "loss": 1.2483,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3055927214343056,
+      "grad_norm": 0.5785844922065735,
+      "learning_rate": 0.00047174102479538853,
+      "loss": 1.1172,
+      "step": 1713
+    },
+    {
+      "epoch": 0.30577111765230575,
+      "grad_norm": 0.559877336025238,
+      "learning_rate": 0.000471708658638194,
+      "loss": 1.1775,
+      "step": 1714
+    },
+    {
+      "epoch": 0.30594951387030594,
+      "grad_norm": 0.5484994649887085,
+      "learning_rate": 0.00047167627506809686,
+      "loss": 1.0763,
+      "step": 1715
+    },
+    {
+      "epoch": 0.30612791008830614,
+      "grad_norm": 0.6489263772964478,
+      "learning_rate": 0.0004716438740876404,
+      "loss": 1.0646,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3063063063063063,
+      "grad_norm": 0.5778820514678955,
+      "learning_rate": 0.0004716114556993695,
+      "loss": 1.0459,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3064847025243065,
+      "grad_norm": 2.8568458557128906,
+      "learning_rate": 0.00047157901990583026,
+      "loss": 0.9765,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3066630987423067,
+      "grad_norm": 0.6178241968154907,
+      "learning_rate": 0.0004715465667095701,
+      "loss": 1.2187,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3068414949603068,
+      "grad_norm": 0.5349613428115845,
+      "learning_rate": 0.0004715140961131379,
+      "loss": 1.063,
+      "step": 1720
+    },
+    {
+      "epoch": 0.307019891178307,
+      "grad_norm": 0.556450366973877,
+      "learning_rate": 0.00047148160811908395,
+      "loss": 1.0956,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3071982873963072,
+      "grad_norm": 0.4912685453891754,
+      "learning_rate": 0.0004714491027299599,
+      "loss": 0.8442,
+      "step": 1722
+    },
+    {
+      "epoch": 0.30737668361430737,
+      "grad_norm": 0.5774837136268616,
+      "learning_rate": 0.0004714165799483185,
+      "loss": 1.2867,
+      "step": 1723
+    },
+    {
+      "epoch": 0.30755507983230757,
+      "grad_norm": 0.5329927206039429,
+      "learning_rate": 0.0004713840397767142,
+      "loss": 0.8711,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3077334760503077,
+      "grad_norm": 0.5508580803871155,
+      "learning_rate": 0.00047135148221770273,
+      "loss": 1.1184,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3079118722683079,
+      "grad_norm": 0.6747270822525024,
+      "learning_rate": 0.0004713189072738411,
+      "loss": 1.3457,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3080902684863081,
+      "grad_norm": 0.5561022162437439,
+      "learning_rate": 0.0004712863149476877,
+      "loss": 1.0277,
+      "step": 1727
+    },
+    {
+      "epoch": 0.30826866470430825,
+      "grad_norm": 0.6054466962814331,
+      "learning_rate": 0.0004712537052418024,
+      "loss": 1.3957,
+      "step": 1728
+    },
+    {
+      "epoch": 0.30844706092230845,
+      "grad_norm": 84.19719696044922,
+      "learning_rate": 0.00047122107815874626,
+      "loss": 1.2228,
+      "step": 1729
+    },
+    {
+      "epoch": 0.30862545714030865,
+      "grad_norm": 0.6073416471481323,
+      "learning_rate": 0.00047118843370108187,
+      "loss": 1.0421,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3088038533583088,
+      "grad_norm": 0.7347621917724609,
+      "learning_rate": 0.00047115577187137304,
+      "loss": 1.2824,
+      "step": 1731
+    },
+    {
+      "epoch": 0.308982249576309,
+      "grad_norm": 0.5585038661956787,
+      "learning_rate": 0.00047112309267218513,
+      "loss": 1.0725,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3091606457943092,
+      "grad_norm": 0.5505072474479675,
+      "learning_rate": 0.00047109039610608464,
+      "loss": 1.1923,
+      "step": 1733
+    },
+    {
+      "epoch": 0.30933904201230933,
+      "grad_norm": 2.2346675395965576,
+      "learning_rate": 0.0004710576821756395,
+      "loss": 1.3058,
+      "step": 1734
+    },
+    {
+      "epoch": 0.3095174382303095,
+      "grad_norm": 0.5852141976356506,
+      "learning_rate": 0.0004710249508834192,
+      "loss": 1.3095,
+      "step": 1735
+    },
+    {
+      "epoch": 0.30969583444830967,
+      "grad_norm": 3.471550464630127,
+      "learning_rate": 0.00047099220223199444,
+      "loss": 1.2149,
+      "step": 1736
+    },
+    {
+      "epoch": 0.30987423066630987,
+      "grad_norm": 4.984081268310547,
+      "learning_rate": 0.0004709594362239371,
+      "loss": 1.286,
+      "step": 1737
+    },
+    {
+      "epoch": 0.31005262688431007,
+      "grad_norm": 0.8435764908790588,
+      "learning_rate": 0.0004709266528618208,
+      "loss": 0.8668,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3102310231023102,
+      "grad_norm": 0.6224226355552673,
+      "learning_rate": 0.0004708938521482202,
+      "loss": 1.2869,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3104094193203104,
+      "grad_norm": 0.569222092628479,
+      "learning_rate": 0.00047086103408571157,
+      "loss": 1.2534,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3105878155383106,
+      "grad_norm": 5.230443000793457,
+      "learning_rate": 0.0004708281986768724,
+      "loss": 1.1547,
+      "step": 1741
+    },
+    {
+      "epoch": 0.31076621175631075,
+      "grad_norm": 0.726719319820404,
+      "learning_rate": 0.00047079534592428144,
+      "loss": 1.1936,
+      "step": 1742
+    },
+    {
+      "epoch": 0.31094460797431095,
+      "grad_norm": 0.794158399105072,
+      "learning_rate": 0.0004707624758305191,
+      "loss": 1.0309,
+      "step": 1743
+    },
+    {
+      "epoch": 0.31112300419231115,
+      "grad_norm": 0.580272376537323,
+      "learning_rate": 0.0004707295883981668,
+      "loss": 1.0352,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3113014004103113,
+      "grad_norm": 0.6194749474525452,
+      "learning_rate": 0.00047069668362980776,
+      "loss": 1.2419,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3114797966283115,
+      "grad_norm": 0.6249500513076782,
+      "learning_rate": 0.0004706637615280261,
+      "loss": 1.0573,
+      "step": 1746
+    },
+    {
+      "epoch": 0.31165819284631163,
+      "grad_norm": 0.6039573550224304,
+      "learning_rate": 0.0004706308220954075,
+      "loss": 1.0936,
+      "step": 1747
+    },
+    {
+      "epoch": 0.31183658906431183,
+      "grad_norm": 0.570054829120636,
+      "learning_rate": 0.00047059786533453917,
+      "loss": 0.9585,
+      "step": 1748
+    },
+    {
+      "epoch": 0.31201498528231203,
+      "grad_norm": 0.5929118394851685,
+      "learning_rate": 0.0004705648912480094,
+      "loss": 0.9652,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3121933815003122,
+      "grad_norm": 0.6323100328445435,
+      "learning_rate": 0.000470531899838408,
+      "loss": 1.3066,
+      "step": 1750
+    },
+    {
+      "epoch": 0.31237177771831237,
+      "grad_norm": 0.5521016120910645,
+      "learning_rate": 0.00047049889110832604,
+      "loss": 1.082,
+      "step": 1751
+    },
+    {
+      "epoch": 0.31255017393631257,
+      "grad_norm": 0.5617358684539795,
+      "learning_rate": 0.00047046586506035613,
+      "loss": 1.2556,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3127285701543127,
+      "grad_norm": 1.5406019687652588,
+      "learning_rate": 0.000470432821697092,
+      "loss": 1.303,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3129069663723129,
+      "grad_norm": 0.5704224109649658,
+      "learning_rate": 0.0004703997610211289,
+      "loss": 1.0045,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3130853625903131,
+      "grad_norm": 0.5109691023826599,
+      "learning_rate": 0.00047036668303506346,
+      "loss": 0.9764,
+      "step": 1755
+    },
+    {
+      "epoch": 0.31326375880831325,
+      "grad_norm": 0.5933413505554199,
+      "learning_rate": 0.0004703335877414936,
+      "loss": 1.1017,
+      "step": 1756
+    },
+    {
+      "epoch": 0.31344215502631345,
+      "grad_norm": 0.6716773509979248,
+      "learning_rate": 0.0004703004751430185,
+      "loss": 1.3598,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3136205512443136,
+      "grad_norm": 0.5351828932762146,
+      "learning_rate": 0.0004702673452422389,
+      "loss": 0.9856,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3137989474623138,
+      "grad_norm": 0.6115684509277344,
+      "learning_rate": 0.0004702341980417568,
+      "loss": 1.1048,
+      "step": 1759
+    },
+    {
+      "epoch": 0.313977343680314,
+      "grad_norm": 0.5649376511573792,
+      "learning_rate": 0.00047020103354417553,
+      "loss": 1.3311,
+      "step": 1760
+    },
+    {
+      "epoch": 0.31415573989831413,
+      "grad_norm": 0.5363413095474243,
+      "learning_rate": 0.0004701678517520999,
+      "loss": 1.1215,
+      "step": 1761
+    },
+    {
+      "epoch": 0.31433413611631433,
+      "grad_norm": 0.5385435223579407,
+      "learning_rate": 0.00047013465266813594,
+      "loss": 1.0061,
+      "step": 1762
+    },
+    {
+      "epoch": 0.31451253233431453,
+      "grad_norm": 0.6086006760597229,
+      "learning_rate": 0.0004701014362948911,
+      "loss": 1.2379,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3146909285523147,
+      "grad_norm": 0.49953147768974304,
+      "learning_rate": 0.0004700682026349741,
+      "loss": 1.0193,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3148693247703149,
+      "grad_norm": 0.5452679395675659,
+      "learning_rate": 0.0004700349516909952,
+      "loss": 1.2415,
+      "step": 1765
+    },
+    {
+      "epoch": 0.31504772098831507,
+      "grad_norm": 0.5797132253646851,
+      "learning_rate": 0.00047000168346556594,
+      "loss": 1.3095,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3152261172063152,
+      "grad_norm": 0.586613655090332,
+      "learning_rate": 0.00046996839796129903,
+      "loss": 1.0955,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3154045134243154,
+      "grad_norm": 0.5819451808929443,
+      "learning_rate": 0.0004699350951808089,
+      "loss": 1.2054,
+      "step": 1768
+    },
+    {
+      "epoch": 0.31558290964231556,
+      "grad_norm": 0.6394988298416138,
+      "learning_rate": 0.00046990177512671097,
+      "loss": 1.3011,
+      "step": 1769
+    },
+    {
+      "epoch": 0.31576130586031576,
+      "grad_norm": 0.5528479218482971,
+      "learning_rate": 0.00046986843780162223,
+      "loss": 1.3098,
+      "step": 1770
+    },
+    {
+      "epoch": 0.31593970207831595,
+      "grad_norm": 0.5443571209907532,
+      "learning_rate": 0.0004698350832081611,
+      "loss": 1.3117,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3161180982963161,
+      "grad_norm": 0.4671507477760315,
+      "learning_rate": 0.00046980171134894714,
+      "loss": 1.0019,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3162964945143163,
+      "grad_norm": 0.5098404884338379,
+      "learning_rate": 0.00046976832222660127,
+      "loss": 0.9793,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3164748907323165,
+      "grad_norm": 0.7493408918380737,
+      "learning_rate": 0.000469734915843746,
+      "loss": 1.1057,
+      "step": 1774
+    },
+    {
+      "epoch": 0.31665328695031664,
+      "grad_norm": 0.5108188986778259,
+      "learning_rate": 0.00046970149220300496,
+      "loss": 1.0821,
+      "step": 1775
+    },
+    {
+      "epoch": 0.31683168316831684,
+      "grad_norm": 0.4858883321285248,
+      "learning_rate": 0.0004696680513070033,
+      "loss": 0.9252,
+      "step": 1776
+    },
+    {
+      "epoch": 0.31701007938631703,
+      "grad_norm": 0.5160908102989197,
+      "learning_rate": 0.00046963459315836744,
+      "loss": 1.2089,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3171884756043172,
+      "grad_norm": 0.5474086403846741,
+      "learning_rate": 0.0004696011177597251,
+      "loss": 0.9082,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3173668718223174,
+      "grad_norm": 0.5507884621620178,
+      "learning_rate": 0.0004695676251137055,
+      "loss": 1.2827,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3175452680403175,
+      "grad_norm": 0.5097134113311768,
+      "learning_rate": 0.0004695341152229391,
+      "loss": 0.9908,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3177236642583177,
+      "grad_norm": 0.5379379391670227,
+      "learning_rate": 0.00046950058809005775,
+      "loss": 1.1212,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3179020604763179,
+      "grad_norm": 0.5245591998100281,
+      "learning_rate": 0.00046946704371769466,
+      "loss": 1.0469,
+      "step": 1782
+    },
+    {
+      "epoch": 0.31808045669431806,
+      "grad_norm": 0.5721558928489685,
+      "learning_rate": 0.0004694334821084845,
+      "loss": 1.5169,
+      "step": 1783
+    },
+    {
+      "epoch": 0.31825885291231826,
+      "grad_norm": 0.4801085889339447,
+      "learning_rate": 0.00046939990326506296,
+      "loss": 0.9409,
+      "step": 1784
+    },
+    {
+      "epoch": 0.31843724913031846,
+      "grad_norm": 0.5646408796310425,
+      "learning_rate": 0.0004693663071900675,
+      "loss": 1.1511,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3186156453483186,
+      "grad_norm": 0.6158486008644104,
+      "learning_rate": 0.00046933269388613663,
+      "loss": 1.3416,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3187940415663188,
+      "grad_norm": 0.5548638701438904,
+      "learning_rate": 0.0004692990633559104,
+      "loss": 1.0945,
+      "step": 1787
+    },
+    {
+      "epoch": 0.318972437784319,
+      "grad_norm": 0.5320894718170166,
+      "learning_rate": 0.0004692654156020302,
+      "loss": 1.1061,
+      "step": 1788
+    },
+    {
+      "epoch": 0.31915083400231914,
+      "grad_norm": 0.5542237162590027,
+      "learning_rate": 0.0004692317506271385,
+      "loss": 1.2178,
+      "step": 1789
+    },
+    {
+      "epoch": 0.31932923022031934,
+      "grad_norm": 0.49565958976745605,
+      "learning_rate": 0.0004691980684338795,
+      "loss": 0.9038,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3195076264383195,
+      "grad_norm": 0.4972078502178192,
+      "learning_rate": 0.00046916436902489847,
+      "loss": 1.0761,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3196860226563197,
+      "grad_norm": 0.5097312927246094,
+      "learning_rate": 0.00046913065240284226,
+      "loss": 0.9506,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3198644188743199,
+      "grad_norm": 0.5239170789718628,
+      "learning_rate": 0.00046909691857035884,
+      "loss": 1.1015,
+      "step": 1793
+    },
+    {
+      "epoch": 0.32004281509232,
+      "grad_norm": 0.5465912818908691,
+      "learning_rate": 0.0004690631675300978,
+      "loss": 1.04,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3202212113103202,
+      "grad_norm": 0.5346068143844604,
+      "learning_rate": 0.00046902939928470977,
+      "loss": 1.0701,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3203996075283204,
+      "grad_norm": 0.5574382543563843,
+      "learning_rate": 0.00046899561383684707,
+      "loss": 1.3175,
+      "step": 1796
+    },
+    {
+      "epoch": 0.32057800374632056,
+      "grad_norm": 0.5112574100494385,
+      "learning_rate": 0.00046896181118916304,
+      "loss": 0.9372,
+      "step": 1797
+    },
+    {
+      "epoch": 0.32075639996432076,
+      "grad_norm": 0.5132248997688293,
+      "learning_rate": 0.00046892799134431263,
+      "loss": 1.0925,
+      "step": 1798
+    },
+    {
+      "epoch": 0.32093479618232096,
+      "grad_norm": 0.5472792983055115,
+      "learning_rate": 0.000468894154304952,
+      "loss": 0.8669,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3211131924003211,
+      "grad_norm": 0.5460713505744934,
+      "learning_rate": 0.0004688603000737386,
+      "loss": 0.9067,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3212915886183213,
+      "grad_norm": 0.5631927847862244,
+      "learning_rate": 0.00046882642865333146,
+      "loss": 1.0512,
+      "step": 1801
+    },
+    {
+      "epoch": 0.32146998483632144,
+      "grad_norm": 0.5665701627731323,
+      "learning_rate": 0.00046879254004639085,
+      "loss": 1.1049,
+      "step": 1802
+    },
+    {
+      "epoch": 0.32164838105432164,
+      "grad_norm": 0.6018973588943481,
+      "learning_rate": 0.00046875863425557823,
+      "loss": 1.1779,
+      "step": 1803
+    },
+    {
+      "epoch": 0.32182677727232184,
+      "grad_norm": 0.5643223524093628,
+      "learning_rate": 0.00046872471128355664,
+      "loss": 1.3245,
+      "step": 1804
+    },
+    {
+      "epoch": 0.322005173490322,
+      "grad_norm": 0.5305078625679016,
+      "learning_rate": 0.0004686907711329903,
+      "loss": 1.1447,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3221835697083222,
+      "grad_norm": 0.5465999245643616,
+      "learning_rate": 0.000468656813806545,
+      "loss": 1.1083,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3223619659263224,
+      "grad_norm": 0.5518834590911865,
+      "learning_rate": 0.0004686228393068875,
+      "loss": 1.4034,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3225403621443225,
+      "grad_norm": 0.5887888669967651,
+      "learning_rate": 0.0004685888476366864,
+      "loss": 1.2044,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3227187583623227,
+      "grad_norm": 0.5048173069953918,
+      "learning_rate": 0.0004685548387986113,
+      "loss": 1.0136,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3228971545803229,
+      "grad_norm": 0.5357009172439575,
+      "learning_rate": 0.0004685208127953331,
+      "loss": 1.1239,
+      "step": 1810
+    },
+    {
+      "epoch": 0.32307555079832306,
+      "grad_norm": 0.4944056272506714,
+      "learning_rate": 0.00046848676962952434,
+      "loss": 1.1098,
+      "step": 1811
+    },
+    {
+      "epoch": 0.32325394701632326,
+      "grad_norm": 0.5399764776229858,
+      "learning_rate": 0.00046845270930385876,
+      "loss": 1.2677,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3234323432343234,
+      "grad_norm": 0.541485607624054,
+      "learning_rate": 0.0004684186318210114,
+      "loss": 0.9894,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3236107394523236,
+      "grad_norm": 0.4978035092353821,
+      "learning_rate": 0.0004683845371836586,
+      "loss": 1.0137,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3237891356703238,
+      "grad_norm": 0.560565710067749,
+      "learning_rate": 0.0004683504253944783,
+      "loss": 1.1744,
+      "step": 1815
+    },
+    {
+      "epoch": 0.32396753188832395,
+      "grad_norm": 0.6290241479873657,
+      "learning_rate": 0.00046831629645614954,
+      "loss": 1.3022,
+      "step": 1816
+    },
+    {
+      "epoch": 0.32414592810632414,
+      "grad_norm": 0.5639137029647827,
+      "learning_rate": 0.0004682821503713528,
+      "loss": 1.3467,
+      "step": 1817
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.9618589282035828,
+      "learning_rate": 0.0004682479871427699,
+      "loss": 1.1614,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3245027205423245,
+      "grad_norm": 0.5054482817649841,
+      "learning_rate": 0.000468213806773084,
+      "loss": 1.0439,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3246811167603247,
+      "grad_norm": 0.604265570640564,
+      "learning_rate": 0.00046817960926497966,
+      "loss": 1.2124,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3248595129783249,
+      "grad_norm": 0.6163999438285828,
+      "learning_rate": 0.00046814539462114267,
+      "loss": 1.3959,
+      "step": 1821
+    },
+    {
+      "epoch": 0.325037909196325,
+      "grad_norm": 0.5077233910560608,
+      "learning_rate": 0.00046811116284426027,
+      "loss": 0.9646,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3252163054143252,
+      "grad_norm": 1.0014747381210327,
+      "learning_rate": 0.000468076913937021,
+      "loss": 1.2883,
+      "step": 1823
+    },
+    {
+      "epoch": 0.32539470163232537,
+      "grad_norm": 0.5301506519317627,
+      "learning_rate": 0.0004680426479021147,
+      "loss": 1.2094,
+      "step": 1824
+    },
+    {
+      "epoch": 0.32557309785032557,
+      "grad_norm": 0.4948747456073761,
+      "learning_rate": 0.00046800836474223274,
+      "loss": 0.9282,
+      "step": 1825
+    },
+    {
+      "epoch": 0.32575149406832576,
+      "grad_norm": 0.5325037837028503,
+      "learning_rate": 0.0004679740644600676,
+      "loss": 1.316,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3259298902863259,
+      "grad_norm": 1.2007477283477783,
+      "learning_rate": 0.0004679397470583133,
+      "loss": 1.1872,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3261082865043261,
+      "grad_norm": 0.5436804294586182,
+      "learning_rate": 0.00046790541253966493,
+      "loss": 1.32,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3262866827223263,
+      "grad_norm": 1.843799114227295,
+      "learning_rate": 0.0004678710609068193,
+      "loss": 1.0474,
+      "step": 1829
+    },
+    {
+      "epoch": 0.32646507894032645,
+      "grad_norm": 0.5451673269271851,
+      "learning_rate": 0.0004678366921624743,
+      "loss": 0.938,
+      "step": 1830
+    },
+    {
+      "epoch": 0.32664347515832665,
+      "grad_norm": 0.6391348242759705,
+      "learning_rate": 0.0004678023063093293,
+      "loss": 1.0066,
+      "step": 1831
+    },
+    {
+      "epoch": 0.32682187137632684,
+      "grad_norm": 0.8246572613716125,
+      "learning_rate": 0.0004677679033500848,
+      "loss": 1.2115,
+      "step": 1832
+    },
+    {
+      "epoch": 0.327000267594327,
+      "grad_norm": 0.5913776159286499,
+      "learning_rate": 0.0004677334832874429,
+      "loss": 1.0055,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3271786638123272,
+      "grad_norm": 0.7291193008422852,
+      "learning_rate": 0.00046769904612410694,
+      "loss": 1.2235,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3273570600303274,
+      "grad_norm": 0.5839000940322876,
+      "learning_rate": 0.0004676645918627816,
+      "loss": 1.0966,
+      "step": 1835
+    },
+    {
+      "epoch": 0.32753545624832753,
+      "grad_norm": 0.6921905875205994,
+      "learning_rate": 0.0004676301205061728,
+      "loss": 1.0023,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3277138524663277,
+      "grad_norm": 0.5414957404136658,
+      "learning_rate": 0.0004675956320569881,
+      "loss": 1.0105,
+      "step": 1837
+    },
+    {
+      "epoch": 0.32789224868432787,
+      "grad_norm": 0.6554968357086182,
+      "learning_rate": 0.000467561126517936,
+      "loss": 1.3459,
+      "step": 1838
+    },
+    {
+      "epoch": 0.32807064490232807,
+      "grad_norm": 0.6246145367622375,
+      "learning_rate": 0.00046752660389172673,
+      "loss": 1.2114,
+      "step": 1839
+    },
+    {
+      "epoch": 0.32824904112032827,
+      "grad_norm": 0.5389582514762878,
+      "learning_rate": 0.0004674920641810716,
+      "loss": 0.8482,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3284274373383284,
+      "grad_norm": 0.5214718580245972,
+      "learning_rate": 0.0004674575073886833,
+      "loss": 0.8899,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3286058335563286,
+      "grad_norm": 0.6277864575386047,
+      "learning_rate": 0.00046742293351727596,
+      "loss": 1.1186,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3287842297743288,
+      "grad_norm": 0.4919874668121338,
+      "learning_rate": 0.00046738834256956495,
+      "loss": 0.8872,
+      "step": 1843
+    },
+    {
+      "epoch": 0.32896262599232895,
+      "grad_norm": 0.545361340045929,
+      "learning_rate": 0.00046735373454826715,
+      "loss": 1.0465,
+      "step": 1844
+    },
+    {
+      "epoch": 0.32914102221032915,
+      "grad_norm": 1.7328273057937622,
+      "learning_rate": 0.00046731910945610044,
+      "loss": 1.1124,
+      "step": 1845
+    },
+    {
+      "epoch": 0.32931941842832935,
+      "grad_norm": 0.6040815711021423,
+      "learning_rate": 0.0004672844672957846,
+      "loss": 1.1342,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3294978146463295,
+      "grad_norm": 0.4832199513912201,
+      "learning_rate": 0.00046724980807004,
+      "loss": 0.825,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3296762108643297,
+      "grad_norm": 0.7503164410591125,
+      "learning_rate": 0.00046721513178158905,
+      "loss": 0.9456,
+      "step": 1848
+    },
+    {
+      "epoch": 0.32985460708232983,
+      "grad_norm": 0.5634304881095886,
+      "learning_rate": 0.0004671804384331551,
+      "loss": 1.1604,
+      "step": 1849
+    },
+    {
+      "epoch": 0.33003300330033003,
+      "grad_norm": 0.5407414436340332,
+      "learning_rate": 0.00046714572802746305,
+      "loss": 1.0941,
+      "step": 1850
+    },
+    {
+      "epoch": 0.33021139951833023,
+      "grad_norm": 0.5838576555252075,
+      "learning_rate": 0.0004671110005672389,
+      "loss": 1.3654,
+      "step": 1851
+    },
+    {
+      "epoch": 0.33038979573633037,
+      "grad_norm": 0.5957556962966919,
+      "learning_rate": 0.00046707625605521016,
+      "loss": 1.153,
+      "step": 1852
+    },
+    {
+      "epoch": 0.33056819195433057,
+      "grad_norm": 0.5474072098731995,
+      "learning_rate": 0.0004670414944941057,
+      "loss": 1.1032,
+      "step": 1853
+    },
+    {
+      "epoch": 0.33074658817233077,
+      "grad_norm": 0.48801934719085693,
+      "learning_rate": 0.00046700671588665574,
+      "loss": 0.8788,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3309249843903309,
+      "grad_norm": 0.6753059029579163,
+      "learning_rate": 0.00046697192023559167,
+      "loss": 1.2628,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3311033806083311,
+      "grad_norm": 0.5072078108787537,
+      "learning_rate": 0.00046693710754364625,
+      "loss": 0.9892,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3312817768263313,
+      "grad_norm": 0.6533268690109253,
+      "learning_rate": 0.0004669022778135539,
+      "loss": 1.1148,
+      "step": 1857
+    },
+    {
+      "epoch": 0.33146017304433145,
+      "grad_norm": 0.5230735540390015,
+      "learning_rate": 0.0004668674310480499,
+      "loss": 0.8552,
+      "step": 1858
+    },
+    {
+      "epoch": 0.33163856926233165,
+      "grad_norm": 0.5872694253921509,
+      "learning_rate": 0.00046683256724987124,
+      "loss": 1.1711,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3318169654803318,
+      "grad_norm": 0.6011821627616882,
+      "learning_rate": 0.000466797686421756,
+      "loss": 1.1871,
+      "step": 1860
+    },
+    {
+      "epoch": 0.331995361698332,
+      "grad_norm": 0.643999457359314,
+      "learning_rate": 0.0004667627885664438,
+      "loss": 1.0446,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3321737579163322,
+      "grad_norm": 0.5912252068519592,
+      "learning_rate": 0.00046672787368667553,
+      "loss": 1.1304,
+      "step": 1862
+    },
+    {
+      "epoch": 0.33235215413433233,
+      "grad_norm": 0.7412044405937195,
+      "learning_rate": 0.0004666929417851933,
+      "loss": 1.3677,
+      "step": 1863
+    },
+    {
+      "epoch": 0.33253055035233253,
+      "grad_norm": 0.5437285900115967,
+      "learning_rate": 0.0004666579928647406,
+      "loss": 1.039,
+      "step": 1864
+    },
+    {
+      "epoch": 0.33270894657033273,
+      "grad_norm": 0.5716599822044373,
+      "learning_rate": 0.0004666230269280625,
+      "loss": 1.1614,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3328873427883329,
+      "grad_norm": 0.48691585659980774,
+      "learning_rate": 0.00046658804397790496,
+      "loss": 0.9387,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3330657390063331,
+      "grad_norm": 0.5562944412231445,
+      "learning_rate": 0.00046655304401701565,
+      "loss": 1.1568,
+      "step": 1867
+    },
+    {
+      "epoch": 0.33324413522433327,
+      "grad_norm": 0.5540555119514465,
+      "learning_rate": 0.0004665180270481435,
+      "loss": 1.283,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3334225314423334,
+      "grad_norm": 0.44283947348594666,
+      "learning_rate": 0.0004664829930740387,
+      "loss": 0.9378,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3336009276603336,
+      "grad_norm": 0.7194570899009705,
+      "learning_rate": 0.00046644794209745277,
+      "loss": 1.3644,
+      "step": 1870
+    },
+    {
+      "epoch": 0.33377932387833376,
+      "grad_norm": 0.5517498254776001,
+      "learning_rate": 0.00046641287412113857,
+      "loss": 1.2408,
+      "step": 1871
+    },
+    {
+      "epoch": 0.33395772009633395,
+      "grad_norm": 0.504384458065033,
+      "learning_rate": 0.00046637778914785044,
+      "loss": 0.9018,
+      "step": 1872
+    },
+    {
+      "epoch": 0.33413611631433415,
+      "grad_norm": 1.1955726146697998,
+      "learning_rate": 0.0004663426871803438,
+      "loss": 0.9939,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3343145125323343,
+      "grad_norm": 0.5495529174804688,
+      "learning_rate": 0.00046630756822137567,
+      "loss": 1.2006,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3344929087503345,
+      "grad_norm": 0.5532088875770569,
+      "learning_rate": 0.00046627243227370413,
+      "loss": 1.0194,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3346713049683347,
+      "grad_norm": 0.4987606406211853,
+      "learning_rate": 0.00046623727934008886,
+      "loss": 0.9675,
+      "step": 1876
+    },
+    {
+      "epoch": 0.33484970118633484,
+      "grad_norm": 0.6175581216812134,
+      "learning_rate": 0.0004662021094232908,
+      "loss": 1.1119,
+      "step": 1877
+    },
+    {
+      "epoch": 0.33502809740433503,
+      "grad_norm": 0.5381768345832825,
+      "learning_rate": 0.00046616692252607204,
+      "loss": 1.0028,
+      "step": 1878
+    },
+    {
+      "epoch": 0.33520649362233523,
+      "grad_norm": 0.5099619030952454,
+      "learning_rate": 0.00046613171865119617,
+      "loss": 0.9536,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3353848898403354,
+      "grad_norm": 0.5081475377082825,
+      "learning_rate": 0.0004660964978014282,
+      "loss": 1.0749,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3355632860583356,
+      "grad_norm": 0.5307644009590149,
+      "learning_rate": 0.00046606125997953425,
+      "loss": 1.3082,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3357416822763357,
+      "grad_norm": 0.5692281126976013,
+      "learning_rate": 0.0004660260051882819,
+      "loss": 1.2991,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3359200784943359,
+      "grad_norm": 0.5353856682777405,
+      "learning_rate": 0.00046599073343044014,
+      "loss": 1.0892,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3360984747123361,
+      "grad_norm": 0.5255775451660156,
+      "learning_rate": 0.000465955444708779,
+      "loss": 1.2572,
+      "step": 1884
+    },
+    {
+      "epoch": 0.33627687093033626,
+      "grad_norm": 0.5284912586212158,
+      "learning_rate": 0.0004659201390260703,
+      "loss": 1.1411,
+      "step": 1885
+    },
+    {
+      "epoch": 0.33645526714833646,
+      "grad_norm": 0.5372970700263977,
+      "learning_rate": 0.0004658848163850867,
+      "loss": 0.9713,
+      "step": 1886
+    },
+    {
+      "epoch": 0.33663366336633666,
+      "grad_norm": 0.4853118062019348,
+      "learning_rate": 0.00046584947678860264,
+      "loss": 0.839,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3368120595843368,
+      "grad_norm": 0.543500542640686,
+      "learning_rate": 0.0004658141202393935,
+      "loss": 1.2338,
+      "step": 1888
+    },
+    {
+      "epoch": 0.336990455802337,
+      "grad_norm": 0.5506289005279541,
+      "learning_rate": 0.00046577874674023624,
+      "loss": 0.9966,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3371688520203372,
+      "grad_norm": 0.81171053647995,
+      "learning_rate": 0.00046574335629390894,
+      "loss": 0.9698,
+      "step": 1890
+    },
+    {
+      "epoch": 0.33734724823833734,
+      "grad_norm": 0.5851067304611206,
+      "learning_rate": 0.0004657079489031915,
+      "loss": 1.2608,
+      "step": 1891
+    },
+    {
+      "epoch": 0.33752564445633754,
+      "grad_norm": 0.5415493845939636,
+      "learning_rate": 0.0004656725245708644,
+      "loss": 1.2209,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3377040406743377,
+      "grad_norm": 1.5110948085784912,
+      "learning_rate": 0.0004656370832997101,
+      "loss": 1.1676,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3378824368923379,
+      "grad_norm": 0.4858531951904297,
+      "learning_rate": 0.00046560162509251204,
+      "loss": 0.8358,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3380608331103381,
+      "grad_norm": 0.5556557774543762,
+      "learning_rate": 0.00046556614995205516,
+      "loss": 1.1761,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3382392293283382,
+      "grad_norm": 0.5492879748344421,
+      "learning_rate": 0.00046553065788112563,
+      "loss": 1.0874,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3384176255463384,
+      "grad_norm": 0.6402779817581177,
+      "learning_rate": 0.000465495148882511,
+      "loss": 1.186,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3385960217643386,
+      "grad_norm": 0.5240576863288879,
+      "learning_rate": 0.0004654596229590001,
+      "loss": 1.0237,
+      "step": 1898
+    },
+    {
+      "epoch": 0.33877441798233876,
+      "grad_norm": 0.5123072862625122,
+      "learning_rate": 0.0004654240801133831,
+      "loss": 1.2273,
+      "step": 1899
+    },
+    {
+      "epoch": 0.33895281420033896,
+      "grad_norm": 0.5299893617630005,
+      "learning_rate": 0.0004653885203484515,
+      "loss": 1.009,
+      "step": 1900
+    },
+    {
+      "epoch": 0.33913121041833916,
+      "grad_norm": 0.548248291015625,
+      "learning_rate": 0.0004653529436669983,
+      "loss": 1.0276,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3393096066363393,
+      "grad_norm": 0.5219201445579529,
+      "learning_rate": 0.00046531735007181754,
+      "loss": 1.1695,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3394880028543395,
+      "grad_norm": 0.4921947419643402,
+      "learning_rate": 0.00046528173956570474,
+      "loss": 0.9397,
+      "step": 1903
+    },
+    {
+      "epoch": 0.33966639907233964,
+      "grad_norm": 0.6388809680938721,
+      "learning_rate": 0.00046524611215145685,
+      "loss": 1.3252,
+      "step": 1904
+    },
+    {
+      "epoch": 0.33984479529033984,
+      "grad_norm": 0.5055054426193237,
+      "learning_rate": 0.0004652104678318718,
+      "loss": 1.0557,
+      "step": 1905
+    },
+    {
+      "epoch": 0.34002319150834004,
+      "grad_norm": 0.5897945761680603,
+      "learning_rate": 0.0004651748066097493,
+      "loss": 1.3774,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3402015877263402,
+      "grad_norm": 0.4601035714149475,
+      "learning_rate": 0.00046513912848789013,
+      "loss": 0.8346,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3403799839443404,
+      "grad_norm": 0.515330970287323,
+      "learning_rate": 0.00046510343346909636,
+      "loss": 1.0378,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3405583801623406,
+      "grad_norm": 0.5221341252326965,
+      "learning_rate": 0.0004650677215561714,
+      "loss": 1.2198,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3407367763803407,
+      "grad_norm": 0.6407569050788879,
+      "learning_rate": 0.00046503199275192025,
+      "loss": 1.2767,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3409151725983409,
+      "grad_norm": 0.5762472748756409,
+      "learning_rate": 0.0004649962470591489,
+      "loss": 1.2465,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3410935688163411,
+      "grad_norm": 0.514964759349823,
+      "learning_rate": 0.00046496048448066484,
+      "loss": 1.1349,
+      "step": 1912
+    },
+    {
+      "epoch": 0.34127196503434126,
+      "grad_norm": 0.482735276222229,
+      "learning_rate": 0.00046492470501927676,
+      "loss": 0.9251,
+      "step": 1913
+    },
+    {
+      "epoch": 0.34145036125234146,
+      "grad_norm": 0.49272486567497253,
+      "learning_rate": 0.0004648889086777949,
+      "loss": 0.8477,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3416287574703416,
+      "grad_norm": 0.5680924654006958,
+      "learning_rate": 0.00046485309545903064,
+      "loss": 1.2822,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3418071536883418,
+      "grad_norm": 0.7367702722549438,
+      "learning_rate": 0.00046481726536579674,
+      "loss": 1.1015,
+      "step": 1916
+    },
+    {
+      "epoch": 0.341985549906342,
+      "grad_norm": 0.8924338817596436,
+      "learning_rate": 0.0004647814184009072,
+      "loss": 1.113,
+      "step": 1917
+    },
+    {
+      "epoch": 0.34216394612434214,
+      "grad_norm": 0.48184382915496826,
+      "learning_rate": 0.00046474555456717747,
+      "loss": 0.9971,
+      "step": 1918
+    },
+    {
+      "epoch": 0.34234234234234234,
+      "grad_norm": 0.6294741034507751,
+      "learning_rate": 0.0004647096738674243,
+      "loss": 1.1922,
+      "step": 1919
+    },
+    {
+      "epoch": 0.34252073856034254,
+      "grad_norm": 0.5504223108291626,
+      "learning_rate": 0.0004646737763044658,
+      "loss": 1.0714,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3426991347783427,
+      "grad_norm": 0.6702556610107422,
+      "learning_rate": 0.0004646378618811212,
+      "loss": 1.2097,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3428775309963429,
+      "grad_norm": 0.5340100526809692,
+      "learning_rate": 0.0004646019306002114,
+      "loss": 1.2057,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3430559272143431,
+      "grad_norm": 0.49807506799697876,
+      "learning_rate": 0.0004645659824645583,
+      "loss": 0.8643,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3432343234323432,
+      "grad_norm": 0.5424205660820007,
+      "learning_rate": 0.0004645300174769852,
+      "loss": 1.257,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3434127196503434,
+      "grad_norm": 0.9090192317962646,
+      "learning_rate": 0.0004644940356403169,
+      "loss": 1.1699,
+      "step": 1925
+    },
+    {
+      "epoch": 0.34359111586834357,
+      "grad_norm": 0.6018331050872803,
+      "learning_rate": 0.0004644580369573793,
+      "loss": 1.1373,
+      "step": 1926
+    },
+    {
+      "epoch": 0.34376951208634376,
+      "grad_norm": 0.5179253220558167,
+      "learning_rate": 0.00046442202143099986,
+      "loss": 0.8974,
+      "step": 1927
+    },
+    {
+      "epoch": 0.34394790830434396,
+      "grad_norm": 0.5565230846405029,
+      "learning_rate": 0.00046438598906400707,
+      "loss": 1.0234,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3441263045223441,
+      "grad_norm": 0.6058740019798279,
+      "learning_rate": 0.0004643499398592309,
+      "loss": 1.2254,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3443047007403443,
+      "grad_norm": 0.6367807984352112,
+      "learning_rate": 0.0004643138738195028,
+      "loss": 0.9239,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3444830969583445,
+      "grad_norm": 0.49235963821411133,
+      "learning_rate": 0.0004642777909476552,
+      "loss": 1.0013,
+      "step": 1931
+    },
+    {
+      "epoch": 0.34466149317634465,
+      "grad_norm": 0.5881189107894897,
+      "learning_rate": 0.00046424169124652216,
+      "loss": 1.2899,
+      "step": 1932
+    },
+    {
+      "epoch": 0.34483988939434485,
+      "grad_norm": 0.5489835143089294,
+      "learning_rate": 0.00046420557471893887,
+      "loss": 0.8001,
+      "step": 1933
+    },
+    {
+      "epoch": 0.34501828561234504,
+      "grad_norm": 0.5183538198471069,
+      "learning_rate": 0.0004641694413677419,
+      "loss": 0.9889,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3451966818303452,
+      "grad_norm": 0.493448942899704,
+      "learning_rate": 0.0004641332911957692,
+      "loss": 0.9633,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3453750780483454,
+      "grad_norm": 0.6520765423774719,
+      "learning_rate": 0.0004640971242058599,
+      "loss": 1.3168,
+      "step": 1936
+    },
+    {
+      "epoch": 0.34555347426634553,
+      "grad_norm": 0.5311633944511414,
+      "learning_rate": 0.00046406094040085465,
+      "loss": 1.1453,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3457318704843457,
+      "grad_norm": 0.6454037427902222,
+      "learning_rate": 0.0004640247397835953,
+      "loss": 1.2844,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3459102667023459,
+      "grad_norm": 0.5690452456474304,
+      "learning_rate": 0.00046398852235692494,
+      "loss": 1.0391,
+      "step": 1939
+    },
+    {
+      "epoch": 0.34608866292034607,
+      "grad_norm": 0.555842936038971,
+      "learning_rate": 0.0004639522881236881,
+      "loss": 1.3138,
+      "step": 1940
+    },
+    {
+      "epoch": 0.34626705913834627,
+      "grad_norm": 0.5336434841156006,
+      "learning_rate": 0.0004639160370867307,
+      "loss": 1.127,
+      "step": 1941
+    },
+    {
+      "epoch": 0.34644545535634647,
+      "grad_norm": 0.6537164449691772,
+      "learning_rate": 0.00046387976924889976,
+      "loss": 1.3808,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3466238515743466,
+      "grad_norm": 0.5720651149749756,
+      "learning_rate": 0.00046384348461304373,
+      "loss": 1.4115,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3468022477923468,
+      "grad_norm": 0.5567306280136108,
+      "learning_rate": 0.00046380718318201247,
+      "loss": 1.0522,
+      "step": 1944
+    },
+    {
+      "epoch": 0.346980644010347,
+      "grad_norm": 0.658480167388916,
+      "learning_rate": 0.00046377086495865707,
+      "loss": 1.1822,
+      "step": 1945
+    },
+    {
+      "epoch": 0.34715904022834715,
+      "grad_norm": 0.5562303066253662,
+      "learning_rate": 0.00046373452994582994,
+      "loss": 1.2199,
+      "step": 1946
+    },
+    {
+      "epoch": 0.34733743644634735,
+      "grad_norm": 0.585731029510498,
+      "learning_rate": 0.00046369817814638476,
+      "loss": 1.0057,
+      "step": 1947
+    },
+    {
+      "epoch": 0.3475158326643475,
+      "grad_norm": 0.5658949017524719,
+      "learning_rate": 0.0004636618095631767,
+      "loss": 1.1701,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3476942288823477,
+      "grad_norm": 1.204848051071167,
+      "learning_rate": 0.00046362542419906195,
+      "loss": 1.1882,
+      "step": 1949
+    },
+    {
+      "epoch": 0.3478726251003479,
+      "grad_norm": 0.5047358274459839,
+      "learning_rate": 0.0004635890220568984,
+      "loss": 1.0479,
+      "step": 1950
+    },
+    {
+      "epoch": 0.34805102131834803,
+      "grad_norm": 0.6019084453582764,
+      "learning_rate": 0.0004635526031395449,
+      "loss": 1.2855,
+      "step": 1951
+    },
+    {
+      "epoch": 0.34822941753634823,
+      "grad_norm": 0.5449895858764648,
+      "learning_rate": 0.00046351616744986194,
+      "loss": 1.2736,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3484078137543484,
+      "grad_norm": 0.5767678022384644,
+      "learning_rate": 0.00046347971499071104,
+      "loss": 1.0872,
+      "step": 1953
+    },
+    {
+      "epoch": 0.34858620997234857,
+      "grad_norm": 0.589439868927002,
+      "learning_rate": 0.0004634432457649552,
+      "loss": 1.1026,
+      "step": 1954
+    },
+    {
+      "epoch": 0.34876460619034877,
+      "grad_norm": 0.534095823764801,
+      "learning_rate": 0.00046340675977545867,
+      "loss": 1.1419,
+      "step": 1955
+    },
+    {
+      "epoch": 0.34894300240834897,
+      "grad_norm": 0.5593207478523254,
+      "learning_rate": 0.00046337025702508704,
+      "loss": 1.1605,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3491213986263491,
+      "grad_norm": 0.5366277694702148,
+      "learning_rate": 0.00046333373751670734,
+      "loss": 1.039,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3492997948443493,
+      "grad_norm": 0.5536123514175415,
+      "learning_rate": 0.00046329720125318766,
+      "loss": 1.1518,
+      "step": 1958
+    },
+    {
+      "epoch": 0.34947819106234945,
+      "grad_norm": 0.4309159219264984,
+      "learning_rate": 0.0004632606482373976,
+      "loss": 0.8568,
+      "step": 1959
+    },
+    {
+      "epoch": 0.34965658728034965,
+      "grad_norm": 0.7334818243980408,
+      "learning_rate": 0.0004632240784722079,
+      "loss": 1.2216,
+      "step": 1960
+    },
+    {
+      "epoch": 0.34983498349834985,
+      "grad_norm": 0.5555823445320129,
+      "learning_rate": 0.0004631874919604909,
+      "loss": 1.1513,
+      "step": 1961
+    },
+    {
+      "epoch": 0.35001337971635,
+      "grad_norm": 0.5582665205001831,
+      "learning_rate": 0.00046315088870512003,
+      "loss": 1.096,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3501917759343502,
+      "grad_norm": 0.5533860921859741,
+      "learning_rate": 0.00046311426870897,
+      "loss": 1.1806,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3503701721523504,
+      "grad_norm": 0.6270779371261597,
+      "learning_rate": 0.00046307763197491717,
+      "loss": 1.1347,
+      "step": 1964
+    },
+    {
+      "epoch": 0.35054856837035053,
+      "grad_norm": 0.5797653794288635,
+      "learning_rate": 0.0004630409785058387,
+      "loss": 1.176,
+      "step": 1965
+    },
+    {
+      "epoch": 0.35072696458835073,
+      "grad_norm": 0.5614036917686462,
+      "learning_rate": 0.0004630043083046135,
+      "loss": 1.2459,
+      "step": 1966
+    },
+    {
+      "epoch": 0.35090536080635093,
+      "grad_norm": 0.5455766916275024,
+      "learning_rate": 0.0004629676213741216,
+      "loss": 1.2499,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3510837570243511,
+      "grad_norm": 0.5755292177200317,
+      "learning_rate": 0.00046293091771724433,
+      "loss": 0.9262,
+      "step": 1968
+    },
+    {
+      "epoch": 0.35126215324235127,
+      "grad_norm": 0.5601824522018433,
+      "learning_rate": 0.00046289419733686445,
+      "loss": 1.2802,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3514405494603514,
+      "grad_norm": 0.5256120562553406,
+      "learning_rate": 0.00046285746023586586,
+      "loss": 1.0818,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3516189456783516,
+      "grad_norm": 0.5001375675201416,
+      "learning_rate": 0.000462820706417134,
+      "loss": 0.8263,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3517973418963518,
+      "grad_norm": 0.6029314994812012,
+      "learning_rate": 0.0004627839358835554,
+      "loss": 1.1909,
+      "step": 1972
+    },
+    {
+      "epoch": 0.35197573811435195,
+      "grad_norm": 0.5109052658081055,
+      "learning_rate": 0.00046274714863801813,
+      "loss": 0.9954,
+      "step": 1973
+    },
+    {
+      "epoch": 0.35215413433235215,
+      "grad_norm": 0.6349180936813354,
+      "learning_rate": 0.0004627103446834113,
+      "loss": 1.1535,
+      "step": 1974
+    },
+    {
+      "epoch": 0.35233253055035235,
+      "grad_norm": 0.5841493606567383,
+      "learning_rate": 0.0004626735240226255,
+      "loss": 0.9702,
+      "step": 1975
+    },
+    {
+      "epoch": 0.3525109267683525,
+      "grad_norm": 0.5837200880050659,
+      "learning_rate": 0.00046263668665855276,
+      "loss": 1.2831,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3526893229863527,
+      "grad_norm": 0.5340337157249451,
+      "learning_rate": 0.000462599832594086,
+      "loss": 1.0643,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3528677192043529,
+      "grad_norm": 0.5065892338752747,
+      "learning_rate": 0.00046256296183212,
+      "loss": 0.9666,
+      "step": 1978
+    },
+    {
+      "epoch": 0.35304611542235304,
+      "grad_norm": 0.5460461378097534,
+      "learning_rate": 0.00046252607437555037,
+      "loss": 1.0391,
+      "step": 1979
+    },
+    {
+      "epoch": 0.35322451164035323,
+      "grad_norm": 0.5758165717124939,
+      "learning_rate": 0.0004624891702272744,
+      "loss": 1.0593,
+      "step": 1980
+    },
+    {
+      "epoch": 0.3534029078583534,
+      "grad_norm": 0.739166259765625,
+      "learning_rate": 0.0004624522493901904,
+      "loss": 1.0089,
+      "step": 1981
+    },
+    {
+      "epoch": 0.3535813040763536,
+      "grad_norm": 0.5259382724761963,
+      "learning_rate": 0.0004624153118671981,
+      "loss": 0.8894,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3537597002943538,
+      "grad_norm": 0.47911664843559265,
+      "learning_rate": 0.00046237835766119867,
+      "loss": 0.8527,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3539380965123539,
+      "grad_norm": 0.5415805578231812,
+      "learning_rate": 0.0004623413867750944,
+      "loss": 1.1951,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3541164927303541,
+      "grad_norm": 0.5259605050086975,
+      "learning_rate": 0.00046230439921178903,
+      "loss": 1.1803,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3542948889483543,
+      "grad_norm": 0.5450373291969299,
+      "learning_rate": 0.00046226739497418745,
+      "loss": 1.0821,
+      "step": 1986
+    },
+    {
+      "epoch": 0.35447328516635446,
+      "grad_norm": 0.5040337443351746,
+      "learning_rate": 0.0004622303740651961,
+      "loss": 1.0438,
+      "step": 1987
+    },
+    {
+      "epoch": 0.35465168138435466,
+      "grad_norm": 0.531369686126709,
+      "learning_rate": 0.00046219333648772246,
+      "loss": 1.0472,
+      "step": 1988
+    },
+    {
+      "epoch": 0.35483007760235485,
+      "grad_norm": 0.5246288180351257,
+      "learning_rate": 0.00046215628224467546,
+      "loss": 1.0166,
+      "step": 1989
+    },
+    {
+      "epoch": 0.355008473820355,
+      "grad_norm": 0.5832285284996033,
+      "learning_rate": 0.0004621192113389654,
+      "loss": 1.3904,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3551868700383552,
+      "grad_norm": 0.5046032667160034,
+      "learning_rate": 0.0004620821237735037,
+      "loss": 0.9671,
+      "step": 1991
+    },
+    {
+      "epoch": 0.35536526625635534,
+      "grad_norm": 0.505187451839447,
+      "learning_rate": 0.00046204501955120333,
+      "loss": 1.1395,
+      "step": 1992
+    },
+    {
+      "epoch": 0.35554366247435554,
+      "grad_norm": 0.5264232158660889,
+      "learning_rate": 0.00046200789867497836,
+      "loss": 1.1737,
+      "step": 1993
+    },
+    {
+      "epoch": 0.35572205869235574,
+      "grad_norm": 0.6432974338531494,
+      "learning_rate": 0.00046197076114774425,
+      "loss": 1.4899,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3559004549103559,
+      "grad_norm": 0.5268731713294983,
+      "learning_rate": 0.00046193360697241773,
+      "loss": 1.3145,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3560788511283561,
+      "grad_norm": 0.5824031829833984,
+      "learning_rate": 0.000461896436151917,
+      "loss": 1.0268,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3562572473463563,
+      "grad_norm": 0.5190456509590149,
+      "learning_rate": 0.0004618592486891613,
+      "loss": 0.9869,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3564356435643564,
+      "grad_norm": 0.5111797451972961,
+      "learning_rate": 0.0004618220445870714,
+      "loss": 0.8515,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3566140397823566,
+      "grad_norm": 0.6227758526802063,
+      "learning_rate": 0.0004617848238485693,
+      "loss": 1.2615,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3567924360003568,
+      "grad_norm": 0.5635523200035095,
+      "learning_rate": 0.0004617475864765782,
+      "loss": 1.1813,
+      "step": 2000
+    },
+    {
+      "epoch": 0.35697083221835696,
+      "grad_norm": 0.6283045411109924,
+      "learning_rate": 0.0004617103324740228,
+      "loss": 1.0593,
+      "step": 2001
+    },
+    {
+      "epoch": 0.35714922843635716,
+      "grad_norm": 0.5376402735710144,
+      "learning_rate": 0.00046167306184382906,
+      "loss": 0.9929,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3573276246543573,
+      "grad_norm": 0.5043478608131409,
+      "learning_rate": 0.00046163577458892415,
+      "loss": 1.1796,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3575060208723575,
+      "grad_norm": 0.5074126720428467,
+      "learning_rate": 0.00046159847071223646,
+      "loss": 0.8408,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3576844170903577,
+      "grad_norm": 0.6142756938934326,
+      "learning_rate": 0.00046156115021669605,
+      "loss": 1.159,
+      "step": 2005
+    },
+    {
+      "epoch": 0.35786281330835784,
+      "grad_norm": 0.4879589378833771,
+      "learning_rate": 0.00046152381310523384,
+      "loss": 0.8117,
+      "step": 2006
+    },
+    {
+      "epoch": 0.35804120952635804,
+      "grad_norm": 0.5088807940483093,
+      "learning_rate": 0.0004614864593807825,
+      "loss": 1.0544,
+      "step": 2007
+    },
+    {
+      "epoch": 0.35821960574435824,
+      "grad_norm": 0.5416921377182007,
+      "learning_rate": 0.00046144908904627557,
+      "loss": 1.1541,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3583980019623584,
+      "grad_norm": 0.5295444130897522,
+      "learning_rate": 0.0004614117021046482,
+      "loss": 1.1063,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3585763981803586,
+      "grad_norm": 0.5223750472068787,
+      "learning_rate": 0.0004613742985588367,
+      "loss": 1.0807,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3587547943983588,
+      "grad_norm": 0.5627544522285461,
+      "learning_rate": 0.0004613368784117787,
+      "loss": 1.0165,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3589331906163589,
+      "grad_norm": 0.5210590958595276,
+      "learning_rate": 0.0004612994416664134,
+      "loss": 1.0561,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3591115868343591,
+      "grad_norm": 0.5011810660362244,
+      "learning_rate": 0.00046126198832568065,
+      "loss": 1.175,
+      "step": 2013
+    },
+    {
+      "epoch": 0.35928998305235926,
+      "grad_norm": 0.47756415605545044,
+      "learning_rate": 0.0004612245183925224,
+      "loss": 1.052,
+      "step": 2014
+    },
+    {
+      "epoch": 0.35946837927035946,
+      "grad_norm": 0.5775620937347412,
+      "learning_rate": 0.0004611870318698813,
+      "loss": 1.0646,
+      "step": 2015
+    },
+    {
+      "epoch": 0.35964677548835966,
+      "grad_norm": 0.5540309548377991,
+      "learning_rate": 0.00046114952876070167,
+      "loss": 1.2858,
+      "step": 2016
+    },
+    {
+      "epoch": 0.3598251717063598,
+      "grad_norm": 0.5066989660263062,
+      "learning_rate": 0.0004611120090679289,
+      "loss": 0.9978,
+      "step": 2017
+    },
+    {
+      "epoch": 0.36000356792436,
+      "grad_norm": 1.06368887424469,
+      "learning_rate": 0.00046107447279450977,
+      "loss": 1.2269,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3601819641423602,
+      "grad_norm": 0.5219504833221436,
+      "learning_rate": 0.00046103691994339236,
+      "loss": 0.9725,
+      "step": 2019
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 0.5700012445449829,
+      "learning_rate": 0.00046099935051752605,
+      "loss": 1.2358,
+      "step": 2020
+    },
+    {
+      "epoch": 0.36053875657836054,
+      "grad_norm": 0.514971911907196,
+      "learning_rate": 0.0004609617645198616,
+      "loss": 1.0931,
+      "step": 2021
+    },
+    {
+      "epoch": 0.36071715279636074,
+      "grad_norm": 0.5299590229988098,
+      "learning_rate": 0.00046092416195335094,
+      "loss": 1.1742,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3608955490143609,
+      "grad_norm": 0.5127617120742798,
+      "learning_rate": 0.00046088654282094735,
+      "loss": 1.0115,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3610739452323611,
+      "grad_norm": 0.5403224229812622,
+      "learning_rate": 0.00046084890712560546,
+      "loss": 1.2271,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3612523414503612,
+      "grad_norm": 0.5359017252922058,
+      "learning_rate": 0.0004608112548702811,
+      "loss": 1.1065,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3614307376683614,
+      "grad_norm": 0.5404418110847473,
+      "learning_rate": 0.00046077358605793156,
+      "loss": 1.0741,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3616091338863616,
+      "grad_norm": 0.541814386844635,
+      "learning_rate": 0.0004607359006915153,
+      "loss": 1.2468,
+      "step": 2027
+    },
+    {
+      "epoch": 0.36178753010436177,
+      "grad_norm": 0.5587732791900635,
+      "learning_rate": 0.000460698198773992,
+      "loss": 1.0027,
+      "step": 2028
+    },
+    {
+      "epoch": 0.36196592632236196,
+      "grad_norm": 0.5569438338279724,
+      "learning_rate": 0.00046066048030832284,
+      "loss": 1.045,
+      "step": 2029
+    },
+    {
+      "epoch": 0.36214432254036216,
+      "grad_norm": 0.5397847294807434,
+      "learning_rate": 0.0004606227452974703,
+      "loss": 1.1198,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3623227187583623,
+      "grad_norm": 0.4841654598712921,
+      "learning_rate": 0.0004605849937443979,
+      "loss": 1.0022,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3625011149763625,
+      "grad_norm": 0.5069999694824219,
+      "learning_rate": 0.00046054722565207077,
+      "loss": 1.1917,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3626795111943627,
+      "grad_norm": 0.5341327786445618,
+      "learning_rate": 0.0004605094410234551,
+      "loss": 0.9297,
+      "step": 2033
+    },
+    {
+      "epoch": 0.36285790741236285,
+      "grad_norm": 0.4698069095611572,
+      "learning_rate": 0.00046047163986151854,
+      "loss": 0.7909,
+      "step": 2034
+    },
+    {
+      "epoch": 0.36303630363036304,
+      "grad_norm": 0.6396573185920715,
+      "learning_rate": 0.0004604338221692299,
+      "loss": 1.26,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3632146998483632,
+      "grad_norm": 0.5331511497497559,
+      "learning_rate": 0.0004603959879495595,
+      "loss": 1.0438,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3633930960663634,
+      "grad_norm": 0.6419988870620728,
+      "learning_rate": 0.00046035813720547876,
+      "loss": 1.232,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3635714922843636,
+      "grad_norm": 0.5257839560508728,
+      "learning_rate": 0.0004603202699399604,
+      "loss": 0.8447,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3637498885023637,
+      "grad_norm": 0.49433434009552,
+      "learning_rate": 0.0004602823861559785,
+      "loss": 0.9731,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3639282847203639,
+      "grad_norm": 0.5541132688522339,
+      "learning_rate": 0.00046024448585650857,
+      "loss": 1.0478,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3641066809383641,
+      "grad_norm": 0.5009471774101257,
+      "learning_rate": 0.00046020656904452716,
+      "loss": 1.0364,
+      "step": 2041
+    },
+    {
+      "epoch": 0.36428507715636427,
+      "grad_norm": 0.5421225428581238,
+      "learning_rate": 0.0004601686357230123,
+      "loss": 1.0979,
+      "step": 2042
+    },
+    {
+      "epoch": 0.36446347337436447,
+      "grad_norm": 0.49714601039886475,
+      "learning_rate": 0.0004601306858949432,
+      "loss": 0.9815,
+      "step": 2043
+    },
+    {
+      "epoch": 0.36464186959236466,
+      "grad_norm": 0.5659156441688538,
+      "learning_rate": 0.0004600927195633005,
+      "loss": 1.0267,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3648202658103648,
+      "grad_norm": 0.5777513980865479,
+      "learning_rate": 0.000460054736731066,
+      "loss": 1.1066,
+      "step": 2045
+    },
+    {
+      "epoch": 0.364998662028365,
+      "grad_norm": 0.4904979169368744,
+      "learning_rate": 0.00046001673740122287,
+      "loss": 1.0805,
+      "step": 2046
+    },
+    {
+      "epoch": 0.36517705824636515,
+      "grad_norm": 0.5406635403633118,
+      "learning_rate": 0.0004599787215767556,
+      "loss": 0.9333,
+      "step": 2047
+    },
+    {
+      "epoch": 0.36535545446436535,
+      "grad_norm": 0.5242641568183899,
+      "learning_rate": 0.0004599406892606498,
+      "loss": 0.9977,
+      "step": 2048
+    },
+    {
+      "epoch": 0.36553385068236555,
+      "grad_norm": 0.538411021232605,
+      "learning_rate": 0.00045990264045589263,
+      "loss": 1.0446,
+      "step": 2049
+    },
+    {
+      "epoch": 0.3657122469003657,
+      "grad_norm": 0.5017095804214478,
+      "learning_rate": 0.0004598645751654724,
+      "loss": 0.9025,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3658906431183659,
+      "grad_norm": 0.5771719813346863,
+      "learning_rate": 0.0004598264933923788,
+      "loss": 1.1323,
+      "step": 2051
+    },
+    {
+      "epoch": 0.3660690393363661,
+      "grad_norm": 0.5927982926368713,
+      "learning_rate": 0.0004597883951396027,
+      "loss": 1.4047,
+      "step": 2052
+    },
+    {
+      "epoch": 0.36624743555436623,
+      "grad_norm": 0.5014179944992065,
+      "learning_rate": 0.00045975028041013633,
+      "loss": 0.9932,
+      "step": 2053
+    },
+    {
+      "epoch": 0.36642583177236643,
+      "grad_norm": 0.5538795590400696,
+      "learning_rate": 0.00045971214920697325,
+      "loss": 1.128,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3666042279903666,
+      "grad_norm": 0.5837451219558716,
+      "learning_rate": 0.00045967400153310814,
+      "loss": 1.1844,
+      "step": 2055
+    },
+    {
+      "epoch": 0.36678262420836677,
+      "grad_norm": 0.4952576458454132,
+      "learning_rate": 0.00045963583739153715,
+      "loss": 0.9711,
+      "step": 2056
+    },
+    {
+      "epoch": 0.36696102042636697,
+      "grad_norm": 0.4757169187068939,
+      "learning_rate": 0.00045959765678525776,
+      "loss": 1.0064,
+      "step": 2057
+    },
+    {
+      "epoch": 0.36713941664436717,
+      "grad_norm": 0.5288332104682922,
+      "learning_rate": 0.00045955945971726855,
+      "loss": 1.4253,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3673178128623673,
+      "grad_norm": 0.5419899225234985,
+      "learning_rate": 0.00045952124619056957,
+      "loss": 1.2097,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3674962090803675,
+      "grad_norm": 0.5925987958908081,
+      "learning_rate": 0.00045948301620816215,
+      "loss": 1.1952,
+      "step": 2060
+    },
+    {
+      "epoch": 0.36767460529836765,
+      "grad_norm": 0.5184533596038818,
+      "learning_rate": 0.00045944476977304873,
+      "loss": 1.0255,
+      "step": 2061
+    },
+    {
+      "epoch": 0.36785300151636785,
+      "grad_norm": 0.5200556516647339,
+      "learning_rate": 0.0004594065068882332,
+      "loss": 1.1104,
+      "step": 2062
+    },
+    {
+      "epoch": 0.36803139773436805,
+      "grad_norm": 0.5943669080734253,
+      "learning_rate": 0.0004593682275567208,
+      "loss": 1.1162,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3682097939523682,
+      "grad_norm": 1.1349503993988037,
+      "learning_rate": 0.00045932993178151785,
+      "loss": 1.0188,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3683881901703684,
+      "grad_norm": 0.5693401098251343,
+      "learning_rate": 0.00045929161956563216,
+      "loss": 1.0564,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3685665863883686,
+      "grad_norm": 0.5493849515914917,
+      "learning_rate": 0.00045925329091207266,
+      "loss": 0.9755,
+      "step": 2066
+    },
+    {
+      "epoch": 0.36874498260636873,
+      "grad_norm": 0.5377547740936279,
+      "learning_rate": 0.00045921494582384985,
+      "loss": 1.0967,
+      "step": 2067
+    },
+    {
+      "epoch": 0.36892337882436893,
+      "grad_norm": 0.5204565525054932,
+      "learning_rate": 0.00045917658430397526,
+      "loss": 0.9365,
+      "step": 2068
+    },
+    {
+      "epoch": 0.36910177504236913,
+      "grad_norm": 0.5643702149391174,
+      "learning_rate": 0.0004591382063554617,
+      "loss": 1.2203,
+      "step": 2069
+    },
+    {
+      "epoch": 0.36928017126036927,
+      "grad_norm": 0.49709245562553406,
+      "learning_rate": 0.00045909981198132336,
+      "loss": 0.9819,
+      "step": 2070
+    },
+    {
+      "epoch": 0.36945856747836947,
+      "grad_norm": 0.5403639674186707,
+      "learning_rate": 0.00045906140118457577,
+      "loss": 1.0955,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3696369636963696,
+      "grad_norm": 0.643334686756134,
+      "learning_rate": 0.00045902297396823577,
+      "loss": 1.3619,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3698153599143698,
+      "grad_norm": 0.5306541919708252,
+      "learning_rate": 0.0004589845303353213,
+      "loss": 0.9215,
+      "step": 2073
+    },
+    {
+      "epoch": 0.36999375613237,
+      "grad_norm": 0.5806733965873718,
+      "learning_rate": 0.00045894607028885173,
+      "loss": 1.0277,
+      "step": 2074
+    },
+    {
+      "epoch": 0.37017215235037015,
+      "grad_norm": 0.5571582317352295,
+      "learning_rate": 0.0004589075938318478,
+      "loss": 0.845,
+      "step": 2075
+    },
+    {
+      "epoch": 0.37035054856837035,
+      "grad_norm": 0.56407231092453,
+      "learning_rate": 0.00045886910096733127,
+      "loss": 1.0461,
+      "step": 2076
+    },
+    {
+      "epoch": 0.37052894478637055,
+      "grad_norm": 0.5526430010795593,
+      "learning_rate": 0.0004588305916983254,
+      "loss": 0.915,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3707073410043707,
+      "grad_norm": 0.5798369646072388,
+      "learning_rate": 0.00045879206602785486,
+      "loss": 1.0021,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3708857372223709,
+      "grad_norm": 0.5774779915809631,
+      "learning_rate": 0.0004587535239589452,
+      "loss": 1.1147,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3710641334403711,
+      "grad_norm": 0.5789695978164673,
+      "learning_rate": 0.00045871496549462364,
+      "loss": 1.1118,
+      "step": 2080
+    },
+    {
+      "epoch": 0.37124252965837123,
+      "grad_norm": 0.4506717622280121,
+      "learning_rate": 0.0004586763906379184,
+      "loss": 0.8033,
+      "step": 2081
+    },
+    {
+      "epoch": 0.37142092587637143,
+      "grad_norm": 0.4841929078102112,
+      "learning_rate": 0.0004586377993918594,
+      "loss": 1.1138,
+      "step": 2082
+    },
+    {
+      "epoch": 0.3715993220943716,
+      "grad_norm": 0.5651434659957886,
+      "learning_rate": 0.0004585991917594774,
+      "loss": 1.1166,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3717777183123718,
+      "grad_norm": 0.5341058969497681,
+      "learning_rate": 0.0004585605677438046,
+      "loss": 1.315,
+      "step": 2084
+    },
+    {
+      "epoch": 0.371956114530372,
+      "grad_norm": 1.9402390718460083,
+      "learning_rate": 0.0004585219273478745,
+      "loss": 1.2177,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3721345107483721,
+      "grad_norm": 0.5437238812446594,
+      "learning_rate": 0.00045848327057472206,
+      "loss": 0.9626,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3723129069663723,
+      "grad_norm": 0.5549441576004028,
+      "learning_rate": 0.0004584445974273832,
+      "loss": 1.0923,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3724913031843725,
+      "grad_norm": 0.5701561570167542,
+      "learning_rate": 0.0004584059079088954,
+      "loss": 1.1516,
+      "step": 2088
+    },
+    {
+      "epoch": 0.37266969940237266,
+      "grad_norm": 0.5430149435997009,
+      "learning_rate": 0.00045836720202229727,
+      "loss": 1.2877,
+      "step": 2089
+    },
+    {
+      "epoch": 0.37284809562037285,
+      "grad_norm": 0.555594801902771,
+      "learning_rate": 0.00045832847977062875,
+      "loss": 0.9959,
+      "step": 2090
+    },
+    {
+      "epoch": 0.37302649183837305,
+      "grad_norm": 0.5235658288002014,
+      "learning_rate": 0.0004582897411569311,
+      "loss": 1.0502,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3732048880563732,
+      "grad_norm": 0.6449779868125916,
+      "learning_rate": 0.0004582509861842468,
+      "loss": 1.1183,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3733832842743734,
+      "grad_norm": 0.592603325843811,
+      "learning_rate": 0.0004582122148556196,
+      "loss": 1.1536,
+      "step": 2093
+    },
+    {
+      "epoch": 0.37356168049237354,
+      "grad_norm": 0.4949917197227478,
+      "learning_rate": 0.0004581734271740948,
+      "loss": 1.0221,
+      "step": 2094
+    },
+    {
+      "epoch": 0.37374007671037374,
+      "grad_norm": 0.6672910451889038,
+      "learning_rate": 0.0004581346231427185,
+      "loss": 1.064,
+      "step": 2095
+    },
+    {
+      "epoch": 0.37391847292837394,
+      "grad_norm": 0.4738200604915619,
+      "learning_rate": 0.00045809580276453843,
+      "loss": 1.0283,
+      "step": 2096
+    },
+    {
+      "epoch": 0.3740968691463741,
+      "grad_norm": 1.385906457901001,
+      "learning_rate": 0.00045805696604260364,
+      "loss": 0.9435,
+      "step": 2097
+    },
+    {
+      "epoch": 0.3742752653643743,
+      "grad_norm": 0.6020336747169495,
+      "learning_rate": 0.0004580181129799643,
+      "loss": 1.0175,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3744536615823745,
+      "grad_norm": 0.7528032660484314,
+      "learning_rate": 0.0004579792435796718,
+      "loss": 1.2405,
+      "step": 2099
+    },
+    {
+      "epoch": 0.3746320578003746,
+      "grad_norm": 0.5541552305221558,
+      "learning_rate": 0.00045794035784477903,
+      "loss": 1.0247,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3748104540183748,
+      "grad_norm": 0.5061083436012268,
+      "learning_rate": 0.00045790145577834007,
+      "loss": 0.8521,
+      "step": 2101
+    },
+    {
+      "epoch": 0.374988850236375,
+      "grad_norm": 1.0418980121612549,
+      "learning_rate": 0.0004578625373834102,
+      "loss": 1.3496,
+      "step": 2102
+    },
+    {
+      "epoch": 0.37516724645437516,
+      "grad_norm": 0.5358394384384155,
+      "learning_rate": 0.00045782360266304615,
+      "loss": 0.9427,
+      "step": 2103
+    },
+    {
+      "epoch": 0.37534564267237536,
+      "grad_norm": 0.5550572276115417,
+      "learning_rate": 0.0004577846516203057,
+      "loss": 1.29,
+      "step": 2104
+    },
+    {
+      "epoch": 0.3755240388903755,
+      "grad_norm": 0.5666816830635071,
+      "learning_rate": 0.0004577456842582482,
+      "loss": 0.9784,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3757024351083757,
+      "grad_norm": 0.6263391375541687,
+      "learning_rate": 0.000457706700579934,
+      "loss": 1.3776,
+      "step": 2106
+    },
+    {
+      "epoch": 0.3758808313263759,
+      "grad_norm": 0.5329514145851135,
+      "learning_rate": 0.000457667700588425,
+      "loss": 0.9403,
+      "step": 2107
+    },
+    {
+      "epoch": 0.37605922754437604,
+      "grad_norm": 0.6623142957687378,
+      "learning_rate": 0.00045762868428678405,
+      "loss": 1.1171,
+      "step": 2108
+    },
+    {
+      "epoch": 0.37623762376237624,
+      "grad_norm": 0.5798327326774597,
+      "learning_rate": 0.00045758965167807565,
+      "loss": 1.1791,
+      "step": 2109
+    },
+    {
+      "epoch": 0.37641601998037644,
+      "grad_norm": 0.5603600144386292,
+      "learning_rate": 0.00045755060276536543,
+      "loss": 1.0435,
+      "step": 2110
+    },
+    {
+      "epoch": 0.3765944161983766,
+      "grad_norm": 0.5544630885124207,
+      "learning_rate": 0.00045751153755172006,
+      "loss": 1.1783,
+      "step": 2111
+    },
+    {
+      "epoch": 0.3767728124163768,
+      "grad_norm": 0.633995771408081,
+      "learning_rate": 0.0004574724560402078,
+      "loss": 1.1487,
+      "step": 2112
+    },
+    {
+      "epoch": 0.376951208634377,
+      "grad_norm": 0.6015441417694092,
+      "learning_rate": 0.00045743335823389826,
+      "loss": 1.0372,
+      "step": 2113
+    },
+    {
+      "epoch": 0.3771296048523771,
+      "grad_norm": 0.5980955958366394,
+      "learning_rate": 0.00045739424413586194,
+      "loss": 1.2749,
+      "step": 2114
+    },
+    {
+      "epoch": 0.3773080010703773,
+      "grad_norm": 0.5361245274543762,
+      "learning_rate": 0.00045735511374917095,
+      "loss": 1.0022,
+      "step": 2115
+    },
+    {
+      "epoch": 0.37748639728837746,
+      "grad_norm": 0.5869659781455994,
+      "learning_rate": 0.0004573159670768986,
+      "loss": 1.214,
+      "step": 2116
+    },
+    {
+      "epoch": 0.37766479350637766,
+      "grad_norm": 0.6182494759559631,
+      "learning_rate": 0.00045727680412211937,
+      "loss": 0.9911,
+      "step": 2117
+    },
+    {
+      "epoch": 0.37784318972437786,
+      "grad_norm": 0.5352988243103027,
+      "learning_rate": 0.0004572376248879091,
+      "loss": 1.0801,
+      "step": 2118
+    },
+    {
+      "epoch": 0.378021585942378,
+      "grad_norm": 0.5704354047775269,
+      "learning_rate": 0.000457198429377345,
+      "loss": 0.9513,
+      "step": 2119
+    },
+    {
+      "epoch": 0.3781999821603782,
+      "grad_norm": 0.5778263807296753,
+      "learning_rate": 0.00045715921759350545,
+      "loss": 1.1593,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 0.5286616683006287,
+      "learning_rate": 0.0004571199895394701,
+      "loss": 1.2545,
+      "step": 2121
+    },
+    {
+      "epoch": 0.37855677459637854,
+      "grad_norm": 0.5010552406311035,
+      "learning_rate": 0.00045708074521831984,
+      "loss": 0.8689,
+      "step": 2122
+    },
+    {
+      "epoch": 0.37873517081437874,
+      "grad_norm": 0.8215789198875427,
+      "learning_rate": 0.000457041484633137,
+      "loss": 1.1025,
+      "step": 2123
+    },
+    {
+      "epoch": 0.37891356703237894,
+      "grad_norm": 0.8685697317123413,
+      "learning_rate": 0.000457002207787005,
+      "loss": 1.0335,
+      "step": 2124
+    },
+    {
+      "epoch": 0.3790919632503791,
+      "grad_norm": 0.5928163528442383,
+      "learning_rate": 0.00045696291468300874,
+      "loss": 1.1333,
+      "step": 2125
+    },
+    {
+      "epoch": 0.3792703594683793,
+      "grad_norm": 0.5363757014274597,
+      "learning_rate": 0.00045692360532423423,
+      "loss": 1.3034,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3794487556863794,
+      "grad_norm": 0.6252255439758301,
+      "learning_rate": 0.00045688427971376876,
+      "loss": 1.1585,
+      "step": 2127
+    },
+    {
+      "epoch": 0.3796271519043796,
+      "grad_norm": 0.6210330128669739,
+      "learning_rate": 0.00045684493785470105,
+      "loss": 0.9278,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3798055481223798,
+      "grad_norm": 0.8249781131744385,
+      "learning_rate": 0.00045680557975012086,
+      "loss": 1.4528,
+      "step": 2129
+    },
+    {
+      "epoch": 0.37998394434037996,
+      "grad_norm": 0.5406894683837891,
+      "learning_rate": 0.00045676620540311953,
+      "loss": 0.9968,
+      "step": 2130
+    },
+    {
+      "epoch": 0.38016234055838016,
+      "grad_norm": 0.631753146648407,
+      "learning_rate": 0.00045672681481678936,
+      "loss": 1.05,
+      "step": 2131
+    },
+    {
+      "epoch": 0.38034073677638036,
+      "grad_norm": 0.5768477916717529,
+      "learning_rate": 0.000456687407994224,
+      "loss": 1.1387,
+      "step": 2132
+    },
+    {
+      "epoch": 0.3805191329943805,
+      "grad_norm": 0.5343456268310547,
+      "learning_rate": 0.00045664798493851873,
+      "loss": 1.0884,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3806975292123807,
+      "grad_norm": 0.5370036363601685,
+      "learning_rate": 0.00045660854565276955,
+      "loss": 1.0014,
+      "step": 2134
+    },
+    {
+      "epoch": 0.3808759254303809,
+      "grad_norm": 0.5855540633201599,
+      "learning_rate": 0.0004565690901400742,
+      "loss": 1.2283,
+      "step": 2135
+    },
+    {
+      "epoch": 0.38105432164838104,
+      "grad_norm": 0.5994966626167297,
+      "learning_rate": 0.00045652961840353135,
+      "loss": 0.801,
+      "step": 2136
+    },
+    {
+      "epoch": 0.38123271786638124,
+      "grad_norm": 0.7402109503746033,
+      "learning_rate": 0.0004564901304462411,
+      "loss": 0.9229,
+      "step": 2137
+    },
+    {
+      "epoch": 0.3814111140843814,
+      "grad_norm": 0.8662267327308655,
+      "learning_rate": 0.0004564506262713049,
+      "loss": 1.0499,
+      "step": 2138
+    },
+    {
+      "epoch": 0.3815895103023816,
+      "grad_norm": 0.5050637125968933,
+      "learning_rate": 0.0004564111058818254,
+      "loss": 0.9188,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3817679065203818,
+      "grad_norm": 0.5128204822540283,
+      "learning_rate": 0.0004563715692809064,
+      "loss": 0.9959,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3819463027383819,
+      "grad_norm": 0.5767316818237305,
+      "learning_rate": 0.00045633201647165325,
+      "loss": 1.1759,
+      "step": 2141
+    },
+    {
+      "epoch": 0.3821246989563821,
+      "grad_norm": 0.5560585856437683,
+      "learning_rate": 0.0004562924474571722,
+      "loss": 1.2575,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3823030951743823,
+      "grad_norm": 0.5354772210121155,
+      "learning_rate": 0.0004562528622405712,
+      "loss": 1.0975,
+      "step": 2143
+    },
+    {
+      "epoch": 0.38248149139238247,
+      "grad_norm": 0.5775118470191956,
+      "learning_rate": 0.0004562132608249591,
+      "loss": 1.0457,
+      "step": 2144
+    },
+    {
+      "epoch": 0.38265988761038267,
+      "grad_norm": 0.5361456274986267,
+      "learning_rate": 0.00045617364321344625,
+      "loss": 1.0162,
+      "step": 2145
+    },
+    {
+      "epoch": 0.38283828382838286,
+      "grad_norm": 0.6208562254905701,
+      "learning_rate": 0.00045613400940914417,
+      "loss": 1.0852,
+      "step": 2146
+    },
+    {
+      "epoch": 0.383016680046383,
+      "grad_norm": 0.5321409106254578,
+      "learning_rate": 0.0004560943594151657,
+      "loss": 1.0978,
+      "step": 2147
+    },
+    {
+      "epoch": 0.3831950762643832,
+      "grad_norm": 0.4865065813064575,
+      "learning_rate": 0.00045605469323462493,
+      "loss": 0.8982,
+      "step": 2148
+    },
+    {
+      "epoch": 0.38337347248238335,
+      "grad_norm": 0.5791817903518677,
+      "learning_rate": 0.0004560150108706372,
+      "loss": 1.0734,
+      "step": 2149
+    },
+    {
+      "epoch": 0.38355186870038355,
+      "grad_norm": 0.5767939686775208,
+      "learning_rate": 0.0004559753123263193,
+      "loss": 1.112,
+      "step": 2150
+    },
+    {
+      "epoch": 0.38373026491838375,
+      "grad_norm": 0.5261261463165283,
+      "learning_rate": 0.0004559355976047889,
+      "loss": 1.0849,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3839086611363839,
+      "grad_norm": 0.5370740294456482,
+      "learning_rate": 0.00045589586670916526,
+      "loss": 0.8849,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3840870573543841,
+      "grad_norm": 0.4868060350418091,
+      "learning_rate": 0.0004558561196425689,
+      "loss": 0.9771,
+      "step": 2153
+    },
+    {
+      "epoch": 0.3842654535723843,
+      "grad_norm": 0.5003844499588013,
+      "learning_rate": 0.0004558163564081215,
+      "loss": 0.9858,
+      "step": 2154
+    },
+    {
+      "epoch": 0.38444384979038443,
+      "grad_norm": 0.4873872399330139,
+      "learning_rate": 0.000455776577008946,
+      "loss": 1.082,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3846222460083846,
+      "grad_norm": 0.5657439231872559,
+      "learning_rate": 0.0004557367814481668,
+      "loss": 1.1766,
+      "step": 2156
+    },
+    {
+      "epoch": 0.3848006422263848,
+      "grad_norm": 0.5301082134246826,
+      "learning_rate": 0.00045569696972890916,
+      "loss": 0.9749,
+      "step": 2157
+    },
+    {
+      "epoch": 0.38497903844438497,
+      "grad_norm": 0.5159010291099548,
+      "learning_rate": 0.0004556571418543001,
+      "loss": 1.0766,
+      "step": 2158
+    },
+    {
+      "epoch": 0.38515743466238517,
+      "grad_norm": 0.5375173687934875,
+      "learning_rate": 0.00045561729782746767,
+      "loss": 1.0563,
+      "step": 2159
+    },
+    {
+      "epoch": 0.3853358308803853,
+      "grad_norm": 0.5973328351974487,
+      "learning_rate": 0.0004555774376515411,
+      "loss": 1.3044,
+      "step": 2160
+    },
+    {
+      "epoch": 0.3855142270983855,
+      "grad_norm": 0.4792105555534363,
+      "learning_rate": 0.00045553756132965105,
+      "loss": 0.8944,
+      "step": 2161
+    },
+    {
+      "epoch": 0.3856926233163857,
+      "grad_norm": 0.5174294114112854,
+      "learning_rate": 0.00045549766886492945,
+      "loss": 1.0259,
+      "step": 2162
+    },
+    {
+      "epoch": 0.38587101953438585,
+      "grad_norm": 0.48922815918922424,
+      "learning_rate": 0.0004554577602605093,
+      "loss": 0.8496,
+      "step": 2163
+    },
+    {
+      "epoch": 0.38604941575238605,
+      "grad_norm": 0.49794623255729675,
+      "learning_rate": 0.0004554178355195251,
+      "loss": 0.8427,
+      "step": 2164
+    },
+    {
+      "epoch": 0.38622781197038625,
+      "grad_norm": 0.5163681507110596,
+      "learning_rate": 0.00045537789464511247,
+      "loss": 1.0703,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3864062081883864,
+      "grad_norm": 0.5273317694664001,
+      "learning_rate": 0.00045533793764040845,
+      "loss": 1.0474,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3865846044063866,
+      "grad_norm": 0.563606858253479,
+      "learning_rate": 0.0004552979645085511,
+      "loss": 1.3796,
+      "step": 2167
+    },
+    {
+      "epoch": 0.3867630006243868,
+      "grad_norm": 0.5003736615180969,
+      "learning_rate": 0.0004552579752526799,
+      "loss": 0.8776,
+      "step": 2168
+    },
+    {
+      "epoch": 0.38694139684238693,
+      "grad_norm": 0.5471118688583374,
+      "learning_rate": 0.0004552179698759358,
+      "loss": 1.2153,
+      "step": 2169
+    },
+    {
+      "epoch": 0.38711979306038713,
+      "grad_norm": 0.5452420115470886,
+      "learning_rate": 0.0004551779483814605,
+      "loss": 1.1194,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3872981892783873,
+      "grad_norm": 0.5115416049957275,
+      "learning_rate": 0.0004551379107723976,
+      "loss": 1.2556,
+      "step": 2171
+    },
+    {
+      "epoch": 0.38747658549638747,
+      "grad_norm": 0.5057587027549744,
+      "learning_rate": 0.0004550978570518913,
+      "loss": 0.9426,
+      "step": 2172
+    },
+    {
+      "epoch": 0.38765498171438767,
+      "grad_norm": 0.5084266662597656,
+      "learning_rate": 0.00045505778722308764,
+      "loss": 0.9323,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3878333779323878,
+      "grad_norm": 0.5315088629722595,
+      "learning_rate": 0.00045501770128913364,
+      "loss": 1.1016,
+      "step": 2174
+    },
+    {
+      "epoch": 0.388011774150388,
+      "grad_norm": 0.6192378401756287,
+      "learning_rate": 0.0004549775992531776,
+      "loss": 1.3612,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3881901703683882,
+      "grad_norm": 0.49113729596138,
+      "learning_rate": 0.0004549374811183692,
+      "loss": 0.984,
+      "step": 2176
+    },
+    {
+      "epoch": 0.38836856658638835,
+      "grad_norm": 0.5000951886177063,
+      "learning_rate": 0.0004548973468878591,
+      "loss": 1.0827,
+      "step": 2177
+    },
+    {
+      "epoch": 0.38854696280438855,
+      "grad_norm": 0.5333259701728821,
+      "learning_rate": 0.00045485719656479957,
+      "loss": 1.0745,
+      "step": 2178
+    },
+    {
+      "epoch": 0.38872535902238875,
+      "grad_norm": 0.5190644860267639,
+      "learning_rate": 0.000454817030152344,
+      "loss": 1.1067,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3889037552403889,
+      "grad_norm": 0.5464503765106201,
+      "learning_rate": 0.00045477684765364703,
+      "loss": 1.1484,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3890821514583891,
+      "grad_norm": 0.5542097091674805,
+      "learning_rate": 0.0004547366490718645,
+      "loss": 1.1321,
+      "step": 2181
+    },
+    {
+      "epoch": 0.38926054767638923,
+      "grad_norm": 0.5173625349998474,
+      "learning_rate": 0.0004546964344101537,
+      "loss": 1.2251,
+      "step": 2182
+    },
+    {
+      "epoch": 0.38943894389438943,
+      "grad_norm": 0.5054041743278503,
+      "learning_rate": 0.0004546562036716732,
+      "loss": 0.9788,
+      "step": 2183
+    },
+    {
+      "epoch": 0.38961734011238963,
+      "grad_norm": 0.48386964201927185,
+      "learning_rate": 0.0004546159568595823,
+      "loss": 1.0653,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3897957363303898,
+      "grad_norm": 0.5080329775810242,
+      "learning_rate": 0.00045457569397704226,
+      "loss": 1.1174,
+      "step": 2185
+    },
+    {
+      "epoch": 0.38997413254839,
+      "grad_norm": 1.7605923414230347,
+      "learning_rate": 0.0004545354150272153,
+      "loss": 1.0394,
+      "step": 2186
+    },
+    {
+      "epoch": 0.39015252876639017,
+      "grad_norm": 0.45817258954048157,
+      "learning_rate": 0.0004544951200132648,
+      "loss": 0.8228,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3903309249843903,
+      "grad_norm": 0.5302280783653259,
+      "learning_rate": 0.00045445480893835567,
+      "loss": 1.1017,
+      "step": 2188
+    },
+    {
+      "epoch": 0.3905093212023905,
+      "grad_norm": 0.5459249019622803,
+      "learning_rate": 0.0004544144818056537,
+      "loss": 1.2306,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3906877174203907,
+      "grad_norm": 0.6084957122802734,
+      "learning_rate": 0.0004543741386183264,
+      "loss": 1.0647,
+      "step": 2190
+    },
+    {
+      "epoch": 0.39086611363839086,
+      "grad_norm": 0.6476156711578369,
+      "learning_rate": 0.00045433377937954215,
+      "loss": 1.2749,
+      "step": 2191
+    },
+    {
+      "epoch": 0.39104450985639105,
+      "grad_norm": 0.5217106938362122,
+      "learning_rate": 0.00045429340409247084,
+      "loss": 0.9928,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3912229060743912,
+      "grad_norm": 0.5622875690460205,
+      "learning_rate": 0.0004542530127602834,
+      "loss": 1.0067,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3914013022923914,
+      "grad_norm": 1.472040057182312,
+      "learning_rate": 0.00045421260538615235,
+      "loss": 1.1993,
+      "step": 2194
+    },
+    {
+      "epoch": 0.3915796985103916,
+      "grad_norm": 0.5135729908943176,
+      "learning_rate": 0.00045417218197325106,
+      "loss": 0.9331,
+      "step": 2195
+    },
+    {
+      "epoch": 0.39175809472839174,
+      "grad_norm": 0.662834644317627,
+      "learning_rate": 0.00045413174252475455,
+      "loss": 1.1606,
+      "step": 2196
+    },
+    {
+      "epoch": 0.39193649094639194,
+      "grad_norm": 0.9087008237838745,
+      "learning_rate": 0.00045409128704383873,
+      "loss": 1.2032,
+      "step": 2197
+    },
+    {
+      "epoch": 0.39211488716439213,
+      "grad_norm": 0.5726129412651062,
+      "learning_rate": 0.0004540508155336811,
+      "loss": 1.1915,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3922932833823923,
+      "grad_norm": 0.6241844296455383,
+      "learning_rate": 0.0004540103279974602,
+      "loss": 1.2713,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3924716796003925,
+      "grad_norm": 0.5253695249557495,
+      "learning_rate": 0.000453969824438356,
+      "loss": 0.9503,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3926500758183927,
+      "grad_norm": 0.4767884612083435,
+      "learning_rate": 0.00045392930485954955,
+      "loss": 0.9004,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3928284720363928,
+      "grad_norm": 0.5421171188354492,
+      "learning_rate": 0.0004538887692642232,
+      "loss": 1.057,
+      "step": 2202
+    },
+    {
+      "epoch": 0.393006868254393,
+      "grad_norm": 0.6273959279060364,
+      "learning_rate": 0.0004538482176555607,
+      "loss": 1.0529,
+      "step": 2203
+    },
+    {
+      "epoch": 0.39318526447239316,
+      "grad_norm": 0.5183041095733643,
+      "learning_rate": 0.00045380765003674684,
+      "loss": 1.0245,
+      "step": 2204
+    },
+    {
+      "epoch": 0.39336366069039336,
+      "grad_norm": 0.5289639830589294,
+      "learning_rate": 0.00045376706641096786,
+      "loss": 1.1824,
+      "step": 2205
+    },
+    {
+      "epoch": 0.39354205690839356,
+      "grad_norm": 0.520853579044342,
+      "learning_rate": 0.00045372646678141127,
+      "loss": 1.0964,
+      "step": 2206
+    },
+    {
+      "epoch": 0.3937204531263937,
+      "grad_norm": 0.6077725291252136,
+      "learning_rate": 0.0004536858511512656,
+      "loss": 1.155,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3938988493443939,
+      "grad_norm": 0.5906806588172913,
+      "learning_rate": 0.0004536452195237208,
+      "loss": 1.3433,
+      "step": 2208
+    },
+    {
+      "epoch": 0.3940772455623941,
+      "grad_norm": 1.7862430810928345,
+      "learning_rate": 0.00045360457190196814,
+      "loss": 1.2213,
+      "step": 2209
+    },
+    {
+      "epoch": 0.39425564178039424,
+      "grad_norm": 0.530116856098175,
+      "learning_rate": 0.0004535639082892,
+      "loss": 1.0639,
+      "step": 2210
+    },
+    {
+      "epoch": 0.39443403799839444,
+      "grad_norm": 0.5256516933441162,
+      "learning_rate": 0.00045352322868861004,
+      "loss": 1.1793,
+      "step": 2211
+    },
+    {
+      "epoch": 0.39461243421639464,
+      "grad_norm": 0.46268409490585327,
+      "learning_rate": 0.00045348253310339336,
+      "loss": 0.9103,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3947908304343948,
+      "grad_norm": 0.5721630454063416,
+      "learning_rate": 0.0004534418215367461,
+      "loss": 1.3037,
+      "step": 2213
+    },
+    {
+      "epoch": 0.394969226652395,
+      "grad_norm": 0.550315797328949,
+      "learning_rate": 0.0004534010939918657,
+      "loss": 1.0648,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3951476228703951,
+      "grad_norm": 0.48966914415359497,
+      "learning_rate": 0.0004533603504719509,
+      "loss": 0.9247,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3953260190883953,
+      "grad_norm": 0.5156149864196777,
+      "learning_rate": 0.0004533195909802017,
+      "loss": 1.1591,
+      "step": 2216
+    },
+    {
+      "epoch": 0.3955044153063955,
+      "grad_norm": 0.5864170789718628,
+      "learning_rate": 0.0004532788155198193,
+      "loss": 0.967,
+      "step": 2217
+    },
+    {
+      "epoch": 0.39568281152439566,
+      "grad_norm": 0.4873606562614441,
+      "learning_rate": 0.00045323802409400626,
+      "loss": 0.9788,
+      "step": 2218
+    },
+    {
+      "epoch": 0.39586120774239586,
+      "grad_norm": 0.5299291014671326,
+      "learning_rate": 0.00045319721670596623,
+      "loss": 1.16,
+      "step": 2219
+    },
+    {
+      "epoch": 0.39603960396039606,
+      "grad_norm": 0.5208775997161865,
+      "learning_rate": 0.00045315639335890423,
+      "loss": 1.1583,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3962180001783962,
+      "grad_norm": 0.5344860553741455,
+      "learning_rate": 0.00045311555405602656,
+      "loss": 1.0034,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 0.5340254306793213,
+      "learning_rate": 0.00045307469880054063,
+      "loss": 1.0769,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3965747926143966,
+      "grad_norm": 0.48012998700141907,
+      "learning_rate": 0.00045303382759565524,
+      "loss": 0.9031,
+      "step": 2223
+    },
+    {
+      "epoch": 0.39675318883239674,
+      "grad_norm": 0.511830747127533,
+      "learning_rate": 0.0004529929404445805,
+      "loss": 0.9729,
+      "step": 2224
+    },
+    {
+      "epoch": 0.39693158505039694,
+      "grad_norm": 0.5484462976455688,
+      "learning_rate": 0.00045295203735052747,
+      "loss": 0.9895,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3971099812683971,
+      "grad_norm": 0.5423020720481873,
+      "learning_rate": 0.0004529111183167088,
+      "loss": 1.1713,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3972883774863973,
+      "grad_norm": 0.5270297527313232,
+      "learning_rate": 0.00045287018334633824,
+      "loss": 1.0229,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3974667737043975,
+      "grad_norm": 0.5546413064002991,
+      "learning_rate": 0.00045282923244263076,
+      "loss": 1.2354,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3976451699223976,
+      "grad_norm": 0.5324910283088684,
+      "learning_rate": 0.00045278826560880267,
+      "loss": 1.225,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3978235661403978,
+      "grad_norm": 0.48066118359565735,
+      "learning_rate": 0.00045274728284807144,
+      "loss": 0.8824,
+      "step": 2230
+    },
+    {
+      "epoch": 0.398001962358398,
+      "grad_norm": 0.5693602561950684,
+      "learning_rate": 0.00045270628416365586,
+      "loss": 1.0692,
+      "step": 2231
+    },
+    {
+      "epoch": 0.39818035857639816,
+      "grad_norm": 0.5462480783462524,
+      "learning_rate": 0.00045266526955877595,
+      "loss": 1.0537,
+      "step": 2232
+    },
+    {
+      "epoch": 0.39835875479439836,
+      "grad_norm": 0.510565459728241,
+      "learning_rate": 0.000452624239036653,
+      "loss": 1.0072,
+      "step": 2233
+    },
+    {
+      "epoch": 0.39853715101239856,
+      "grad_norm": 0.4951585829257965,
+      "learning_rate": 0.0004525831926005095,
+      "loss": 0.9697,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3987155472303987,
+      "grad_norm": 0.49053049087524414,
+      "learning_rate": 0.0004525421302535693,
+      "loss": 0.9513,
+      "step": 2235
+    },
+    {
+      "epoch": 0.3988939434483989,
+      "grad_norm": 1.259932041168213,
+      "learning_rate": 0.0004525010519990572,
+      "loss": 0.9762,
+      "step": 2236
+    },
+    {
+      "epoch": 0.39907233966639905,
+      "grad_norm": 0.5876604914665222,
+      "learning_rate": 0.00045245995784019973,
+      "loss": 1.1267,
+      "step": 2237
+    },
+    {
+      "epoch": 0.39925073588439924,
+      "grad_norm": 2.0633914470672607,
+      "learning_rate": 0.00045241884778022423,
+      "loss": 1.0779,
+      "step": 2238
+    },
+    {
+      "epoch": 0.39942913210239944,
+      "grad_norm": 0.4787190854549408,
+      "learning_rate": 0.0004523777218223596,
+      "loss": 0.9998,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3996075283203996,
+      "grad_norm": 2.782982110977173,
+      "learning_rate": 0.0004523365799698358,
+      "loss": 1.0978,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3997859245383998,
+      "grad_norm": 0.5323127508163452,
+      "learning_rate": 0.00045229542222588405,
+      "loss": 1.0,
+      "step": 2241
+    },
+    {
+      "epoch": 0.3999643207564,
+      "grad_norm": 0.49209773540496826,
+      "learning_rate": 0.0004522542485937369,
+      "loss": 0.9856,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4001427169744001,
+      "grad_norm": 0.9652454257011414,
+      "learning_rate": 0.000452213059076628,
+      "loss": 1.0638,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4003211131924003,
+      "grad_norm": 0.5381683111190796,
+      "learning_rate": 0.00045217185367779265,
+      "loss": 1.0115,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4004995094104005,
+      "grad_norm": 0.48560935258865356,
+      "learning_rate": 0.0004521306324004668,
+      "loss": 1.0621,
+      "step": 2245
+    },
+    {
+      "epoch": 0.40067790562840067,
+      "grad_norm": 0.5368553996086121,
+      "learning_rate": 0.00045208939524788805,
+      "loss": 0.9944,
+      "step": 2246
+    },
+    {
+      "epoch": 0.40085630184640086,
+      "grad_norm": 0.5911176204681396,
+      "learning_rate": 0.00045204814222329513,
+      "loss": 1.1529,
+      "step": 2247
+    },
+    {
+      "epoch": 0.401034698064401,
+      "grad_norm": 0.4951234459877014,
+      "learning_rate": 0.0004520068733299282,
+      "loss": 1.0468,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4012130942824012,
+      "grad_norm": 0.5037215948104858,
+      "learning_rate": 0.0004519655885710283,
+      "loss": 0.9121,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4013914905004014,
+      "grad_norm": 0.6309484243392944,
+      "learning_rate": 0.000451924287949838,
+      "loss": 1.2108,
+      "step": 2250
+    },
+    {
+      "epoch": 0.40156988671840155,
+      "grad_norm": 0.5557146072387695,
+      "learning_rate": 0.000451882971469601,
+      "loss": 1.1713,
+      "step": 2251
+    },
+    {
+      "epoch": 0.40174828293640175,
+      "grad_norm": 0.8298324346542358,
+      "learning_rate": 0.0004518416391335623,
+      "loss": 1.1566,
+      "step": 2252
+    },
+    {
+      "epoch": 0.40192667915440194,
+      "grad_norm": 4.0279459953308105,
+      "learning_rate": 0.00045180029094496813,
+      "loss": 1.1309,
+      "step": 2253
+    },
+    {
+      "epoch": 0.4021050753724021,
+      "grad_norm": 0.5673984885215759,
+      "learning_rate": 0.0004517589269070659,
+      "loss": 1.0474,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4022834715904023,
+      "grad_norm": 0.4788002669811249,
+      "learning_rate": 0.0004517175470231044,
+      "loss": 0.8327,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4024618678084025,
+      "grad_norm": 0.549028217792511,
+      "learning_rate": 0.0004516761512963337,
+      "loss": 1.1673,
+      "step": 2256
+    },
+    {
+      "epoch": 0.40264026402640263,
+      "grad_norm": 0.7566819787025452,
+      "learning_rate": 0.0004516347397300047,
+      "loss": 1.103,
+      "step": 2257
+    },
+    {
+      "epoch": 0.4028186602444028,
+      "grad_norm": 0.5571449995040894,
+      "learning_rate": 0.00045159331232737,
+      "loss": 1.4132,
+      "step": 2258
+    },
+    {
+      "epoch": 0.40299705646240297,
+      "grad_norm": 0.47221845388412476,
+      "learning_rate": 0.00045155186909168345,
+      "loss": 0.8425,
+      "step": 2259
+    },
+    {
+      "epoch": 0.40317545268040317,
+      "grad_norm": 0.956231415271759,
+      "learning_rate": 0.0004515104100261997,
+      "loss": 1.0242,
+      "step": 2260
+    },
+    {
+      "epoch": 0.40335384889840337,
+      "grad_norm": 0.5245352387428284,
+      "learning_rate": 0.0004514689351341751,
+      "loss": 1.2424,
+      "step": 2261
+    },
+    {
+      "epoch": 0.4035322451164035,
+      "grad_norm": 0.5289427638053894,
+      "learning_rate": 0.00045142744441886706,
+      "loss": 0.9463,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4037106413344037,
+      "grad_norm": 0.5312831997871399,
+      "learning_rate": 0.00045138593788353424,
+      "loss": 1.0078,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4038890375524039,
+      "grad_norm": 0.5654627680778503,
+      "learning_rate": 0.00045134441553143647,
+      "loss": 1.2575,
+      "step": 2264
+    },
+    {
+      "epoch": 0.40406743377040405,
+      "grad_norm": 0.5695852637290955,
+      "learning_rate": 0.00045130287736583493,
+      "loss": 1.0575,
+      "step": 2265
+    },
+    {
+      "epoch": 0.40424582998840425,
+      "grad_norm": 0.5789728164672852,
+      "learning_rate": 0.00045126132338999203,
+      "loss": 1.0646,
+      "step": 2266
+    },
+    {
+      "epoch": 0.40442422620640445,
+      "grad_norm": 0.526038408279419,
+      "learning_rate": 0.0004512197536071715,
+      "loss": 1.035,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4046026224244046,
+      "grad_norm": 0.537186324596405,
+      "learning_rate": 0.000451178168020638,
+      "loss": 1.0141,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4047810186424048,
+      "grad_norm": 0.5389789342880249,
+      "learning_rate": 0.0004511365666336578,
+      "loss": 1.0923,
+      "step": 2269
+    },
+    {
+      "epoch": 0.404959414860405,
+      "grad_norm": 0.5258745551109314,
+      "learning_rate": 0.00045109494944949827,
+      "loss": 1.277,
+      "step": 2270
+    },
+    {
+      "epoch": 0.40513781107840513,
+      "grad_norm": 0.5116296410560608,
+      "learning_rate": 0.0004510533164714278,
+      "loss": 0.9776,
+      "step": 2271
+    },
+    {
+      "epoch": 0.40531620729640533,
+      "grad_norm": 0.4564659297466278,
+      "learning_rate": 0.0004510116677027165,
+      "loss": 1.0638,
+      "step": 2272
+    },
+    {
+      "epoch": 0.40549460351440547,
+      "grad_norm": 0.5655273199081421,
+      "learning_rate": 0.00045097000314663527,
+      "loss": 1.0859,
+      "step": 2273
+    },
+    {
+      "epoch": 0.40567299973240567,
+      "grad_norm": 0.5012032389640808,
+      "learning_rate": 0.0004509283228064565,
+      "loss": 0.9221,
+      "step": 2274
+    },
+    {
+      "epoch": 0.40585139595040587,
+      "grad_norm": 0.48415374755859375,
+      "learning_rate": 0.00045088662668545375,
+      "loss": 1.0236,
+      "step": 2275
+    },
+    {
+      "epoch": 0.406029792168406,
+      "grad_norm": 0.47426506876945496,
+      "learning_rate": 0.00045084491478690177,
+      "loss": 1.0458,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4062081883864062,
+      "grad_norm": 0.4942657947540283,
+      "learning_rate": 0.0004508031871140765,
+      "loss": 1.1705,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4063865846044064,
+      "grad_norm": 0.5193724036216736,
+      "learning_rate": 0.0004507614436702555,
+      "loss": 1.0264,
+      "step": 2278
+    },
+    {
+      "epoch": 0.40656498082240655,
+      "grad_norm": 0.4963577091693878,
+      "learning_rate": 0.000450719684458717,
+      "loss": 0.8334,
+      "step": 2279
+    },
+    {
+      "epoch": 0.40674337704040675,
+      "grad_norm": 1.6158602237701416,
+      "learning_rate": 0.0004506779094827409,
+      "loss": 1.128,
+      "step": 2280
+    },
+    {
+      "epoch": 0.40692177325840695,
+      "grad_norm": 1.3288167715072632,
+      "learning_rate": 0.00045063611874560815,
+      "loss": 1.006,
+      "step": 2281
+    },
+    {
+      "epoch": 0.4071001694764071,
+      "grad_norm": 0.5889169573783875,
+      "learning_rate": 0.000450594312250601,
+      "loss": 1.0914,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4072785656944073,
+      "grad_norm": 0.562699019908905,
+      "learning_rate": 0.00045055249000100283,
+      "loss": 1.0634,
+      "step": 2283
+    },
+    {
+      "epoch": 0.40745696191240743,
+      "grad_norm": 0.5045656561851501,
+      "learning_rate": 0.00045051065200009844,
+      "loss": 1.0903,
+      "step": 2284
+    },
+    {
+      "epoch": 0.40763535813040763,
+      "grad_norm": 0.5693420767784119,
+      "learning_rate": 0.0004504687982511737,
+      "loss": 1.0265,
+      "step": 2285
+    },
+    {
+      "epoch": 0.40781375434840783,
+      "grad_norm": 0.556315004825592,
+      "learning_rate": 0.00045042692875751585,
+      "loss": 1.0908,
+      "step": 2286
+    },
+    {
+      "epoch": 0.407992150566408,
+      "grad_norm": 1.0736392736434937,
+      "learning_rate": 0.00045038504352241324,
+      "loss": 1.2133,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4081705467844082,
+      "grad_norm": 0.6816602349281311,
+      "learning_rate": 0.00045034314254915555,
+      "loss": 0.9846,
+      "step": 2288
+    },
+    {
+      "epoch": 0.40834894300240837,
+      "grad_norm": 0.6575304865837097,
+      "learning_rate": 0.0004503012258410336,
+      "loss": 1.014,
+      "step": 2289
+    },
+    {
+      "epoch": 0.4085273392204085,
+      "grad_norm": 4.394046306610107,
+      "learning_rate": 0.00045025929340133963,
+      "loss": 1.0483,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4087057354384087,
+      "grad_norm": 1.7720059156417847,
+      "learning_rate": 0.000450217345233367,
+      "loss": 0.9815,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4088841316564089,
+      "grad_norm": 0.8023939728736877,
+      "learning_rate": 0.00045017538134041013,
+      "loss": 0.9606,
+      "step": 2292
+    },
+    {
+      "epoch": 0.40906252787440905,
+      "grad_norm": 0.7400693893432617,
+      "learning_rate": 0.000450133401725765,
+      "loss": 1.2293,
+      "step": 2293
+    },
+    {
+      "epoch": 0.40924092409240925,
+      "grad_norm": 0.6982590556144714,
+      "learning_rate": 0.0004500914063927286,
+      "loss": 1.056,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4094193203104094,
+      "grad_norm": 0.7190932035446167,
+      "learning_rate": 0.00045004939534459923,
+      "loss": 1.333,
+      "step": 2295
+    },
+    {
+      "epoch": 0.4095977165284096,
+      "grad_norm": 0.5682306885719299,
+      "learning_rate": 0.0004500073685846765,
+      "loss": 1.0389,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4097761127464098,
+      "grad_norm": 0.5392178297042847,
+      "learning_rate": 0.0004499653261162611,
+      "loss": 0.8848,
+      "step": 2297
+    },
+    {
+      "epoch": 0.40995450896440994,
+      "grad_norm": 0.7155362963676453,
+      "learning_rate": 0.0004499232679426549,
+      "loss": 1.0928,
+      "step": 2298
+    },
+    {
+      "epoch": 0.41013290518241013,
+      "grad_norm": 0.5469278693199158,
+      "learning_rate": 0.00044988119406716144,
+      "loss": 1.0405,
+      "step": 2299
+    },
+    {
+      "epoch": 0.41031130140041033,
+      "grad_norm": 0.5402587056159973,
+      "learning_rate": 0.000449839104493085,
+      "loss": 0.8677,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4104896976184105,
+      "grad_norm": 0.5472204089164734,
+      "learning_rate": 0.0004497969992237312,
+      "loss": 0.8406,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4106680938364107,
+      "grad_norm": 0.529851496219635,
+      "learning_rate": 0.00044975487826240715,
+      "loss": 1.058,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4108464900544109,
+      "grad_norm": 0.5373932719230652,
+      "learning_rate": 0.00044971274161242084,
+      "loss": 1.0915,
+      "step": 2303
+    },
+    {
+      "epoch": 0.411024886272411,
+      "grad_norm": 0.5543463826179504,
+      "learning_rate": 0.0004496705892770818,
+      "loss": 1.0177,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4112032824904112,
+      "grad_norm": 0.5836469531059265,
+      "learning_rate": 0.0004496284212597006,
+      "loss": 0.9446,
+      "step": 2305
+    },
+    {
+      "epoch": 0.41138167870841136,
+      "grad_norm": 0.6079695820808411,
+      "learning_rate": 0.00044958623756358905,
+      "loss": 0.9742,
+      "step": 2306
+    },
+    {
+      "epoch": 0.41156007492641156,
+      "grad_norm": 0.4826619625091553,
+      "learning_rate": 0.00044954403819206037,
+      "loss": 1.0292,
+      "step": 2307
+    },
+    {
+      "epoch": 0.41173847114441176,
+      "grad_norm": 0.6502671241760254,
+      "learning_rate": 0.00044950182314842875,
+      "loss": 1.0641,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4119168673624119,
+      "grad_norm": 0.4860910475254059,
+      "learning_rate": 0.0004494595924360098,
+      "loss": 0.9264,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4120952635804121,
+      "grad_norm": 0.6258044838905334,
+      "learning_rate": 0.00044941734605812033,
+      "loss": 0.9347,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4122736597984123,
+      "grad_norm": 0.5750169157981873,
+      "learning_rate": 0.00044937508401807826,
+      "loss": 1.2065,
+      "step": 2311
+    },
+    {
+      "epoch": 0.41245205601641244,
+      "grad_norm": 0.560260534286499,
+      "learning_rate": 0.0004493328063192029,
+      "loss": 1.2164,
+      "step": 2312
+    },
+    {
+      "epoch": 0.41263045223441264,
+      "grad_norm": 0.5267819166183472,
+      "learning_rate": 0.0004492905129648147,
+      "loss": 1.1227,
+      "step": 2313
+    },
+    {
+      "epoch": 0.41280884845241284,
+      "grad_norm": 0.5849962830543518,
+      "learning_rate": 0.0004492482039582354,
+      "loss": 1.2503,
+      "step": 2314
+    },
+    {
+      "epoch": 0.412987244670413,
+      "grad_norm": 0.5180212259292603,
+      "learning_rate": 0.00044920587930278796,
+      "loss": 0.92,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4131656408884132,
+      "grad_norm": 0.4263732135295868,
+      "learning_rate": 0.0004491635390017964,
+      "loss": 0.8326,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4133440371064133,
+      "grad_norm": 0.5409104824066162,
+      "learning_rate": 0.0004491211830585862,
+      "loss": 1.1766,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4135224333244135,
+      "grad_norm": 0.5557611584663391,
+      "learning_rate": 0.000449078811476484,
+      "loss": 1.0369,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4137008295424137,
+      "grad_norm": 0.5291704535484314,
+      "learning_rate": 0.0004490364242588176,
+      "loss": 1.024,
+      "step": 2319
+    },
+    {
+      "epoch": 0.41387922576041386,
+      "grad_norm": 0.4819130301475525,
+      "learning_rate": 0.0004489940214089161,
+      "loss": 1.0032,
+      "step": 2320
+    },
+    {
+      "epoch": 0.41405762197841406,
+      "grad_norm": 0.5592676401138306,
+      "learning_rate": 0.0004489516029301098,
+      "loss": 1.1959,
+      "step": 2321
+    },
+    {
+      "epoch": 0.41423601819641426,
+      "grad_norm": 0.612910270690918,
+      "learning_rate": 0.0004489091688257303,
+      "loss": 1.1389,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4144144144144144,
+      "grad_norm": 0.47932949662208557,
+      "learning_rate": 0.00044886671909911014,
+      "loss": 0.9523,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4145928106324146,
+      "grad_norm": 0.49534985423088074,
+      "learning_rate": 0.0004488242537535835,
+      "loss": 1.1314,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4147712068504148,
+      "grad_norm": 0.4879145920276642,
+      "learning_rate": 0.00044878177279248553,
+      "loss": 1.0072,
+      "step": 2325
+    },
+    {
+      "epoch": 0.41494960306841494,
+      "grad_norm": 0.6173017621040344,
+      "learning_rate": 0.00044873927621915267,
+      "loss": 1.3328,
+      "step": 2326
+    },
+    {
+      "epoch": 0.41512799928641514,
+      "grad_norm": 0.4791456162929535,
+      "learning_rate": 0.00044869676403692254,
+      "loss": 1.0026,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4153063955044153,
+      "grad_norm": 0.4833122491836548,
+      "learning_rate": 0.0004486542362491341,
+      "loss": 0.9301,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4154847917224155,
+      "grad_norm": 0.6367433667182922,
+      "learning_rate": 0.00044861169285912746,
+      "loss": 1.2782,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4156631879404157,
+      "grad_norm": 0.47810354828834534,
+      "learning_rate": 0.0004485691338702439,
+      "loss": 0.8764,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4158415841584158,
+      "grad_norm": 0.5335472226142883,
+      "learning_rate": 0.000448526559285826,
+      "loss": 1.033,
+      "step": 2331
+    },
+    {
+      "epoch": 0.416019980376416,
+      "grad_norm": 0.5508213043212891,
+      "learning_rate": 0.00044848396910921763,
+      "loss": 1.2068,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4161983765944162,
+      "grad_norm": 0.5242919325828552,
+      "learning_rate": 0.00044844136334376366,
+      "loss": 1.1176,
+      "step": 2333
+    },
+    {
+      "epoch": 0.41637677281241636,
+      "grad_norm": 0.49891209602355957,
+      "learning_rate": 0.0004483987419928104,
+      "loss": 0.8453,
+      "step": 2334
+    },
+    {
+      "epoch": 0.41655516903041656,
+      "grad_norm": 0.4879932999610901,
+      "learning_rate": 0.0004483561050597054,
+      "loss": 0.8443,
+      "step": 2335
+    },
+    {
+      "epoch": 0.41673356524841676,
+      "grad_norm": 0.49955081939697266,
+      "learning_rate": 0.00044831345254779724,
+      "loss": 0.9135,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4169119614664169,
+      "grad_norm": 0.5931764841079712,
+      "learning_rate": 0.0004482707844604359,
+      "loss": 0.8997,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4170903576844171,
+      "grad_norm": 0.882546603679657,
+      "learning_rate": 0.00044822810080097245,
+      "loss": 1.1035,
+      "step": 2338
+    },
+    {
+      "epoch": 0.41726875390241724,
+      "grad_norm": 0.5271819233894348,
+      "learning_rate": 0.00044818540157275924,
+      "loss": 1.0151,
+      "step": 2339
+    },
+    {
+      "epoch": 0.41744715012041744,
+      "grad_norm": 0.5230399370193481,
+      "learning_rate": 0.00044814268677914983,
+      "loss": 1.21,
+      "step": 2340
+    },
+    {
+      "epoch": 0.41762554633841764,
+      "grad_norm": 0.5473104119300842,
+      "learning_rate": 0.0004480999564234991,
+      "loss": 1.2911,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4178039425564178,
+      "grad_norm": 0.5277518033981323,
+      "learning_rate": 0.0004480572105091631,
+      "loss": 1.1298,
+      "step": 2342
+    },
+    {
+      "epoch": 0.417982338774418,
+      "grad_norm": 1.1111489534378052,
+      "learning_rate": 0.00044801444903949894,
+      "loss": 1.1337,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4181607349924182,
+      "grad_norm": 0.5263912677764893,
+      "learning_rate": 0.00044797167201786526,
+      "loss": 0.9498,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4183391312104183,
+      "grad_norm": 0.4941869378089905,
+      "learning_rate": 0.00044792887944762155,
+      "loss": 0.9439,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4185175274284185,
+      "grad_norm": 0.5406994223594666,
+      "learning_rate": 0.0004478860713321289,
+      "loss": 1.0799,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4186959236464187,
+      "grad_norm": 0.4971858263015747,
+      "learning_rate": 0.0004478432476747493,
+      "loss": 0.909,
+      "step": 2347
+    },
+    {
+      "epoch": 0.41887431986441886,
+      "grad_norm": 0.5575050115585327,
+      "learning_rate": 0.0004478004084788462,
+      "loss": 0.8755,
+      "step": 2348
+    },
+    {
+      "epoch": 0.41905271608241906,
+      "grad_norm": 0.6695868372917175,
+      "learning_rate": 0.00044775755374778413,
+      "loss": 1.1526,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4192311123004192,
+      "grad_norm": 0.5547947883605957,
+      "learning_rate": 0.00044771468348492896,
+      "loss": 1.1636,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4194095085184194,
+      "grad_norm": 1.2415190935134888,
+      "learning_rate": 0.00044767179769364754,
+      "loss": 0.9455,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4195879047364196,
+      "grad_norm": 0.5791499614715576,
+      "learning_rate": 0.0004476288963773082,
+      "loss": 0.9658,
+      "step": 2352
+    },
+    {
+      "epoch": 0.41976630095441975,
+      "grad_norm": 0.576765239238739,
+      "learning_rate": 0.00044758597953928044,
+      "loss": 1.0575,
+      "step": 2353
+    },
+    {
+      "epoch": 0.41994469717241995,
+      "grad_norm": 0.5667539238929749,
+      "learning_rate": 0.0004475430471829348,
+      "loss": 1.1187,
+      "step": 2354
+    },
+    {
+      "epoch": 0.42012309339042014,
+      "grad_norm": 0.5511680245399475,
+      "learning_rate": 0.00044750009931164336,
+      "loss": 0.9857,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4203014896084203,
+      "grad_norm": 2.5863170623779297,
+      "learning_rate": 0.00044745713592877904,
+      "loss": 0.9495,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4204798858264205,
+      "grad_norm": 0.5761575102806091,
+      "learning_rate": 0.00044741415703771615,
+      "loss": 0.8793,
+      "step": 2357
+    },
+    {
+      "epoch": 0.4206582820444207,
+      "grad_norm": 2.08255672454834,
+      "learning_rate": 0.0004473711626418304,
+      "loss": 1.1166,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4208366782624208,
+      "grad_norm": 1.4676316976547241,
+      "learning_rate": 0.00044732815274449856,
+      "loss": 1.1489,
+      "step": 2359
+    },
+    {
+      "epoch": 0.421015074480421,
+      "grad_norm": 0.5451486706733704,
+      "learning_rate": 0.00044728512734909845,
+      "loss": 0.9492,
+      "step": 2360
+    },
+    {
+      "epoch": 0.42119347069842117,
+      "grad_norm": 0.650875985622406,
+      "learning_rate": 0.0004472420864590093,
+      "loss": 1.0389,
+      "step": 2361
+    },
+    {
+      "epoch": 0.42137186691642137,
+      "grad_norm": 0.5705146193504333,
+      "learning_rate": 0.00044719903007761153,
+      "loss": 0.9989,
+      "step": 2362
+    },
+    {
+      "epoch": 0.42155026313442157,
+      "grad_norm": 0.4973379671573639,
+      "learning_rate": 0.00044715595820828694,
+      "loss": 1.0576,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4217286593524217,
+      "grad_norm": 0.5207223892211914,
+      "learning_rate": 0.0004471128708544181,
+      "loss": 1.2406,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4219070555704219,
+      "grad_norm": 0.5396470427513123,
+      "learning_rate": 0.00044706976801938927,
+      "loss": 1.1947,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4220854517884221,
+      "grad_norm": 0.5786265730857849,
+      "learning_rate": 0.0004470266497065856,
+      "loss": 1.1694,
+      "step": 2366
+    },
+    {
+      "epoch": 0.42226384800642225,
+      "grad_norm": 0.5583718419075012,
+      "learning_rate": 0.0004469835159193937,
+      "loss": 1.1314,
+      "step": 2367
+    },
+    {
+      "epoch": 0.42244224422442245,
+      "grad_norm": 0.4932156205177307,
+      "learning_rate": 0.0004469403666612013,
+      "loss": 0.8307,
+      "step": 2368
+    },
+    {
+      "epoch": 0.42262064044242265,
+      "grad_norm": 0.48325487971305847,
+      "learning_rate": 0.0004468972019353972,
+      "loss": 1.0107,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4227990366604228,
+      "grad_norm": 0.583512008190155,
+      "learning_rate": 0.0004468540217453715,
+      "loss": 0.897,
+      "step": 2370
+    },
+    {
+      "epoch": 0.422977432878423,
+      "grad_norm": 0.5151441097259521,
+      "learning_rate": 0.0004468108260945157,
+      "loss": 0.955,
+      "step": 2371
+    },
+    {
+      "epoch": 0.42315582909642313,
+      "grad_norm": 0.5649638772010803,
+      "learning_rate": 0.00044676761498622236,
+      "loss": 1.0448,
+      "step": 2372
+    },
+    {
+      "epoch": 0.42333422531442333,
+      "grad_norm": 0.7743201851844788,
+      "learning_rate": 0.0004467243884238852,
+      "loss": 1.0063,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4235126215324235,
+      "grad_norm": 0.535677433013916,
+      "learning_rate": 0.0004466811464108992,
+      "loss": 1.2873,
+      "step": 2374
+    },
+    {
+      "epoch": 0.42369101775042367,
+      "grad_norm": 0.9543548822402954,
+      "learning_rate": 0.00044663788895066064,
+      "loss": 0.9269,
+      "step": 2375
+    },
+    {
+      "epoch": 0.42386941396842387,
+      "grad_norm": 0.5472463965415955,
+      "learning_rate": 0.00044659461604656687,
+      "loss": 1.1469,
+      "step": 2376
+    },
+    {
+      "epoch": 0.42404781018642407,
+      "grad_norm": 0.5283387303352356,
+      "learning_rate": 0.00044655132770201657,
+      "loss": 0.9037,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4242262064044242,
+      "grad_norm": 0.8139501810073853,
+      "learning_rate": 0.00044650802392040957,
+      "loss": 0.9288,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4244046026224244,
+      "grad_norm": 0.4796660542488098,
+      "learning_rate": 0.0004464647047051469,
+      "loss": 1.0693,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4245829988404246,
+      "grad_norm": 0.648838222026825,
+      "learning_rate": 0.0004464213700596309,
+      "loss": 1.3307,
+      "step": 2380
+    },
+    {
+      "epoch": 0.42476139505842475,
+      "grad_norm": 0.5429944396018982,
+      "learning_rate": 0.0004463780199872651,
+      "loss": 1.1117,
+      "step": 2381
+    },
+    {
+      "epoch": 0.42493979127642495,
+      "grad_norm": 0.47701987624168396,
+      "learning_rate": 0.000446334654491454,
+      "loss": 0.9715,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4251181874944251,
+      "grad_norm": 0.479390412569046,
+      "learning_rate": 0.00044629127357560366,
+      "loss": 0.7807,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4252965837124253,
+      "grad_norm": 0.5513942241668701,
+      "learning_rate": 0.00044624787724312123,
+      "loss": 0.945,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4254749799304255,
+      "grad_norm": 0.5833590626716614,
+      "learning_rate": 0.0004462044654974149,
+      "loss": 1.0181,
+      "step": 2385
+    },
+    {
+      "epoch": 0.42565337614842563,
+      "grad_norm": 0.5582590699195862,
+      "learning_rate": 0.00044616103834189426,
+      "loss": 1.2099,
+      "step": 2386
+    },
+    {
+      "epoch": 0.42583177236642583,
+      "grad_norm": 0.7029179930686951,
+      "learning_rate": 0.0004461175957799701,
+      "loss": 1.0159,
+      "step": 2387
+    },
+    {
+      "epoch": 0.42601016858442603,
+      "grad_norm": 0.6166037321090698,
+      "learning_rate": 0.0004460741378150544,
+      "loss": 1.157,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4261885648024262,
+      "grad_norm": 0.4319639801979065,
+      "learning_rate": 0.0004460306644505603,
+      "loss": 1.0279,
+      "step": 2389
+    },
+    {
+      "epoch": 0.42636696102042637,
+      "grad_norm": 0.5709377527236938,
+      "learning_rate": 0.00044598717568990214,
+      "loss": 1.1857,
+      "step": 2390
+    },
+    {
+      "epoch": 0.42654535723842657,
+      "grad_norm": 0.5123762488365173,
+      "learning_rate": 0.00044594367153649554,
+      "loss": 0.8416,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4267237534564267,
+      "grad_norm": 0.5814945101737976,
+      "learning_rate": 0.0004459001519937573,
+      "loss": 0.8822,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4269021496744269,
+      "grad_norm": 0.5209176540374756,
+      "learning_rate": 0.0004458566170651055,
+      "loss": 0.8742,
+      "step": 2393
+    },
+    {
+      "epoch": 0.42708054589242705,
+      "grad_norm": 0.5272184610366821,
+      "learning_rate": 0.00044581306675395917,
+      "loss": 0.9116,
+      "step": 2394
+    },
+    {
+      "epoch": 0.42725894211042725,
+      "grad_norm": 0.6671530604362488,
+      "learning_rate": 0.00044576950106373894,
+      "loss": 1.5091,
+      "step": 2395
+    },
+    {
+      "epoch": 0.42743733832842745,
+      "grad_norm": 0.5785043835639954,
+      "learning_rate": 0.00044572591999786625,
+      "loss": 0.9558,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4276157345464276,
+      "grad_norm": 0.5078967213630676,
+      "learning_rate": 0.0004456823235597641,
+      "loss": 0.9042,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4277941307644278,
+      "grad_norm": 0.5760866403579712,
+      "learning_rate": 0.0004456387117528564,
+      "loss": 1.1446,
+      "step": 2398
+    },
+    {
+      "epoch": 0.427972526982428,
+      "grad_norm": 0.5377951264381409,
+      "learning_rate": 0.00044559508458056853,
+      "loss": 1.185,
+      "step": 2399
+    },
+    {
+      "epoch": 0.42815092320042814,
+      "grad_norm": 0.5057874917984009,
+      "learning_rate": 0.0004455514420463269,
+      "loss": 1.0324,
+      "step": 2400
+    },
+    {
+      "epoch": 0.42832931941842833,
+      "grad_norm": 0.45790576934814453,
+      "learning_rate": 0.0004455077841535591,
+      "loss": 1.1008,
+      "step": 2401
+    },
+    {
+      "epoch": 0.42850771563642853,
+      "grad_norm": 0.5488193035125732,
+      "learning_rate": 0.0004454641109056941,
+      "loss": 1.0925,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4286861118544287,
+      "grad_norm": 0.5256093740463257,
+      "learning_rate": 0.0004454204223061619,
+      "loss": 1.1341,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4288645080724289,
+      "grad_norm": 0.4891919493675232,
+      "learning_rate": 0.00044537671835839386,
+      "loss": 1.0355,
+      "step": 2404
+    },
+    {
+      "epoch": 0.429042904290429,
+      "grad_norm": 0.4800609052181244,
+      "learning_rate": 0.0004453329990658225,
+      "loss": 0.9841,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4292213005084292,
+      "grad_norm": 0.5242645740509033,
+      "learning_rate": 0.00044528926443188137,
+      "loss": 0.9167,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4293996967264294,
+      "grad_norm": 0.7704287767410278,
+      "learning_rate": 0.0004452455144600055,
+      "loss": 1.1246,
+      "step": 2407
+    },
+    {
+      "epoch": 0.42957809294442956,
+      "grad_norm": 0.5126969814300537,
+      "learning_rate": 0.0004452017491536309,
+      "loss": 1.1712,
+      "step": 2408
+    },
+    {
+      "epoch": 0.42975648916242976,
+      "grad_norm": 0.5316195487976074,
+      "learning_rate": 0.00044515796851619493,
+      "loss": 1.017,
+      "step": 2409
+    },
+    {
+      "epoch": 0.42993488538042995,
+      "grad_norm": 0.5696223974227905,
+      "learning_rate": 0.0004451141725511362,
+      "loss": 1.0804,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4301132815984301,
+      "grad_norm": 0.45580875873565674,
+      "learning_rate": 0.00044507036126189414,
+      "loss": 0.9472,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4302916778164303,
+      "grad_norm": 0.579031765460968,
+      "learning_rate": 0.00044502653465191,
+      "loss": 1.0372,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4304700740344305,
+      "grad_norm": 0.5286597609519958,
+      "learning_rate": 0.00044498269272462567,
+      "loss": 1.2533,
+      "step": 2413
+    },
+    {
+      "epoch": 0.43064847025243064,
+      "grad_norm": 0.6354557275772095,
+      "learning_rate": 0.00044493883548348456,
+      "loss": 1.0672,
+      "step": 2414
+    },
+    {
+      "epoch": 0.43082686647043084,
+      "grad_norm": 0.6015318036079407,
+      "learning_rate": 0.0004448949629319313,
+      "loss": 1.1149,
+      "step": 2415
+    },
+    {
+      "epoch": 0.431005262688431,
+      "grad_norm": 0.5161615014076233,
+      "learning_rate": 0.0004448510750734114,
+      "loss": 1.1315,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4311836589064312,
+      "grad_norm": 0.5163938403129578,
+      "learning_rate": 0.000444807171911372,
+      "loss": 1.0624,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4313620551244314,
+      "grad_norm": 0.558363676071167,
+      "learning_rate": 0.00044476325344926113,
+      "loss": 1.3461,
+      "step": 2418
+    },
+    {
+      "epoch": 0.4315404513424315,
+      "grad_norm": 0.5142703652381897,
+      "learning_rate": 0.00044471931969052817,
+      "loss": 0.9696,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4317188475604317,
+      "grad_norm": 0.5983433127403259,
+      "learning_rate": 0.00044467537063862353,
+      "loss": 1.0887,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4318972437784319,
+      "grad_norm": 0.5289214253425598,
+      "learning_rate": 0.00044463140629699916,
+      "loss": 1.2276,
+      "step": 2421
+    },
+    {
+      "epoch": 0.43207563999643206,
+      "grad_norm": 0.5035809874534607,
+      "learning_rate": 0.00044458742666910785,
+      "loss": 1.048,
+      "step": 2422
+    },
+    {
+      "epoch": 0.43225403621443226,
+      "grad_norm": 0.5544365048408508,
+      "learning_rate": 0.0004445434317584038,
+      "loss": 1.0358,
+      "step": 2423
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.4954850673675537,
+      "learning_rate": 0.00044449942156834236,
+      "loss": 0.9533,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4326108286504326,
+      "grad_norm": 0.5363656282424927,
+      "learning_rate": 0.0004444553961023801,
+      "loss": 1.0444,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4327892248684328,
+      "grad_norm": 0.4978641867637634,
+      "learning_rate": 0.00044441135536397455,
+      "loss": 1.0043,
+      "step": 2426
+    },
+    {
+      "epoch": 0.43296762108643294,
+      "grad_norm": 0.536332905292511,
+      "learning_rate": 0.0004443672993565849,
+      "loss": 0.8316,
+      "step": 2427
+    },
+    {
+      "epoch": 0.43314601730443314,
+      "grad_norm": 0.47319361567497253,
+      "learning_rate": 0.0004443232280836712,
+      "loss": 0.9772,
+      "step": 2428
+    },
+    {
+      "epoch": 0.43332441352243334,
+      "grad_norm": 0.47262993454933167,
+      "learning_rate": 0.0004442791415486948,
+      "loss": 0.8884,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4335028097404335,
+      "grad_norm": 0.5313016176223755,
+      "learning_rate": 0.0004442350397551182,
+      "loss": 1.0347,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4336812059584337,
+      "grad_norm": 0.47843775153160095,
+      "learning_rate": 0.0004441909227064052,
+      "loss": 0.9594,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4338596021764339,
+      "grad_norm": 0.5179790258407593,
+      "learning_rate": 0.00044414679040602066,
+      "loss": 0.9443,
+      "step": 2432
+    },
+    {
+      "epoch": 0.434037998394434,
+      "grad_norm": 0.5228581428527832,
+      "learning_rate": 0.0004441026428574308,
+      "loss": 0.8851,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4342163946124342,
+      "grad_norm": 0.5385475754737854,
+      "learning_rate": 0.0004440584800641029,
+      "loss": 1.1273,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4343947908304344,
+      "grad_norm": 0.5584160685539246,
+      "learning_rate": 0.0004440143020295054,
+      "loss": 1.1722,
+      "step": 2435
+    },
+    {
+      "epoch": 0.43457318704843456,
+      "grad_norm": 0.5520399808883667,
+      "learning_rate": 0.0004439701087571082,
+      "loss": 1.1687,
+      "step": 2436
+    },
+    {
+      "epoch": 0.43475158326643476,
+      "grad_norm": 0.5202842950820923,
+      "learning_rate": 0.000443925900250382,
+      "loss": 1.1505,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4349299794844349,
+      "grad_norm": 0.49009421467781067,
+      "learning_rate": 0.0004438816765127992,
+      "loss": 1.0714,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4351083757024351,
+      "grad_norm": 0.5388837456703186,
+      "learning_rate": 0.0004438374375478329,
+      "loss": 0.9845,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4352867719204353,
+      "grad_norm": 0.5120062232017517,
+      "learning_rate": 0.0004437931833589577,
+      "loss": 0.8921,
+      "step": 2440
+    },
+    {
+      "epoch": 0.43546516813843544,
+      "grad_norm": 0.4710652530193329,
+      "learning_rate": 0.00044374891394964925,
+      "loss": 0.9668,
+      "step": 2441
+    },
+    {
+      "epoch": 0.43564356435643564,
+      "grad_norm": 1.251098871231079,
+      "learning_rate": 0.0004437046293233845,
+      "loss": 1.2214,
+      "step": 2442
+    },
+    {
+      "epoch": 0.43582196057443584,
+      "grad_norm": 0.5326117873191833,
+      "learning_rate": 0.00044366032948364145,
+      "loss": 1.2166,
+      "step": 2443
+    },
+    {
+      "epoch": 0.436000356792436,
+      "grad_norm": 0.558892011642456,
+      "learning_rate": 0.0004436160144338995,
+      "loss": 1.0941,
+      "step": 2444
+    },
+    {
+      "epoch": 0.4361787530104362,
+      "grad_norm": 0.5412200093269348,
+      "learning_rate": 0.0004435716841776391,
+      "loss": 0.9558,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4363571492284364,
+      "grad_norm": 0.50641268491745,
+      "learning_rate": 0.00044352733871834193,
+      "loss": 0.87,
+      "step": 2446
+    },
+    {
+      "epoch": 0.4365355454464365,
+      "grad_norm": 0.5298489332199097,
+      "learning_rate": 0.0004434829780594909,
+      "loss": 1.0362,
+      "step": 2447
+    },
+    {
+      "epoch": 0.4367139416644367,
+      "grad_norm": 0.5438408851623535,
+      "learning_rate": 0.00044343860220456995,
+      "loss": 0.8945,
+      "step": 2448
+    },
+    {
+      "epoch": 0.43689233788243687,
+      "grad_norm": 0.5930485725402832,
+      "learning_rate": 0.00044339421115706444,
+      "loss": 1.1371,
+      "step": 2449
+    },
+    {
+      "epoch": 0.43707073410043706,
+      "grad_norm": 0.5204629302024841,
+      "learning_rate": 0.00044334980492046085,
+      "loss": 1.047,
+      "step": 2450
+    },
+    {
+      "epoch": 0.43724913031843726,
+      "grad_norm": 0.46508142352104187,
+      "learning_rate": 0.0004433053834982468,
+      "loss": 0.9044,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4374275265364374,
+      "grad_norm": 0.5729008913040161,
+      "learning_rate": 0.00044326094689391105,
+      "loss": 1.0573,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4376059227544376,
+      "grad_norm": 0.5557507276535034,
+      "learning_rate": 0.00044321649511094375,
+      "loss": 1.1433,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4377843189724378,
+      "grad_norm": 0.4726797044277191,
+      "learning_rate": 0.00044317202815283605,
+      "loss": 1.0822,
+      "step": 2454
+    },
+    {
+      "epoch": 0.43796271519043795,
+      "grad_norm": 0.5183228254318237,
+      "learning_rate": 0.00044312754602308035,
+      "loss": 1.1219,
+      "step": 2455
+    },
+    {
+      "epoch": 0.43814111140843814,
+      "grad_norm": 0.5738768577575684,
+      "learning_rate": 0.00044308304872517037,
+      "loss": 1.2338,
+      "step": 2456
+    },
+    {
+      "epoch": 0.43831950762643834,
+      "grad_norm": 0.5397445559501648,
+      "learning_rate": 0.0004430385362626008,
+      "loss": 1.1709,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4384979038444385,
+      "grad_norm": 0.5338584780693054,
+      "learning_rate": 0.0004429940086388676,
+      "loss": 1.1662,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4386763000624387,
+      "grad_norm": 0.5330877304077148,
+      "learning_rate": 0.00044294946585746815,
+      "loss": 0.9068,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4388546962804388,
+      "grad_norm": 0.5154666304588318,
+      "learning_rate": 0.0004429049079219006,
+      "loss": 0.8542,
+      "step": 2460
+    },
+    {
+      "epoch": 0.439033092498439,
+      "grad_norm": 0.5239487886428833,
+      "learning_rate": 0.00044286033483566456,
+      "loss": 1.1177,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4392114887164392,
+      "grad_norm": 1.6719294786453247,
+      "learning_rate": 0.00044281574660226086,
+      "loss": 0.7778,
+      "step": 2462
+    },
+    {
+      "epoch": 0.43938988493443937,
+      "grad_norm": 0.515557587146759,
+      "learning_rate": 0.0004427711432251914,
+      "loss": 0.9718,
+      "step": 2463
+    },
+    {
+      "epoch": 0.43956828115243957,
+      "grad_norm": 0.51800537109375,
+      "learning_rate": 0.0004427265247079594,
+      "loss": 0.8828,
+      "step": 2464
+    },
+    {
+      "epoch": 0.43974667737043976,
+      "grad_norm": 0.7687791585922241,
+      "learning_rate": 0.00044268189105406896,
+      "loss": 1.1033,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4399250735884399,
+      "grad_norm": 0.6673977971076965,
+      "learning_rate": 0.00044263724226702573,
+      "loss": 1.1026,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4401034698064401,
+      "grad_norm": 1.0776582956314087,
+      "learning_rate": 0.0004425925783503364,
+      "loss": 1.1335,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4402818660244403,
+      "grad_norm": 0.5958316326141357,
+      "learning_rate": 0.00044254789930750884,
+      "loss": 0.968,
+      "step": 2468
+    },
+    {
+      "epoch": 0.44046026224244045,
+      "grad_norm": 0.5925278663635254,
+      "learning_rate": 0.0004425032051420522,
+      "loss": 1.2383,
+      "step": 2469
+    },
+    {
+      "epoch": 0.44063865846044065,
+      "grad_norm": 0.5400102138519287,
+      "learning_rate": 0.00044245849585747656,
+      "loss": 1.0723,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4408170546784408,
+      "grad_norm": 0.5745581388473511,
+      "learning_rate": 0.00044241377145729356,
+      "loss": 1.1578,
+      "step": 2471
+    },
+    {
+      "epoch": 0.440995450896441,
+      "grad_norm": 0.514392077922821,
+      "learning_rate": 0.00044236903194501566,
+      "loss": 0.9727,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4411738471144412,
+      "grad_norm": 1.0300344228744507,
+      "learning_rate": 0.00044232427732415687,
+      "loss": 1.1997,
+      "step": 2473
+    },
+    {
+      "epoch": 0.44135224333244133,
+      "grad_norm": 0.45754095911979675,
+      "learning_rate": 0.0004422795075982321,
+      "loss": 0.9891,
+      "step": 2474
+    },
+    {
+      "epoch": 0.44153063955044153,
+      "grad_norm": 0.5444067120552063,
+      "learning_rate": 0.0004422347227707575,
+      "loss": 1.0424,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4417090357684417,
+      "grad_norm": 2.9430816173553467,
+      "learning_rate": 0.0004421899228452505,
+      "loss": 1.0375,
+      "step": 2476
+    },
+    {
+      "epoch": 0.44188743198644187,
+      "grad_norm": 0.570728600025177,
+      "learning_rate": 0.00044214510782522966,
+      "loss": 1.257,
+      "step": 2477
+    },
+    {
+      "epoch": 0.44206582820444207,
+      "grad_norm": 0.5782069563865662,
+      "learning_rate": 0.0004421002777142148,
+      "loss": 1.0603,
+      "step": 2478
+    },
+    {
+      "epoch": 0.44224422442244227,
+      "grad_norm": 0.505310595035553,
+      "learning_rate": 0.00044205543251572664,
+      "loss": 0.9608,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4424226206404424,
+      "grad_norm": 0.5967997312545776,
+      "learning_rate": 0.00044201057223328767,
+      "loss": 1.2066,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4426010168584426,
+      "grad_norm": 0.5399075746536255,
+      "learning_rate": 0.00044196569687042085,
+      "loss": 1.0442,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4427794130764428,
+      "grad_norm": 0.5512840151786804,
+      "learning_rate": 0.0004419208064306509,
+      "loss": 1.1372,
+      "step": 2482
+    },
+    {
+      "epoch": 0.44295780929444295,
+      "grad_norm": 0.540708601474762,
+      "learning_rate": 0.0004418759009175033,
+      "loss": 1.2523,
+      "step": 2483
+    },
+    {
+      "epoch": 0.44313620551244315,
+      "grad_norm": 0.5288528800010681,
+      "learning_rate": 0.0004418309803345051,
+      "loss": 1.1503,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4433146017304433,
+      "grad_norm": 0.48563152551651,
+      "learning_rate": 0.0004417860446851841,
+      "loss": 0.9179,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4434929979484435,
+      "grad_norm": 0.4692551791667938,
+      "learning_rate": 0.00044174109397306983,
+      "loss": 0.9259,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4436713941664437,
+      "grad_norm": 0.5046235918998718,
+      "learning_rate": 0.0004416961282016926,
+      "loss": 0.8989,
+      "step": 2487
+    },
+    {
+      "epoch": 0.44384979038444383,
+      "grad_norm": 0.5326201915740967,
+      "learning_rate": 0.00044165114737458377,
+      "loss": 1.1311,
+      "step": 2488
+    },
+    {
+      "epoch": 0.44402818660244403,
+      "grad_norm": 0.4889252185821533,
+      "learning_rate": 0.00044160615149527643,
+      "loss": 0.9787,
+      "step": 2489
+    },
+    {
+      "epoch": 0.44420658282044423,
+      "grad_norm": 0.5168465971946716,
+      "learning_rate": 0.0004415611405673044,
+      "loss": 1.1548,
+      "step": 2490
+    },
+    {
+      "epoch": 0.44438497903844437,
+      "grad_norm": 0.49787530303001404,
+      "learning_rate": 0.0004415161145942028,
+      "loss": 1.1097,
+      "step": 2491
+    },
+    {
+      "epoch": 0.44456337525644457,
+      "grad_norm": 0.47516435384750366,
+      "learning_rate": 0.000441471073579508,
+      "loss": 1.0759,
+      "step": 2492
+    },
+    {
+      "epoch": 0.44474177147444477,
+      "grad_norm": 0.5186420679092407,
+      "learning_rate": 0.0004414260175267574,
+      "loss": 0.9333,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4449201676924449,
+      "grad_norm": 0.538212239742279,
+      "learning_rate": 0.0004413809464394899,
+      "loss": 1.0112,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4450985639104451,
+      "grad_norm": 0.6391366124153137,
+      "learning_rate": 0.0004413358603212451,
+      "loss": 1.0763,
+      "step": 2495
+    },
+    {
+      "epoch": 0.44527696012844525,
+      "grad_norm": 0.49085238575935364,
+      "learning_rate": 0.00044129075917556415,
+      "loss": 0.9748,
+      "step": 2496
+    },
+    {
+      "epoch": 0.44545535634644545,
+      "grad_norm": 3.5137956142425537,
+      "learning_rate": 0.00044124564300598943,
+      "loss": 0.9487,
+      "step": 2497
+    },
+    {
+      "epoch": 0.44563375256444565,
+      "grad_norm": 0.5351426601409912,
+      "learning_rate": 0.00044120051181606413,
+      "loss": 1.0537,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4458121487824458,
+      "grad_norm": 0.5824923515319824,
+      "learning_rate": 0.0004411553656093329,
+      "loss": 0.9824,
+      "step": 2499
+    },
+    {
+      "epoch": 0.445990545000446,
+      "grad_norm": 0.5528644323348999,
+      "learning_rate": 0.00044111020438934153,
+      "loss": 1.1647,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4461689412184462,
+      "grad_norm": 0.6510967016220093,
+      "learning_rate": 0.0004410650281596369,
+      "loss": 1.101,
+      "step": 2501
+    },
+    {
+      "epoch": 0.44634733743644633,
+      "grad_norm": 0.5696078538894653,
+      "learning_rate": 0.00044101983692376723,
+      "loss": 1.0418,
+      "step": 2502
+    },
+    {
+      "epoch": 0.44652573365444653,
+      "grad_norm": 0.5028705596923828,
+      "learning_rate": 0.0004409746306852818,
+      "loss": 0.8853,
+      "step": 2503
+    },
+    {
+      "epoch": 0.44670412987244673,
+      "grad_norm": 0.5344279408454895,
+      "learning_rate": 0.00044092940944773104,
+      "loss": 1.0267,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4468825260904469,
+      "grad_norm": 0.5304036140441895,
+      "learning_rate": 0.00044088417321466657,
+      "loss": 1.0281,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4470609223084471,
+      "grad_norm": 0.5835813879966736,
+      "learning_rate": 0.0004408389219896413,
+      "loss": 1.1255,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4472393185264472,
+      "grad_norm": 0.48519715666770935,
+      "learning_rate": 0.0004407936557762093,
+      "loss": 1.0046,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4474177147444474,
+      "grad_norm": 0.5301933288574219,
+      "learning_rate": 0.0004407483745779256,
+      "loss": 1.0998,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4475961109624476,
+      "grad_norm": 0.4589798152446747,
+      "learning_rate": 0.00044070307839834665,
+      "loss": 0.7962,
+      "step": 2509
+    },
+    {
+      "epoch": 0.44777450718044776,
+      "grad_norm": 0.496530681848526,
+      "learning_rate": 0.00044065776724103,
+      "loss": 0.7726,
+      "step": 2510
+    },
+    {
+      "epoch": 0.44795290339844795,
+      "grad_norm": 0.4880942702293396,
+      "learning_rate": 0.00044061244110953437,
+      "loss": 0.8395,
+      "step": 2511
+    },
+    {
+      "epoch": 0.44813129961644815,
+      "grad_norm": 0.5023302435874939,
+      "learning_rate": 0.0004405671000074196,
+      "loss": 0.9507,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4483096958344483,
+      "grad_norm": 0.561840832233429,
+      "learning_rate": 0.00044052174393824687,
+      "loss": 1.0694,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4484880920524485,
+      "grad_norm": 0.5529274344444275,
+      "learning_rate": 0.00044047637290557835,
+      "loss": 1.221,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4486664882704487,
+      "grad_norm": 0.44090861082077026,
+      "learning_rate": 0.00044043098691297746,
+      "loss": 1.0016,
+      "step": 2515
+    },
+    {
+      "epoch": 0.44884488448844884,
+      "grad_norm": 0.5670448541641235,
+      "learning_rate": 0.0004403855859640088,
+      "loss": 0.9971,
+      "step": 2516
+    },
+    {
+      "epoch": 0.44902328070644904,
+      "grad_norm": 0.45753970742225647,
+      "learning_rate": 0.00044034017006223817,
+      "loss": 0.7955,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4492016769244492,
+      "grad_norm": 0.7095730900764465,
+      "learning_rate": 0.0004402947392112324,
+      "loss": 0.7677,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4493800731424494,
+      "grad_norm": 0.5540285110473633,
+      "learning_rate": 0.00044024929341455977,
+      "loss": 1.34,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4495584693604496,
+      "grad_norm": 0.5628980994224548,
+      "learning_rate": 0.00044020383267578955,
+      "loss": 1.0199,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4497368655784497,
+      "grad_norm": 0.5465995073318481,
+      "learning_rate": 0.0004401583569984921,
+      "loss": 1.1564,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4499152617964499,
+      "grad_norm": 0.4945712089538574,
+      "learning_rate": 0.00044011286638623916,
+      "loss": 1.0976,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4500936580144501,
+      "grad_norm": 0.5624112486839294,
+      "learning_rate": 0.00044006736084260345,
+      "loss": 1.0677,
+      "step": 2523
+    },
+    {
+      "epoch": 0.45027205423245026,
+      "grad_norm": 0.5418776869773865,
+      "learning_rate": 0.0004400218403711591,
+      "loss": 1.2596,
+      "step": 2524
+    },
+    {
+      "epoch": 0.45045045045045046,
+      "grad_norm": 0.4920550584793091,
+      "learning_rate": 0.0004399763049754811,
+      "loss": 1.0434,
+      "step": 2525
+    },
+    {
+      "epoch": 0.45062884666845066,
+      "grad_norm": 0.5869379639625549,
+      "learning_rate": 0.0004399307546591459,
+      "loss": 1.0799,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4508072428864508,
+      "grad_norm": 0.619750440120697,
+      "learning_rate": 0.0004398851894257309,
+      "loss": 1.1675,
+      "step": 2527
+    },
+    {
+      "epoch": 0.450985639104451,
+      "grad_norm": 0.5530658960342407,
+      "learning_rate": 0.00043983960927881493,
+      "loss": 0.9814,
+      "step": 2528
+    },
+    {
+      "epoch": 0.45116403532245114,
+      "grad_norm": 0.5448163151741028,
+      "learning_rate": 0.0004397940142219776,
+      "loss": 1.2051,
+      "step": 2529
+    },
+    {
+      "epoch": 0.45134243154045134,
+      "grad_norm": 0.5526171922683716,
+      "learning_rate": 0.00043974840425880027,
+      "loss": 1.2632,
+      "step": 2530
+    },
+    {
+      "epoch": 0.45152082775845154,
+      "grad_norm": 0.46508967876434326,
+      "learning_rate": 0.0004397027793928648,
+      "loss": 0.859,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4516992239764517,
+      "grad_norm": 0.9926976561546326,
+      "learning_rate": 0.00043965713962775465,
+      "loss": 0.924,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4518776201944519,
+      "grad_norm": 0.5205312967300415,
+      "learning_rate": 0.00043961148496705443,
+      "loss": 0.8696,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4520560164124521,
+      "grad_norm": 0.5645661354064941,
+      "learning_rate": 0.0004395658154143498,
+      "loss": 1.0412,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4522344126304522,
+      "grad_norm": 0.6434807181358337,
+      "learning_rate": 0.00043952013097322754,
+      "loss": 1.1406,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4524128088484524,
+      "grad_norm": 0.5284314751625061,
+      "learning_rate": 0.00043947443164727573,
+      "loss": 0.9631,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4525912050664526,
+      "grad_norm": 0.5045761466026306,
+      "learning_rate": 0.00043942871744008375,
+      "loss": 0.9317,
+      "step": 2537
+    },
+    {
+      "epoch": 0.45276960128445276,
+      "grad_norm": 0.5030120015144348,
+      "learning_rate": 0.00043938298835524166,
+      "loss": 0.8189,
+      "step": 2538
+    },
+    {
+      "epoch": 0.45294799750245296,
+      "grad_norm": 0.4574146270751953,
+      "learning_rate": 0.0004393372443963412,
+      "loss": 0.7983,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4531263937204531,
+      "grad_norm": 0.5439005494117737,
+      "learning_rate": 0.00043929148556697505,
+      "loss": 1.0398,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4533047899384533,
+      "grad_norm": 0.5667563080787659,
+      "learning_rate": 0.0004392457118707371,
+      "loss": 1.1755,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4534831861564535,
+      "grad_norm": 0.5080005526542664,
+      "learning_rate": 0.0004391999233112224,
+      "loss": 1.0836,
+      "step": 2542
+    },
+    {
+      "epoch": 0.45366158237445364,
+      "grad_norm": 0.4875447750091553,
+      "learning_rate": 0.00043915411989202706,
+      "loss": 0.8928,
+      "step": 2543
+    },
+    {
+      "epoch": 0.45383997859245384,
+      "grad_norm": 0.5923610925674438,
+      "learning_rate": 0.0004391083016167486,
+      "loss": 1.0488,
+      "step": 2544
+    },
+    {
+      "epoch": 0.45401837481045404,
+      "grad_norm": 0.49262282252311707,
+      "learning_rate": 0.00043906246848898544,
+      "loss": 0.9689,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4541967710284542,
+      "grad_norm": 0.4940873086452484,
+      "learning_rate": 0.0004390166205123375,
+      "loss": 1.0168,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4543751672464544,
+      "grad_norm": 0.44292840361595154,
+      "learning_rate": 0.00043897075769040543,
+      "loss": 0.6536,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4545535634644546,
+      "grad_norm": 0.4958019256591797,
+      "learning_rate": 0.00043892488002679144,
+      "loss": 0.918,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4547319596824547,
+      "grad_norm": 8.831953048706055,
+      "learning_rate": 0.0004388789875250986,
+      "loss": 1.2642,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4549103559004549,
+      "grad_norm": 0.8993814587593079,
+      "learning_rate": 0.0004388330801889314,
+      "loss": 1.0951,
+      "step": 2550
+    },
+    {
+      "epoch": 0.45508875211845506,
+      "grad_norm": 0.47165387868881226,
+      "learning_rate": 0.00043878715802189526,
+      "loss": 0.9744,
+      "step": 2551
+    },
+    {
+      "epoch": 0.45526714833645526,
+      "grad_norm": 0.4588290750980377,
+      "learning_rate": 0.0004387412210275971,
+      "loss": 0.7959,
+      "step": 2552
+    },
+    {
+      "epoch": 0.45544554455445546,
+      "grad_norm": 0.5076047778129578,
+      "learning_rate": 0.00043869526920964466,
+      "loss": 1.1043,
+      "step": 2553
+    },
+    {
+      "epoch": 0.4556239407724556,
+      "grad_norm": 0.5313244462013245,
+      "learning_rate": 0.0004386493025716469,
+      "loss": 1.1507,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4558023369904558,
+      "grad_norm": 0.7201479077339172,
+      "learning_rate": 0.00043860332111721407,
+      "loss": 1.0751,
+      "step": 2555
+    },
+    {
+      "epoch": 0.455980733208456,
+      "grad_norm": 0.503410816192627,
+      "learning_rate": 0.0004385573248499576,
+      "loss": 1.0373,
+      "step": 2556
+    },
+    {
+      "epoch": 0.45615912942645614,
+      "grad_norm": 0.5491756200790405,
+      "learning_rate": 0.00043851131377349004,
+      "loss": 1.0991,
+      "step": 2557
+    },
+    {
+      "epoch": 0.45633752564445634,
+      "grad_norm": 0.5328734517097473,
+      "learning_rate": 0.0004384652878914249,
+      "loss": 1.0833,
+      "step": 2558
+    },
+    {
+      "epoch": 0.45651592186245654,
+      "grad_norm": 0.564434826374054,
+      "learning_rate": 0.00043841924720737724,
+      "loss": 1.1307,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4566943180804567,
+      "grad_norm": 0.5131104588508606,
+      "learning_rate": 0.000438373191724963,
+      "loss": 0.9291,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4568727142984569,
+      "grad_norm": 0.4851999282836914,
+      "learning_rate": 0.0004383271214477993,
+      "loss": 0.8158,
+      "step": 2561
+    },
+    {
+      "epoch": 0.457051110516457,
+      "grad_norm": 0.5580843091011047,
+      "learning_rate": 0.00043828103637950455,
+      "loss": 1.1642,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4572295067344572,
+      "grad_norm": 0.5292948484420776,
+      "learning_rate": 0.00043823493652369824,
+      "loss": 1.0218,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4574079029524574,
+      "grad_norm": 0.4889855682849884,
+      "learning_rate": 0.000438188821884001,
+      "loss": 0.9337,
+      "step": 2564
+    },
+    {
+      "epoch": 0.45758629917045757,
+      "grad_norm": 0.5168807506561279,
+      "learning_rate": 0.0004381426924640346,
+      "loss": 1.0195,
+      "step": 2565
+    },
+    {
+      "epoch": 0.45776469538845777,
+      "grad_norm": 0.5251139998435974,
+      "learning_rate": 0.0004380965482674222,
+      "loss": 1.0455,
+      "step": 2566
+    },
+    {
+      "epoch": 0.45794309160645796,
+      "grad_norm": 0.49821674823760986,
+      "learning_rate": 0.00043805038929778785,
+      "loss": 0.9241,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4581214878244581,
+      "grad_norm": 0.48444879055023193,
+      "learning_rate": 0.0004380042155587568,
+      "loss": 1.0078,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4582998840424583,
+      "grad_norm": 0.4438982903957367,
+      "learning_rate": 0.00043795802705395555,
+      "loss": 0.7366,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4584782802604585,
+      "grad_norm": 0.48436489701271057,
+      "learning_rate": 0.0004379118237870118,
+      "loss": 0.9042,
+      "step": 2570
+    },
+    {
+      "epoch": 0.45865667647845865,
+      "grad_norm": 0.5146360397338867,
+      "learning_rate": 0.0004378656057615542,
+      "loss": 0.9988,
+      "step": 2571
+    },
+    {
+      "epoch": 0.45883507269645885,
+      "grad_norm": 0.5639342665672302,
+      "learning_rate": 0.00043781937298121275,
+      "loss": 1.2857,
+      "step": 2572
+    },
+    {
+      "epoch": 0.459013468914459,
+      "grad_norm": 0.5131158828735352,
+      "learning_rate": 0.00043777312544961865,
+      "loss": 1.1494,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4591918651324592,
+      "grad_norm": 0.5281994342803955,
+      "learning_rate": 0.0004377268631704041,
+      "loss": 1.1338,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4593702613504594,
+      "grad_norm": 0.6242533922195435,
+      "learning_rate": 0.0004376805861472024,
+      "loss": 1.1631,
+      "step": 2575
+    },
+    {
+      "epoch": 0.45954865756845953,
+      "grad_norm": 0.48754405975341797,
+      "learning_rate": 0.0004376342943836483,
+      "loss": 1.016,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4597270537864597,
+      "grad_norm": 0.5026589632034302,
+      "learning_rate": 0.0004375879878833775,
+      "loss": 1.0022,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4599054500044599,
+      "grad_norm": 0.4625113606452942,
+      "learning_rate": 0.0004375416666500268,
+      "loss": 0.942,
+      "step": 2578
+    },
+    {
+      "epoch": 0.46008384622246007,
+      "grad_norm": 0.552786111831665,
+      "learning_rate": 0.00043749533068723436,
+      "loss": 1.1234,
+      "step": 2579
+    },
+    {
+      "epoch": 0.46026224244046027,
+      "grad_norm": 0.5415485501289368,
+      "learning_rate": 0.0004374489799986393,
+      "loss": 0.8123,
+      "step": 2580
+    },
+    {
+      "epoch": 0.46044063865846047,
+      "grad_norm": 0.49887794256210327,
+      "learning_rate": 0.000437402614587882,
+      "loss": 1.0032,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4606190348764606,
+      "grad_norm": 0.5717129111289978,
+      "learning_rate": 0.00043735623445860397,
+      "loss": 1.2778,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4607974310944608,
+      "grad_norm": 0.4855397939682007,
+      "learning_rate": 0.00043730983961444794,
+      "loss": 1.0836,
+      "step": 2583
+    },
+    {
+      "epoch": 0.46097582731246095,
+      "grad_norm": 0.46733468770980835,
+      "learning_rate": 0.0004372634300590578,
+      "loss": 0.9161,
+      "step": 2584
+    },
+    {
+      "epoch": 0.46115422353046115,
+      "grad_norm": 0.5448064804077148,
+      "learning_rate": 0.0004372170057960783,
+      "loss": 1.0957,
+      "step": 2585
+    },
+    {
+      "epoch": 0.46133261974846135,
+      "grad_norm": 0.5003434419631958,
+      "learning_rate": 0.0004371705668291558,
+      "loss": 1.1793,
+      "step": 2586
+    },
+    {
+      "epoch": 0.4615110159664615,
+      "grad_norm": 0.5096003413200378,
+      "learning_rate": 0.00043712411316193755,
+      "loss": 1.29,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4616894121844617,
+      "grad_norm": 0.5056552290916443,
+      "learning_rate": 0.00043707764479807194,
+      "loss": 1.0382,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4618678084024619,
+      "grad_norm": 0.4839148223400116,
+      "learning_rate": 0.0004370311617412086,
+      "loss": 0.9115,
+      "step": 2589
+    },
+    {
+      "epoch": 0.46204620462046203,
+      "grad_norm": 0.49606168270111084,
+      "learning_rate": 0.0004369846639949984,
+      "loss": 1.1419,
+      "step": 2590
+    },
+    {
+      "epoch": 0.46222460083846223,
+      "grad_norm": 0.4850931763648987,
+      "learning_rate": 0.000436938151563093,
+      "loss": 1.0411,
+      "step": 2591
+    },
+    {
+      "epoch": 0.46240299705646243,
+      "grad_norm": 0.5619013905525208,
+      "learning_rate": 0.0004368916244491458,
+      "loss": 1.0051,
+      "step": 2592
+    },
+    {
+      "epoch": 0.46258139327446257,
+      "grad_norm": 0.5425794720649719,
+      "learning_rate": 0.00043684508265681065,
+      "loss": 1.3862,
+      "step": 2593
+    },
+    {
+      "epoch": 0.46275978949246277,
+      "grad_norm": 0.5217241644859314,
+      "learning_rate": 0.00043679852618974327,
+      "loss": 1.2932,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4629381857104629,
+      "grad_norm": 0.5080055594444275,
+      "learning_rate": 0.0004367519550515999,
+      "loss": 1.1496,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4631165819284631,
+      "grad_norm": 0.5273397564888,
+      "learning_rate": 0.0004367053692460385,
+      "loss": 1.1263,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4632949781464633,
+      "grad_norm": 0.5620000958442688,
+      "learning_rate": 0.0004366587687767176,
+      "loss": 0.9767,
+      "step": 2597
+    },
+    {
+      "epoch": 0.46347337436446345,
+      "grad_norm": 0.49073854088783264,
+      "learning_rate": 0.0004366121536472974,
+      "loss": 1.2086,
+      "step": 2598
+    },
+    {
+      "epoch": 0.46365177058246365,
+      "grad_norm": 0.5440914034843445,
+      "learning_rate": 0.000436565523861439,
+      "loss": 1.248,
+      "step": 2599
+    },
+    {
+      "epoch": 0.46383016680046385,
+      "grad_norm": 0.5432100892066956,
+      "learning_rate": 0.00043651887942280454,
+      "loss": 0.9665,
+      "step": 2600
+    },
+    {
+      "epoch": 0.464008563018464,
+      "grad_norm": 0.5001043081283569,
+      "learning_rate": 0.00043647222033505763,
+      "loss": 0.8812,
+      "step": 2601
+    },
+    {
+      "epoch": 0.4641869592364642,
+      "grad_norm": 0.4610103964805603,
+      "learning_rate": 0.0004364255466018627,
+      "loss": 0.9032,
+      "step": 2602
+    },
+    {
+      "epoch": 0.4643653554544644,
+      "grad_norm": 0.5426071286201477,
+      "learning_rate": 0.0004363788582268857,
+      "loss": 1.1128,
+      "step": 2603
+    },
+    {
+      "epoch": 0.46454375167246453,
+      "grad_norm": 0.5425922274589539,
+      "learning_rate": 0.00043633215521379326,
+      "loss": 1.134,
+      "step": 2604
+    },
+    {
+      "epoch": 0.46472214789046473,
+      "grad_norm": 0.5755911469459534,
+      "learning_rate": 0.0004362854375662536,
+      "loss": 1.345,
+      "step": 2605
+    },
+    {
+      "epoch": 0.4649005441084649,
+      "grad_norm": 0.5014548301696777,
+      "learning_rate": 0.0004362387052879358,
+      "loss": 1.0121,
+      "step": 2606
+    },
+    {
+      "epoch": 0.4650789403264651,
+      "grad_norm": 0.5047395825386047,
+      "learning_rate": 0.00043619195838251023,
+      "loss": 1.0983,
+      "step": 2607
+    },
+    {
+      "epoch": 0.46525733654446527,
+      "grad_norm": 0.5320248603820801,
+      "learning_rate": 0.0004361451968536484,
+      "loss": 0.9432,
+      "step": 2608
+    },
+    {
+      "epoch": 0.4654357327624654,
+      "grad_norm": 0.5131525993347168,
+      "learning_rate": 0.0004360984207050229,
+      "loss": 0.9492,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4656141289804656,
+      "grad_norm": 0.57234787940979,
+      "learning_rate": 0.0004360516299403075,
+      "loss": 1.3292,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4657925251984658,
+      "grad_norm": 0.502707302570343,
+      "learning_rate": 0.0004360048245631772,
+      "loss": 1.0897,
+      "step": 2611
+    },
+    {
+      "epoch": 0.46597092141646596,
+      "grad_norm": 0.5083783268928528,
+      "learning_rate": 0.00043595800457730795,
+      "loss": 0.9891,
+      "step": 2612
+    },
+    {
+      "epoch": 0.46614931763446615,
+      "grad_norm": 0.5372927188873291,
+      "learning_rate": 0.00043591116998637717,
+      "loss": 1.1991,
+      "step": 2613
+    },
+    {
+      "epoch": 0.46632771385246635,
+      "grad_norm": 0.46498778462409973,
+      "learning_rate": 0.00043586432079406297,
+      "loss": 0.8831,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4665061100704665,
+      "grad_norm": 0.5247387886047363,
+      "learning_rate": 0.0004358174570040451,
+      "loss": 1.2432,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4666845062884667,
+      "grad_norm": 0.5532581210136414,
+      "learning_rate": 0.000435770578620004,
+      "loss": 1.1609,
+      "step": 2616
+    },
+    {
+      "epoch": 0.46686290250646684,
+      "grad_norm": 0.5167216062545776,
+      "learning_rate": 0.0004357236856456217,
+      "loss": 1.0657,
+      "step": 2617
+    },
+    {
+      "epoch": 0.46704129872446704,
+      "grad_norm": 0.48156222701072693,
+      "learning_rate": 0.000435676778084581,
+      "loss": 1.1393,
+      "step": 2618
+    },
+    {
+      "epoch": 0.46721969494246723,
+      "grad_norm": 0.5826752185821533,
+      "learning_rate": 0.0004356298559405661,
+      "loss": 1.1713,
+      "step": 2619
+    },
+    {
+      "epoch": 0.4673980911604674,
+      "grad_norm": 0.5204340815544128,
+      "learning_rate": 0.00043558291921726215,
+      "loss": 1.1773,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4675764873784676,
+      "grad_norm": 0.5820581316947937,
+      "learning_rate": 0.00043553596791835557,
+      "loss": 0.9531,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4677548835964678,
+      "grad_norm": 0.535647988319397,
+      "learning_rate": 0.000435489002047534,
+      "loss": 0.8793,
+      "step": 2622
+    },
+    {
+      "epoch": 0.4679332798144679,
+      "grad_norm": 0.4820907413959503,
+      "learning_rate": 0.000435442021608486,
+      "loss": 1.0139,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4681116760324681,
+      "grad_norm": 0.5509294867515564,
+      "learning_rate": 0.0004353950266049014,
+      "loss": 1.0987,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4682900722504683,
+      "grad_norm": 0.554036021232605,
+      "learning_rate": 0.00043534801704047115,
+      "loss": 1.0175,
+      "step": 2625
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 0.5267329812049866,
+      "learning_rate": 0.0004353009929188875,
+      "loss": 1.0262,
+      "step": 2626
+    },
+    {
+      "epoch": 0.46864686468646866,
+      "grad_norm": 0.5234902501106262,
+      "learning_rate": 0.0004352539542438436,
+      "loss": 0.8124,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4688252609044688,
+      "grad_norm": 0.6302539706230164,
+      "learning_rate": 0.00043520690101903374,
+      "loss": 1.3212,
+      "step": 2628
+    },
+    {
+      "epoch": 0.469003657122469,
+      "grad_norm": 0.5413920879364014,
+      "learning_rate": 0.00043515983324815365,
+      "loss": 1.1461,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4691820533404692,
+      "grad_norm": 0.4936441481113434,
+      "learning_rate": 0.0004351127509349,
+      "loss": 0.9603,
+      "step": 2630
+    },
+    {
+      "epoch": 0.46936044955846934,
+      "grad_norm": 0.4682817757129669,
+      "learning_rate": 0.00043506565408297045,
+      "loss": 0.8536,
+      "step": 2631
+    },
+    {
+      "epoch": 0.46953884577646954,
+      "grad_norm": 0.6088268160820007,
+      "learning_rate": 0.0004350185426960641,
+      "loss": 1.1274,
+      "step": 2632
+    },
+    {
+      "epoch": 0.46971724199446974,
+      "grad_norm": 0.5248781442642212,
+      "learning_rate": 0.00043497141677788107,
+      "loss": 1.1042,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4698956382124699,
+      "grad_norm": 0.48056891560554504,
+      "learning_rate": 0.0004349242763321225,
+      "loss": 1.0877,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4700740344304701,
+      "grad_norm": 0.48219484090805054,
+      "learning_rate": 0.0004348771213624909,
+      "loss": 1.0188,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4702524306484703,
+      "grad_norm": 0.5074059963226318,
+      "learning_rate": 0.0004348299518726897,
+      "loss": 0.8514,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4704308268664704,
+      "grad_norm": 0.48974233865737915,
+      "learning_rate": 0.00043478276786642364,
+      "loss": 1.0564,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4706092230844706,
+      "grad_norm": 0.45620694756507874,
+      "learning_rate": 0.0004347355693473985,
+      "loss": 0.8984,
+      "step": 2638
+    },
+    {
+      "epoch": 0.47078761930247076,
+      "grad_norm": 0.5297985672950745,
+      "learning_rate": 0.00043468835631932124,
+      "loss": 0.9644,
+      "step": 2639
+    },
+    {
+      "epoch": 0.47096601552047096,
+      "grad_norm": 0.5142080187797546,
+      "learning_rate": 0.0004346411287859,
+      "loss": 1.0379,
+      "step": 2640
+    },
+    {
+      "epoch": 0.47114441173847116,
+      "grad_norm": 0.5011320114135742,
+      "learning_rate": 0.00043459388675084386,
+      "loss": 1.1448,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4713228079564713,
+      "grad_norm": 0.5459404587745667,
+      "learning_rate": 0.00043454663021786337,
+      "loss": 1.0494,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4715012041744715,
+      "grad_norm": 0.4870483875274658,
+      "learning_rate": 0.00043449935919066997,
+      "loss": 1.1025,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4716796003924717,
+      "grad_norm": 0.4572905898094177,
+      "learning_rate": 0.00043445207367297624,
+      "loss": 1.0561,
+      "step": 2644
+    },
+    {
+      "epoch": 0.47185799661047184,
+      "grad_norm": 0.5108131766319275,
+      "learning_rate": 0.00043440477366849607,
+      "loss": 0.8723,
+      "step": 2645
+    },
+    {
+      "epoch": 0.47203639282847204,
+      "grad_norm": 0.49937039613723755,
+      "learning_rate": 0.00043435745918094437,
+      "loss": 1.0159,
+      "step": 2646
+    },
+    {
+      "epoch": 0.47221478904647224,
+      "grad_norm": 0.5099616050720215,
+      "learning_rate": 0.00043431013021403707,
+      "loss": 0.9302,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4723931852644724,
+      "grad_norm": 0.5088136792182922,
+      "learning_rate": 0.0004342627867714915,
+      "loss": 1.054,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4725715814824726,
+      "grad_norm": 0.48349374532699585,
+      "learning_rate": 0.0004342154288570259,
+      "loss": 1.0577,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4727499777004727,
+      "grad_norm": 0.528062105178833,
+      "learning_rate": 0.0004341680564743599,
+      "loss": 1.0324,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4729283739184729,
+      "grad_norm": 1.0809184312820435,
+      "learning_rate": 0.0004341206696272139,
+      "loss": 1.1805,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4731067701364731,
+      "grad_norm": 0.5047591328620911,
+      "learning_rate": 0.00043407326831930985,
+      "loss": 1.0002,
+      "step": 2652
+    },
+    {
+      "epoch": 0.47328516635447326,
+      "grad_norm": 0.5000717639923096,
+      "learning_rate": 0.00043402585255437035,
+      "loss": 1.0015,
+      "step": 2653
+    },
+    {
+      "epoch": 0.47346356257247346,
+      "grad_norm": 0.608333170413971,
+      "learning_rate": 0.0004339784223361197,
+      "loss": 0.9044,
+      "step": 2654
+    },
+    {
+      "epoch": 0.47364195879047366,
+      "grad_norm": 0.5130841135978699,
+      "learning_rate": 0.00043393097766828293,
+      "loss": 1.1026,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4738203550084738,
+      "grad_norm": 0.5287538170814514,
+      "learning_rate": 0.0004338835185545863,
+      "loss": 1.0682,
+      "step": 2656
+    },
+    {
+      "epoch": 0.473998751226474,
+      "grad_norm": 0.5100160837173462,
+      "learning_rate": 0.00043383604499875727,
+      "loss": 1.231,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4741771474444742,
+      "grad_norm": 0.4938991963863373,
+      "learning_rate": 0.0004337885570045244,
+      "loss": 1.0292,
+      "step": 2658
+    },
+    {
+      "epoch": 0.47435554366247434,
+      "grad_norm": 0.45146873593330383,
+      "learning_rate": 0.0004337410545756173,
+      "loss": 0.8997,
+      "step": 2659
+    },
+    {
+      "epoch": 0.47453393988047454,
+      "grad_norm": 0.45835426449775696,
+      "learning_rate": 0.0004336935377157668,
+      "loss": 0.773,
+      "step": 2660
+    },
+    {
+      "epoch": 0.4747123360984747,
+      "grad_norm": 0.4573782980442047,
+      "learning_rate": 0.000433646006428705,
+      "loss": 0.8717,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4748907323164749,
+      "grad_norm": 0.4976518154144287,
+      "learning_rate": 0.00043359846071816484,
+      "loss": 0.862,
+      "step": 2662
+    },
+    {
+      "epoch": 0.4750691285344751,
+      "grad_norm": 0.5328902006149292,
+      "learning_rate": 0.0004335509005878806,
+      "loss": 0.9372,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4752475247524752,
+      "grad_norm": 0.5134164690971375,
+      "learning_rate": 0.0004335033260415876,
+      "loss": 0.9136,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4754259209704754,
+      "grad_norm": 0.5302520990371704,
+      "learning_rate": 0.00043345573708302235,
+      "loss": 1.0568,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4756043171884756,
+      "grad_norm": 0.4945065379142761,
+      "learning_rate": 0.0004334081337159225,
+      "loss": 0.8435,
+      "step": 2666
+    },
+    {
+      "epoch": 0.47578271340647577,
+      "grad_norm": 1.0435439348220825,
+      "learning_rate": 0.0004333605159440266,
+      "loss": 1.0573,
+      "step": 2667
+    },
+    {
+      "epoch": 0.47596110962447596,
+      "grad_norm": 2.6927244663238525,
+      "learning_rate": 0.0004333128837710748,
+      "loss": 1.0532,
+      "step": 2668
+    },
+    {
+      "epoch": 0.47613950584247616,
+      "grad_norm": 0.5720566511154175,
+      "learning_rate": 0.000433265237200808,
+      "loss": 1.0378,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4763179020604763,
+      "grad_norm": 0.49117255210876465,
+      "learning_rate": 0.00043321757623696836,
+      "loss": 1.0109,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4764962982784765,
+      "grad_norm": 0.5019311904907227,
+      "learning_rate": 0.00043316990088329907,
+      "loss": 1.0556,
+      "step": 2671
+    },
+    {
+      "epoch": 0.47667469449647665,
+      "grad_norm": 1.1491711139678955,
+      "learning_rate": 0.00043312221114354463,
+      "loss": 1.0185,
+      "step": 2672
+    },
+    {
+      "epoch": 0.47685309071447685,
+      "grad_norm": 1.0251749753952026,
+      "learning_rate": 0.00043307450702145047,
+      "loss": 1.1779,
+      "step": 2673
+    },
+    {
+      "epoch": 0.47703148693247704,
+      "grad_norm": 0.5985758900642395,
+      "learning_rate": 0.00043302678852076337,
+      "loss": 1.1387,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4772098831504772,
+      "grad_norm": 0.6510484218597412,
+      "learning_rate": 0.000432979055645231,
+      "loss": 1.1599,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4773882793684774,
+      "grad_norm": 0.5804511904716492,
+      "learning_rate": 0.0004329313083986024,
+      "loss": 1.0075,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4775666755864776,
+      "grad_norm": 0.5480517745018005,
+      "learning_rate": 0.00043288354678462757,
+      "loss": 1.1907,
+      "step": 2677
+    },
+    {
+      "epoch": 0.47774507180447773,
+      "grad_norm": 0.5310084223747253,
+      "learning_rate": 0.0004328357708070576,
+      "loss": 0.9193,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4779234680224779,
+      "grad_norm": 0.5491518974304199,
+      "learning_rate": 0.00043278798046964494,
+      "loss": 1.2297,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4781018642404781,
+      "grad_norm": 0.4863007664680481,
+      "learning_rate": 0.0004327401757761429,
+      "loss": 0.9668,
+      "step": 2680
+    },
+    {
+      "epoch": 0.47828026045847827,
+      "grad_norm": 0.5505911111831665,
+      "learning_rate": 0.00043269235673030614,
+      "loss": 1.0962,
+      "step": 2681
+    },
+    {
+      "epoch": 0.47845865667647847,
+      "grad_norm": 0.4990256428718567,
+      "learning_rate": 0.00043264452333589034,
+      "loss": 0.8941,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4786370528944786,
+      "grad_norm": 0.5696044564247131,
+      "learning_rate": 0.0004325966755966522,
+      "loss": 1.1817,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4788154491124788,
+      "grad_norm": 0.5234004259109497,
+      "learning_rate": 0.00043254881351634976,
+      "loss": 1.0851,
+      "step": 2684
+    },
+    {
+      "epoch": 0.478993845330479,
+      "grad_norm": 0.5903022289276123,
+      "learning_rate": 0.0004325009370987421,
+      "loss": 1.2182,
+      "step": 2685
+    },
+    {
+      "epoch": 0.47917224154847915,
+      "grad_norm": 0.5186327695846558,
+      "learning_rate": 0.0004324530463475893,
+      "loss": 0.9566,
+      "step": 2686
+    },
+    {
+      "epoch": 0.47935063776647935,
+      "grad_norm": 0.5512902736663818,
+      "learning_rate": 0.00043240514126665274,
+      "loss": 0.9735,
+      "step": 2687
+    },
+    {
+      "epoch": 0.47952903398447955,
+      "grad_norm": 0.5918879508972168,
+      "learning_rate": 0.00043235722185969497,
+      "loss": 1.1593,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4797074302024797,
+      "grad_norm": 0.5881208777427673,
+      "learning_rate": 0.0004323092881304794,
+      "loss": 1.0421,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4798858264204799,
+      "grad_norm": 0.5843618512153625,
+      "learning_rate": 0.00043226134008277084,
+      "loss": 1.199,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4800642226384801,
+      "grad_norm": 0.5014129877090454,
+      "learning_rate": 0.0004322133777203351,
+      "loss": 0.852,
+      "step": 2691
+    },
+    {
+      "epoch": 0.48024261885648023,
+      "grad_norm": 0.5492891073226929,
+      "learning_rate": 0.000432165401046939,
+      "loss": 0.9382,
+      "step": 2692
+    },
+    {
+      "epoch": 0.48042101507448043,
+      "grad_norm": 0.5708451271057129,
+      "learning_rate": 0.00043211741006635076,
+      "loss": 0.9182,
+      "step": 2693
+    },
+    {
+      "epoch": 0.48059941129248057,
+      "grad_norm": 0.5316191911697388,
+      "learning_rate": 0.00043206940478233947,
+      "loss": 1.05,
+      "step": 2694
+    },
+    {
+      "epoch": 0.48077780751048077,
+      "grad_norm": 0.5057091116905212,
+      "learning_rate": 0.0004320213851986755,
+      "loss": 0.8626,
+      "step": 2695
+    },
+    {
+      "epoch": 0.48095620372848097,
+      "grad_norm": 0.5974101424217224,
+      "learning_rate": 0.00043197335131913025,
+      "loss": 1.1793,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4811345999464811,
+      "grad_norm": 0.5790625214576721,
+      "learning_rate": 0.00043192530314747625,
+      "loss": 1.0063,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4813129961644813,
+      "grad_norm": 0.4502911865711212,
+      "learning_rate": 0.0004318772406874873,
+      "loss": 0.8672,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4814913923824815,
+      "grad_norm": 0.5690745711326599,
+      "learning_rate": 0.00043182916394293817,
+      "loss": 0.9999,
+      "step": 2699
+    },
+    {
+      "epoch": 0.48166978860048165,
+      "grad_norm": 0.584483802318573,
+      "learning_rate": 0.00043178107291760463,
+      "loss": 1.0083,
+      "step": 2700
+    },
+    {
+      "epoch": 0.48184818481848185,
+      "grad_norm": 0.48927435278892517,
+      "learning_rate": 0.00043173296761526395,
+      "loss": 0.7844,
+      "step": 2701
+    },
+    {
+      "epoch": 0.48202658103648205,
+      "grad_norm": 0.6203662157058716,
+      "learning_rate": 0.0004316848480396941,
+      "loss": 0.9332,
+      "step": 2702
+    },
+    {
+      "epoch": 0.4822049772544822,
+      "grad_norm": 0.5448145866394043,
+      "learning_rate": 0.00043163671419467454,
+      "loss": 1.0309,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4823833734724824,
+      "grad_norm": 0.540557324886322,
+      "learning_rate": 0.0004315885660839857,
+      "loss": 0.9622,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4825617696904826,
+      "grad_norm": 0.4740718603134155,
+      "learning_rate": 0.0004315404037114089,
+      "loss": 1.003,
+      "step": 2705
+    },
+    {
+      "epoch": 0.48274016590848273,
+      "grad_norm": 0.5453110933303833,
+      "learning_rate": 0.0004314922270807269,
+      "loss": 1.2003,
+      "step": 2706
+    },
+    {
+      "epoch": 0.48291856212648293,
+      "grad_norm": 0.4645708501338959,
+      "learning_rate": 0.0004314440361957235,
+      "loss": 0.9038,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4830969583444831,
+      "grad_norm": 0.48761603236198425,
+      "learning_rate": 0.00043139583106018367,
+      "loss": 1.1861,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4832753545624833,
+      "grad_norm": 0.48210960626602173,
+      "learning_rate": 0.0004313476116778933,
+      "loss": 1.0031,
+      "step": 2709
+    },
+    {
+      "epoch": 0.48345375078048347,
+      "grad_norm": 0.4890519380569458,
+      "learning_rate": 0.00043129937805263944,
+      "loss": 0.9203,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4836321469984836,
+      "grad_norm": 0.5152431726455688,
+      "learning_rate": 0.00043125113018821054,
+      "loss": 1.001,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4838105432164838,
+      "grad_norm": 0.5150320529937744,
+      "learning_rate": 0.00043120286808839587,
+      "loss": 0.9304,
+      "step": 2712
+    },
+    {
+      "epoch": 0.483988939434484,
+      "grad_norm": 0.4525447487831116,
+      "learning_rate": 0.0004311545917569859,
+      "loss": 0.8478,
+      "step": 2713
+    },
+    {
+      "epoch": 0.48416733565248415,
+      "grad_norm": 0.47689542174339294,
+      "learning_rate": 0.0004311063011977723,
+      "loss": 1.032,
+      "step": 2714
+    },
+    {
+      "epoch": 0.48434573187048435,
+      "grad_norm": 0.5542166829109192,
+      "learning_rate": 0.0004310579964145477,
+      "loss": 1.2234,
+      "step": 2715
+    },
+    {
+      "epoch": 0.48452412808848455,
+      "grad_norm": 0.513331949710846,
+      "learning_rate": 0.00043100967741110593,
+      "loss": 0.9973,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4847025243064847,
+      "grad_norm": 0.4602411389350891,
+      "learning_rate": 0.000430961344191242,
+      "loss": 0.8423,
+      "step": 2717
+    },
+    {
+      "epoch": 0.4848809205244849,
+      "grad_norm": 0.5340333580970764,
+      "learning_rate": 0.000430912996758752,
+      "loss": 1.2688,
+      "step": 2718
+    },
+    {
+      "epoch": 0.48505931674248504,
+      "grad_norm": 0.47053262591362,
+      "learning_rate": 0.00043086463511743313,
+      "loss": 0.8783,
+      "step": 2719
+    },
+    {
+      "epoch": 0.48523771296048523,
+      "grad_norm": 0.5520641207695007,
+      "learning_rate": 0.0004308162592710836,
+      "loss": 1.1515,
+      "step": 2720
+    },
+    {
+      "epoch": 0.48541610917848543,
+      "grad_norm": 0.48937782645225525,
+      "learning_rate": 0.0004307678692235029,
+      "loss": 1.0731,
+      "step": 2721
+    },
+    {
+      "epoch": 0.4855945053964856,
+      "grad_norm": 0.5234431624412537,
+      "learning_rate": 0.0004307194649784915,
+      "loss": 1.081,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4857729016144858,
+      "grad_norm": 0.5052714943885803,
+      "learning_rate": 0.0004306710465398511,
+      "loss": 1.2472,
+      "step": 2723
+    },
+    {
+      "epoch": 0.485951297832486,
+      "grad_norm": 0.466281920671463,
+      "learning_rate": 0.00043062261391138454,
+      "loss": 0.875,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4861296940504861,
+      "grad_norm": 0.47504422068595886,
+      "learning_rate": 0.00043057416709689554,
+      "loss": 1.0092,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4863080902684863,
+      "grad_norm": 0.46300631761550903,
+      "learning_rate": 0.00043052570610018913,
+      "loss": 0.9929,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 0.4238641560077667,
+      "learning_rate": 0.0004304772309250715,
+      "loss": 0.8075,
+      "step": 2727
+    },
+    {
+      "epoch": 0.48666488270448666,
+      "grad_norm": 0.509427011013031,
+      "learning_rate": 0.00043042874157534985,
+      "loss": 0.9091,
+      "step": 2728
+    },
+    {
+      "epoch": 0.48684327892248686,
+      "grad_norm": 1.843088150024414,
+      "learning_rate": 0.0004303802380548324,
+      "loss": 1.6363,
+      "step": 2729
+    },
+    {
+      "epoch": 0.487021675140487,
+      "grad_norm": 0.5678585767745972,
+      "learning_rate": 0.0004303317203673287,
+      "loss": 1.1911,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4872000713584872,
+      "grad_norm": 0.49592193961143494,
+      "learning_rate": 0.00043028318851664934,
+      "loss": 0.8901,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4873784675764874,
+      "grad_norm": 0.4731305241584778,
+      "learning_rate": 0.0004302346425066059,
+      "loss": 0.9217,
+      "step": 2732
+    },
+    {
+      "epoch": 0.48755686379448754,
+      "grad_norm": 0.6168728470802307,
+      "learning_rate": 0.0004301860823410112,
+      "loss": 1.081,
+      "step": 2733
+    },
+    {
+      "epoch": 0.48773526001248774,
+      "grad_norm": 0.48209506273269653,
+      "learning_rate": 0.00043013750802367915,
+      "loss": 0.7949,
+      "step": 2734
+    },
+    {
+      "epoch": 0.48791365623048794,
+      "grad_norm": 0.7890462875366211,
+      "learning_rate": 0.00043008891955842465,
+      "loss": 1.1409,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4880920524484881,
+      "grad_norm": 2.6161890029907227,
+      "learning_rate": 0.00043004031694906397,
+      "loss": 1.1929,
+      "step": 2736
+    },
+    {
+      "epoch": 0.4882704486664883,
+      "grad_norm": 0.7953986525535583,
+      "learning_rate": 0.0004299917001994143,
+      "loss": 0.956,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4884488448844885,
+      "grad_norm": 1.4645198583602905,
+      "learning_rate": 0.00042994306931329394,
+      "loss": 1.0847,
+      "step": 2738
+    },
+    {
+      "epoch": 0.4886272411024886,
+      "grad_norm": 0.5251899361610413,
+      "learning_rate": 0.00042989442429452237,
+      "loss": 1.0796,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4888056373204888,
+      "grad_norm": 0.4881310760974884,
+      "learning_rate": 0.0004298457651469201,
+      "loss": 0.9101,
+      "step": 2740
+    },
+    {
+      "epoch": 0.48898403353848896,
+      "grad_norm": 0.7399401068687439,
+      "learning_rate": 0.0004297970918743088,
+      "loss": 1.1056,
+      "step": 2741
+    },
+    {
+      "epoch": 0.48916242975648916,
+      "grad_norm": 0.531498908996582,
+      "learning_rate": 0.00042974840448051135,
+      "loss": 1.104,
+      "step": 2742
+    },
+    {
+      "epoch": 0.48934082597448936,
+      "grad_norm": 0.5090386867523193,
+      "learning_rate": 0.00042969970296935153,
+      "loss": 0.7713,
+      "step": 2743
+    },
+    {
+      "epoch": 0.4895192221924895,
+      "grad_norm": 0.5436068177223206,
+      "learning_rate": 0.00042965098734465434,
+      "loss": 0.9884,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4896976184104897,
+      "grad_norm": 0.5477145314216614,
+      "learning_rate": 0.00042960225761024597,
+      "loss": 1.1816,
+      "step": 2745
+    },
+    {
+      "epoch": 0.4898760146284899,
+      "grad_norm": 0.5642074942588806,
+      "learning_rate": 0.00042955351376995355,
+      "loss": 1.057,
+      "step": 2746
+    },
+    {
+      "epoch": 0.49005441084649004,
+      "grad_norm": 0.5721654295921326,
+      "learning_rate": 0.0004295047558276054,
+      "loss": 0.8484,
+      "step": 2747
+    },
+    {
+      "epoch": 0.49023280706449024,
+      "grad_norm": 0.4718813896179199,
+      "learning_rate": 0.000429455983787031,
+      "loss": 0.8456,
+      "step": 2748
+    },
+    {
+      "epoch": 0.49041120328249044,
+      "grad_norm": 0.4769846498966217,
+      "learning_rate": 0.0004294071976520608,
+      "loss": 0.7766,
+      "step": 2749
+    },
+    {
+      "epoch": 0.4905895995004906,
+      "grad_norm": 0.571636438369751,
+      "learning_rate": 0.0004293583974265266,
+      "loss": 0.921,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4907679957184908,
+      "grad_norm": 0.5464839935302734,
+      "learning_rate": 0.000429309583114261,
+      "loss": 1.1658,
+      "step": 2751
+    },
+    {
+      "epoch": 0.4909463919364909,
+      "grad_norm": 0.5281213521957397,
+      "learning_rate": 0.00042926075471909787,
+      "loss": 1.1353,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4911247881544911,
+      "grad_norm": 0.5068462491035461,
+      "learning_rate": 0.0004292119122448723,
+      "loss": 0.9014,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4913031843724913,
+      "grad_norm": 0.532954752445221,
+      "learning_rate": 0.0004291630556954202,
+      "loss": 1.0867,
+      "step": 2754
+    },
+    {
+      "epoch": 0.49148158059049146,
+      "grad_norm": 0.48287615180015564,
+      "learning_rate": 0.0004291141850745788,
+      "loss": 0.9296,
+      "step": 2755
+    },
+    {
+      "epoch": 0.49165997680849166,
+      "grad_norm": 0.4885706603527069,
+      "learning_rate": 0.0004290653003861864,
+      "loss": 0.9466,
+      "step": 2756
+    },
+    {
+      "epoch": 0.49183837302649186,
+      "grad_norm": 0.5004509091377258,
+      "learning_rate": 0.0004290164016340824,
+      "loss": 0.9552,
+      "step": 2757
+    },
+    {
+      "epoch": 0.492016769244492,
+      "grad_norm": 0.5846733450889587,
+      "learning_rate": 0.0004289674888221072,
+      "loss": 1.3886,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4921951654624922,
+      "grad_norm": 0.4464746415615082,
+      "learning_rate": 0.00042891856195410237,
+      "loss": 0.7713,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4923735616804924,
+      "grad_norm": 0.48179852962493896,
+      "learning_rate": 0.0004288696210339108,
+      "loss": 0.9217,
+      "step": 2760
+    },
+    {
+      "epoch": 0.49255195789849254,
+      "grad_norm": 0.5003469586372375,
+      "learning_rate": 0.0004288206660653762,
+      "loss": 0.8922,
+      "step": 2761
+    },
+    {
+      "epoch": 0.49273035411649274,
+      "grad_norm": 0.5266236066818237,
+      "learning_rate": 0.00042877169705234335,
+      "loss": 0.9248,
+      "step": 2762
+    },
+    {
+      "epoch": 0.4929087503344929,
+      "grad_norm": 0.5324473977088928,
+      "learning_rate": 0.00042872271399865835,
+      "loss": 1.0357,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4930871465524931,
+      "grad_norm": 0.7425055503845215,
+      "learning_rate": 0.0004286737169081684,
+      "loss": 1.026,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4932655427704933,
+      "grad_norm": 0.8448666334152222,
+      "learning_rate": 0.0004286247057847215,
+      "loss": 0.9976,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4934439389884934,
+      "grad_norm": 0.5740126371383667,
+      "learning_rate": 0.0004285756806321671,
+      "loss": 1.1577,
+      "step": 2766
+    },
+    {
+      "epoch": 0.4936223352064936,
+      "grad_norm": 0.5506383180618286,
+      "learning_rate": 0.0004285266414543556,
+      "loss": 1.0981,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4938007314244938,
+      "grad_norm": 0.49879491329193115,
+      "learning_rate": 0.00042847758825513847,
+      "loss": 0.8611,
+      "step": 2768
+    },
+    {
+      "epoch": 0.49397912764249396,
+      "grad_norm": 0.5577161908149719,
+      "learning_rate": 0.0004284285210383685,
+      "loss": 0.8755,
+      "step": 2769
+    },
+    {
+      "epoch": 0.49415752386049416,
+      "grad_norm": 1.8733638525009155,
+      "learning_rate": 0.00042837943980789914,
+      "loss": 1.0863,
+      "step": 2770
+    },
+    {
+      "epoch": 0.49433592007849436,
+      "grad_norm": 0.620151162147522,
+      "learning_rate": 0.00042833034456758533,
+      "loss": 1.0697,
+      "step": 2771
+    },
+    {
+      "epoch": 0.4945143162964945,
+      "grad_norm": 0.6251631379127502,
+      "learning_rate": 0.00042828123532128305,
+      "loss": 1.163,
+      "step": 2772
+    },
+    {
+      "epoch": 0.4946927125144947,
+      "grad_norm": 0.55173259973526,
+      "learning_rate": 0.0004282321120728493,
+      "loss": 1.0081,
+      "step": 2773
+    },
+    {
+      "epoch": 0.49487110873249485,
+      "grad_norm": 0.5649335384368896,
+      "learning_rate": 0.0004281829748261421,
+      "loss": 1.2093,
+      "step": 2774
+    },
+    {
+      "epoch": 0.49504950495049505,
+      "grad_norm": 0.5768599510192871,
+      "learning_rate": 0.0004281338235850208,
+      "loss": 1.1961,
+      "step": 2775
+    },
+    {
+      "epoch": 0.49522790116849524,
+      "grad_norm": 0.4870094358921051,
+      "learning_rate": 0.0004280846583533456,
+      "loss": 1.0099,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4954062973864954,
+      "grad_norm": 0.5303916335105896,
+      "learning_rate": 0.00042803547913497795,
+      "loss": 0.934,
+      "step": 2777
+    },
+    {
+      "epoch": 0.4955846936044956,
+      "grad_norm": 1.3991183042526245,
+      "learning_rate": 0.00042798628593378044,
+      "loss": 1.0462,
+      "step": 2778
+    },
+    {
+      "epoch": 0.4957630898224958,
+      "grad_norm": 0.5671733617782593,
+      "learning_rate": 0.00042793707875361667,
+      "loss": 1.0136,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4959414860404959,
+      "grad_norm": 0.5110854506492615,
+      "learning_rate": 0.00042788785759835117,
+      "loss": 0.9371,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4961198822584961,
+      "grad_norm": 0.48012104630470276,
+      "learning_rate": 0.00042783862247185,
+      "loss": 1.0671,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4962982784764963,
+      "grad_norm": 1.609937310218811,
+      "learning_rate": 0.0004277893733779798,
+      "loss": 1.1734,
+      "step": 2782
+    },
+    {
+      "epoch": 0.49647667469449647,
+      "grad_norm": 0.49931612610816956,
+      "learning_rate": 0.0004277401103206089,
+      "loss": 0.8801,
+      "step": 2783
+    },
+    {
+      "epoch": 0.49665507091249667,
+      "grad_norm": 0.5576810240745544,
+      "learning_rate": 0.0004276908333036061,
+      "loss": 0.998,
+      "step": 2784
+    },
+    {
+      "epoch": 0.4968334671304968,
+      "grad_norm": 0.5244370102882385,
+      "learning_rate": 0.00042764154233084184,
+      "loss": 1.1198,
+      "step": 2785
+    },
+    {
+      "epoch": 0.497011863348497,
+      "grad_norm": 0.5110731720924377,
+      "learning_rate": 0.00042759223740618723,
+      "loss": 1.0418,
+      "step": 2786
+    },
+    {
+      "epoch": 0.4971902595664972,
+      "grad_norm": 0.5166656374931335,
+      "learning_rate": 0.0004275429185335147,
+      "loss": 1.0333,
+      "step": 2787
+    },
+    {
+      "epoch": 0.49736865578449735,
+      "grad_norm": 0.558217465877533,
+      "learning_rate": 0.00042749358571669783,
+      "loss": 1.0004,
+      "step": 2788
+    },
+    {
+      "epoch": 0.49754705200249755,
+      "grad_norm": 0.5724664926528931,
+      "learning_rate": 0.000427444238959611,
+      "loss": 1.0876,
+      "step": 2789
+    },
+    {
+      "epoch": 0.49772544822049775,
+      "grad_norm": 0.5535562038421631,
+      "learning_rate": 0.00042739487826613006,
+      "loss": 1.0584,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4979038444384979,
+      "grad_norm": 0.6063182950019836,
+      "learning_rate": 0.0004273455036401317,
+      "loss": 0.9144,
+      "step": 2791
+    },
+    {
+      "epoch": 0.4980822406564981,
+      "grad_norm": 0.6180481314659119,
+      "learning_rate": 0.00042729611508549384,
+      "loss": 1.1755,
+      "step": 2792
+    },
+    {
+      "epoch": 0.4982606368744983,
+      "grad_norm": 0.524019718170166,
+      "learning_rate": 0.0004272467126060954,
+      "loss": 1.0019,
+      "step": 2793
+    },
+    {
+      "epoch": 0.49843903309249843,
+      "grad_norm": 0.49826425313949585,
+      "learning_rate": 0.00042719729620581637,
+      "loss": 0.9266,
+      "step": 2794
+    },
+    {
+      "epoch": 0.4986174293104986,
+      "grad_norm": 0.5382309556007385,
+      "learning_rate": 0.000427147865888538,
+      "loss": 1.105,
+      "step": 2795
+    },
+    {
+      "epoch": 0.49879582552849877,
+      "grad_norm": 0.4761876165866852,
+      "learning_rate": 0.0004270984216581425,
+      "loss": 0.8737,
+      "step": 2796
+    },
+    {
+      "epoch": 0.49897422174649897,
+      "grad_norm": 0.4881121516227722,
+      "learning_rate": 0.0004270489635185131,
+      "loss": 0.9282,
+      "step": 2797
+    },
+    {
+      "epoch": 0.49915261796449917,
+      "grad_norm": 0.6625795960426331,
+      "learning_rate": 0.00042699949147353435,
+      "loss": 1.1568,
+      "step": 2798
+    },
+    {
+      "epoch": 0.4993310141824993,
+      "grad_norm": 0.5215954184532166,
+      "learning_rate": 0.00042695000552709164,
+      "loss": 1.0541,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4995094104004995,
+      "grad_norm": 0.4769946336746216,
+      "learning_rate": 0.0004269005056830717,
+      "loss": 0.9529,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4996878066184997,
+      "grad_norm": 0.46005573868751526,
+      "learning_rate": 0.00042685099194536216,
+      "loss": 1.0152,
+      "step": 2801
+    },
+    {
+      "epoch": 0.49986620283649985,
+      "grad_norm": 0.8600802421569824,
+      "learning_rate": 0.00042680146431785184,
+      "loss": 0.891,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5000445990545,
+      "grad_norm": 0.47280198335647583,
+      "learning_rate": 0.00042675192280443053,
+      "loss": 1.0193,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5002229952725002,
+      "grad_norm": 0.44590240716934204,
+      "learning_rate": 0.00042670236740898935,
+      "loss": 0.8403,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5004013914905004,
+      "grad_norm": 0.5397332906723022,
+      "learning_rate": 0.00042665279813542024,
+      "loss": 1.0327,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5005797877085005,
+      "grad_norm": 0.49650824069976807,
+      "learning_rate": 0.0004266032149876163,
+      "loss": 0.8893,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5007581839265007,
+      "grad_norm": 0.5721169710159302,
+      "learning_rate": 0.000426553617969472,
+      "loss": 0.8712,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5009365801445009,
+      "grad_norm": 0.4694698750972748,
+      "learning_rate": 0.00042650400708488245,
+      "loss": 0.8884,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5011149763625011,
+      "grad_norm": 0.5037152767181396,
+      "learning_rate": 0.00042645438233774414,
+      "loss": 1.2473,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5012933725805013,
+      "grad_norm": 0.4780758321285248,
+      "learning_rate": 0.00042640474373195457,
+      "loss": 0.9399,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5014717687985015,
+      "grad_norm": 0.5098987817764282,
+      "learning_rate": 0.00042635509127141236,
+      "loss": 1.0123,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5016501650165016,
+      "grad_norm": 0.4265352487564087,
+      "learning_rate": 0.0004263054249600172,
+      "loss": 0.704,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5018285612345018,
+      "grad_norm": 0.45242688059806824,
+      "learning_rate": 0.0004262557448016697,
+      "loss": 0.8694,
+      "step": 2813
+    },
+    {
+      "epoch": 0.502006957452502,
+      "grad_norm": 0.503936767578125,
+      "learning_rate": 0.00042620605080027197,
+      "loss": 0.9914,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5021853536705022,
+      "grad_norm": 0.45868226885795593,
+      "learning_rate": 0.0004261563429597268,
+      "loss": 0.9463,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5023637498885024,
+      "grad_norm": 0.5536309480667114,
+      "learning_rate": 0.0004261066212839383,
+      "loss": 1.2046,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5025421461065025,
+      "grad_norm": 0.43166589736938477,
+      "learning_rate": 0.00042605688577681156,
+      "loss": 0.793,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5027205423245027,
+      "grad_norm": 0.5023233294487,
+      "learning_rate": 0.00042600713644225274,
+      "loss": 1.0378,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5028989385425029,
+      "grad_norm": 0.5103394985198975,
+      "learning_rate": 0.0004259573732841692,
+      "loss": 1.0274,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5030773347605031,
+      "grad_norm": 0.4885903298854828,
+      "learning_rate": 0.0004259075963064692,
+      "loss": 0.9873,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5032557309785033,
+      "grad_norm": 0.46850115060806274,
+      "learning_rate": 0.0004258578055130623,
+      "loss": 1.0053,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5034341271965035,
+      "grad_norm": 1.0507224798202515,
+      "learning_rate": 0.0004258080009078591,
+      "loss": 0.9356,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5036125234145036,
+      "grad_norm": 0.48565196990966797,
+      "learning_rate": 0.0004257581824947711,
+      "loss": 0.9913,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5037909196325038,
+      "grad_norm": 0.4672144651412964,
+      "learning_rate": 0.00042570835027771114,
+      "loss": 1.0402,
+      "step": 2824
+    },
+    {
+      "epoch": 0.503969315850504,
+      "grad_norm": 0.5116181373596191,
+      "learning_rate": 0.00042565850426059295,
+      "loss": 1.0377,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5041477120685042,
+      "grad_norm": 0.5363556742668152,
+      "learning_rate": 0.0004256086444473314,
+      "loss": 1.0358,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5043261082865044,
+      "grad_norm": 0.4890466332435608,
+      "learning_rate": 0.0004255587708418425,
+      "loss": 1.0442,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 0.50138258934021,
+      "learning_rate": 0.0004255088834480433,
+      "loss": 1.009,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5046829007225047,
+      "grad_norm": 0.5070481300354004,
+      "learning_rate": 0.00042545898226985186,
+      "loss": 1.0232,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5048612969405049,
+      "grad_norm": 0.5242338180541992,
+      "learning_rate": 0.00042540906731118746,
+      "loss": 1.099,
+      "step": 2830
+    },
+    {
+      "epoch": 0.505039693158505,
+      "grad_norm": 0.5118682980537415,
+      "learning_rate": 0.00042535913857597046,
+      "loss": 1.0284,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5052180893765053,
+      "grad_norm": 0.4739261269569397,
+      "learning_rate": 0.00042530919606812215,
+      "loss": 1.0342,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5053964855945055,
+      "grad_norm": 0.4524725675582886,
+      "learning_rate": 0.00042525923979156507,
+      "loss": 0.8644,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5055748818125055,
+      "grad_norm": 0.499525249004364,
+      "learning_rate": 0.00042520926975022266,
+      "loss": 1.0513,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5057532780305057,
+      "grad_norm": 0.4836816191673279,
+      "learning_rate": 0.00042515928594801964,
+      "loss": 0.8886,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5059316742485059,
+      "grad_norm": 0.5527642965316772,
+      "learning_rate": 0.00042510928838888163,
+      "loss": 1.2704,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5061100704665061,
+      "grad_norm": 0.5419808030128479,
+      "learning_rate": 0.0004250592770767355,
+      "loss": 1.1396,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5062884666845063,
+      "grad_norm": 0.4811457395553589,
+      "learning_rate": 0.0004250092520155091,
+      "loss": 0.9047,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5064668629025064,
+      "grad_norm": 0.4570630192756653,
+      "learning_rate": 0.0004249592132091313,
+      "loss": 0.8595,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5066452591205066,
+      "grad_norm": 0.5125581622123718,
+      "learning_rate": 0.0004249091606615322,
+      "loss": 1.2342,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5068236553385068,
+      "grad_norm": 0.5075557231903076,
+      "learning_rate": 0.0004248590943766429,
+      "loss": 1.032,
+      "step": 2841
+    },
+    {
+      "epoch": 0.507002051556507,
+      "grad_norm": 0.7127680778503418,
+      "learning_rate": 0.00042480901435839566,
+      "loss": 1.1435,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5071804477745072,
+      "grad_norm": 0.5281385779380798,
+      "learning_rate": 0.0004247589206107236,
+      "loss": 1.1247,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5073588439925074,
+      "grad_norm": 0.6512944102287292,
+      "learning_rate": 0.00042470881313756107,
+      "loss": 1.1724,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5075372402105075,
+      "grad_norm": 0.6242824196815491,
+      "learning_rate": 0.0004246586919428436,
+      "loss": 1.0296,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5077156364285077,
+      "grad_norm": 0.5622745752334595,
+      "learning_rate": 0.0004246085570305076,
+      "loss": 0.9505,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5078940326465079,
+      "grad_norm": 0.5400723218917847,
+      "learning_rate": 0.0004245584084044907,
+      "loss": 1.081,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5080724288645081,
+      "grad_norm": 0.8070915937423706,
+      "learning_rate": 0.00042450824606873145,
+      "loss": 0.9311,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5082508250825083,
+      "grad_norm": 0.6803063154220581,
+      "learning_rate": 0.00042445807002716967,
+      "loss": 1.2538,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5084292213005084,
+      "grad_norm": 0.6261130571365356,
+      "learning_rate": 0.0004244078802837462,
+      "loss": 1.1796,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5086076175185086,
+      "grad_norm": 0.5159333348274231,
+      "learning_rate": 0.00042435767684240286,
+      "loss": 1.074,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5087860137365088,
+      "grad_norm": 0.5446671843528748,
+      "learning_rate": 0.0004243074597070826,
+      "loss": 0.9219,
+      "step": 2852
+    },
+    {
+      "epoch": 0.508964409954509,
+      "grad_norm": 0.5136865377426147,
+      "learning_rate": 0.00042425722888172937,
+      "loss": 0.966,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5091428061725092,
+      "grad_norm": 0.7158051133155823,
+      "learning_rate": 0.00042420698437028846,
+      "loss": 0.9874,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5093212023905094,
+      "grad_norm": 0.4794641137123108,
+      "learning_rate": 0.000424156726176706,
+      "loss": 0.9166,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5094995986085095,
+      "grad_norm": 0.5229628086090088,
+      "learning_rate": 0.0004241064543049292,
+      "loss": 0.9187,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5096779948265097,
+      "grad_norm": 0.5953057408332825,
+      "learning_rate": 0.00042405616875890634,
+      "loss": 1.205,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5098563910445099,
+      "grad_norm": 0.629094660282135,
+      "learning_rate": 0.0004240058695425869,
+      "loss": 1.5532,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5100347872625101,
+      "grad_norm": 0.5075649619102478,
+      "learning_rate": 0.0004239555566599214,
+      "loss": 1.0137,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5102131834805103,
+      "grad_norm": 0.5661181807518005,
+      "learning_rate": 0.00042390523011486133,
+      "loss": 0.9655,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5103915796985103,
+      "grad_norm": 0.5415096282958984,
+      "learning_rate": 0.00042385488991135927,
+      "loss": 1.1261,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5105699759165105,
+      "grad_norm": 0.5538334846496582,
+      "learning_rate": 0.00042380453605336897,
+      "loss": 0.8759,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5107483721345107,
+      "grad_norm": 0.5417324900627136,
+      "learning_rate": 0.00042375416854484527,
+      "loss": 0.9578,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5109267683525109,
+      "grad_norm": 0.5347568392753601,
+      "learning_rate": 0.0004237037873897439,
+      "loss": 1.0563,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5111051645705111,
+      "grad_norm": 0.5405890941619873,
+      "learning_rate": 0.00042365339259202184,
+      "loss": 1.1051,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5112835607885113,
+      "grad_norm": 0.4867878258228302,
+      "learning_rate": 0.00042360298415563706,
+      "loss": 0.9752,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5114619570065114,
+      "grad_norm": 0.519904375076294,
+      "learning_rate": 0.0004235525620845486,
+      "loss": 1.0394,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5116403532245116,
+      "grad_norm": 0.6319023966789246,
+      "learning_rate": 0.00042350212638271655,
+      "loss": 1.163,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5118187494425118,
+      "grad_norm": 0.5480625033378601,
+      "learning_rate": 0.00042345167705410227,
+      "loss": 1.1205,
+      "step": 2869
+    },
+    {
+      "epoch": 0.511997145660512,
+      "grad_norm": 0.5435265898704529,
+      "learning_rate": 0.00042340121410266784,
+      "loss": 0.9511,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5121755418785122,
+      "grad_norm": 0.4731312096118927,
+      "learning_rate": 0.00042335073753237666,
+      "loss": 0.8017,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5123539380965123,
+      "grad_norm": 0.49716416001319885,
+      "learning_rate": 0.00042330024734719317,
+      "loss": 0.8897,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5125323343145125,
+      "grad_norm": 0.48113831877708435,
+      "learning_rate": 0.00042324974355108285,
+      "loss": 0.9106,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5127107305325127,
+      "grad_norm": 0.4926007390022278,
+      "learning_rate": 0.0004231992261480122,
+      "loss": 0.9383,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5128891267505129,
+      "grad_norm": 0.537855327129364,
+      "learning_rate": 0.000423148695141949,
+      "loss": 1.0949,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5130675229685131,
+      "grad_norm": 0.5071426033973694,
+      "learning_rate": 0.0004230981505368616,
+      "loss": 1.0838,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5132459191865133,
+      "grad_norm": 0.5214564800262451,
+      "learning_rate": 0.00042304759233672,
+      "loss": 1.0188,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5134243154045134,
+      "grad_norm": 0.48505401611328125,
+      "learning_rate": 0.000422997020545495,
+      "loss": 1.0605,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5136027116225136,
+      "grad_norm": 0.5197986364364624,
+      "learning_rate": 0.0004229464351671585,
+      "loss": 0.8301,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5137811078405138,
+      "grad_norm": 0.4954800605773926,
+      "learning_rate": 0.00042289583620568326,
+      "loss": 1.0674,
+      "step": 2880
+    },
+    {
+      "epoch": 0.513959504058514,
+      "grad_norm": 0.5165004730224609,
+      "learning_rate": 0.00042284522366504355,
+      "loss": 0.9053,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5141379002765142,
+      "grad_norm": 0.6076773405075073,
+      "learning_rate": 0.00042279459754921436,
+      "loss": 1.08,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5143162964945143,
+      "grad_norm": 0.7019060850143433,
+      "learning_rate": 0.0004227439578621718,
+      "loss": 0.994,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5144946927125145,
+      "grad_norm": 0.5617488026618958,
+      "learning_rate": 0.00042269330460789314,
+      "loss": 1.1718,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5146730889305147,
+      "grad_norm": 0.46552005410194397,
+      "learning_rate": 0.0004226426377903566,
+      "loss": 0.8396,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5148514851485149,
+      "grad_norm": 0.49444034695625305,
+      "learning_rate": 0.00042259195741354167,
+      "loss": 0.8349,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5150298813665151,
+      "grad_norm": 0.46994128823280334,
+      "learning_rate": 0.0004225412634814287,
+      "loss": 0.9429,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5152082775845153,
+      "grad_norm": 0.5488457083702087,
+      "learning_rate": 0.0004224905559979991,
+      "loss": 0.9858,
+      "step": 2888
+    },
+    {
+      "epoch": 0.5153866738025153,
+      "grad_norm": 0.5592421889305115,
+      "learning_rate": 0.0004224398349672354,
+      "loss": 1.0363,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5155650700205155,
+      "grad_norm": 0.4703425168991089,
+      "learning_rate": 0.00042238910039312134,
+      "loss": 0.9297,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5157434662385157,
+      "grad_norm": 0.9002471566200256,
+      "learning_rate": 0.00042233835227964146,
+      "loss": 0.8405,
+      "step": 2891
+    },
+    {
+      "epoch": 0.515921862456516,
+      "grad_norm": 0.5256131291389465,
+      "learning_rate": 0.0004222875906307816,
+      "loss": 0.9992,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5161002586745161,
+      "grad_norm": 0.47402289509773254,
+      "learning_rate": 0.0004222368154505285,
+      "loss": 0.7949,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5162786548925162,
+      "grad_norm": 0.5556320548057556,
+      "learning_rate": 0.0004221860267428701,
+      "loss": 1.0749,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5164570511105164,
+      "grad_norm": 0.5696842670440674,
+      "learning_rate": 0.0004221352245117952,
+      "loss": 0.9672,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5166354473285166,
+      "grad_norm": 0.586262047290802,
+      "learning_rate": 0.00042208440876129384,
+      "loss": 1.2136,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5168138435465168,
+      "grad_norm": 0.49055540561676025,
+      "learning_rate": 0.000422033579495357,
+      "loss": 0.9203,
+      "step": 2897
+    },
+    {
+      "epoch": 0.516992239764517,
+      "grad_norm": 0.5185585618019104,
+      "learning_rate": 0.00042198273671797693,
+      "loss": 0.9001,
+      "step": 2898
+    },
+    {
+      "epoch": 0.5171706359825172,
+      "grad_norm": 0.5147084593772888,
+      "learning_rate": 0.00042193188043314675,
+      "loss": 1.0424,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5173490322005173,
+      "grad_norm": 0.5919014811515808,
+      "learning_rate": 0.0004218810106448606,
+      "loss": 1.3816,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5175274284185175,
+      "grad_norm": 0.494163453578949,
+      "learning_rate": 0.0004218301273571139,
+      "loss": 0.8105,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5177058246365177,
+      "grad_norm": 0.558854877948761,
+      "learning_rate": 0.000421779230573903,
+      "loss": 1.0603,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5178842208545179,
+      "grad_norm": 0.47792795300483704,
+      "learning_rate": 0.00042172832029922514,
+      "loss": 0.9657,
+      "step": 2903
+    },
+    {
+      "epoch": 0.5180626170725181,
+      "grad_norm": 0.4979441463947296,
+      "learning_rate": 0.0004216773965370789,
+      "loss": 0.8444,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5182410132905182,
+      "grad_norm": 0.5534027814865112,
+      "learning_rate": 0.00042162645929146394,
+      "loss": 1.0369,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5184194095085184,
+      "grad_norm": 0.525301992893219,
+      "learning_rate": 0.0004215755085663806,
+      "loss": 1.0527,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5185978057265186,
+      "grad_norm": 0.5665169954299927,
+      "learning_rate": 0.0004215245443658307,
+      "loss": 1.0721,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5187762019445188,
+      "grad_norm": 0.4932265281677246,
+      "learning_rate": 0.0004214735666938169,
+      "loss": 1.0579,
+      "step": 2908
+    },
+    {
+      "epoch": 0.518954598162519,
+      "grad_norm": 0.47899937629699707,
+      "learning_rate": 0.0004214225755543429,
+      "loss": 0.9099,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5191329943805192,
+      "grad_norm": 0.459412157535553,
+      "learning_rate": 0.00042137157095141367,
+      "loss": 0.8198,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5193113905985193,
+      "grad_norm": 0.4750489890575409,
+      "learning_rate": 0.00042132055288903505,
+      "loss": 0.9241,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5194897868165195,
+      "grad_norm": 0.4672413468360901,
+      "learning_rate": 0.0004212695213712138,
+      "loss": 0.7955,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5196681830345197,
+      "grad_norm": 0.44326889514923096,
+      "learning_rate": 0.0004212184764019581,
+      "loss": 0.8813,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5198465792525199,
+      "grad_norm": 0.5115212798118591,
+      "learning_rate": 0.00042116741798527694,
+      "loss": 1.0218,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5200249754705201,
+      "grad_norm": 0.471718430519104,
+      "learning_rate": 0.0004211163461251804,
+      "loss": 0.9756,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5202033716885202,
+      "grad_norm": 0.4659128785133362,
+      "learning_rate": 0.0004210652608256798,
+      "loss": 0.9583,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5203817679065204,
+      "grad_norm": 0.5611960887908936,
+      "learning_rate": 0.00042101416209078707,
+      "loss": 1.1081,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5205601641245206,
+      "grad_norm": 0.5797203183174133,
+      "learning_rate": 0.00042096304992451575,
+      "loss": 1.1466,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5207385603425208,
+      "grad_norm": 0.522687554359436,
+      "learning_rate": 0.00042091192433088,
+      "loss": 0.9893,
+      "step": 2919
+    },
+    {
+      "epoch": 0.520916956560521,
+      "grad_norm": 0.5880950093269348,
+      "learning_rate": 0.00042086078531389524,
+      "loss": 1.0184,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5210953527785211,
+      "grad_norm": 1.1611589193344116,
+      "learning_rate": 0.000420809632877578,
+      "loss": 1.0494,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5212737489965212,
+      "grad_norm": 0.5259320735931396,
+      "learning_rate": 0.00042075846702594567,
+      "loss": 1.0708,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5214521452145214,
+      "grad_norm": 0.5445079803466797,
+      "learning_rate": 0.0004207072877630168,
+      "loss": 1.2258,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5216305414325216,
+      "grad_norm": 0.43372058868408203,
+      "learning_rate": 0.00042065609509281106,
+      "loss": 0.9371,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5218089376505218,
+      "grad_norm": 0.4435960650444031,
+      "learning_rate": 0.000420604889019349,
+      "loss": 0.8437,
+      "step": 2925
+    },
+    {
+      "epoch": 0.521987333868522,
+      "grad_norm": 0.6090930104255676,
+      "learning_rate": 0.00042055366954665244,
+      "loss": 1.4102,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5221657300865221,
+      "grad_norm": 0.5087897777557373,
+      "learning_rate": 0.000420502436678744,
+      "loss": 0.9809,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5223441263045223,
+      "grad_norm": 0.4611089825630188,
+      "learning_rate": 0.0004204511904196476,
+      "loss": 0.9503,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5225225225225225,
+      "grad_norm": 0.5042673945426941,
+      "learning_rate": 0.000420399930773388,
+      "loss": 0.865,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5227009187405227,
+      "grad_norm": 0.4814876914024353,
+      "learning_rate": 0.00042034865774399124,
+      "loss": 1.1542,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5228793149585229,
+      "grad_norm": 0.47714847326278687,
+      "learning_rate": 0.0004202973713354842,
+      "loss": 0.9069,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5230577111765231,
+      "grad_norm": 0.5801002979278564,
+      "learning_rate": 0.0004202460715518948,
+      "loss": 1.1885,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5232361073945232,
+      "grad_norm": 0.45451635122299194,
+      "learning_rate": 0.0004201947583972523,
+      "loss": 0.8125,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5234145036125234,
+      "grad_norm": 0.5364030003547668,
+      "learning_rate": 0.00042014343187558666,
+      "loss": 0.9351,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5235928998305236,
+      "grad_norm": 0.5005396604537964,
+      "learning_rate": 0.0004200920919909292,
+      "loss": 1.0081,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5237712960485238,
+      "grad_norm": 0.5246373414993286,
+      "learning_rate": 0.00042004073874731196,
+      "loss": 1.0374,
+      "step": 2936
+    },
+    {
+      "epoch": 0.523949692266524,
+      "grad_norm": 0.5010136365890503,
+      "learning_rate": 0.0004199893721487682,
+      "loss": 1.0534,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5241280884845241,
+      "grad_norm": 0.9451296329498291,
+      "learning_rate": 0.00041993799219933235,
+      "loss": 1.0364,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5243064847025243,
+      "grad_norm": 0.5468368530273438,
+      "learning_rate": 0.0004198865989030398,
+      "loss": 1.0716,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5244848809205245,
+      "grad_norm": 0.46857789158821106,
+      "learning_rate": 0.00041983519226392686,
+      "loss": 1.0187,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5246632771385247,
+      "grad_norm": 0.4949086606502533,
+      "learning_rate": 0.00041978377228603093,
+      "loss": 0.8706,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5248416733565249,
+      "grad_norm": 0.4794335961341858,
+      "learning_rate": 0.00041973233897339067,
+      "loss": 0.9308,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5250200695745251,
+      "grad_norm": 0.4741996228694916,
+      "learning_rate": 0.0004196808923300455,
+      "loss": 1.0586,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5251984657925252,
+      "grad_norm": 0.49954500794410706,
+      "learning_rate": 0.0004196294323600361,
+      "loss": 0.9257,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5253768620105254,
+      "grad_norm": 0.5138105154037476,
+      "learning_rate": 0.000419577959067404,
+      "loss": 0.9921,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5255552582285256,
+      "grad_norm": 0.4848662316799164,
+      "learning_rate": 0.00041952647245619204,
+      "loss": 1.0388,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5257336544465258,
+      "grad_norm": 0.4525747001171112,
+      "learning_rate": 0.00041947497253044385,
+      "loss": 0.9866,
+      "step": 2947
+    },
+    {
+      "epoch": 0.525912050664526,
+      "grad_norm": 0.5498698949813843,
+      "learning_rate": 0.0004194234592942043,
+      "loss": 1.1586,
+      "step": 2948
+    },
+    {
+      "epoch": 0.526090446882526,
+      "grad_norm": 0.4874838590621948,
+      "learning_rate": 0.0004193719327515192,
+      "loss": 1.0105,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5262688431005262,
+      "grad_norm": 0.5352873206138611,
+      "learning_rate": 0.0004193203929064353,
+      "loss": 1.0994,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5264472393185264,
+      "grad_norm": 0.4859972596168518,
+      "learning_rate": 0.0004192688397630006,
+      "loss": 0.9533,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5266256355365266,
+      "grad_norm": 0.5137537717819214,
+      "learning_rate": 0.0004192172733252641,
+      "loss": 1.0008,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5268040317545268,
+      "grad_norm": 0.47844600677490234,
+      "learning_rate": 0.00041916569359727574,
+      "loss": 1.0954,
+      "step": 2953
+    },
+    {
+      "epoch": 0.526982427972527,
+      "grad_norm": 0.52599036693573,
+      "learning_rate": 0.00041911410058308667,
+      "loss": 1.0523,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5271608241905271,
+      "grad_norm": 0.47040775418281555,
+      "learning_rate": 0.0004190624942867489,
+      "loss": 0.8066,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5273392204085273,
+      "grad_norm": 0.5250223278999329,
+      "learning_rate": 0.0004190108747123156,
+      "loss": 1.2687,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5275176166265275,
+      "grad_norm": 0.5163100361824036,
+      "learning_rate": 0.0004189592418638408,
+      "loss": 1.1398,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5276960128445277,
+      "grad_norm": 0.47173264622688293,
+      "learning_rate": 0.00041890759574538,
+      "loss": 0.9907,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5278744090625279,
+      "grad_norm": 0.5113806128501892,
+      "learning_rate": 0.0004188559363609893,
+      "loss": 1.2076,
+      "step": 2959
+    },
+    {
+      "epoch": 0.528052805280528,
+      "grad_norm": 0.47075924277305603,
+      "learning_rate": 0.0004188042637147259,
+      "loss": 0.8586,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5282312014985282,
+      "grad_norm": 0.4917936325073242,
+      "learning_rate": 0.00041875257781064833,
+      "loss": 0.913,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5284095977165284,
+      "grad_norm": 0.49223649501800537,
+      "learning_rate": 0.000418700878652816,
+      "loss": 1.0004,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5285879939345286,
+      "grad_norm": 0.4502630829811096,
+      "learning_rate": 0.0004186491662452892,
+      "loss": 0.8698,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5287663901525288,
+      "grad_norm": 0.5051430463790894,
+      "learning_rate": 0.00041859744059212945,
+      "loss": 1.093,
+      "step": 2964
+    },
+    {
+      "epoch": 0.528944786370529,
+      "grad_norm": 0.7281007170677185,
+      "learning_rate": 0.0004185457016973993,
+      "loss": 1.0823,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5291231825885291,
+      "grad_norm": 0.4868094027042389,
+      "learning_rate": 0.00041849394956516227,
+      "loss": 1.0735,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5293015788065293,
+      "grad_norm": 0.4833069443702698,
+      "learning_rate": 0.0004184421841994829,
+      "loss": 0.8735,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5294799750245295,
+      "grad_norm": 0.5559689402580261,
+      "learning_rate": 0.0004183904056044269,
+      "loss": 1.0569,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5296583712425297,
+      "grad_norm": 0.5828696489334106,
+      "learning_rate": 0.0004183386137840609,
+      "loss": 1.126,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5298367674605299,
+      "grad_norm": 0.48288482427597046,
+      "learning_rate": 0.0004182868087424526,
+      "loss": 1.1427,
+      "step": 2970
+    },
+    {
+      "epoch": 0.53001516367853,
+      "grad_norm": 0.4690031111240387,
+      "learning_rate": 0.0004182349904836708,
+      "loss": 0.9756,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5301935598965302,
+      "grad_norm": 0.5362717509269714,
+      "learning_rate": 0.00041818315901178527,
+      "loss": 1.1166,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5303719561145304,
+      "grad_norm": 0.5287902355194092,
+      "learning_rate": 0.0004181313143308667,
+      "loss": 1.1939,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5305503523325306,
+      "grad_norm": 0.4559158682823181,
+      "learning_rate": 0.0004180794564449872,
+      "loss": 0.7655,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5307287485505308,
+      "grad_norm": 0.5267648696899414,
+      "learning_rate": 0.0004180275853582194,
+      "loss": 0.9355,
+      "step": 2975
+    },
+    {
+      "epoch": 0.530907144768531,
+      "grad_norm": 0.46050286293029785,
+      "learning_rate": 0.00041797570107463737,
+      "loss": 0.931,
+      "step": 2976
+    },
+    {
+      "epoch": 0.531085540986531,
+      "grad_norm": 0.4590972363948822,
+      "learning_rate": 0.0004179238035983161,
+      "loss": 0.9246,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5312639372045312,
+      "grad_norm": 0.5887408256530762,
+      "learning_rate": 0.00041787189293333155,
+      "loss": 0.9576,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5314423334225314,
+      "grad_norm": 0.5028809309005737,
+      "learning_rate": 0.00041781996908376077,
+      "loss": 1.0257,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5316207296405316,
+      "grad_norm": 0.4969048500061035,
+      "learning_rate": 0.00041776803205368187,
+      "loss": 1.0092,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5317991258585318,
+      "grad_norm": 0.5167542099952698,
+      "learning_rate": 0.00041771608184717384,
+      "loss": 1.0632,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5319775220765319,
+      "grad_norm": 0.4737110733985901,
+      "learning_rate": 0.00041766411846831696,
+      "loss": 0.8571,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5321559182945321,
+      "grad_norm": 0.5081421732902527,
+      "learning_rate": 0.00041761214192119234,
+      "loss": 1.216,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5323343145125323,
+      "grad_norm": 0.47318655252456665,
+      "learning_rate": 0.0004175601522098823,
+      "loss": 0.939,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5325127107305325,
+      "grad_norm": 0.5063821077346802,
+      "learning_rate": 0.0004175081493384699,
+      "loss": 0.9156,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5326911069485327,
+      "grad_norm": 0.483039915561676,
+      "learning_rate": 0.00041745613331103964,
+      "loss": 0.9842,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5328695031665329,
+      "grad_norm": 0.4720330536365509,
+      "learning_rate": 0.0004174041041316767,
+      "loss": 1.0204,
+      "step": 2987
+    },
+    {
+      "epoch": 0.533047899384533,
+      "grad_norm": 11.117390632629395,
+      "learning_rate": 0.0004173520618044675,
+      "loss": 1.2214,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5332262956025332,
+      "grad_norm": 0.5497750043869019,
+      "learning_rate": 0.00041730000633349927,
+      "loss": 1.0853,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5334046918205334,
+      "grad_norm": 0.5697376132011414,
+      "learning_rate": 0.00041724793772286066,
+      "loss": 1.0129,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5335830880385336,
+      "grad_norm": 0.527197003364563,
+      "learning_rate": 0.0004171958559766409,
+      "loss": 1.2534,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5337614842565338,
+      "grad_norm": 1.886084794998169,
+      "learning_rate": 0.0004171437610989306,
+      "loss": 1.1865,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5339398804745339,
+      "grad_norm": 0.5286929607391357,
+      "learning_rate": 0.00041709165309382123,
+      "loss": 1.2268,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5341182766925341,
+      "grad_norm": 0.5276958346366882,
+      "learning_rate": 0.0004170395319654054,
+      "loss": 1.024,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5342966729105343,
+      "grad_norm": 0.44607049226760864,
+      "learning_rate": 0.0004169873977177765,
+      "loss": 0.8801,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5344750691285345,
+      "grad_norm": 0.4721534252166748,
+      "learning_rate": 0.0004169352503550293,
+      "loss": 0.9466,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5346534653465347,
+      "grad_norm": 0.5831915736198425,
+      "learning_rate": 0.00041688308988125944,
+      "loss": 1.2001,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5348318615645349,
+      "grad_norm": 0.49683353304862976,
+      "learning_rate": 0.00041683091630056334,
+      "loss": 0.9686,
+      "step": 2998
+    },
+    {
+      "epoch": 0.535010257782535,
+      "grad_norm": 0.6138126850128174,
+      "learning_rate": 0.000416778729617039,
+      "loss": 1.0234,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5351886540005352,
+      "grad_norm": 0.506582498550415,
+      "learning_rate": 0.000416726529834785,
+      "loss": 0.9276,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5353670502185354,
+      "grad_norm": 0.5279721617698669,
+      "learning_rate": 0.000416674316957901,
+      "loss": 1.1615,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5355454464365356,
+      "grad_norm": 0.48710203170776367,
+      "learning_rate": 0.000416622090990488,
+      "loss": 1.0638,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5357238426545358,
+      "grad_norm": 0.5252576470375061,
+      "learning_rate": 0.00041656985193664763,
+      "loss": 0.9231,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5359022388725359,
+      "grad_norm": 0.529524028301239,
+      "learning_rate": 0.00041651759980048276,
+      "loss": 0.9436,
+      "step": 3004
+    },
+    {
+      "epoch": 0.536080635090536,
+      "grad_norm": 0.4857736825942993,
+      "learning_rate": 0.00041646533458609725,
+      "loss": 0.8424,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5362590313085362,
+      "grad_norm": 0.5015032291412354,
+      "learning_rate": 0.00041641305629759595,
+      "loss": 1.0656,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5364374275265364,
+      "grad_norm": 0.4542873501777649,
+      "learning_rate": 0.0004163607649390849,
+      "loss": 1.0041,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5366158237445366,
+      "grad_norm": 0.5027053952217102,
+      "learning_rate": 0.000416308460514671,
+      "loss": 1.209,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5367942199625368,
+      "grad_norm": 0.4859675168991089,
+      "learning_rate": 0.00041625614302846206,
+      "loss": 0.9146,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5369726161805369,
+      "grad_norm": 0.5812960863113403,
+      "learning_rate": 0.0004162038124845673,
+      "loss": 1.044,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5371510123985371,
+      "grad_norm": 0.5171328186988831,
+      "learning_rate": 0.00041615146888709654,
+      "loss": 0.9578,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5373294086165373,
+      "grad_norm": 0.5038743615150452,
+      "learning_rate": 0.000416099112240161,
+      "loss": 0.9988,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5375078048345375,
+      "grad_norm": 0.5556420087814331,
+      "learning_rate": 0.0004160467425478726,
+      "loss": 1.0005,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5376862010525377,
+      "grad_norm": 0.45622992515563965,
+      "learning_rate": 0.0004159943598143445,
+      "loss": 0.9728,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5378645972705378,
+      "grad_norm": 0.5222346782684326,
+      "learning_rate": 0.00041594196404369076,
+      "loss": 1.1148,
+      "step": 3015
+    },
+    {
+      "epoch": 0.538042993488538,
+      "grad_norm": 0.6195734143257141,
+      "learning_rate": 0.0004158895552400267,
+      "loss": 1.0752,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5382213897065382,
+      "grad_norm": 0.5263445973396301,
+      "learning_rate": 0.0004158371334074683,
+      "loss": 1.023,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5383997859245384,
+      "grad_norm": 0.4883100688457489,
+      "learning_rate": 0.00041578469855013277,
+      "loss": 0.8856,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5385781821425386,
+      "grad_norm": 0.517374575138092,
+      "learning_rate": 0.0004157322506721384,
+      "loss": 1.1114,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5387565783605388,
+      "grad_norm": 0.5005238652229309,
+      "learning_rate": 0.00041567978977760444,
+      "loss": 1.1022,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5389349745785389,
+      "grad_norm": 0.46176883578300476,
+      "learning_rate": 0.00041562731587065093,
+      "loss": 0.758,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5391133707965391,
+      "grad_norm": 0.43858736753463745,
+      "learning_rate": 0.00041557482895539943,
+      "loss": 0.8694,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5392917670145393,
+      "grad_norm": 0.4803151488304138,
+      "learning_rate": 0.0004155223290359721,
+      "loss": 0.8272,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5394701632325395,
+      "grad_norm": 0.5557406544685364,
+      "learning_rate": 0.0004154698161164923,
+      "loss": 1.0889,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5396485594505397,
+      "grad_norm": 2.459235668182373,
+      "learning_rate": 0.0004154172902010843,
+      "loss": 1.1591,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5398269556685398,
+      "grad_norm": 0.5599625706672668,
+      "learning_rate": 0.0004153647512938735,
+      "loss": 1.1661,
+      "step": 3026
+    },
+    {
+      "epoch": 0.54000535188654,
+      "grad_norm": 0.5514055490493774,
+      "learning_rate": 0.00041531219939898635,
+      "loss": 1.0657,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5401837481045402,
+      "grad_norm": 0.8572493195533752,
+      "learning_rate": 0.0004152596345205502,
+      "loss": 0.923,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5403621443225404,
+      "grad_norm": 0.497321754693985,
+      "learning_rate": 0.00041520705666269343,
+      "loss": 0.9463,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.5045643448829651,
+      "learning_rate": 0.0004151544658295455,
+      "loss": 0.9698,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5407189367585408,
+      "grad_norm": 0.519391655921936,
+      "learning_rate": 0.00041510186202523697,
+      "loss": 0.8543,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5408973329765409,
+      "grad_norm": 1.3143996000289917,
+      "learning_rate": 0.0004150492452538992,
+      "loss": 0.9187,
+      "step": 3032
+    },
+    {
+      "epoch": 0.541075729194541,
+      "grad_norm": 0.5469959378242493,
+      "learning_rate": 0.0004149966155196648,
+      "loss": 1.3002,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5412541254125413,
+      "grad_norm": 0.4593210518360138,
+      "learning_rate": 0.0004149439728266671,
+      "loss": 0.9523,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5414325216305415,
+      "grad_norm": 0.4348868727684021,
+      "learning_rate": 0.0004148913171790408,
+      "loss": 0.7697,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5416109178485417,
+      "grad_norm": 0.5638412833213806,
+      "learning_rate": 0.00041483864858092145,
+      "loss": 1.1035,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5417893140665417,
+      "grad_norm": 0.4964679479598999,
+      "learning_rate": 0.00041478596703644553,
+      "loss": 1.0256,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5419677102845419,
+      "grad_norm": 0.5189842581748962,
+      "learning_rate": 0.0004147332725497507,
+      "loss": 0.9451,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5421461065025421,
+      "grad_norm": 0.5478013157844543,
+      "learning_rate": 0.0004146805651249755,
+      "loss": 0.938,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5423245027205423,
+      "grad_norm": 0.5358539819717407,
+      "learning_rate": 0.0004146278447662597,
+      "loss": 0.8084,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5425028989385425,
+      "grad_norm": 0.5389429330825806,
+      "learning_rate": 0.00041457511147774374,
+      "loss": 1.1265,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5426812951565427,
+      "grad_norm": 0.4505546987056732,
+      "learning_rate": 0.0004145223652635693,
+      "loss": 0.8022,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5428596913745428,
+      "grad_norm": 0.5185169577598572,
+      "learning_rate": 0.00041446960612787916,
+      "loss": 1.1086,
+      "step": 3043
+    },
+    {
+      "epoch": 0.543038087592543,
+      "grad_norm": 0.4975489377975464,
+      "learning_rate": 0.00041441683407481683,
+      "loss": 0.9063,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5432164838105432,
+      "grad_norm": 0.48402321338653564,
+      "learning_rate": 0.0004143640491085272,
+      "loss": 0.9616,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5433948800285434,
+      "grad_norm": 0.47887566685676575,
+      "learning_rate": 0.0004143112512331558,
+      "loss": 0.9543,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5435732762465436,
+      "grad_norm": 0.5627540946006775,
+      "learning_rate": 0.00041425844045284957,
+      "loss": 1.1141,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5437516724645437,
+      "grad_norm": 0.4868026077747345,
+      "learning_rate": 0.0004142056167717561,
+      "loss": 0.8547,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5439300686825439,
+      "grad_norm": 0.4439375102519989,
+      "learning_rate": 0.000414152780194024,
+      "loss": 0.7576,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5441084649005441,
+      "grad_norm": 0.5482441782951355,
+      "learning_rate": 0.00041409993072380333,
+      "loss": 1.1234,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5442868611185443,
+      "grad_norm": 0.5158576965332031,
+      "learning_rate": 0.00041404706836524463,
+      "loss": 0.9731,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5444652573365445,
+      "grad_norm": 0.44664448499679565,
+      "learning_rate": 0.0004139941931224998,
+      "loss": 0.9086,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5446436535545447,
+      "grad_norm": 0.6744524240493774,
+      "learning_rate": 0.0004139413049997216,
+      "loss": 1.1988,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5448220497725448,
+      "grad_norm": 0.47927117347717285,
+      "learning_rate": 0.0004138884040010639,
+      "loss": 1.06,
+      "step": 3054
+    },
+    {
+      "epoch": 0.545000445990545,
+      "grad_norm": 0.4434390962123871,
+      "learning_rate": 0.00041383549013068147,
+      "loss": 0.9022,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5451788422085452,
+      "grad_norm": 0.5459398627281189,
+      "learning_rate": 0.0004137825633927301,
+      "loss": 1.183,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5453572384265454,
+      "grad_norm": 0.48765432834625244,
+      "learning_rate": 0.00041372962379136676,
+      "loss": 1.0203,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5455356346445456,
+      "grad_norm": 0.8096850514411926,
+      "learning_rate": 0.00041367667133074916,
+      "loss": 1.0457,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5457140308625457,
+      "grad_norm": 0.4843488037586212,
+      "learning_rate": 0.0004136237060150363,
+      "loss": 1.0277,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5458924270805459,
+      "grad_norm": 0.48028862476348877,
+      "learning_rate": 0.0004135707278483879,
+      "loss": 0.9519,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5460708232985461,
+      "grad_norm": 0.7017949223518372,
+      "learning_rate": 0.00041351773683496497,
+      "loss": 0.9328,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5462492195165463,
+      "grad_norm": 1.0810225009918213,
+      "learning_rate": 0.0004134647329789293,
+      "loss": 1.1302,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5464276157345465,
+      "grad_norm": 0.5587711930274963,
+      "learning_rate": 0.0004134117162844439,
+      "loss": 1.0982,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5466060119525467,
+      "grad_norm": 0.579770028591156,
+      "learning_rate": 0.00041335868675567263,
+      "loss": 1.14,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5467844081705467,
+      "grad_norm": 0.7189714908599854,
+      "learning_rate": 0.0004133056443967804,
+      "loss": 1.0545,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5469628043885469,
+      "grad_norm": 0.7987565398216248,
+      "learning_rate": 0.0004132525892119331,
+      "loss": 1.0576,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5471412006065471,
+      "grad_norm": 0.6334807276725769,
+      "learning_rate": 0.00041319952120529767,
+      "loss": 1.0614,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5473195968245473,
+      "grad_norm": 27.262441635131836,
+      "learning_rate": 0.00041314644038104216,
+      "loss": 1.2708,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5474979930425475,
+      "grad_norm": 0.5688011646270752,
+      "learning_rate": 0.00041309334674333544,
+      "loss": 0.8942,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5476763892605476,
+      "grad_norm": 0.5215781331062317,
+      "learning_rate": 0.00041304024029634737,
+      "loss": 0.895,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5478547854785478,
+      "grad_norm": 0.5432342886924744,
+      "learning_rate": 0.00041298712104424903,
+      "loss": 0.9245,
+      "step": 3071
+    },
+    {
+      "epoch": 0.548033181696548,
+      "grad_norm": 0.49601709842681885,
+      "learning_rate": 0.0004129339889912123,
+      "loss": 0.9943,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5482115779145482,
+      "grad_norm": 0.5069965124130249,
+      "learning_rate": 0.0004128808441414103,
+      "loss": 0.8476,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5483899741325484,
+      "grad_norm": 0.7115893959999084,
+      "learning_rate": 0.0004128276864990168,
+      "loss": 1.1841,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5485683703505486,
+      "grad_norm": 1.3543672561645508,
+      "learning_rate": 0.0004127745160682068,
+      "loss": 1.011,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5487467665685487,
+      "grad_norm": 0.6685739159584045,
+      "learning_rate": 0.0004127213328531565,
+      "loss": 1.1114,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5489251627865489,
+      "grad_norm": 1.056970477104187,
+      "learning_rate": 0.0004126681368580427,
+      "loss": 1.0183,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5491035590045491,
+      "grad_norm": 0.7646797299385071,
+      "learning_rate": 0.00041261492808704336,
+      "loss": 1.1184,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5492819552225493,
+      "grad_norm": 0.5978376269340515,
+      "learning_rate": 0.00041256170654433767,
+      "loss": 0.8961,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5494603514405495,
+      "grad_norm": 0.5183614492416382,
+      "learning_rate": 0.0004125084722341054,
+      "loss": 0.8404,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5496387476585496,
+      "grad_norm": 0.9243518114089966,
+      "learning_rate": 0.0004124552251605277,
+      "loss": 1.2223,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5498171438765498,
+      "grad_norm": 0.5044293403625488,
+      "learning_rate": 0.0004124019653277865,
+      "loss": 0.8062,
+      "step": 3082
+    },
+    {
+      "epoch": 0.54999554009455,
+      "grad_norm": 0.8046514391899109,
+      "learning_rate": 0.0004123486927400649,
+      "loss": 0.9456,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5501739363125502,
+      "grad_norm": 0.5840319991111755,
+      "learning_rate": 0.0004122954074015468,
+      "loss": 1.2686,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5503523325305504,
+      "grad_norm": 1.5194361209869385,
+      "learning_rate": 0.0004122421093164172,
+      "loss": 1.1059,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5505307287485506,
+      "grad_norm": 0.5864779353141785,
+      "learning_rate": 0.0004121887984888622,
+      "loss": 1.0505,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5507091249665507,
+      "grad_norm": 1.8019944429397583,
+      "learning_rate": 0.00041213547492306875,
+      "loss": 1.0039,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5508875211845509,
+      "grad_norm": 0.5829388499259949,
+      "learning_rate": 0.00041208213862322485,
+      "loss": 1.1364,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5510659174025511,
+      "grad_norm": 0.7362989783287048,
+      "learning_rate": 0.0004120287895935196,
+      "loss": 1.0522,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5512443136205513,
+      "grad_norm": 0.4867869019508362,
+      "learning_rate": 0.00041197542783814287,
+      "loss": 0.7512,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5514227098385515,
+      "grad_norm": 0.6978092193603516,
+      "learning_rate": 0.0004119220533612857,
+      "loss": 1.1723,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5516011060565515,
+      "grad_norm": 0.5516607761383057,
+      "learning_rate": 0.00041186866616714024,
+      "loss": 1.0468,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5517795022745517,
+      "grad_norm": 0.7430167198181152,
+      "learning_rate": 0.0004118152662598994,
+      "loss": 1.1032,
+      "step": 3093
+    },
+    {
+      "epoch": 0.551957898492552,
+      "grad_norm": 0.48890167474746704,
+      "learning_rate": 0.0004117618536437571,
+      "loss": 0.7917,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5521362947105521,
+      "grad_norm": 0.5151326060295105,
+      "learning_rate": 0.00041170842832290844,
+      "loss": 0.8906,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5523146909285523,
+      "grad_norm": 0.4648483991622925,
+      "learning_rate": 0.0004116549903015495,
+      "loss": 0.9603,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5524930871465525,
+      "grad_norm": 0.7839336395263672,
+      "learning_rate": 0.00041160153958387714,
+      "loss": 1.0892,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5526714833645526,
+      "grad_norm": 0.48735418915748596,
+      "learning_rate": 0.0004115480761740893,
+      "loss": 0.9049,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5528498795825528,
+      "grad_norm": 0.5337851047515869,
+      "learning_rate": 0.0004114946000763852,
+      "loss": 1.0263,
+      "step": 3099
+    },
+    {
+      "epoch": 0.553028275800553,
+      "grad_norm": 0.7853549122810364,
+      "learning_rate": 0.0004114411112949647,
+      "loss": 1.3071,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5532066720185532,
+      "grad_norm": 0.5989134907722473,
+      "learning_rate": 0.0004113876098340288,
+      "loss": 1.1487,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5533850682365534,
+      "grad_norm": 0.5571014881134033,
+      "learning_rate": 0.00041133409569777936,
+      "loss": 0.8642,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5535634644545535,
+      "grad_norm": 0.5321959853172302,
+      "learning_rate": 0.0004112805688904196,
+      "loss": 0.9383,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5537418606725537,
+      "grad_norm": 0.7023414373397827,
+      "learning_rate": 0.00041122702941615334,
+      "loss": 1.0434,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5539202568905539,
+      "grad_norm": 0.6236468553543091,
+      "learning_rate": 0.00041117347727918555,
+      "loss": 0.9331,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5540986531085541,
+      "grad_norm": 0.5762714743614197,
+      "learning_rate": 0.00041111991248372215,
+      "loss": 0.9617,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5542770493265543,
+      "grad_norm": 0.9596529603004456,
+      "learning_rate": 0.00041106633503397016,
+      "loss": 1.0038,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5544554455445545,
+      "grad_norm": 0.46744972467422485,
+      "learning_rate": 0.00041101274493413764,
+      "loss": 0.8293,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5546338417625546,
+      "grad_norm": 0.514829158782959,
+      "learning_rate": 0.0004109591421884334,
+      "loss": 0.9105,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5548122379805548,
+      "grad_norm": 0.5770788788795471,
+      "learning_rate": 0.0004109055268010674,
+      "loss": 1.164,
+      "step": 3110
+    },
+    {
+      "epoch": 0.554990634198555,
+      "grad_norm": 9.369170188903809,
+      "learning_rate": 0.00041085189877625053,
+      "loss": 1.8313,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5551690304165552,
+      "grad_norm": 0.4911796748638153,
+      "learning_rate": 0.0004107982581181947,
+      "loss": 0.8055,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5553474266345554,
+      "grad_norm": 0.5030617117881775,
+      "learning_rate": 0.00041074460483111287,
+      "loss": 0.9508,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5555258228525555,
+      "grad_norm": 0.49075013399124146,
+      "learning_rate": 0.000410690938919219,
+      "loss": 0.9634,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5557042190705557,
+      "grad_norm": 0.5361286401748657,
+      "learning_rate": 0.000410637260386728,
+      "loss": 0.8745,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5558826152885559,
+      "grad_norm": 0.5814827084541321,
+      "learning_rate": 0.00041058356923785565,
+      "loss": 1.1225,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5560610115065561,
+      "grad_norm": 0.48355334997177124,
+      "learning_rate": 0.0004105298654768189,
+      "loss": 0.758,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5562394077245563,
+      "grad_norm": 0.6016453504562378,
+      "learning_rate": 0.0004104761491078355,
+      "loss": 1.1899,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5564178039425565,
+      "grad_norm": 0.7221107482910156,
+      "learning_rate": 0.0004104224201351245,
+      "loss": 1.1253,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5565962001605566,
+      "grad_norm": 0.5367442965507507,
+      "learning_rate": 0.00041036867856290567,
+      "loss": 0.8561,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5567745963785568,
+      "grad_norm": 0.7140209078788757,
+      "learning_rate": 0.00041031492439539975,
+      "loss": 0.961,
+      "step": 3121
+    },
+    {
+      "epoch": 0.556952992596557,
+      "grad_norm": 0.7320627570152283,
+      "learning_rate": 0.0004102611576368287,
+      "loss": 1.2107,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5571313888145571,
+      "grad_norm": 0.5843806862831116,
+      "learning_rate": 0.0004102073782914153,
+      "loss": 0.844,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5573097850325573,
+      "grad_norm": 0.7023711204528809,
+      "learning_rate": 0.00041015358636338343,
+      "loss": 1.1463,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5574881812505574,
+      "grad_norm": 0.5064725875854492,
+      "learning_rate": 0.0004100997818569577,
+      "loss": 0.9558,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5576665774685576,
+      "grad_norm": 21.4315128326416,
+      "learning_rate": 0.00041004596477636405,
+      "loss": 1.645,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5578449736865578,
+      "grad_norm": 0.8198938965797424,
+      "learning_rate": 0.0004099921351258292,
+      "loss": 0.9687,
+      "step": 3127
+    },
+    {
+      "epoch": 0.558023369904558,
+      "grad_norm": 8.272482872009277,
+      "learning_rate": 0.00040993829290958086,
+      "loss": 1.1445,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5582017661225582,
+      "grad_norm": 0.8829731941223145,
+      "learning_rate": 0.0004098844381318478,
+      "loss": 1.2275,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5583801623405584,
+      "grad_norm": 1.6794313192367554,
+      "learning_rate": 0.00040983057079685984,
+      "loss": 1.118,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5585585585585585,
+      "grad_norm": 1.177979826927185,
+      "learning_rate": 0.0004097766909088476,
+      "loss": 1.0017,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5587369547765587,
+      "grad_norm": 2.6292262077331543,
+      "learning_rate": 0.0004097227984720429,
+      "loss": 0.9328,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5589153509945589,
+      "grad_norm": 1.0025197267532349,
+      "learning_rate": 0.0004096688934906782,
+      "loss": 1.0593,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5590937472125591,
+      "grad_norm": 0.5532941222190857,
+      "learning_rate": 0.0004096149759689874,
+      "loss": 1.1529,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5592721434305593,
+      "grad_norm": 0.6335615515708923,
+      "learning_rate": 0.00040956104591120503,
+      "loss": 1.0475,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5594505396485594,
+      "grad_norm": 0.663866400718689,
+      "learning_rate": 0.00040950710332156683,
+      "loss": 0.8865,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5596289358665596,
+      "grad_norm": 0.5401614308357239,
+      "learning_rate": 0.00040945314820430934,
+      "loss": 0.9367,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5598073320845598,
+      "grad_norm": 0.5373167991638184,
+      "learning_rate": 0.0004093991805636702,
+      "loss": 0.8718,
+      "step": 3138
+    },
+    {
+      "epoch": 0.55998572830256,
+      "grad_norm": 0.5260427594184875,
+      "learning_rate": 0.00040934520040388807,
+      "loss": 0.911,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5601641245205602,
+      "grad_norm": 0.5970147252082825,
+      "learning_rate": 0.00040929120772920243,
+      "loss": 1.0448,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5603425207385604,
+      "grad_norm": 0.5675042271614075,
+      "learning_rate": 0.0004092372025438539,
+      "loss": 1.1786,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5605209169565605,
+      "grad_norm": 0.5005537867546082,
+      "learning_rate": 0.0004091831848520839,
+      "loss": 0.827,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5606993131745607,
+      "grad_norm": 0.5521116256713867,
+      "learning_rate": 0.00040912915465813525,
+      "loss": 0.8586,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5608777093925609,
+      "grad_norm": 0.5322579145431519,
+      "learning_rate": 0.0004090751119662511,
+      "loss": 0.9931,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5610561056105611,
+      "grad_norm": 0.5480323433876038,
+      "learning_rate": 0.00040902105678067627,
+      "loss": 1.031,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5612345018285613,
+      "grad_norm": 0.5208452343940735,
+      "learning_rate": 0.00040896698910565597,
+      "loss": 1.0623,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5614128980465614,
+      "grad_norm": 0.5295253992080688,
+      "learning_rate": 0.00040891290894543676,
+      "loss": 0.9008,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5615912942645616,
+      "grad_norm": 0.521321713924408,
+      "learning_rate": 0.00040885881630426616,
+      "loss": 1.0494,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5617696904825618,
+      "grad_norm": 0.49271395802497864,
+      "learning_rate": 0.0004088047111863924,
+      "loss": 0.8965,
+      "step": 3149
+    },
+    {
+      "epoch": 0.561948086700562,
+      "grad_norm": 0.5301600098609924,
+      "learning_rate": 0.000408750593596065,
+      "loss": 1.0071,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5621264829185622,
+      "grad_norm": 0.5512881875038147,
+      "learning_rate": 0.0004086964635375342,
+      "loss": 1.0579,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5623048791365624,
+      "grad_norm": 0.5429875254631042,
+      "learning_rate": 0.00040864232101505153,
+      "loss": 0.9885,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5624832753545624,
+      "grad_norm": 0.4952228367328644,
+      "learning_rate": 0.00040858816603286924,
+      "loss": 0.8444,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5626616715725626,
+      "grad_norm": 0.6742690205574036,
+      "learning_rate": 0.00040853399859524066,
+      "loss": 1.0294,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5628400677905628,
+      "grad_norm": 0.5073679685592651,
+      "learning_rate": 0.00040847981870642004,
+      "loss": 1.0393,
+      "step": 3155
+    },
+    {
+      "epoch": 0.563018464008563,
+      "grad_norm": 0.5249845385551453,
+      "learning_rate": 0.0004084256263706626,
+      "loss": 1.0371,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5631968602265632,
+      "grad_norm": 0.5406726002693176,
+      "learning_rate": 0.00040837142159222466,
+      "loss": 1.096,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5633752564445633,
+      "grad_norm": 0.4763408899307251,
+      "learning_rate": 0.0004083172043753635,
+      "loss": 0.948,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5635536526625635,
+      "grad_norm": 0.5455912947654724,
+      "learning_rate": 0.0004082629747243371,
+      "loss": 0.9939,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5637320488805637,
+      "grad_norm": 0.5416297316551208,
+      "learning_rate": 0.00040820873264340484,
+      "loss": 0.823,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5639104450985639,
+      "grad_norm": 0.5497506856918335,
+      "learning_rate": 0.0004081544781368268,
+      "loss": 0.8989,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5640888413165641,
+      "grad_norm": 0.8488923907279968,
+      "learning_rate": 0.0004081002112088641,
+      "loss": 0.869,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5642672375345643,
+      "grad_norm": 0.539082944393158,
+      "learning_rate": 0.0004080459318637789,
+      "loss": 1.0027,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5644456337525644,
+      "grad_norm": 0.6145626306533813,
+      "learning_rate": 0.0004079916401058342,
+      "loss": 1.1312,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5646240299705646,
+      "grad_norm": 0.5471964478492737,
+      "learning_rate": 0.00040793733593929405,
+      "loss": 1.0311,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5648024261885648,
+      "grad_norm": 1.0628180503845215,
+      "learning_rate": 0.00040788301936842353,
+      "loss": 0.9842,
+      "step": 3166
+    },
+    {
+      "epoch": 0.564980822406565,
+      "grad_norm": 0.6143336892127991,
+      "learning_rate": 0.00040782869039748847,
+      "loss": 1.0282,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5651592186245652,
+      "grad_norm": 0.4468381702899933,
+      "learning_rate": 0.0004077743490307562,
+      "loss": 0.6927,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5653376148425653,
+      "grad_norm": 0.524698793888092,
+      "learning_rate": 0.0004077199952724944,
+      "loss": 1.0529,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5655160110605655,
+      "grad_norm": 0.4760201573371887,
+      "learning_rate": 0.0004076656291269719,
+      "loss": 0.8662,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5656944072785657,
+      "grad_norm": 0.5162889957427979,
+      "learning_rate": 0.00040761125059845887,
+      "loss": 0.9204,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5658728034965659,
+      "grad_norm": 0.5113767981529236,
+      "learning_rate": 0.00040755685969122603,
+      "loss": 1.0988,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5660511997145661,
+      "grad_norm": 0.5655469298362732,
+      "learning_rate": 0.0004075024564095452,
+      "loss": 0.918,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5662295959325663,
+      "grad_norm": 0.5044064521789551,
+      "learning_rate": 0.0004074480407576892,
+      "loss": 1.1065,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5664079921505664,
+      "grad_norm": 0.5524999499320984,
+      "learning_rate": 0.0004073936127399319,
+      "loss": 1.0733,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5665863883685666,
+      "grad_norm": 0.4410164952278137,
+      "learning_rate": 0.000407339172360548,
+      "loss": 0.84,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5667647845865668,
+      "grad_norm": 0.5362854599952698,
+      "learning_rate": 0.0004072847196238131,
+      "loss": 0.9016,
+      "step": 3177
+    },
+    {
+      "epoch": 0.566943180804567,
+      "grad_norm": 1.1655226945877075,
+      "learning_rate": 0.0004072302545340041,
+      "loss": 1.0215,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5671215770225672,
+      "grad_norm": 5.327235698699951,
+      "learning_rate": 0.00040717577709539857,
+      "loss": 1.6297,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5672999732405672,
+      "grad_norm": 0.6369319558143616,
+      "learning_rate": 0.00040712128731227513,
+      "loss": 1.1747,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5674783694585674,
+      "grad_norm": 0.5702166557312012,
+      "learning_rate": 0.0004070667851889134,
+      "loss": 1.0747,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5676567656765676,
+      "grad_norm": 0.6028847694396973,
+      "learning_rate": 0.0004070122707295939,
+      "loss": 0.7787,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5678351618945678,
+      "grad_norm": 1.2918931245803833,
+      "learning_rate": 0.0004069577439385982,
+      "loss": 1.7838,
+      "step": 3183
+    },
+    {
+      "epoch": 0.568013558112568,
+      "grad_norm": 0.5654813647270203,
+      "learning_rate": 0.00040690320482020893,
+      "loss": 0.9124,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5681919543305682,
+      "grad_norm": 0.902948796749115,
+      "learning_rate": 0.00040684865337870945,
+      "loss": 0.8613,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5683703505485683,
+      "grad_norm": 1.1347112655639648,
+      "learning_rate": 0.00040679408961838426,
+      "loss": 1.1453,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5685487467665685,
+      "grad_norm": 0.5079895853996277,
+      "learning_rate": 0.0004067395135435187,
+      "loss": 0.8407,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5687271429845687,
+      "grad_norm": 0.7240801453590393,
+      "learning_rate": 0.0004066849251583992,
+      "loss": 0.8808,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5689055392025689,
+      "grad_norm": 0.5793579816818237,
+      "learning_rate": 0.0004066303244673132,
+      "loss": 0.915,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5690839354205691,
+      "grad_norm": 1.3618862628936768,
+      "learning_rate": 0.00040657571147454877,
+      "loss": 0.9031,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5692623316385692,
+      "grad_norm": 14.210494995117188,
+      "learning_rate": 0.0004065210861843954,
+      "loss": 1.538,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5694407278565694,
+      "grad_norm": 2.7108030319213867,
+      "learning_rate": 0.0004064664486011433,
+      "loss": 0.8688,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5696191240745696,
+      "grad_norm": 1.628589391708374,
+      "learning_rate": 0.0004064117987290836,
+      "loss": 1.0493,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5697975202925698,
+      "grad_norm": 4.802014350891113,
+      "learning_rate": 0.0004063571365725086,
+      "loss": 1.0736,
+      "step": 3194
+    },
+    {
+      "epoch": 0.56997591651057,
+      "grad_norm": 0.9495110511779785,
+      "learning_rate": 0.00040630246213571136,
+      "loss": 0.8673,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5701543127285702,
+      "grad_norm": 1.0658725500106812,
+      "learning_rate": 0.000406247775422986,
+      "loss": 1.0411,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5703327089465703,
+      "grad_norm": 0.6458380222320557,
+      "learning_rate": 0.00040619307643862757,
+      "loss": 1.1536,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5705111051645705,
+      "grad_norm": 91.1878662109375,
+      "learning_rate": 0.00040613836518693213,
+      "loss": 1.0854,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5706895013825707,
+      "grad_norm": 1.0638850927352905,
+      "learning_rate": 0.0004060836416721968,
+      "loss": 0.9599,
+      "step": 3199
+    },
+    {
+      "epoch": 0.5708678976005709,
+      "grad_norm": 0.7722949981689453,
+      "learning_rate": 0.00040602890589871933,
+      "loss": 0.9146,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5710462938185711,
+      "grad_norm": 1.2300078868865967,
+      "learning_rate": 0.0004059741578707987,
+      "loss": 0.9866,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5712246900365712,
+      "grad_norm": 0.5897446870803833,
+      "learning_rate": 0.00040591939759273486,
+      "loss": 0.8663,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5714030862545714,
+      "grad_norm": 1.1152317523956299,
+      "learning_rate": 0.0004058646250688287,
+      "loss": 1.1437,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5715814824725716,
+      "grad_norm": 0.5559517741203308,
+      "learning_rate": 0.00040580984030338187,
+      "loss": 0.7943,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5717598786905718,
+      "grad_norm": 0.8801344633102417,
+      "learning_rate": 0.0004057550433006972,
+      "loss": 1.3015,
+      "step": 3205
+    },
+    {
+      "epoch": 0.571938274908572,
+      "grad_norm": 0.9479907155036926,
+      "learning_rate": 0.00040570023406507857,
+      "loss": 0.8974,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5721166711265722,
+      "grad_norm": 0.6277695298194885,
+      "learning_rate": 0.0004056454126008305,
+      "loss": 1.0662,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5722950673445723,
+      "grad_norm": 0.476595014333725,
+      "learning_rate": 0.0004055905789122587,
+      "loss": 0.7801,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5724734635625724,
+      "grad_norm": 0.6994301080703735,
+      "learning_rate": 0.00040553573300366986,
+      "loss": 1.223,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5726518597805726,
+      "grad_norm": 0.8489440679550171,
+      "learning_rate": 0.0004054808748793714,
+      "loss": 0.863,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5728302559985728,
+      "grad_norm": 0.8950226902961731,
+      "learning_rate": 0.00040542600454367193,
+      "loss": 0.9557,
+      "step": 3211
+    },
+    {
+      "epoch": 0.573008652216573,
+      "grad_norm": 0.6158692240715027,
+      "learning_rate": 0.000405371122000881,
+      "loss": 1.0143,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5731870484345731,
+      "grad_norm": 27.874290466308594,
+      "learning_rate": 0.00040531622725530894,
+      "loss": 0.8735,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5733654446525733,
+      "grad_norm": 33.128963470458984,
+      "learning_rate": 0.0004052613203112673,
+      "loss": 1.7168,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5735438408705735,
+      "grad_norm": 2.284363031387329,
+      "learning_rate": 0.0004052064011730684,
+      "loss": 0.9891,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5737222370885737,
+      "grad_norm": 0.8609527945518494,
+      "learning_rate": 0.0004051514698450255,
+      "loss": 0.9398,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5739006333065739,
+      "grad_norm": 0.9100723266601562,
+      "learning_rate": 0.0004050965263314529,
+      "loss": 0.7602,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5740790295245741,
+      "grad_norm": 8.203483581542969,
+      "learning_rate": 0.0004050415706366659,
+      "loss": 0.9475,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5742574257425742,
+      "grad_norm": 0.828872561454773,
+      "learning_rate": 0.0004049866027649807,
+      "loss": 0.9817,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5744358219605744,
+      "grad_norm": 0.49633899331092834,
+      "learning_rate": 0.00040493162272071427,
+      "loss": 0.5957,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5746142181785746,
+      "grad_norm": 4.844064712524414,
+      "learning_rate": 0.000404876630508185,
+      "loss": 1.3244,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5747926143965748,
+      "grad_norm": 0.8198102116584778,
+      "learning_rate": 0.00040482162613171167,
+      "loss": 1.0323,
+      "step": 3222
+    },
+    {
+      "epoch": 0.574971010614575,
+      "grad_norm": 1.0812023878097534,
+      "learning_rate": 0.00040476660959561464,
+      "loss": 1.2488,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5751494068325751,
+      "grad_norm": 0.8954762816429138,
+      "learning_rate": 0.0004047115809042146,
+      "loss": 0.7903,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5753278030505753,
+      "grad_norm": 2.91465163230896,
+      "learning_rate": 0.0004046565400618336,
+      "loss": 1.1885,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5755061992685755,
+      "grad_norm": 1.1171669960021973,
+      "learning_rate": 0.0004046014870727944,
+      "loss": 1.0482,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5756845954865757,
+      "grad_norm": 0.8723791837692261,
+      "learning_rate": 0.0004045464219414211,
+      "loss": 0.8968,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5758629917045759,
+      "grad_norm": 0.8948122262954712,
+      "learning_rate": 0.0004044913446720382,
+      "loss": 0.9063,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5760413879225761,
+      "grad_norm": 0.7297793626785278,
+      "learning_rate": 0.0004044362552689716,
+      "loss": 0.9518,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5762197841405762,
+      "grad_norm": 0.7813421487808228,
+      "learning_rate": 0.000404381153736548,
+      "loss": 0.923,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5763981803585764,
+      "grad_norm": 0.5343239307403564,
+      "learning_rate": 0.00040432604007909504,
+      "loss": 0.9697,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 0.5333415269851685,
+      "learning_rate": 0.0004042709143009412,
+      "loss": 0.7906,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5767549727945768,
+      "grad_norm": 0.5309287309646606,
+      "learning_rate": 0.0004042157764064163,
+      "loss": 0.8805,
+      "step": 3233
+    },
+    {
+      "epoch": 0.576933369012577,
+      "grad_norm": 0.8856096863746643,
+      "learning_rate": 0.00040416062639985053,
+      "loss": 0.9808,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5771117652305772,
+      "grad_norm": 0.6096811890602112,
+      "learning_rate": 0.0004041054642855756,
+      "loss": 0.8629,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5772901614485773,
+      "grad_norm": 0.5578605532646179,
+      "learning_rate": 0.0004040502900679237,
+      "loss": 1.0188,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5774685576665775,
+      "grad_norm": 0.9917730689048767,
+      "learning_rate": 0.0004039951037512284,
+      "loss": 0.9557,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5776469538845777,
+      "grad_norm": 0.8114546537399292,
+      "learning_rate": 0.00040393990533982397,
+      "loss": 0.9911,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5778253501025779,
+      "grad_norm": 0.6181529760360718,
+      "learning_rate": 0.0004038846948380456,
+      "loss": 0.9745,
+      "step": 3239
+    },
+    {
+      "epoch": 0.578003746320578,
+      "grad_norm": 1.0224014520645142,
+      "learning_rate": 0.00040382947225022945,
+      "loss": 1.1888,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5781821425385781,
+      "grad_norm": 0.6258061528205872,
+      "learning_rate": 0.0004037742375807127,
+      "loss": 0.9421,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5783605387565783,
+      "grad_norm": 1.0917785167694092,
+      "learning_rate": 0.00040371899083383367,
+      "loss": 1.0569,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5785389349745785,
+      "grad_norm": 0.6143338084220886,
+      "learning_rate": 0.00040366373201393115,
+      "loss": 1.0274,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5787173311925787,
+      "grad_norm": 1.4802119731903076,
+      "learning_rate": 0.00040360846112534533,
+      "loss": 1.0424,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5788957274105789,
+      "grad_norm": 7.478890895843506,
+      "learning_rate": 0.000403553178172417,
+      "loss": 2.1221,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5790741236285791,
+      "grad_norm": 0.7259086966514587,
+      "learning_rate": 0.0004034978831594881,
+      "loss": 0.9648,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5792525198465792,
+      "grad_norm": 0.8427548408508301,
+      "learning_rate": 0.00040344257609090155,
+      "loss": 0.8322,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5794309160645794,
+      "grad_norm": 0.5752832889556885,
+      "learning_rate": 0.0004033872569710011,
+      "loss": 1.046,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5796093122825796,
+      "grad_norm": 0.8446690440177917,
+      "learning_rate": 0.0004033319258041316,
+      "loss": 1.0119,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5797877085005798,
+      "grad_norm": 0.5777229070663452,
+      "learning_rate": 0.0004032765825946385,
+      "loss": 0.9398,
+      "step": 3250
+    },
+    {
+      "epoch": 0.57996610471858,
+      "grad_norm": 1.039317011833191,
+      "learning_rate": 0.0004032212273468686,
+      "loss": 0.9415,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5801445009365801,
+      "grad_norm": 0.9003728628158569,
+      "learning_rate": 0.0004031658600651694,
+      "loss": 1.0614,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5803228971545803,
+      "grad_norm": 0.5967848896980286,
+      "learning_rate": 0.0004031104807538896,
+      "loss": 0.8657,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5805012933725805,
+      "grad_norm": 0.9425020813941956,
+      "learning_rate": 0.0004030550894173783,
+      "loss": 1.0047,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5806796895905807,
+      "grad_norm": 0.7595664262771606,
+      "learning_rate": 0.0004029996860599864,
+      "loss": 0.9598,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5808580858085809,
+      "grad_norm": 0.629330039024353,
+      "learning_rate": 0.0004029442706860649,
+      "loss": 1.0411,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5810364820265811,
+      "grad_norm": 0.588629961013794,
+      "learning_rate": 0.0004028888432999661,
+      "loss": 1.0998,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5812148782445812,
+      "grad_norm": 0.8526011109352112,
+      "learning_rate": 0.0004028334039060434,
+      "loss": 1.2057,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5813932744625814,
+      "grad_norm": 1.3286303281784058,
+      "learning_rate": 0.00040277795250865094,
+      "loss": 0.8182,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5815716706805816,
+      "grad_norm": 0.985548734664917,
+      "learning_rate": 0.0004027224891121438,
+      "loss": 1.0537,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5817500668985818,
+      "grad_norm": 0.6656898856163025,
+      "learning_rate": 0.0004026670137208782,
+      "loss": 0.7234,
+      "step": 3261
+    },
+    {
+      "epoch": 0.581928463116582,
+      "grad_norm": 1.2689619064331055,
+      "learning_rate": 0.00040261152633921097,
+      "loss": 1.115,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5821068593345821,
+      "grad_norm": 1.3990188837051392,
+      "learning_rate": 0.00040255602697150005,
+      "loss": 1.0482,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5822852555525823,
+      "grad_norm": 0.6655217409133911,
+      "learning_rate": 0.00040250051562210456,
+      "loss": 1.0004,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5824636517705825,
+      "grad_norm": 0.6590464115142822,
+      "learning_rate": 0.0004024449922953841,
+      "loss": 1.2517,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5826420479885827,
+      "grad_norm": 1.0634833574295044,
+      "learning_rate": 0.0004023894569956996,
+      "loss": 1.582,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5828204442065829,
+      "grad_norm": 0.7259156703948975,
+      "learning_rate": 0.00040233390972741276,
+      "loss": 1.0507,
+      "step": 3267
+    },
+    {
+      "epoch": 0.582998840424583,
+      "grad_norm": 0.5403652191162109,
+      "learning_rate": 0.00040227835049488615,
+      "loss": 0.8231,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5831772366425831,
+      "grad_norm": 0.6394296288490295,
+      "learning_rate": 0.0004022227793024834,
+      "loss": 0.9511,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5833556328605833,
+      "grad_norm": 0.8414486050605774,
+      "learning_rate": 0.0004021671961545691,
+      "loss": 1.097,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5835340290785835,
+      "grad_norm": 1.2080618143081665,
+      "learning_rate": 0.0004021116010555087,
+      "loss": 0.9097,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5837124252965837,
+      "grad_norm": 0.6970852613449097,
+      "learning_rate": 0.00040205599400966864,
+      "loss": 0.9379,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5838908215145839,
+      "grad_norm": 0.7079569101333618,
+      "learning_rate": 0.00040200037502141617,
+      "loss": 1.0276,
+      "step": 3273
+    },
+    {
+      "epoch": 0.584069217732584,
+      "grad_norm": 0.6691045761108398,
+      "learning_rate": 0.0004019447440951197,
+      "loss": 1.1062,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5842476139505842,
+      "grad_norm": 0.7754026055335999,
+      "learning_rate": 0.0004018891012351484,
+      "loss": 1.0128,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5844260101685844,
+      "grad_norm": 0.774585485458374,
+      "learning_rate": 0.0004018334464458725,
+      "loss": 0.9005,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5846044063865846,
+      "grad_norm": 0.6378253698348999,
+      "learning_rate": 0.000401777779731663,
+      "loss": 1.1857,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5847828026045848,
+      "grad_norm": 0.6450658440589905,
+      "learning_rate": 0.00040172210109689206,
+      "loss": 0.7742,
+      "step": 3278
+    },
+    {
+      "epoch": 0.584961198822585,
+      "grad_norm": 0.6561940908432007,
+      "learning_rate": 0.00040166641054593255,
+      "loss": 1.1851,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5851395950405851,
+      "grad_norm": 0.5854855179786682,
+      "learning_rate": 0.0004016107080831584,
+      "loss": 0.9683,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5853179912585853,
+      "grad_norm": 0.6039174199104309,
+      "learning_rate": 0.00040155499371294454,
+      "loss": 0.9314,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5854963874765855,
+      "grad_norm": 0.8320233225822449,
+      "learning_rate": 0.0004014992674396666,
+      "loss": 0.7973,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5856747836945857,
+      "grad_norm": 0.6969453692436218,
+      "learning_rate": 0.00040144352926770147,
+      "loss": 0.9077,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5858531799125859,
+      "grad_norm": 0.9570133686065674,
+      "learning_rate": 0.0004013877792014267,
+      "loss": 1.0072,
+      "step": 3284
+    },
+    {
+      "epoch": 0.586031576130586,
+      "grad_norm": 0.6976554989814758,
+      "learning_rate": 0.0004013320172452209,
+      "loss": 1.054,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5862099723485862,
+      "grad_norm": 0.6149062514305115,
+      "learning_rate": 0.00040127624340346356,
+      "loss": 0.9388,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5863883685665864,
+      "grad_norm": 0.517910897731781,
+      "learning_rate": 0.0004012204576805352,
+      "loss": 0.9243,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5865667647845866,
+      "grad_norm": 0.5356748700141907,
+      "learning_rate": 0.0004011646600808172,
+      "loss": 0.9906,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5867451610025868,
+      "grad_norm": 0.5061875581741333,
+      "learning_rate": 0.00040110885060869173,
+      "loss": 1.0385,
+      "step": 3289
+    },
+    {
+      "epoch": 0.586923557220587,
+      "grad_norm": 0.6880626082420349,
+      "learning_rate": 0.00040105302926854224,
+      "loss": 0.8258,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5871019534385871,
+      "grad_norm": 0.4615683853626251,
+      "learning_rate": 0.00040099719606475286,
+      "loss": 0.8729,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5872803496565873,
+      "grad_norm": 0.7369007468223572,
+      "learning_rate": 0.0004009413510017087,
+      "loss": 0.9377,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5874587458745875,
+      "grad_norm": 0.5646438598632812,
+      "learning_rate": 0.0004008854940837957,
+      "loss": 1.1106,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5876371420925877,
+      "grad_norm": 0.5093971490859985,
+      "learning_rate": 0.000400829625315401,
+      "loss": 0.9263,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5878155383105879,
+      "grad_norm": 0.8065229058265686,
+      "learning_rate": 0.00040077374470091237,
+      "loss": 0.7977,
+      "step": 3295
+    },
+    {
+      "epoch": 0.587993934528588,
+      "grad_norm": 1.8008029460906982,
+      "learning_rate": 0.0004007178522447188,
+      "loss": 1.1636,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5881723307465881,
+      "grad_norm": 0.5000921487808228,
+      "learning_rate": 0.00040066194795120984,
+      "loss": 1.0567,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5883507269645883,
+      "grad_norm": 0.5407942533493042,
+      "learning_rate": 0.0004006060318247764,
+      "loss": 1.1229,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5885291231825885,
+      "grad_norm": 0.8123578429222107,
+      "learning_rate": 0.00040055010386981006,
+      "loss": 1.0442,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5887075194005887,
+      "grad_norm": 0.5218117237091064,
+      "learning_rate": 0.00040049416409070326,
+      "loss": 0.8096,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5888859156185889,
+      "grad_norm": 0.5112317204475403,
+      "learning_rate": 0.0004004382124918497,
+      "loss": 0.9776,
+      "step": 3301
+    },
+    {
+      "epoch": 0.589064311836589,
+      "grad_norm": 0.539185643196106,
+      "learning_rate": 0.00040038224907764356,
+      "loss": 1.0191,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5892427080545892,
+      "grad_norm": 0.5724341869354248,
+      "learning_rate": 0.0004003262738524804,
+      "loss": 1.1246,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5894211042725894,
+      "grad_norm": 0.568626880645752,
+      "learning_rate": 0.00040027028682075626,
+      "loss": 1.0875,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5895995004905896,
+      "grad_norm": 0.48848995566368103,
+      "learning_rate": 0.00040021428798686854,
+      "loss": 0.8293,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5897778967085898,
+      "grad_norm": 0.5772598385810852,
+      "learning_rate": 0.00040015827735521525,
+      "loss": 0.8439,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5899562929265899,
+      "grad_norm": 1.8581256866455078,
+      "learning_rate": 0.0004001022549301955,
+      "loss": 1.0762,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5901346891445901,
+      "grad_norm": 0.6715808510780334,
+      "learning_rate": 0.00040004622071620924,
+      "loss": 0.8739,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5903130853625903,
+      "grad_norm": 0.5865873694419861,
+      "learning_rate": 0.00039999017471765736,
+      "loss": 1.2157,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5904914815805905,
+      "grad_norm": 0.6115093231201172,
+      "learning_rate": 0.0003999341169389417,
+      "loss": 0.9827,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5906698777985907,
+      "grad_norm": 1.7687232494354248,
+      "learning_rate": 0.0003998780473844651,
+      "loss": 0.7466,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5908482740165909,
+      "grad_norm": 0.5366143584251404,
+      "learning_rate": 0.00039982196605863095,
+      "loss": 0.6974,
+      "step": 3312
+    },
+    {
+      "epoch": 0.591026670234591,
+      "grad_norm": 0.48479849100112915,
+      "learning_rate": 0.0003997658729658442,
+      "loss": 0.8932,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5912050664525912,
+      "grad_norm": 0.5348507761955261,
+      "learning_rate": 0.0003997097681105103,
+      "loss": 1.0747,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5913834626705914,
+      "grad_norm": 0.5859421491622925,
+      "learning_rate": 0.00039965365149703555,
+      "loss": 0.9456,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5915618588885916,
+      "grad_norm": 0.5688529014587402,
+      "learning_rate": 0.00039959752312982745,
+      "loss": 0.985,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5917402551065918,
+      "grad_norm": 0.5745177865028381,
+      "learning_rate": 0.00039954138301329426,
+      "loss": 1.0356,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5919186513245919,
+      "grad_norm": 0.45974186062812805,
+      "learning_rate": 0.00039948523115184516,
+      "loss": 0.8968,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5920970475425921,
+      "grad_norm": 0.5509001612663269,
+      "learning_rate": 0.00039942906754989035,
+      "loss": 1.0219,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5922754437605923,
+      "grad_norm": 3.9084794521331787,
+      "learning_rate": 0.0003993728922118408,
+      "loss": 1.1745,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5924538399785925,
+      "grad_norm": 20.251436233520508,
+      "learning_rate": 0.0003993167051421087,
+      "loss": 0.842,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5926322361965927,
+      "grad_norm": 0.7031044363975525,
+      "learning_rate": 0.0003992605063451068,
+      "loss": 0.9488,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5928106324145929,
+      "grad_norm": 0.5919948816299438,
+      "learning_rate": 0.00039920429582524896,
+      "loss": 1.0003,
+      "step": 3323
+    },
+    {
+      "epoch": 0.592989028632593,
+      "grad_norm": 1.6224722862243652,
+      "learning_rate": 0.00039914807358694995,
+      "loss": 1.2006,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5931674248505932,
+      "grad_norm": 0.5909742712974548,
+      "learning_rate": 0.00039909183963462535,
+      "loss": 0.9017,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5933458210685933,
+      "grad_norm": 0.5360785126686096,
+      "learning_rate": 0.0003990355939726919,
+      "loss": 0.9806,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5935242172865935,
+      "grad_norm": 0.6022999286651611,
+      "learning_rate": 0.00039897933660556703,
+      "loss": 1.1784,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5937026135045937,
+      "grad_norm": 0.5530278086662292,
+      "learning_rate": 0.0003989230675376691,
+      "loss": 0.7749,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5938810097225938,
+      "grad_norm": 0.48187515139579773,
+      "learning_rate": 0.0003988667867734176,
+      "loss": 0.8854,
+      "step": 3329
+    },
+    {
+      "epoch": 0.594059405940594,
+      "grad_norm": 0.5293228030204773,
+      "learning_rate": 0.0003988104943172327,
+      "loss": 0.9786,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5942378021585942,
+      "grad_norm": 1.1052613258361816,
+      "learning_rate": 0.00039875419017353564,
+      "loss": 1.1643,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5944161983765944,
+      "grad_norm": 0.663966953754425,
+      "learning_rate": 0.00039869787434674853,
+      "loss": 1.2176,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 0.6256305575370789,
+      "learning_rate": 0.0003986415468412943,
+      "loss": 1.0519,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5947729908125948,
+      "grad_norm": 14.08318042755127,
+      "learning_rate": 0.00039858520766159703,
+      "loss": 1.3631,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5949513870305949,
+      "grad_norm": 0.6284887790679932,
+      "learning_rate": 0.00039852885681208134,
+      "loss": 0.8746,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5951297832485951,
+      "grad_norm": 0.9674341678619385,
+      "learning_rate": 0.00039847249429717326,
+      "loss": 1.2364,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5953081794665953,
+      "grad_norm": 0.6534430980682373,
+      "learning_rate": 0.00039841612012129937,
+      "loss": 1.1151,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5954865756845955,
+      "grad_norm": 0.5643439888954163,
+      "learning_rate": 0.0003983597342888872,
+      "loss": 0.8678,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5956649719025957,
+      "grad_norm": 0.5396862626075745,
+      "learning_rate": 0.0003983033368043654,
+      "loss": 0.9284,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5958433681205958,
+      "grad_norm": 0.6234003305435181,
+      "learning_rate": 0.00039824692767216337,
+      "loss": 1.0596,
+      "step": 3340
+    },
+    {
+      "epoch": 0.596021764338596,
+      "grad_norm": 0.5749251246452332,
+      "learning_rate": 0.00039819050689671143,
+      "loss": 0.8237,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5962001605565962,
+      "grad_norm": 0.5611945390701294,
+      "learning_rate": 0.0003981340744824408,
+      "loss": 0.9064,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5963785567745964,
+      "grad_norm": 0.6729212403297424,
+      "learning_rate": 0.0003980776304337838,
+      "loss": 1.1313,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5965569529925966,
+      "grad_norm": 0.5482896566390991,
+      "learning_rate": 0.0003980211747551733,
+      "loss": 0.9448,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5967353492105968,
+      "grad_norm": 44.00550842285156,
+      "learning_rate": 0.0003979647074510435,
+      "loss": 1.6368,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5969137454285969,
+      "grad_norm": 0.5960306525230408,
+      "learning_rate": 0.00039790822852582927,
+      "loss": 1.0672,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5970921416465971,
+      "grad_norm": 6.083531856536865,
+      "learning_rate": 0.00039785173798396637,
+      "loss": 1.4715,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5972705378645973,
+      "grad_norm": 0.8834306597709656,
+      "learning_rate": 0.00039779523582989163,
+      "loss": 1.1152,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5974489340825975,
+      "grad_norm": 1.775227427482605,
+      "learning_rate": 0.0003977387220680427,
+      "loss": 1.1541,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5976273303005977,
+      "grad_norm": 0.7069253325462341,
+      "learning_rate": 0.00039768219670285805,
+      "loss": 0.889,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5978057265185978,
+      "grad_norm": 1.1338763236999512,
+      "learning_rate": 0.00039762565973877726,
+      "loss": 0.8112,
+      "step": 3351
+    },
+    {
+      "epoch": 0.597984122736598,
+      "grad_norm": 1.5388410091400146,
+      "learning_rate": 0.00039756911118024065,
+      "loss": 1.0269,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5981625189545982,
+      "grad_norm": 0.7311750650405884,
+      "learning_rate": 0.0003975125510316896,
+      "loss": 0.9498,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5983409151725984,
+      "grad_norm": 2.2909646034240723,
+      "learning_rate": 0.0003974559792975663,
+      "loss": 0.9101,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5985193113905986,
+      "grad_norm": 0.8901395201683044,
+      "learning_rate": 0.0003973993959823137,
+      "loss": 0.8863,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5986977076085988,
+      "grad_norm": 0.6024214029312134,
+      "learning_rate": 0.00039734280109037613,
+      "loss": 1.0322,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5988761038265988,
+      "grad_norm": 0.6770859956741333,
+      "learning_rate": 0.0003972861946261983,
+      "loss": 1.1135,
+      "step": 3357
+    },
+    {
+      "epoch": 0.599054500044599,
+      "grad_norm": 0.73415607213974,
+      "learning_rate": 0.0003972295765942261,
+      "loss": 0.9596,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5992328962625992,
+      "grad_norm": 0.6480002403259277,
+      "learning_rate": 0.00039717294699890627,
+      "loss": 0.9741,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5994112924805994,
+      "grad_norm": 1.2426977157592773,
+      "learning_rate": 0.0003971163058446866,
+      "loss": 1.0937,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5995896886985996,
+      "grad_norm": 0.8961933851242065,
+      "learning_rate": 0.0003970596531360156,
+      "loss": 0.9281,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5997680849165997,
+      "grad_norm": 0.7000132203102112,
+      "learning_rate": 0.00039700298887734273,
+      "loss": 1.0097,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5999464811345999,
+      "grad_norm": 1.2038791179656982,
+      "learning_rate": 0.0003969463130731183,
+      "loss": 0.9242,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6001248773526001,
+      "grad_norm": 1.661060094833374,
+      "learning_rate": 0.00039688962572779373,
+      "loss": 1.1603,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6003032735706003,
+      "grad_norm": 1.0910298824310303,
+      "learning_rate": 0.0003968329268458212,
+      "loss": 0.9369,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6004816697886005,
+      "grad_norm": 1.3606114387512207,
+      "learning_rate": 0.00039677621643165363,
+      "loss": 0.9468,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6006600660066007,
+      "grad_norm": 1.111076831817627,
+      "learning_rate": 0.0003967194944897453,
+      "loss": 1.1383,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6008384622246008,
+      "grad_norm": 1.8882251977920532,
+      "learning_rate": 0.000396662761024551,
+      "loss": 0.8431,
+      "step": 3368
+    },
+    {
+      "epoch": 0.601016858442601,
+      "grad_norm": 1.5878911018371582,
+      "learning_rate": 0.0003966060160405266,
+      "loss": 1.0407,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6011952546606012,
+      "grad_norm": 0.7476792931556702,
+      "learning_rate": 0.00039654925954212873,
+      "loss": 0.7851,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6013736508786014,
+      "grad_norm": 0.8636188507080078,
+      "learning_rate": 0.00039649249153381514,
+      "loss": 1.0617,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6015520470966016,
+      "grad_norm": 0.7655645608901978,
+      "learning_rate": 0.00039643571202004426,
+      "loss": 0.9721,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6017304433146017,
+      "grad_norm": 0.6459742188453674,
+      "learning_rate": 0.0003963789210052755,
+      "loss": 0.8403,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6019088395326019,
+      "grad_norm": 0.7047243714332581,
+      "learning_rate": 0.00039632211849396936,
+      "loss": 1.1643,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6020872357506021,
+      "grad_norm": 0.5050498247146606,
+      "learning_rate": 0.000396265304490587,
+      "loss": 0.7855,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6022656319686023,
+      "grad_norm": 0.9648879766464233,
+      "learning_rate": 0.0003962084789995906,
+      "loss": 0.9511,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6024440281866025,
+      "grad_norm": 0.5397990345954895,
+      "learning_rate": 0.00039615164202544314,
+      "loss": 0.8164,
+      "step": 3377
+    },
+    {
+      "epoch": 0.6026224244046027,
+      "grad_norm": 2.0958774089813232,
+      "learning_rate": 0.0003960947935726086,
+      "loss": 0.8944,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6028008206226028,
+      "grad_norm": 1.3503490686416626,
+      "learning_rate": 0.00039603793364555184,
+      "loss": 0.8005,
+      "step": 3379
+    },
+    {
+      "epoch": 0.602979216840603,
+      "grad_norm": 0.5871055722236633,
+      "learning_rate": 0.00039598106224873866,
+      "loss": 0.9893,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6031576130586032,
+      "grad_norm": 1.3168357610702515,
+      "learning_rate": 0.0003959241793866356,
+      "loss": 0.9926,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6033360092766034,
+      "grad_norm": 0.8099937438964844,
+      "learning_rate": 0.0003958672850637103,
+      "loss": 0.9152,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6035144054946036,
+      "grad_norm": 0.5617466568946838,
+      "learning_rate": 0.0003958103792844313,
+      "loss": 0.9704,
+      "step": 3383
+    },
+    {
+      "epoch": 0.6036928017126036,
+      "grad_norm": 0.758649468421936,
+      "learning_rate": 0.00039575346205326776,
+      "loss": 0.869,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6038711979306038,
+      "grad_norm": 0.5572230219841003,
+      "learning_rate": 0.0003956965333746901,
+      "loss": 1.0961,
+      "step": 3385
+    },
+    {
+      "epoch": 0.604049594148604,
+      "grad_norm": 78.27342224121094,
+      "learning_rate": 0.00039563959325316934,
+      "loss": 1.2738,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6042279903666042,
+      "grad_norm": 2.010958194732666,
+      "learning_rate": 0.00039558264169317766,
+      "loss": 0.9795,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6044063865846044,
+      "grad_norm": 1.1752885580062866,
+      "learning_rate": 0.0003955256786991879,
+      "loss": 1.334,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6045847828026046,
+      "grad_norm": 0.719290018081665,
+      "learning_rate": 0.0003954687042756739,
+      "loss": 0.9268,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6047631790206047,
+      "grad_norm": 0.9729273915290833,
+      "learning_rate": 0.00039541171842711063,
+      "loss": 1.1478,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6049415752386049,
+      "grad_norm": 0.5412781834602356,
+      "learning_rate": 0.00039535472115797345,
+      "loss": 1.0843,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6051199714566051,
+      "grad_norm": 0.6292357444763184,
+      "learning_rate": 0.00039529771247273903,
+      "loss": 1.2036,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6052983676746053,
+      "grad_norm": 0.617896556854248,
+      "learning_rate": 0.0003952406923758849,
+      "loss": 1.0214,
+      "step": 3393
+    },
+    {
+      "epoch": 0.6054767638926055,
+      "grad_norm": 0.4818667471408844,
+      "learning_rate": 0.00039518366087188924,
+      "loss": 1.0368,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6056551601106056,
+      "grad_norm": 0.7148341536521912,
+      "learning_rate": 0.0003951266179652313,
+      "loss": 0.9985,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6058335563286058,
+      "grad_norm": 1.9086631536483765,
+      "learning_rate": 0.0003950695636603912,
+      "loss": 1.0418,
+      "step": 3396
+    },
+    {
+      "epoch": 0.606011952546606,
+      "grad_norm": 0.5101553201675415,
+      "learning_rate": 0.00039501249796185006,
+      "loss": 0.7728,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6061903487646062,
+      "grad_norm": 0.5302074551582336,
+      "learning_rate": 0.00039495542087408976,
+      "loss": 0.8685,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6063687449826064,
+      "grad_norm": 0.6329625844955444,
+      "learning_rate": 0.000394898332401593,
+      "loss": 1.0125,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6065471412006066,
+      "grad_norm": 0.6048588752746582,
+      "learning_rate": 0.0003948412325488436,
+      "loss": 1.1885,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6067255374186067,
+      "grad_norm": 0.5368157029151917,
+      "learning_rate": 0.00039478412132032615,
+      "loss": 1.0989,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6069039336366069,
+      "grad_norm": 0.5177137851715088,
+      "learning_rate": 0.0003947269987205261,
+      "loss": 0.9014,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6070823298546071,
+      "grad_norm": 0.6155733466148376,
+      "learning_rate": 0.00039466986475392987,
+      "loss": 1.1821,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6072607260726073,
+      "grad_norm": 0.49641674757003784,
+      "learning_rate": 0.0003946127194250247,
+      "loss": 1.0431,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6074391222906075,
+      "grad_norm": 0.48299992084503174,
+      "learning_rate": 0.00039455556273829877,
+      "loss": 0.8856,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6076175185086076,
+      "grad_norm": 0.49870428442955017,
+      "learning_rate": 0.0003944983946982412,
+      "loss": 0.9731,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6077959147266078,
+      "grad_norm": 0.47452113032341003,
+      "learning_rate": 0.00039444121530934185,
+      "loss": 0.7636,
+      "step": 3407
+    },
+    {
+      "epoch": 0.607974310944608,
+      "grad_norm": 0.5165370106697083,
+      "learning_rate": 0.0003943840245760916,
+      "loss": 1.1063,
+      "step": 3408
+    },
+    {
+      "epoch": 0.6081527071626082,
+      "grad_norm": 0.5465608835220337,
+      "learning_rate": 0.00039432682250298225,
+      "loss": 1.1446,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6083311033806084,
+      "grad_norm": 0.4796379804611206,
+      "learning_rate": 0.00039426960909450627,
+      "loss": 0.9792,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6085094995986086,
+      "grad_norm": 0.48459067940711975,
+      "learning_rate": 0.00039421238435515736,
+      "loss": 1.1648,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6086878958166086,
+      "grad_norm": 0.5157718062400818,
+      "learning_rate": 0.00039415514828942976,
+      "loss": 1.0494,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6088662920346088,
+      "grad_norm": 0.4796636998653412,
+      "learning_rate": 0.00039409790090181896,
+      "loss": 0.8812,
+      "step": 3413
+    },
+    {
+      "epoch": 0.609044688252609,
+      "grad_norm": 0.5595741868019104,
+      "learning_rate": 0.000394040642196821,
+      "loss": 0.8946,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6092230844706092,
+      "grad_norm": 0.5820662975311279,
+      "learning_rate": 0.00039398337217893295,
+      "loss": 1.1571,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6094014806886094,
+      "grad_norm": 0.4832160770893097,
+      "learning_rate": 0.0003939260908526528,
+      "loss": 0.8372,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6095798769066095,
+      "grad_norm": 0.4839080274105072,
+      "learning_rate": 0.00039386879822247945,
+      "loss": 0.8548,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6097582731246097,
+      "grad_norm": 0.45255133509635925,
+      "learning_rate": 0.00039381149429291263,
+      "loss": 0.9553,
+      "step": 3418
+    },
+    {
+      "epoch": 0.6099366693426099,
+      "grad_norm": 0.4793449342250824,
+      "learning_rate": 0.00039375417906845284,
+      "loss": 0.965,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6101150655606101,
+      "grad_norm": 0.4889264404773712,
+      "learning_rate": 0.0003936968525536018,
+      "loss": 1.0507,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6102934617786103,
+      "grad_norm": 0.5014956593513489,
+      "learning_rate": 0.00039363951475286164,
+      "loss": 0.9827,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6104718579966105,
+      "grad_norm": 0.9582971334457397,
+      "learning_rate": 0.0003935821656707359,
+      "loss": 1.0391,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6106502542146106,
+      "grad_norm": 0.5389576554298401,
+      "learning_rate": 0.00039352480531172873,
+      "loss": 1.0455,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6108286504326108,
+      "grad_norm": 0.4678865373134613,
+      "learning_rate": 0.000393467433680345,
+      "loss": 0.8876,
+      "step": 3424
+    },
+    {
+      "epoch": 0.611007046650611,
+      "grad_norm": 0.581387996673584,
+      "learning_rate": 0.00039341005078109083,
+      "loss": 1.0716,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6111854428686112,
+      "grad_norm": 0.47930246591567993,
+      "learning_rate": 0.000393352656618473,
+      "loss": 0.7716,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6113638390866114,
+      "grad_norm": 0.5663626790046692,
+      "learning_rate": 0.0003932952511969991,
+      "loss": 0.8454,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6115422353046115,
+      "grad_norm": 0.5635462403297424,
+      "learning_rate": 0.0003932378345211779,
+      "loss": 0.9529,
+      "step": 3428
+    },
+    {
+      "epoch": 0.6117206315226117,
+      "grad_norm": 0.46473008394241333,
+      "learning_rate": 0.0003931804065955188,
+      "loss": 0.8475,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6118990277406119,
+      "grad_norm": 0.5133330225944519,
+      "learning_rate": 0.00039312296742453223,
+      "loss": 0.896,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6120774239586121,
+      "grad_norm": 0.5049217343330383,
+      "learning_rate": 0.0003930655170127294,
+      "loss": 1.0923,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6122558201766123,
+      "grad_norm": 0.4742034375667572,
+      "learning_rate": 0.00039300805536462237,
+      "loss": 0.7796,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6124342163946125,
+      "grad_norm": 0.5486430525779724,
+      "learning_rate": 0.0003929505824847243,
+      "loss": 0.92,
+      "step": 3433
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 0.514842689037323,
+      "learning_rate": 0.00039289309837754895,
+      "loss": 1.0119,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6127910088306128,
+      "grad_norm": 0.4459700584411621,
+      "learning_rate": 0.000392835603047611,
+      "loss": 0.6496,
+      "step": 3435
+    },
+    {
+      "epoch": 0.612969405048613,
+      "grad_norm": 0.525751531124115,
+      "learning_rate": 0.00039277809649942644,
+      "loss": 0.8839,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6131478012666132,
+      "grad_norm": 0.5619456171989441,
+      "learning_rate": 0.0003927205787375115,
+      "loss": 0.8713,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6133261974846134,
+      "grad_norm": 0.4977574348449707,
+      "learning_rate": 0.0003926630497663839,
+      "loss": 0.9844,
+      "step": 3438
+    },
+    {
+      "epoch": 0.6135045937026135,
+      "grad_norm": 0.5254572033882141,
+      "learning_rate": 0.0003926055095905616,
+      "loss": 1.085,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6136829899206137,
+      "grad_norm": 0.4378454089164734,
+      "learning_rate": 0.000392547958214564,
+      "loss": 0.7315,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6138613861386139,
+      "grad_norm": 0.5086297392845154,
+      "learning_rate": 0.0003924903956429111,
+      "loss": 1.0405,
+      "step": 3441
+    },
+    {
+      "epoch": 0.614039782356614,
+      "grad_norm": 0.47228503227233887,
+      "learning_rate": 0.00039243282188012387,
+      "loss": 0.8687,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6142181785746142,
+      "grad_norm": 0.46949103474617004,
+      "learning_rate": 0.0003923752369307241,
+      "loss": 0.7208,
+      "step": 3443
+    },
+    {
+      "epoch": 0.6143965747926144,
+      "grad_norm": 0.5485444664955139,
+      "learning_rate": 0.00039231764079923447,
+      "loss": 1.071,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6145749710106145,
+      "grad_norm": 0.610989511013031,
+      "learning_rate": 0.0003922600334901786,
+      "loss": 0.989,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6147533672286147,
+      "grad_norm": 0.48183125257492065,
+      "learning_rate": 0.0003922024150080808,
+      "loss": 0.9965,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6149317634466149,
+      "grad_norm": 0.540808379650116,
+      "learning_rate": 0.00039214478535746665,
+      "loss": 1.0388,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6151101596646151,
+      "grad_norm": 0.4990650415420532,
+      "learning_rate": 0.0003920871445428622,
+      "loss": 1.1055,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6152885558826153,
+      "grad_norm": 0.5104119181632996,
+      "learning_rate": 0.00039202949256879463,
+      "loss": 0.9729,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6154669521006154,
+      "grad_norm": 0.48018768429756165,
+      "learning_rate": 0.0003919718294397917,
+      "loss": 0.9862,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6156453483186156,
+      "grad_norm": 0.48195213079452515,
+      "learning_rate": 0.0003919141551603824,
+      "loss": 1.0078,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6158237445366158,
+      "grad_norm": 0.4913448095321655,
+      "learning_rate": 0.0003918564697350965,
+      "loss": 0.8028,
+      "step": 3452
+    },
+    {
+      "epoch": 0.616002140754616,
+      "grad_norm": 1.1352670192718506,
+      "learning_rate": 0.00039179877316846453,
+      "loss": 1.019,
+      "step": 3453
+    },
+    {
+      "epoch": 0.6161805369726162,
+      "grad_norm": 0.5746923685073853,
+      "learning_rate": 0.0003917410654650179,
+      "loss": 0.8682,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6163589331906164,
+      "grad_norm": 0.7708411812782288,
+      "learning_rate": 0.00039168334662928895,
+      "loss": 1.3231,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6165373294086165,
+      "grad_norm": 0.794039249420166,
+      "learning_rate": 0.00039162561666581096,
+      "loss": 0.8659,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6167157256266167,
+      "grad_norm": 0.5135617256164551,
+      "learning_rate": 0.000391567875579118,
+      "loss": 1.0136,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6168941218446169,
+      "grad_norm": 0.47095632553100586,
+      "learning_rate": 0.00039151012337374495,
+      "loss": 0.8962,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6170725180626171,
+      "grad_norm": 0.5985879302024841,
+      "learning_rate": 0.0003914523600542277,
+      "loss": 1.0213,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6172509142806173,
+      "grad_norm": 0.6508655548095703,
+      "learning_rate": 0.000391394585625103,
+      "loss": 1.2593,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6174293104986174,
+      "grad_norm": 0.5690905451774597,
+      "learning_rate": 0.00039133680009090845,
+      "loss": 0.9674,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6176077067166176,
+      "grad_norm": 0.610589325428009,
+      "learning_rate": 0.0003912790034561824,
+      "loss": 0.8924,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6177861029346178,
+      "grad_norm": 0.7118551135063171,
+      "learning_rate": 0.00039122119572546424,
+      "loss": 1.1188,
+      "step": 3463
+    },
+    {
+      "epoch": 0.617964499152618,
+      "grad_norm": 0.4738007187843323,
+      "learning_rate": 0.0003911633769032941,
+      "loss": 0.8131,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6181428953706182,
+      "grad_norm": 0.5291326642036438,
+      "learning_rate": 0.000391105546994213,
+      "loss": 0.8803,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6183212915886184,
+      "grad_norm": 17.467649459838867,
+      "learning_rate": 0.0003910477060027631,
+      "loss": 1.0413,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6184996878066185,
+      "grad_norm": 0.5017330646514893,
+      "learning_rate": 0.00039098985393348697,
+      "loss": 0.7768,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6186780840246187,
+      "grad_norm": 0.49945586919784546,
+      "learning_rate": 0.00039093199079092843,
+      "loss": 0.8372,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6188564802426189,
+      "grad_norm": 0.5586909651756287,
+      "learning_rate": 0.000390874116579632,
+      "loss": 0.8797,
+      "step": 3469
+    },
+    {
+      "epoch": 0.619034876460619,
+      "grad_norm": 0.5252298712730408,
+      "learning_rate": 0.0003908162313041431,
+      "loss": 0.917,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6192132726786193,
+      "grad_norm": 0.48841506242752075,
+      "learning_rate": 0.00039075833496900794,
+      "loss": 0.8785,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6193916688966193,
+      "grad_norm": 0.4840584099292755,
+      "learning_rate": 0.0003907004275787737,
+      "loss": 0.9556,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6195700651146195,
+      "grad_norm": 0.47886261343955994,
+      "learning_rate": 0.0003906425091379885,
+      "loss": 1.1327,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6197484613326197,
+      "grad_norm": 0.51800936460495,
+      "learning_rate": 0.0003905845796512011,
+      "loss": 0.8071,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6199268575506199,
+      "grad_norm": 0.5068585872650146,
+      "learning_rate": 0.00039052663912296135,
+      "loss": 0.959,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6201052537686201,
+      "grad_norm": 0.4845023453235626,
+      "learning_rate": 0.00039046868755781986,
+      "loss": 0.8652,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6202836499866203,
+      "grad_norm": 0.4877776503562927,
+      "learning_rate": 0.00039041072496032804,
+      "loss": 1.0887,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6204620462046204,
+      "grad_norm": 0.4742729663848877,
+      "learning_rate": 0.0003903527513350383,
+      "loss": 1.0202,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6206404424226206,
+      "grad_norm": 0.4939587712287903,
+      "learning_rate": 0.0003902947666865039,
+      "loss": 0.8183,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6208188386406208,
+      "grad_norm": 0.4751134216785431,
+      "learning_rate": 0.0003902367710192789,
+      "loss": 1.1003,
+      "step": 3480
+    },
+    {
+      "epoch": 0.620997234858621,
+      "grad_norm": 0.45981040596961975,
+      "learning_rate": 0.00039017876433791824,
+      "loss": 0.8677,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6211756310766212,
+      "grad_norm": 0.5187476873397827,
+      "learning_rate": 0.00039012074664697774,
+      "loss": 0.8187,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6213540272946213,
+      "grad_norm": 0.4452672600746155,
+      "learning_rate": 0.0003900627179510141,
+      "loss": 0.8628,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6215324235126215,
+      "grad_norm": 0.5037822127342224,
+      "learning_rate": 0.0003900046782545849,
+      "loss": 0.9356,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6217108197306217,
+      "grad_norm": 0.4531655013561249,
+      "learning_rate": 0.00038994662756224843,
+      "loss": 0.8167,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6218892159486219,
+      "grad_norm": 0.5035321116447449,
+      "learning_rate": 0.00038988856587856413,
+      "loss": 0.8075,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6220676121666221,
+      "grad_norm": 0.4611433148384094,
+      "learning_rate": 0.00038983049320809207,
+      "loss": 1.1105,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6222460083846223,
+      "grad_norm": 0.48017269372940063,
+      "learning_rate": 0.00038977240955539316,
+      "loss": 0.8857,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6224244046026224,
+      "grad_norm": 0.5595654845237732,
+      "learning_rate": 0.0003897143149250295,
+      "loss": 0.8854,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6226028008206226,
+      "grad_norm": 0.5068066716194153,
+      "learning_rate": 0.00038965620932156355,
+      "loss": 0.9231,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6227811970386228,
+      "grad_norm": 0.5294225215911865,
+      "learning_rate": 0.00038959809274955907,
+      "loss": 0.9997,
+      "step": 3491
+    },
+    {
+      "epoch": 0.622959593256623,
+      "grad_norm": 0.5624359846115112,
+      "learning_rate": 0.0003895399652135805,
+      "loss": 0.9754,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6231379894746232,
+      "grad_norm": 0.4356614053249359,
+      "learning_rate": 0.00038948182671819304,
+      "loss": 0.7682,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6233163856926233,
+      "grad_norm": 0.5377020239830017,
+      "learning_rate": 0.00038942367726796297,
+      "loss": 1.2892,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6234947819106235,
+      "grad_norm": 0.5128370523452759,
+      "learning_rate": 0.0003893655168674572,
+      "loss": 0.8542,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6236731781286237,
+      "grad_norm": 0.5239914059638977,
+      "learning_rate": 0.0003893073455212438,
+      "loss": 0.922,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6238515743466239,
+      "grad_norm": 0.5009093284606934,
+      "learning_rate": 0.00038924916323389145,
+      "loss": 1.2072,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6240299705646241,
+      "grad_norm": 0.5097877383232117,
+      "learning_rate": 0.00038919097000996973,
+      "loss": 1.0078,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6242083667826243,
+      "grad_norm": 0.5235947370529175,
+      "learning_rate": 0.0003891327658540491,
+      "loss": 1.0455,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6243867630006243,
+      "grad_norm": 0.4533289670944214,
+      "learning_rate": 0.00038907455077070085,
+      "loss": 0.7731,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6245651592186245,
+      "grad_norm": 0.5486941337585449,
+      "learning_rate": 0.0003890163247644973,
+      "loss": 0.9088,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6247435554366247,
+      "grad_norm": 0.5437402129173279,
+      "learning_rate": 0.0003889580878400115,
+      "loss": 0.9288,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6249219516546249,
+      "grad_norm": 0.49082422256469727,
+      "learning_rate": 0.00038889984000181724,
+      "loss": 0.8177,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6251003478726251,
+      "grad_norm": 0.7093980312347412,
+      "learning_rate": 0.0003888415812544892,
+      "loss": 1.1146,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6252787440906252,
+      "grad_norm": 0.45868930220603943,
+      "learning_rate": 0.00038878331160260317,
+      "loss": 0.9161,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6254571403086254,
+      "grad_norm": 0.527919590473175,
+      "learning_rate": 0.00038872503105073563,
+      "loss": 1.0866,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6256355365266256,
+      "grad_norm": 0.5056113600730896,
+      "learning_rate": 0.0003886667396034638,
+      "loss": 0.8475,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6258139327446258,
+      "grad_norm": 0.46975040435791016,
+      "learning_rate": 0.00038860843726536593,
+      "loss": 0.8076,
+      "step": 3508
+    },
+    {
+      "epoch": 0.625992328962626,
+      "grad_norm": 0.5507307052612305,
+      "learning_rate": 0.00038855012404102104,
+      "loss": 0.9643,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6261707251806262,
+      "grad_norm": 0.6463258862495422,
+      "learning_rate": 0.00038849179993500905,
+      "loss": 0.9585,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6263491213986263,
+      "grad_norm": 0.5084063410758972,
+      "learning_rate": 0.0003884334649519106,
+      "loss": 1.1892,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6265275176166265,
+      "grad_norm": 0.5068933367729187,
+      "learning_rate": 0.0003883751190963075,
+      "loss": 0.8802,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6267059138346267,
+      "grad_norm": 0.5390312075614929,
+      "learning_rate": 0.0003883167623727821,
+      "loss": 1.0255,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6268843100526269,
+      "grad_norm": 0.5518016815185547,
+      "learning_rate": 0.0003882583947859176,
+      "loss": 1.0363,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6270627062706271,
+      "grad_norm": 0.435151606798172,
+      "learning_rate": 0.0003882000163402983,
+      "loss": 0.8027,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6272411024886272,
+      "grad_norm": 0.4693804979324341,
+      "learning_rate": 0.00038814162704050925,
+      "loss": 0.7981,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6274194987066274,
+      "grad_norm": 0.44543004035949707,
+      "learning_rate": 0.0003880832268911363,
+      "loss": 0.7381,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6275978949246276,
+      "grad_norm": 0.47917303442955017,
+      "learning_rate": 0.00038802481589676605,
+      "loss": 0.8535,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6277762911426278,
+      "grad_norm": 0.5061355233192444,
+      "learning_rate": 0.0003879663940619861,
+      "loss": 0.9866,
+      "step": 3519
+    },
+    {
+      "epoch": 0.627954687360628,
+      "grad_norm": 0.5138995051383972,
+      "learning_rate": 0.00038790796139138506,
+      "loss": 0.9151,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6281330835786282,
+      "grad_norm": 0.5253726840019226,
+      "learning_rate": 0.000387849517889552,
+      "loss": 1.0164,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6283114797966283,
+      "grad_norm": 0.4949781000614166,
+      "learning_rate": 0.00038779106356107715,
+      "loss": 1.0074,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6284898760146285,
+      "grad_norm": 0.49461349844932556,
+      "learning_rate": 0.0003877325984105514,
+      "loss": 0.9045,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6286682722326287,
+      "grad_norm": 0.4767039716243744,
+      "learning_rate": 0.00038767412244256673,
+      "loss": 0.7782,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6288466684506289,
+      "grad_norm": 0.5934154391288757,
+      "learning_rate": 0.00038761563566171576,
+      "loss": 1.2435,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6290250646686291,
+      "grad_norm": 0.5259369611740112,
+      "learning_rate": 0.00038755713807259184,
+      "loss": 0.9742,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6292034608866292,
+      "grad_norm": 0.44396549463272095,
+      "learning_rate": 0.0003874986296797896,
+      "loss": 0.7857,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6293818571046293,
+      "grad_norm": 0.5298401713371277,
+      "learning_rate": 0.0003874401104879041,
+      "loss": 0.9236,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6295602533226295,
+      "grad_norm": 0.5219202637672424,
+      "learning_rate": 0.00038738158050153157,
+      "loss": 1.0415,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6297386495406297,
+      "grad_norm": 0.5267150402069092,
+      "learning_rate": 0.0003873230397252687,
+      "loss": 0.9082,
+      "step": 3530
+    },
+    {
+      "epoch": 0.62991704575863,
+      "grad_norm": 0.5192800164222717,
+      "learning_rate": 0.0003872644881637135,
+      "loss": 0.8653,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6300954419766301,
+      "grad_norm": 0.5463857054710388,
+      "learning_rate": 0.0003872059258214644,
+      "loss": 0.9292,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6302738381946302,
+      "grad_norm": 0.659090518951416,
+      "learning_rate": 0.00038714735270312095,
+      "loss": 0.8808,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6304522344126304,
+      "grad_norm": 0.5042756795883179,
+      "learning_rate": 0.0003870887688132834,
+      "loss": 0.8481,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6306306306306306,
+      "grad_norm": 0.4895661771297455,
+      "learning_rate": 0.00038703017415655296,
+      "loss": 0.9983,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6308090268486308,
+      "grad_norm": 0.46938470005989075,
+      "learning_rate": 0.00038697156873753163,
+      "loss": 0.9572,
+      "step": 3536
+    },
+    {
+      "epoch": 0.630987423066631,
+      "grad_norm": 0.4477916359901428,
+      "learning_rate": 0.00038691295256082227,
+      "loss": 0.798,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6311658192846311,
+      "grad_norm": 0.5256043076515198,
+      "learning_rate": 0.0003868543256310284,
+      "loss": 0.9967,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6313442155026313,
+      "grad_norm": 0.5086610913276672,
+      "learning_rate": 0.0003867956879527548,
+      "loss": 1.1852,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6315226117206315,
+      "grad_norm": 0.47662028670310974,
+      "learning_rate": 0.00038673703953060677,
+      "loss": 0.9776,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6317010079386317,
+      "grad_norm": 0.461160272359848,
+      "learning_rate": 0.00038667838036919046,
+      "loss": 0.8206,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6318794041566319,
+      "grad_norm": 0.5197997093200684,
+      "learning_rate": 0.0003866197104731129,
+      "loss": 0.9646,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6320578003746321,
+      "grad_norm": 0.48244959115982056,
+      "learning_rate": 0.0003865610298469821,
+      "loss": 0.7589,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6322361965926322,
+      "grad_norm": 0.4731544852256775,
+      "learning_rate": 0.00038650233849540683,
+      "loss": 0.9059,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6324145928106324,
+      "grad_norm": 0.520389199256897,
+      "learning_rate": 0.00038644363642299665,
+      "loss": 1.0436,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6325929890286326,
+      "grad_norm": 0.49500808119773865,
+      "learning_rate": 0.00038638492363436195,
+      "loss": 0.9817,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6327713852466328,
+      "grad_norm": 0.4835432469844818,
+      "learning_rate": 0.000386326200134114,
+      "loss": 0.9696,
+      "step": 3547
+    },
+    {
+      "epoch": 0.632949781464633,
+      "grad_norm": 0.45955875515937805,
+      "learning_rate": 0.000386267465926865,
+      "loss": 0.9491,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6331281776826331,
+      "grad_norm": 0.5106682777404785,
+      "learning_rate": 0.00038620872101722783,
+      "loss": 0.9027,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6333065739006333,
+      "grad_norm": 0.5530784130096436,
+      "learning_rate": 0.0003861499654098164,
+      "loss": 1.1835,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6334849701186335,
+      "grad_norm": 0.45463332533836365,
+      "learning_rate": 0.0003860911991092452,
+      "loss": 0.8381,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6336633663366337,
+      "grad_norm": 0.5137251615524292,
+      "learning_rate": 0.0003860324221201298,
+      "loss": 0.9746,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6338417625546339,
+      "grad_norm": 0.4892425835132599,
+      "learning_rate": 0.00038597363444708657,
+      "loss": 1.0123,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6340201587726341,
+      "grad_norm": 0.482229083776474,
+      "learning_rate": 0.00038591483609473257,
+      "loss": 0.796,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6341985549906342,
+      "grad_norm": 0.44259580969810486,
+      "learning_rate": 0.0003858560270676858,
+      "loss": 1.0132,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6343769512086344,
+      "grad_norm": 0.5081446170806885,
+      "learning_rate": 0.00038579720737056517,
+      "loss": 0.9065,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6345553474266346,
+      "grad_norm": 0.4761013090610504,
+      "learning_rate": 0.0003857383770079902,
+      "loss": 0.9184,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6347337436446348,
+      "grad_norm": 0.5290868282318115,
+      "learning_rate": 0.00038567953598458163,
+      "loss": 0.9764,
+      "step": 3558
+    },
+    {
+      "epoch": 0.634912139862635,
+      "grad_norm": 0.5481404066085815,
+      "learning_rate": 0.00038562068430496066,
+      "loss": 0.9443,
+      "step": 3559
+    },
+    {
+      "epoch": 0.635090536080635,
+      "grad_norm": 0.48112472891807556,
+      "learning_rate": 0.00038556182197374957,
+      "loss": 1.0576,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6352689322986352,
+      "grad_norm": 0.46322664618492126,
+      "learning_rate": 0.0003855029489955713,
+      "loss": 0.8526,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6354473285166354,
+      "grad_norm": 0.501327633857727,
+      "learning_rate": 0.0003854440653750496,
+      "loss": 1.0425,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6356257247346356,
+      "grad_norm": 0.9595483541488647,
+      "learning_rate": 0.0003853851711168094,
+      "loss": 0.73,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6358041209526358,
+      "grad_norm": 0.44412168860435486,
+      "learning_rate": 0.00038532626622547614,
+      "loss": 0.8409,
+      "step": 3564
+    },
+    {
+      "epoch": 0.635982517170636,
+      "grad_norm": 0.5374020338058472,
+      "learning_rate": 0.0003852673507056761,
+      "loss": 1.0608,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6361609133886361,
+      "grad_norm": 0.5410022735595703,
+      "learning_rate": 0.0003852084245620365,
+      "loss": 1.0212,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6363393096066363,
+      "grad_norm": 0.46647214889526367,
+      "learning_rate": 0.0003851494877991856,
+      "loss": 0.9998,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6365177058246365,
+      "grad_norm": 0.49088791012763977,
+      "learning_rate": 0.0003850905404217519,
+      "loss": 1.0369,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6366961020426367,
+      "grad_norm": 0.5109038352966309,
+      "learning_rate": 0.00038503158243436537,
+      "loss": 1.1344,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6368744982606369,
+      "grad_norm": 0.5097025036811829,
+      "learning_rate": 0.0003849726138416565,
+      "loss": 0.9358,
+      "step": 3570
+    },
+    {
+      "epoch": 0.637052894478637,
+      "grad_norm": 0.4957421123981476,
+      "learning_rate": 0.00038491363464825655,
+      "loss": 1.1196,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6372312906966372,
+      "grad_norm": 0.5967016220092773,
+      "learning_rate": 0.00038485464485879783,
+      "loss": 1.1693,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6374096869146374,
+      "grad_norm": 0.4782984256744385,
+      "learning_rate": 0.0003847956444779133,
+      "loss": 1.0175,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6375880831326376,
+      "grad_norm": 0.4927447438240051,
+      "learning_rate": 0.0003847366335102369,
+      "loss": 1.0854,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6377664793506378,
+      "grad_norm": 0.4766659140586853,
+      "learning_rate": 0.0003846776119604033,
+      "loss": 0.8279,
+      "step": 3575
+    },
+    {
+      "epoch": 0.637944875568638,
+      "grad_norm": 0.46618908643722534,
+      "learning_rate": 0.000384618579833048,
+      "loss": 0.8925,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6381232717866381,
+      "grad_norm": 0.48249977827072144,
+      "learning_rate": 0.0003845595371328074,
+      "loss": 0.7996,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6383016680046383,
+      "grad_norm": 0.463788777589798,
+      "learning_rate": 0.0003845004838643186,
+      "loss": 0.8095,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6384800642226385,
+      "grad_norm": 0.5117577910423279,
+      "learning_rate": 0.0003844414200322197,
+      "loss": 0.939,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6386584604406387,
+      "grad_norm": 1.0510876178741455,
+      "learning_rate": 0.0003843823456411495,
+      "loss": 0.9856,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6388368566586389,
+      "grad_norm": 0.4879765808582306,
+      "learning_rate": 0.00038432326069574776,
+      "loss": 0.8623,
+      "step": 3581
+    },
+    {
+      "epoch": 0.639015252876639,
+      "grad_norm": 0.5112807154655457,
+      "learning_rate": 0.0003842641652006549,
+      "loss": 0.7826,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6391936490946392,
+      "grad_norm": 0.49718165397644043,
+      "learning_rate": 0.0003842050591605122,
+      "loss": 0.9188,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6393720453126394,
+      "grad_norm": 0.4578002989292145,
+      "learning_rate": 0.00038414594257996207,
+      "loss": 0.8357,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6395504415306396,
+      "grad_norm": 0.5022866725921631,
+      "learning_rate": 0.0003840868154636472,
+      "loss": 0.9341,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6397288377486398,
+      "grad_norm": 0.5059497356414795,
+      "learning_rate": 0.00038402767781621163,
+      "loss": 0.9208,
+      "step": 3586
+    },
+    {
+      "epoch": 0.63990723396664,
+      "grad_norm": 0.46369996666908264,
+      "learning_rate": 0.0003839685296422999,
+      "loss": 1.0211,
+      "step": 3587
+    },
+    {
+      "epoch": 0.64008563018464,
+      "grad_norm": 0.46570396423339844,
+      "learning_rate": 0.0003839093709465574,
+      "loss": 0.7812,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6402640264026402,
+      "grad_norm": 0.5327849388122559,
+      "learning_rate": 0.00038385020173363065,
+      "loss": 0.868,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6404424226206404,
+      "grad_norm": 0.5208088755607605,
+      "learning_rate": 0.0003837910220081667,
+      "loss": 1.0375,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6406208188386406,
+      "grad_norm": 0.6217350363731384,
+      "learning_rate": 0.00038373183177481336,
+      "loss": 0.9972,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6407992150566408,
+      "grad_norm": 0.5081048607826233,
+      "learning_rate": 0.00038367263103821956,
+      "loss": 0.9218,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6409776112746409,
+      "grad_norm": 0.540647566318512,
+      "learning_rate": 0.00038361341980303477,
+      "loss": 0.838,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6411560074926411,
+      "grad_norm": 0.4405638873577118,
+      "learning_rate": 0.0003835541980739096,
+      "loss": 0.9328,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6413344037106413,
+      "grad_norm": 0.501441240310669,
+      "learning_rate": 0.00038349496585549504,
+      "loss": 0.9606,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6415127999286415,
+      "grad_norm": 0.6081332564353943,
+      "learning_rate": 0.00038343572315244337,
+      "loss": 1.032,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6416911961466417,
+      "grad_norm": 0.5091875791549683,
+      "learning_rate": 0.00038337646996940746,
+      "loss": 0.7836,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6418695923646419,
+      "grad_norm": 0.48131605982780457,
+      "learning_rate": 0.00038331720631104094,
+      "loss": 0.8946,
+      "step": 3598
+    },
+    {
+      "epoch": 0.642047988582642,
+      "grad_norm": 1.0239105224609375,
+      "learning_rate": 0.00038325793218199844,
+      "loss": 0.7923,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6422263848006422,
+      "grad_norm": 0.5322194695472717,
+      "learning_rate": 0.00038319864758693537,
+      "loss": 1.0488,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6424047810186424,
+      "grad_norm": 0.47548213601112366,
+      "learning_rate": 0.00038313935253050767,
+      "loss": 1.0053,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6425831772366426,
+      "grad_norm": 0.5237056016921997,
+      "learning_rate": 0.00038308004701737263,
+      "loss": 1.1428,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6427615734546428,
+      "grad_norm": 0.6333736181259155,
+      "learning_rate": 0.00038302073105218794,
+      "loss": 1.0623,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6429399696726429,
+      "grad_norm": 0.49249467253685,
+      "learning_rate": 0.00038296140463961226,
+      "loss": 1.0911,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6431183658906431,
+      "grad_norm": 0.479319304227829,
+      "learning_rate": 0.00038290206778430515,
+      "loss": 0.8087,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6432967621086433,
+      "grad_norm": 0.5194647312164307,
+      "learning_rate": 0.00038284272049092673,
+      "loss": 0.8521,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6434751583266435,
+      "grad_norm": 0.47771453857421875,
+      "learning_rate": 0.00038278336276413827,
+      "loss": 1.0094,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6436535545446437,
+      "grad_norm": 1.2369436025619507,
+      "learning_rate": 0.00038272399460860166,
+      "loss": 1.007,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6438319507626439,
+      "grad_norm": 0.4876594841480255,
+      "learning_rate": 0.00038266461602897957,
+      "loss": 1.0556,
+      "step": 3609
+    },
+    {
+      "epoch": 0.644010346980644,
+      "grad_norm": 0.5744562149047852,
+      "learning_rate": 0.0003826052270299356,
+      "loss": 1.0985,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6441887431986442,
+      "grad_norm": 0.4735945165157318,
+      "learning_rate": 0.00038254582761613424,
+      "loss": 1.0116,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6443671394166444,
+      "grad_norm": 0.5286600589752197,
+      "learning_rate": 0.0003824864177922406,
+      "loss": 1.1372,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6445455356346446,
+      "grad_norm": 0.5075108408927917,
+      "learning_rate": 0.0003824269975629207,
+      "loss": 1.0383,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6447239318526448,
+      "grad_norm": 0.5046312212944031,
+      "learning_rate": 0.00038236756693284143,
+      "loss": 1.0341,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6449023280706448,
+      "grad_norm": 0.49919357895851135,
+      "learning_rate": 0.00038230812590667044,
+      "loss": 1.0578,
+      "step": 3615
+    },
+    {
+      "epoch": 0.645080724288645,
+      "grad_norm": 0.4575040340423584,
+      "learning_rate": 0.0003822486744890761,
+      "loss": 0.7656,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6452591205066452,
+      "grad_norm": 0.4908180236816406,
+      "learning_rate": 0.00038218921268472786,
+      "loss": 0.7803,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6454375167246454,
+      "grad_norm": 0.4667602479457855,
+      "learning_rate": 0.00038212974049829564,
+      "loss": 0.9217,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6456159129426456,
+      "grad_norm": 0.49208131432533264,
+      "learning_rate": 0.00038207025793445047,
+      "loss": 1.3218,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6457943091606458,
+      "grad_norm": 0.47899407148361206,
+      "learning_rate": 0.0003820107649978641,
+      "loss": 0.9374,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6459727053786459,
+      "grad_norm": 0.5173959136009216,
+      "learning_rate": 0.00038195126169320915,
+      "loss": 1.1439,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6461511015966461,
+      "grad_norm": 1.7555327415466309,
+      "learning_rate": 0.00038189174802515883,
+      "loss": 1.1495,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6463294978146463,
+      "grad_norm": 0.5820402503013611,
+      "learning_rate": 0.0003818322239983873,
+      "loss": 0.9644,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6465078940326465,
+      "grad_norm": 0.5061475038528442,
+      "learning_rate": 0.0003817726896175697,
+      "loss": 1.0867,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6466862902506467,
+      "grad_norm": 0.5175919532775879,
+      "learning_rate": 0.00038171314488738176,
+      "loss": 0.9003,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6468646864686468,
+      "grad_norm": 0.5276834964752197,
+      "learning_rate": 0.0003816535898125001,
+      "loss": 1.0232,
+      "step": 3626
+    },
+    {
+      "epoch": 0.647043082686647,
+      "grad_norm": 0.4682513177394867,
+      "learning_rate": 0.0003815940243976022,
+      "loss": 0.8202,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6472214789046472,
+      "grad_norm": 0.5589001774787903,
+      "learning_rate": 0.00038153444864736616,
+      "loss": 0.9364,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6473998751226474,
+      "grad_norm": 0.5058655738830566,
+      "learning_rate": 0.00038147486256647113,
+      "loss": 1.0972,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6475782713406476,
+      "grad_norm": 0.44797369837760925,
+      "learning_rate": 0.0003814152661595971,
+      "loss": 0.8263,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6477566675586478,
+      "grad_norm": 0.5152768492698669,
+      "learning_rate": 0.00038135565943142445,
+      "loss": 0.9512,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6479350637766479,
+      "grad_norm": 0.4588833451271057,
+      "learning_rate": 0.00038129604238663494,
+      "loss": 0.8475,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6481134599946481,
+      "grad_norm": 0.5147669911384583,
+      "learning_rate": 0.00038123641502991074,
+      "loss": 0.9934,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6482918562126483,
+      "grad_norm": 0.46724918484687805,
+      "learning_rate": 0.0003811767773659349,
+      "loss": 0.7732,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6484702524306485,
+      "grad_norm": 0.6038320064544678,
+      "learning_rate": 0.00038111712939939153,
+      "loss": 1.3013,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.4923225939273834,
+      "learning_rate": 0.0003810574711349652,
+      "loss": 1.205,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6488270448666488,
+      "grad_norm": 0.4713652431964874,
+      "learning_rate": 0.0003809978025773415,
+      "loss": 0.8677,
+      "step": 3637
+    },
+    {
+      "epoch": 0.649005441084649,
+      "grad_norm": 0.4485718607902527,
+      "learning_rate": 0.00038093812373120675,
+      "loss": 0.7994,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6491838373026492,
+      "grad_norm": 0.5080364942550659,
+      "learning_rate": 0.00038087843460124813,
+      "loss": 0.9032,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6493622335206494,
+      "grad_norm": 0.5731015801429749,
+      "learning_rate": 0.0003808187351921535,
+      "loss": 0.8565,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6495406297386496,
+      "grad_norm": 0.47458168864250183,
+      "learning_rate": 0.00038075902550861176,
+      "loss": 1.0262,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6497190259566498,
+      "grad_norm": 1.2307530641555786,
+      "learning_rate": 0.0003806993055553124,
+      "loss": 0.8389,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6498974221746499,
+      "grad_norm": 0.481864333152771,
+      "learning_rate": 0.00038063957533694594,
+      "loss": 0.9069,
+      "step": 3643
+    },
+    {
+      "epoch": 0.65007581839265,
+      "grad_norm": 0.46057215332984924,
+      "learning_rate": 0.0003805798348582034,
+      "loss": 0.7955,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6502542146106502,
+      "grad_norm": 0.5290510058403015,
+      "learning_rate": 0.0003805200841237767,
+      "loss": 1.1641,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6504326108286504,
+      "grad_norm": 0.4636685252189636,
+      "learning_rate": 0.0003804603231383589,
+      "loss": 0.8106,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6506110070466506,
+      "grad_norm": 0.5197182893753052,
+      "learning_rate": 0.00038040055190664336,
+      "loss": 1.0978,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6507894032646507,
+      "grad_norm": 0.5658067464828491,
+      "learning_rate": 0.0003803407704333246,
+      "loss": 1.0294,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6509677994826509,
+      "grad_norm": 0.4995347261428833,
+      "learning_rate": 0.0003802809787230979,
+      "loss": 1.0974,
+      "step": 3649
+    },
+    {
+      "epoch": 0.6511461957006511,
+      "grad_norm": 0.5178309679031372,
+      "learning_rate": 0.00038022117678065915,
+      "loss": 1.0917,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6513245919186513,
+      "grad_norm": 0.450339674949646,
+      "learning_rate": 0.0003801613646107052,
+      "loss": 0.7781,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6515029881366515,
+      "grad_norm": 0.48041054606437683,
+      "learning_rate": 0.0003801015422179337,
+      "loss": 0.7998,
+      "step": 3652
+    },
+    {
+      "epoch": 0.6516813843546517,
+      "grad_norm": 0.5061138868331909,
+      "learning_rate": 0.00038004170960704306,
+      "loss": 0.9966,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6518597805726518,
+      "grad_norm": 0.4858459234237671,
+      "learning_rate": 0.0003799818667827325,
+      "loss": 0.8406,
+      "step": 3654
+    },
+    {
+      "epoch": 0.652038176790652,
+      "grad_norm": 0.49688011407852173,
+      "learning_rate": 0.00037992201374970205,
+      "loss": 1.0493,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6522165730086522,
+      "grad_norm": 0.47799569368362427,
+      "learning_rate": 0.0003798621505126526,
+      "loss": 0.8846,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6523949692266524,
+      "grad_norm": 0.500594437122345,
+      "learning_rate": 0.0003798022770762857,
+      "loss": 1.0214,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6525733654446526,
+      "grad_norm": 0.5715755224227905,
+      "learning_rate": 0.0003797423934453038,
+      "loss": 1.312,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6527517616626527,
+      "grad_norm": 0.47365105152130127,
+      "learning_rate": 0.00037968249962441015,
+      "loss": 0.9041,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6529301578806529,
+      "grad_norm": 0.5020959973335266,
+      "learning_rate": 0.00037962259561830883,
+      "loss": 0.9089,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6531085540986531,
+      "grad_norm": 0.5108777284622192,
+      "learning_rate": 0.0003795626814317046,
+      "loss": 0.7617,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6532869503166533,
+      "grad_norm": 0.6099095940589905,
+      "learning_rate": 0.0003795027570693032,
+      "loss": 1.2797,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6534653465346535,
+      "grad_norm": 0.5367801189422607,
+      "learning_rate": 0.00037944282253581086,
+      "loss": 0.9654,
+      "step": 3663
+    },
+    {
+      "epoch": 0.6536437427526537,
+      "grad_norm": 0.4482106566429138,
+      "learning_rate": 0.000379382877835935,
+      "loss": 0.8033,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6538221389706538,
+      "grad_norm": 0.48024243116378784,
+      "learning_rate": 0.0003793229229743836,
+      "loss": 0.7399,
+      "step": 3665
+    },
+    {
+      "epoch": 0.654000535188654,
+      "grad_norm": 0.49091434478759766,
+      "learning_rate": 0.00037926295795586546,
+      "loss": 0.9129,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6541789314066542,
+      "grad_norm": 0.428303062915802,
+      "learning_rate": 0.00037920298278509027,
+      "loss": 0.8762,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6543573276246544,
+      "grad_norm": 0.4690133035182953,
+      "learning_rate": 0.00037914299746676837,
+      "loss": 0.748,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6545357238426546,
+      "grad_norm": 0.5279445052146912,
+      "learning_rate": 0.00037908300200561107,
+      "loss": 0.9056,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6547141200606548,
+      "grad_norm": 0.5457646250724792,
+      "learning_rate": 0.0003790229964063303,
+      "loss": 0.8575,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6548925162786549,
+      "grad_norm": 0.4674092233181,
+      "learning_rate": 0.00037896298067363897,
+      "loss": 0.7337,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6550709124966551,
+      "grad_norm": 0.539322555065155,
+      "learning_rate": 0.00037890295481225056,
+      "loss": 1.1518,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6552493087146553,
+      "grad_norm": 0.5358078479766846,
+      "learning_rate": 0.00037884291882687955,
+      "loss": 1.2239,
+      "step": 3673
+    },
+    {
+      "epoch": 0.6554277049326555,
+      "grad_norm": 0.4866465926170349,
+      "learning_rate": 0.0003787828727222412,
+      "loss": 1.1942,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6556061011506557,
+      "grad_norm": 0.5002499222755432,
+      "learning_rate": 0.0003787228165030514,
+      "loss": 0.8127,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6557844973686557,
+      "grad_norm": 0.5849499106407166,
+      "learning_rate": 0.00037866275017402694,
+      "loss": 0.9762,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6559628935866559,
+      "grad_norm": 0.46588608622550964,
+      "learning_rate": 0.0003786026737398857,
+      "loss": 0.8933,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6561412898046561,
+      "grad_norm": 0.48897257447242737,
+      "learning_rate": 0.0003785425872053455,
+      "loss": 0.8101,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6563196860226563,
+      "grad_norm": 0.46128103137016296,
+      "learning_rate": 0.00037848249057512596,
+      "loss": 0.8682,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6564980822406565,
+      "grad_norm": 0.5110675692558289,
+      "learning_rate": 0.00037842238385394684,
+      "loss": 0.8593,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6566764784586567,
+      "grad_norm": 0.4921587407588959,
+      "learning_rate": 0.00037836226704652897,
+      "loss": 0.9287,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6568548746766568,
+      "grad_norm": 0.5147567987442017,
+      "learning_rate": 0.00037830214015759393,
+      "loss": 0.8533,
+      "step": 3682
+    },
+    {
+      "epoch": 0.657033270894657,
+      "grad_norm": 0.4901014268398285,
+      "learning_rate": 0.0003782420031918641,
+      "loss": 0.9394,
+      "step": 3683
+    },
+    {
+      "epoch": 0.6572116671126572,
+      "grad_norm": 0.5078468918800354,
+      "learning_rate": 0.00037818185615406236,
+      "loss": 0.9224,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6573900633306574,
+      "grad_norm": 0.5428446531295776,
+      "learning_rate": 0.0003781216990489129,
+      "loss": 1.1286,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6575684595486576,
+      "grad_norm": 0.5101941227912903,
+      "learning_rate": 0.0003780615318811402,
+      "loss": 0.9386,
+      "step": 3686
+    },
+    {
+      "epoch": 0.6577468557666577,
+      "grad_norm": 0.5293329358100891,
+      "learning_rate": 0.00037800135465547,
+      "loss": 0.9849,
+      "step": 3687
+    },
+    {
+      "epoch": 0.6579252519846579,
+      "grad_norm": 0.47697338461875916,
+      "learning_rate": 0.00037794116737662847,
+      "loss": 0.8925,
+      "step": 3688
+    },
+    {
+      "epoch": 0.6581036482026581,
+      "grad_norm": 0.5124788284301758,
+      "learning_rate": 0.00037788097004934275,
+      "loss": 1.0044,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6582820444206583,
+      "grad_norm": 0.4236263632774353,
+      "learning_rate": 0.00037782076267834063,
+      "loss": 0.8828,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6584604406386585,
+      "grad_norm": 0.4617786109447479,
+      "learning_rate": 0.00037776054526835086,
+      "loss": 0.8969,
+      "step": 3691
+    },
+    {
+      "epoch": 0.6586388368566587,
+      "grad_norm": 0.596630871295929,
+      "learning_rate": 0.0003777003178241028,
+      "loss": 1.0182,
+      "step": 3692
+    },
+    {
+      "epoch": 0.6588172330746588,
+      "grad_norm": 0.4604688882827759,
+      "learning_rate": 0.00037764008035032676,
+      "loss": 0.9566,
+      "step": 3693
+    },
+    {
+      "epoch": 0.658995629292659,
+      "grad_norm": 0.8154518604278564,
+      "learning_rate": 0.00037757983285175367,
+      "loss": 0.8091,
+      "step": 3694
+    },
+    {
+      "epoch": 0.6591740255106592,
+      "grad_norm": 0.5884724259376526,
+      "learning_rate": 0.00037751957533311545,
+      "loss": 0.9988,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6593524217286594,
+      "grad_norm": 0.43609222769737244,
+      "learning_rate": 0.0003774593077991447,
+      "loss": 0.8279,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6595308179466596,
+      "grad_norm": 0.5332170128822327,
+      "learning_rate": 0.0003773990302545748,
+      "loss": 0.8307,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6597092141646597,
+      "grad_norm": 0.4429665505886078,
+      "learning_rate": 0.0003773387427041398,
+      "loss": 0.9021,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6598876103826599,
+      "grad_norm": 0.48425301909446716,
+      "learning_rate": 0.00037727844515257473,
+      "loss": 0.8488,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6600660066006601,
+      "grad_norm": 0.45970770716667175,
+      "learning_rate": 0.00037721813760461544,
+      "loss": 0.9876,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6602444028186603,
+      "grad_norm": 0.47570955753326416,
+      "learning_rate": 0.00037715782006499826,
+      "loss": 0.9802,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6604227990366605,
+      "grad_norm": 0.5924715399742126,
+      "learning_rate": 0.0003770974925384607,
+      "loss": 0.9458,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6606011952546607,
+      "grad_norm": 0.47865262627601624,
+      "learning_rate": 0.0003770371550297407,
+      "loss": 0.7191,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6607795914726607,
+      "grad_norm": 0.6037271618843079,
+      "learning_rate": 0.00037697680754357726,
+      "loss": 1.1374,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6609579876906609,
+      "grad_norm": 0.499804824590683,
+      "learning_rate": 0.00037691645008471,
+      "loss": 0.9418,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6611363839086611,
+      "grad_norm": 0.4647560715675354,
+      "learning_rate": 0.00037685608265787936,
+      "loss": 0.7682,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6613147801266613,
+      "grad_norm": 0.48037344217300415,
+      "learning_rate": 0.0003767957052678266,
+      "loss": 0.6937,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6614931763446615,
+      "grad_norm": 0.5261335968971252,
+      "learning_rate": 0.00037673531791929365,
+      "loss": 1.2998,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6616715725626616,
+      "grad_norm": 0.48943454027175903,
+      "learning_rate": 0.0003766749206170234,
+      "loss": 0.7968,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6618499687806618,
+      "grad_norm": 0.4693853557109833,
+      "learning_rate": 0.0003766145133657594,
+      "loss": 0.9241,
+      "step": 3710
+    },
+    {
+      "epoch": 0.662028364998662,
+      "grad_norm": 0.49054959416389465,
+      "learning_rate": 0.00037655409617024606,
+      "loss": 0.913,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6622067612166622,
+      "grad_norm": 0.49110257625579834,
+      "learning_rate": 0.0003764936690352284,
+      "loss": 1.0838,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6623851574346624,
+      "grad_norm": 0.4667607843875885,
+      "learning_rate": 0.00037643323196545245,
+      "loss": 0.8978,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6625635536526626,
+      "grad_norm": 0.5172199010848999,
+      "learning_rate": 0.0003763727849656648,
+      "loss": 1.2393,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6627419498706627,
+      "grad_norm": 0.4816920757293701,
+      "learning_rate": 0.0003763123280406131,
+      "loss": 1.0024,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6629203460886629,
+      "grad_norm": 0.5231166481971741,
+      "learning_rate": 0.00037625186119504537,
+      "loss": 1.0882,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6630987423066631,
+      "grad_norm": 0.5482428669929504,
+      "learning_rate": 0.000376191384433711,
+      "loss": 1.3842,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6632771385246633,
+      "grad_norm": 0.5781173706054688,
+      "learning_rate": 0.00037613089776135947,
+      "loss": 1.4105,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6634555347426635,
+      "grad_norm": 0.5220258235931396,
+      "learning_rate": 0.0003760704011827415,
+      "loss": 0.9949,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6636339309606636,
+      "grad_norm": 0.46553224325180054,
+      "learning_rate": 0.0003760098947026085,
+      "loss": 0.7985,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6638123271786638,
+      "grad_norm": 0.47272759675979614,
+      "learning_rate": 0.00037594937832571254,
+      "loss": 0.819,
+      "step": 3721
+    },
+    {
+      "epoch": 0.663990723396664,
+      "grad_norm": 0.5201370716094971,
+      "learning_rate": 0.0003758888520568067,
+      "loss": 0.899,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6641691196146642,
+      "grad_norm": 0.52834552526474,
+      "learning_rate": 0.0003758283159006446,
+      "loss": 0.9406,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6643475158326644,
+      "grad_norm": 0.4457343518733978,
+      "learning_rate": 0.00037576776986198064,
+      "loss": 0.841,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6645259120506646,
+      "grad_norm": 0.49147358536720276,
+      "learning_rate": 0.00037570721394557016,
+      "loss": 0.9662,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6647043082686647,
+      "grad_norm": 0.5019490122795105,
+      "learning_rate": 0.00037564664815616924,
+      "loss": 1.1227,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6648827044866649,
+      "grad_norm": 0.4801279902458191,
+      "learning_rate": 0.0003755860724985346,
+      "loss": 0.7102,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6650611007046651,
+      "grad_norm": 0.5752885937690735,
+      "learning_rate": 0.00037552548697742386,
+      "loss": 1.1417,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6652394969226653,
+      "grad_norm": 0.45870140194892883,
+      "learning_rate": 0.00037546489159759545,
+      "loss": 0.763,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6654178931406655,
+      "grad_norm": 0.4713680148124695,
+      "learning_rate": 0.0003754042863638084,
+      "loss": 0.8104,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6655962893586655,
+      "grad_norm": 0.5558474659919739,
+      "learning_rate": 0.0003753436712808227,
+      "loss": 1.0264,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6657746855766657,
+      "grad_norm": 0.5492153167724609,
+      "learning_rate": 0.000375283046353399,
+      "loss": 1.1912,
+      "step": 3732
+    },
+    {
+      "epoch": 0.665953081794666,
+      "grad_norm": 0.4615228474140167,
+      "learning_rate": 0.00037522241158629866,
+      "loss": 0.9509,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6661314780126661,
+      "grad_norm": 0.4563251733779907,
+      "learning_rate": 0.00037516176698428413,
+      "loss": 0.8065,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6663098742306663,
+      "grad_norm": 0.5740782618522644,
+      "learning_rate": 0.0003751011125521182,
+      "loss": 0.9293,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6664882704486665,
+      "grad_norm": 0.49503761529922485,
+      "learning_rate": 0.0003750404482945648,
+      "loss": 0.9861,
+      "step": 3736
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.4828610122203827,
+      "learning_rate": 0.0003749797742163883,
+      "loss": 1.1359,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6668450628846668,
+      "grad_norm": 0.4599243402481079,
+      "learning_rate": 0.00037491909032235423,
+      "loss": 0.7794,
+      "step": 3738
+    },
+    {
+      "epoch": 0.667023459102667,
+      "grad_norm": 0.5307605862617493,
+      "learning_rate": 0.0003748583966172285,
+      "loss": 0.8625,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6672018553206672,
+      "grad_norm": 0.49819859862327576,
+      "learning_rate": 0.000374797693105778,
+      "loss": 0.8321,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6673802515386674,
+      "grad_norm": 0.5101906657218933,
+      "learning_rate": 0.0003747369797927704,
+      "loss": 1.0208,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6675586477566675,
+      "grad_norm": 0.4965439438819885,
+      "learning_rate": 0.0003746762566829742,
+      "loss": 0.9561,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6677370439746677,
+      "grad_norm": 0.4781574308872223,
+      "learning_rate": 0.00037461552378115833,
+      "loss": 0.8842,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6679154401926679,
+      "grad_norm": 0.512837827205658,
+      "learning_rate": 0.00037455478109209284,
+      "loss": 0.9223,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6680938364106681,
+      "grad_norm": 0.484651654958725,
+      "learning_rate": 0.0003744940286205485,
+      "loss": 0.8012,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6682722326286683,
+      "grad_norm": 0.5075222849845886,
+      "learning_rate": 0.00037443326637129674,
+      "loss": 0.9905,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6684506288466685,
+      "grad_norm": 0.48406416177749634,
+      "learning_rate": 0.0003743724943491097,
+      "loss": 1.0252,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6686290250646686,
+      "grad_norm": 0.4906768202781677,
+      "learning_rate": 0.0003743117125587606,
+      "loss": 0.8734,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6688074212826688,
+      "grad_norm": 0.4577210545539856,
+      "learning_rate": 0.00037425092100502297,
+      "loss": 0.9074,
+      "step": 3749
+    },
+    {
+      "epoch": 0.668985817500669,
+      "grad_norm": 0.4762527346611023,
+      "learning_rate": 0.0003741901196926715,
+      "loss": 0.9042,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6691642137186692,
+      "grad_norm": 0.5320841073989868,
+      "learning_rate": 0.00037412930862648153,
+      "loss": 1.1009,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6693426099366694,
+      "grad_norm": 0.4859885275363922,
+      "learning_rate": 0.00037406848781122904,
+      "loss": 0.9112,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6695210061546695,
+      "grad_norm": 0.5073356628417969,
+      "learning_rate": 0.0003740076572516909,
+      "loss": 1.1243,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6696994023726697,
+      "grad_norm": 0.492034912109375,
+      "learning_rate": 0.00037394681695264475,
+      "loss": 1.0717,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6698777985906699,
+      "grad_norm": 0.40963229537010193,
+      "learning_rate": 0.0003738859669188689,
+      "loss": 0.7091,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6700561948086701,
+      "grad_norm": 0.5281137824058533,
+      "learning_rate": 0.00037382510715514255,
+      "loss": 1.0417,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6702345910266703,
+      "grad_norm": 0.48045825958251953,
+      "learning_rate": 0.0003737642376662456,
+      "loss": 0.9008,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6704129872446705,
+      "grad_norm": 0.565067708492279,
+      "learning_rate": 0.0003737033584569586,
+      "loss": 0.9471,
+      "step": 3758
+    },
+    {
+      "epoch": 0.6705913834626706,
+      "grad_norm": 0.5361825227737427,
+      "learning_rate": 0.0003736424695320631,
+      "loss": 1.1245,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6707697796806708,
+      "grad_norm": 0.47022882103919983,
+      "learning_rate": 0.00037358157089634127,
+      "loss": 0.9096,
+      "step": 3760
+    },
+    {
+      "epoch": 0.670948175898671,
+      "grad_norm": 0.49209490418434143,
+      "learning_rate": 0.000373520662554576,
+      "loss": 0.9196,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6711265721166711,
+      "grad_norm": 0.4670359194278717,
+      "learning_rate": 0.0003734597445115511,
+      "loss": 0.7455,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6713049683346713,
+      "grad_norm": 0.520520806312561,
+      "learning_rate": 0.000373398816772051,
+      "loss": 0.9116,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6714833645526714,
+      "grad_norm": 0.5995724201202393,
+      "learning_rate": 0.0003733378793408609,
+      "loss": 1.3693,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6716617607706716,
+      "grad_norm": 0.4709326922893524,
+      "learning_rate": 0.00037327693222276683,
+      "loss": 0.9914,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6718401569886718,
+      "grad_norm": 0.4547140598297119,
+      "learning_rate": 0.00037321597542255554,
+      "loss": 0.9951,
+      "step": 3766
+    },
+    {
+      "epoch": 0.672018553206672,
+      "grad_norm": 0.46117034554481506,
+      "learning_rate": 0.0003731550089450146,
+      "loss": 0.9847,
+      "step": 3767
+    },
+    {
+      "epoch": 0.6721969494246722,
+      "grad_norm": 0.6557008028030396,
+      "learning_rate": 0.00037309403279493227,
+      "loss": 0.9359,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6723753456426724,
+      "grad_norm": 0.4666654169559479,
+      "learning_rate": 0.00037303304697709755,
+      "loss": 0.8247,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6725537418606725,
+      "grad_norm": 0.4964097738265991,
+      "learning_rate": 0.00037297205149630023,
+      "loss": 0.8682,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6727321380786727,
+      "grad_norm": 0.47724971175193787,
+      "learning_rate": 0.000372911046357331,
+      "loss": 0.7927,
+      "step": 3771
+    },
+    {
+      "epoch": 0.6729105342966729,
+      "grad_norm": 0.4240172803401947,
+      "learning_rate": 0.00037285003156498097,
+      "loss": 0.6913,
+      "step": 3772
+    },
+    {
+      "epoch": 0.6730889305146731,
+      "grad_norm": 0.7888692021369934,
+      "learning_rate": 0.00037278900712404235,
+      "loss": 0.734,
+      "step": 3773
+    },
+    {
+      "epoch": 0.6732673267326733,
+      "grad_norm": 0.46591684222221375,
+      "learning_rate": 0.000372727973039308,
+      "loss": 0.8113,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6734457229506734,
+      "grad_norm": 0.47991588711738586,
+      "learning_rate": 0.00037266692931557145,
+      "loss": 0.9792,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6736241191686736,
+      "grad_norm": 0.4709876775741577,
+      "learning_rate": 0.00037260587595762705,
+      "loss": 0.8581,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6738025153866738,
+      "grad_norm": 0.4677605628967285,
+      "learning_rate": 0.0003725448129702699,
+      "loss": 0.9992,
+      "step": 3777
+    },
+    {
+      "epoch": 0.673980911604674,
+      "grad_norm": 0.47087326645851135,
+      "learning_rate": 0.0003724837403582959,
+      "loss": 0.9046,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6741593078226742,
+      "grad_norm": 0.44651949405670166,
+      "learning_rate": 0.0003724226581265016,
+      "loss": 0.6954,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6743377040406744,
+      "grad_norm": 0.4970155656337738,
+      "learning_rate": 0.0003723615662796844,
+      "loss": 0.9304,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6745161002586745,
+      "grad_norm": 0.5014781355857849,
+      "learning_rate": 0.0003723004648226425,
+      "loss": 1.093,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6746944964766747,
+      "grad_norm": 0.43172088265419006,
+      "learning_rate": 0.0003722393537601748,
+      "loss": 0.7743,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6748728926946749,
+      "grad_norm": 0.4725377559661865,
+      "learning_rate": 0.0003721782330970808,
+      "loss": 1.0071,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6750512889126751,
+      "grad_norm": 0.4693569839000702,
+      "learning_rate": 0.0003721171028381609,
+      "loss": 0.9637,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6752296851306753,
+      "grad_norm": 1.366469144821167,
+      "learning_rate": 0.0003720559629882163,
+      "loss": 0.8897,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6754080813486754,
+      "grad_norm": 0.4972679615020752,
+      "learning_rate": 0.0003719948135520489,
+      "loss": 0.8672,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6755864775666756,
+      "grad_norm": 0.47456255555152893,
+      "learning_rate": 0.00037193365453446126,
+      "loss": 0.8747,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6757648737846758,
+      "grad_norm": 0.4588843286037445,
+      "learning_rate": 0.000371872485940257,
+      "loss": 0.7423,
+      "step": 3788
+    },
+    {
+      "epoch": 0.675943270002676,
+      "grad_norm": 0.5166525840759277,
+      "learning_rate": 0.0003718113077742401,
+      "loss": 1.0436,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6761216662206762,
+      "grad_norm": 0.5156573057174683,
+      "learning_rate": 0.0003717501200412154,
+      "loss": 0.8229,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6763000624386764,
+      "grad_norm": 0.4706706702709198,
+      "learning_rate": 0.00037168892274598884,
+      "loss": 0.9703,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6764784586566764,
+      "grad_norm": 0.48749077320098877,
+      "learning_rate": 0.0003716277158933666,
+      "loss": 0.9563,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6766568548746766,
+      "grad_norm": 0.49335891008377075,
+      "learning_rate": 0.00037156649948815585,
+      "loss": 1.1277,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6768352510926768,
+      "grad_norm": 0.45877259969711304,
+      "learning_rate": 0.00037150527353516457,
+      "loss": 0.7111,
+      "step": 3794
+    },
+    {
+      "epoch": 0.677013647310677,
+      "grad_norm": 0.5235643982887268,
+      "learning_rate": 0.00037144403803920136,
+      "loss": 0.9515,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6771920435286772,
+      "grad_norm": 0.43552201986312866,
+      "learning_rate": 0.00037138279300507574,
+      "loss": 0.7094,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6773704397466773,
+      "grad_norm": 0.4871746301651001,
+      "learning_rate": 0.0003713215384375977,
+      "loss": 1.0252,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6775488359646775,
+      "grad_norm": 0.6767199635505676,
+      "learning_rate": 0.00037126027434157826,
+      "loss": 1.0102,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6777272321826777,
+      "grad_norm": 0.514312207698822,
+      "learning_rate": 0.0003711990007218291,
+      "loss": 0.921,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6779056284006779,
+      "grad_norm": 0.5360898375511169,
+      "learning_rate": 0.00037113771758316255,
+      "loss": 0.9598,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6780840246186781,
+      "grad_norm": 0.47847217321395874,
+      "learning_rate": 0.00037107642493039184,
+      "loss": 0.8728,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6782624208366783,
+      "grad_norm": 0.49450331926345825,
+      "learning_rate": 0.0003710151227683307,
+      "loss": 0.7853,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6784408170546784,
+      "grad_norm": 0.47471120953559875,
+      "learning_rate": 0.00037095381110179406,
+      "loss": 1.0686,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6786192132726786,
+      "grad_norm": 0.5599477887153625,
+      "learning_rate": 0.000370892489935597,
+      "loss": 1.1936,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6787976094906788,
+      "grad_norm": 0.525962233543396,
+      "learning_rate": 0.0003708311592745559,
+      "loss": 0.7819,
+      "step": 3805
+    },
+    {
+      "epoch": 0.678976005708679,
+      "grad_norm": 2.3493411540985107,
+      "learning_rate": 0.00037076981912348753,
+      "loss": 0.8769,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6791544019266792,
+      "grad_norm": 0.4811551868915558,
+      "learning_rate": 0.0003707084694872095,
+      "loss": 0.9784,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6793327981446793,
+      "grad_norm": 0.4610568583011627,
+      "learning_rate": 0.0003706471103705402,
+      "loss": 1.0097,
+      "step": 3808
+    },
+    {
+      "epoch": 0.6795111943626795,
+      "grad_norm": 0.5705133080482483,
+      "learning_rate": 0.0003705857417782989,
+      "loss": 1.113,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6796895905806797,
+      "grad_norm": 0.4349512457847595,
+      "learning_rate": 0.00037052436371530517,
+      "loss": 0.7286,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6798679867986799,
+      "grad_norm": 0.5243642330169678,
+      "learning_rate": 0.00037046297618637984,
+      "loss": 0.8463,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6800463830166801,
+      "grad_norm": 0.7079116702079773,
+      "learning_rate": 0.0003704015791963442,
+      "loss": 1.1995,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6802247792346803,
+      "grad_norm": 0.547192394733429,
+      "learning_rate": 0.00037034017275002043,
+      "loss": 0.9706,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6804031754526804,
+      "grad_norm": 1.1593716144561768,
+      "learning_rate": 0.00037027875685223115,
+      "loss": 0.8588,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6805815716706806,
+      "grad_norm": 0.46274659037590027,
+      "learning_rate": 0.0003702173315078001,
+      "loss": 0.9027,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6807599678886808,
+      "grad_norm": 0.5039249658584595,
+      "learning_rate": 0.0003701558967215517,
+      "loss": 1.0564,
+      "step": 3816
+    },
+    {
+      "epoch": 0.680938364106681,
+      "grad_norm": 0.43996962904930115,
+      "learning_rate": 0.00037009445249831075,
+      "loss": 0.9811,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6811167603246812,
+      "grad_norm": 0.5002626180648804,
+      "learning_rate": 0.00037003299884290315,
+      "loss": 1.0044,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6812951565426812,
+      "grad_norm": 0.5452658534049988,
+      "learning_rate": 0.0003699715357601555,
+      "loss": 0.82,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6814735527606814,
+      "grad_norm": 0.5324759483337402,
+      "learning_rate": 0.00036991006325489507,
+      "loss": 0.8513,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6816519489786816,
+      "grad_norm": 0.4961846172809601,
+      "learning_rate": 0.00036984858133194985,
+      "loss": 0.8615,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6818303451966818,
+      "grad_norm": 0.46880999207496643,
+      "learning_rate": 0.0003697870899961487,
+      "loss": 0.907,
+      "step": 3822
+    },
+    {
+      "epoch": 0.682008741414682,
+      "grad_norm": 0.4610176086425781,
+      "learning_rate": 0.0003697255892523211,
+      "loss": 0.9399,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6821871376326822,
+      "grad_norm": 0.5100571513175964,
+      "learning_rate": 0.00036966407910529715,
+      "loss": 0.9109,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6823655338506823,
+      "grad_norm": 0.4634707570075989,
+      "learning_rate": 0.00036960255955990787,
+      "loss": 0.8546,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6825439300686825,
+      "grad_norm": 0.5142379403114319,
+      "learning_rate": 0.0003695410306209851,
+      "loss": 0.9521,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6827223262866827,
+      "grad_norm": 0.4953165352344513,
+      "learning_rate": 0.0003694794922933612,
+      "loss": 0.9574,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6829007225046829,
+      "grad_norm": 0.530571460723877,
+      "learning_rate": 0.0003694179445818694,
+      "loss": 1.0247,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6830791187226831,
+      "grad_norm": 0.49260106682777405,
+      "learning_rate": 0.0003693563874913437,
+      "loss": 0.8785,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6832575149406832,
+      "grad_norm": 0.48824089765548706,
+      "learning_rate": 0.0003692948210266186,
+      "loss": 0.9205,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6834359111586834,
+      "grad_norm": 0.5169446468353271,
+      "learning_rate": 0.0003692332451925296,
+      "loss": 0.7974,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6836143073766836,
+      "grad_norm": 0.5154136419296265,
+      "learning_rate": 0.0003691716599939129,
+      "loss": 0.8386,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6837927035946838,
+      "grad_norm": 0.47398611903190613,
+      "learning_rate": 0.00036911006543560514,
+      "loss": 0.9594,
+      "step": 3833
+    },
+    {
+      "epoch": 0.683971099812684,
+      "grad_norm": 0.4491838812828064,
+      "learning_rate": 0.00036904846152244425,
+      "loss": 0.8647,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6841494960306842,
+      "grad_norm": 0.5062149167060852,
+      "learning_rate": 0.0003689868482592684,
+      "loss": 1.0463,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6843278922486843,
+      "grad_norm": 0.7602079510688782,
+      "learning_rate": 0.00036892522565091666,
+      "loss": 1.035,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6845062884666845,
+      "grad_norm": 0.46411171555519104,
+      "learning_rate": 0.00036886359370222896,
+      "loss": 0.8845,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 1.1559884548187256,
+      "learning_rate": 0.00036880195241804567,
+      "loss": 1.0434,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6848630809026849,
+      "grad_norm": 0.477621853351593,
+      "learning_rate": 0.0003687403018032082,
+      "loss": 1.0325,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6850414771206851,
+      "grad_norm": 0.47134527564048767,
+      "learning_rate": 0.0003686786418625585,
+      "loss": 0.8306,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6852198733386852,
+      "grad_norm": 0.47260260581970215,
+      "learning_rate": 0.0003686169726009393,
+      "loss": 1.0148,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6853982695566854,
+      "grad_norm": 0.4394271671772003,
+      "learning_rate": 0.0003685552940231942,
+      "loss": 0.8118,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6855766657746856,
+      "grad_norm": 0.5549933910369873,
+      "learning_rate": 0.0003684936061341673,
+      "loss": 0.8082,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6857550619926858,
+      "grad_norm": 0.49931079149246216,
+      "learning_rate": 0.00036843190893870356,
+      "loss": 0.9249,
+      "step": 3844
+    },
+    {
+      "epoch": 0.685933458210686,
+      "grad_norm": 0.5112189054489136,
+      "learning_rate": 0.00036837020244164865,
+      "loss": 0.9894,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6861118544286862,
+      "grad_norm": 0.544303297996521,
+      "learning_rate": 0.00036830848664784894,
+      "loss": 1.0843,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6862902506466863,
+      "grad_norm": 0.5149818658828735,
+      "learning_rate": 0.00036824676156215164,
+      "loss": 1.1905,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6864686468646864,
+      "grad_norm": 0.44846704602241516,
+      "learning_rate": 0.00036818502718940463,
+      "loss": 0.7878,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6866470430826866,
+      "grad_norm": 0.5292502641677856,
+      "learning_rate": 0.00036812328353445637,
+      "loss": 1.015,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6868254393006868,
+      "grad_norm": 0.4887588918209076,
+      "learning_rate": 0.00036806153060215627,
+      "loss": 0.9761,
+      "step": 3850
+    },
+    {
+      "epoch": 0.687003835518687,
+      "grad_norm": 0.4837714433670044,
+      "learning_rate": 0.00036799976839735436,
+      "loss": 0.9546,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6871822317366871,
+      "grad_norm": 0.4452599883079529,
+      "learning_rate": 0.00036793799692490145,
+      "loss": 0.9103,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6873606279546873,
+      "grad_norm": 0.4157600402832031,
+      "learning_rate": 0.000367876216189649,
+      "loss": 0.7636,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6875390241726875,
+      "grad_norm": 0.4187622666358948,
+      "learning_rate": 0.0003678144261964492,
+      "loss": 0.7129,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6877174203906877,
+      "grad_norm": 1.3808401823043823,
+      "learning_rate": 0.0003677526269501551,
+      "loss": 1.03,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6878958166086879,
+      "grad_norm": 0.4691815972328186,
+      "learning_rate": 0.00036769081845562033,
+      "loss": 0.8812,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6880742128266881,
+      "grad_norm": 0.5647638440132141,
+      "learning_rate": 0.0003676290007176994,
+      "loss": 1.0311,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6882526090446882,
+      "grad_norm": 0.47513964772224426,
+      "learning_rate": 0.0003675671737412473,
+      "loss": 0.8746,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6884310052626884,
+      "grad_norm": 0.5500690340995789,
+      "learning_rate": 0.00036750533753112004,
+      "loss": 1.1705,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6886094014806886,
+      "grad_norm": 0.45833203196525574,
+      "learning_rate": 0.0003674434920921741,
+      "loss": 0.8107,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6887877976986888,
+      "grad_norm": 0.6918094158172607,
+      "learning_rate": 0.00036738163742926677,
+      "loss": 0.9976,
+      "step": 3861
+    },
+    {
+      "epoch": 0.688966193916689,
+      "grad_norm": 0.49966564774513245,
+      "learning_rate": 0.0003673197735472563,
+      "loss": 0.7383,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6891445901346891,
+      "grad_norm": 0.5262913107872009,
+      "learning_rate": 0.0003672579004510012,
+      "loss": 0.924,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6893229863526893,
+      "grad_norm": 0.5179538726806641,
+      "learning_rate": 0.000367196018145361,
+      "loss": 0.9421,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6895013825706895,
+      "grad_norm": 0.502363920211792,
+      "learning_rate": 0.00036713412663519606,
+      "loss": 1.0026,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6896797787886897,
+      "grad_norm": 0.5639678239822388,
+      "learning_rate": 0.0003670722259253672,
+      "loss": 1.0164,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6898581750066899,
+      "grad_norm": 0.4934401214122772,
+      "learning_rate": 0.0003670103160207361,
+      "loss": 1.126,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6900365712246901,
+      "grad_norm": 0.4546942710876465,
+      "learning_rate": 0.0003669483969261651,
+      "loss": 0.8328,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6902149674426902,
+      "grad_norm": 0.5671041011810303,
+      "learning_rate": 0.00036688646864651745,
+      "loss": 1.0304,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6903933636606904,
+      "grad_norm": 0.5230646133422852,
+      "learning_rate": 0.0003668245311866567,
+      "loss": 0.9212,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6905717598786906,
+      "grad_norm": 0.4050388038158417,
+      "learning_rate": 0.0003667625845514476,
+      "loss": 0.5513,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6907501560966908,
+      "grad_norm": 0.6363729238510132,
+      "learning_rate": 0.00036670062874575535,
+      "loss": 0.9514,
+      "step": 3872
+    },
+    {
+      "epoch": 0.690928552314691,
+      "grad_norm": 0.49325573444366455,
+      "learning_rate": 0.0003666386637744459,
+      "loss": 1.0872,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6911069485326911,
+      "grad_norm": 0.49835440516471863,
+      "learning_rate": 0.00036657668964238613,
+      "loss": 1.1338,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6912853447506913,
+      "grad_norm": 0.4174972176551819,
+      "learning_rate": 0.0003665147063544432,
+      "loss": 0.7673,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6914637409686915,
+      "grad_norm": 0.43042904138565063,
+      "learning_rate": 0.0003664527139154854,
+      "loss": 0.768,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6916421371866917,
+      "grad_norm": 0.46625080704689026,
+      "learning_rate": 0.00036639071233038155,
+      "loss": 0.9156,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6918205334046919,
+      "grad_norm": 0.4662775993347168,
+      "learning_rate": 0.0003663287016040013,
+      "loss": 0.8285,
+      "step": 3878
+    },
+    {
+      "epoch": 0.691998929622692,
+      "grad_norm": 0.4520246982574463,
+      "learning_rate": 0.0003662666817412148,
+      "loss": 0.8158,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6921773258406921,
+      "grad_norm": 0.5535463690757751,
+      "learning_rate": 0.0003662046527468932,
+      "loss": 1.1017,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6923557220586923,
+      "grad_norm": 1.1918361186981201,
+      "learning_rate": 0.00036614261462590824,
+      "loss": 0.793,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6925341182766925,
+      "grad_norm": 0.5035448670387268,
+      "learning_rate": 0.00036608056738313225,
+      "loss": 0.8917,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6927125144946927,
+      "grad_norm": 0.5091261267662048,
+      "learning_rate": 0.00036601851102343843,
+      "loss": 1.0292,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6928909107126929,
+      "grad_norm": 0.45190200209617615,
+      "learning_rate": 0.0003659564455517007,
+      "loss": 1.043,
+      "step": 3884
+    },
+    {
+      "epoch": 0.693069306930693,
+      "grad_norm": 0.5362012386322021,
+      "learning_rate": 0.0003658943709727936,
+      "loss": 1.0457,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6932477031486932,
+      "grad_norm": 0.5333651304244995,
+      "learning_rate": 0.00036583228729159244,
+      "loss": 1.0046,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6934260993666934,
+      "grad_norm": 0.5217045545578003,
+      "learning_rate": 0.0003657701945129734,
+      "loss": 0.9786,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6936044955846936,
+      "grad_norm": 0.5744450688362122,
+      "learning_rate": 0.0003657080926418131,
+      "loss": 1.03,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6937828918026938,
+      "grad_norm": 0.5800135135650635,
+      "learning_rate": 0.0003656459816829889,
+      "loss": 1.0379,
+      "step": 3889
+    },
+    {
+      "epoch": 0.693961288020694,
+      "grad_norm": 0.4809887707233429,
+      "learning_rate": 0.0003655838616413791,
+      "loss": 0.9534,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6941396842386941,
+      "grad_norm": 0.46151599287986755,
+      "learning_rate": 0.0003655217325218626,
+      "loss": 0.9506,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6943180804566943,
+      "grad_norm": 1.1912611722946167,
+      "learning_rate": 0.0003654595943293189,
+      "loss": 0.7983,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6944964766746945,
+      "grad_norm": 0.48154789209365845,
+      "learning_rate": 0.00036539744706862837,
+      "loss": 0.8602,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6946748728926947,
+      "grad_norm": 0.5169458985328674,
+      "learning_rate": 0.000365335290744672,
+      "loss": 1.0327,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6948532691106949,
+      "grad_norm": 0.4628457725048065,
+      "learning_rate": 0.00036527312536233146,
+      "loss": 0.8947,
+      "step": 3895
+    },
+    {
+      "epoch": 0.695031665328695,
+      "grad_norm": 0.5199170708656311,
+      "learning_rate": 0.00036521095092648933,
+      "loss": 0.9375,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6952100615466952,
+      "grad_norm": 0.5095205307006836,
+      "learning_rate": 0.0003651487674420287,
+      "loss": 1.0075,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6953884577646954,
+      "grad_norm": 0.5311485528945923,
+      "learning_rate": 0.0003650865749138334,
+      "loss": 0.8594,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6955668539826956,
+      "grad_norm": 0.5111739635467529,
+      "learning_rate": 0.000365024373346788,
+      "loss": 0.8538,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6957452502006958,
+      "grad_norm": 0.48374149203300476,
+      "learning_rate": 0.0003649621627457779,
+      "loss": 0.9959,
+      "step": 3900
+    },
+    {
+      "epoch": 0.695923646418696,
+      "grad_norm": 0.4373980760574341,
+      "learning_rate": 0.00036489994311568897,
+      "loss": 0.8307,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6961020426366961,
+      "grad_norm": 0.47654423117637634,
+      "learning_rate": 0.0003648377144614079,
+      "loss": 0.8287,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6962804388546963,
+      "grad_norm": 0.5087730884552002,
+      "learning_rate": 0.0003647754767878222,
+      "loss": 1.21,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6964588350726965,
+      "grad_norm": 0.503084123134613,
+      "learning_rate": 0.0003647132300998199,
+      "loss": 0.8816,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6966372312906967,
+      "grad_norm": 0.4780065417289734,
+      "learning_rate": 0.00036465097440229003,
+      "loss": 0.9056,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6968156275086969,
+      "grad_norm": 0.48324069380760193,
+      "learning_rate": 0.0003645887097001218,
+      "loss": 0.9316,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6969940237266969,
+      "grad_norm": 0.4731059968471527,
+      "learning_rate": 0.0003645264359982057,
+      "loss": 0.8059,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6971724199446971,
+      "grad_norm": 0.4973444640636444,
+      "learning_rate": 0.0003644641533014326,
+      "loss": 1.0185,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6973508161626973,
+      "grad_norm": 0.5493341088294983,
+      "learning_rate": 0.00036440186161469407,
+      "loss": 1.1505,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6975292123806975,
+      "grad_norm": 0.5255337357521057,
+      "learning_rate": 0.0003643395609428827,
+      "loss": 1.0592,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6977076085986977,
+      "grad_norm": 0.4478547275066376,
+      "learning_rate": 0.0003642772512908913,
+      "loss": 0.8705,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6978860048166979,
+      "grad_norm": 0.49308907985687256,
+      "learning_rate": 0.00036421493266361384,
+      "loss": 0.745,
+      "step": 3912
+    },
+    {
+      "epoch": 0.698064401034698,
+      "grad_norm": 0.5503551363945007,
+      "learning_rate": 0.00036415260506594463,
+      "loss": 1.0582,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6982427972526982,
+      "grad_norm": 0.5110753178596497,
+      "learning_rate": 0.00036409026850277906,
+      "loss": 0.7994,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6984211934706984,
+      "grad_norm": 0.5047644376754761,
+      "learning_rate": 0.00036402792297901283,
+      "loss": 0.9037,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6985995896886986,
+      "grad_norm": 0.5323673486709595,
+      "learning_rate": 0.0003639655684995426,
+      "loss": 0.8984,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6987779859066988,
+      "grad_norm": 0.49493974447250366,
+      "learning_rate": 0.0003639032050692656,
+      "loss": 0.949,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6989563821246989,
+      "grad_norm": 0.4778236150741577,
+      "learning_rate": 0.00036384083269308,
+      "loss": 0.6483,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6991347783426991,
+      "grad_norm": 0.519985556602478,
+      "learning_rate": 0.00036377845137588435,
+      "loss": 1.0307,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6993131745606993,
+      "grad_norm": 0.49531275033950806,
+      "learning_rate": 0.00036371606112257807,
+      "loss": 0.7647,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6994915707786995,
+      "grad_norm": 0.5061904788017273,
+      "learning_rate": 0.00036365366193806135,
+      "loss": 0.93,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6996699669966997,
+      "grad_norm": 0.45195135474205017,
+      "learning_rate": 0.0003635912538272349,
+      "loss": 0.831,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6998483632146999,
+      "grad_norm": 0.5661010146141052,
+      "learning_rate": 0.00036352883679500027,
+      "loss": 0.9327,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7000267594327,
+      "grad_norm": 0.5504164099693298,
+      "learning_rate": 0.0003634664108462596,
+      "loss": 0.8443,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7002051556507002,
+      "grad_norm": 0.6557819843292236,
+      "learning_rate": 0.0003634039759859158,
+      "loss": 1.0994,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7003835518687004,
+      "grad_norm": 0.48669978976249695,
+      "learning_rate": 0.00036334153221887264,
+      "loss": 0.7105,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7005619480867006,
+      "grad_norm": 0.4523518979549408,
+      "learning_rate": 0.00036327907955003425,
+      "loss": 0.8413,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7007403443047008,
+      "grad_norm": 0.469539999961853,
+      "learning_rate": 0.0003632166179843058,
+      "loss": 0.7825,
+      "step": 3928
+    },
+    {
+      "epoch": 0.7009187405227009,
+      "grad_norm": 0.5297795534133911,
+      "learning_rate": 0.0003631541475265928,
+      "loss": 1.044,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7010971367407011,
+      "grad_norm": 0.46915116906166077,
+      "learning_rate": 0.0003630916681818018,
+      "loss": 1.0134,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7012755329587013,
+      "grad_norm": 0.7894530296325684,
+      "learning_rate": 0.0003630291799548398,
+      "loss": 1.0422,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7014539291767015,
+      "grad_norm": 0.4228399395942688,
+      "learning_rate": 0.00036296668285061464,
+      "loss": 0.8499,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7016323253947017,
+      "grad_norm": 0.5037341713905334,
+      "learning_rate": 0.00036290417687403483,
+      "loss": 0.9448,
+      "step": 3933
+    },
+    {
+      "epoch": 0.7018107216127019,
+      "grad_norm": 0.43407323956489563,
+      "learning_rate": 0.00036284166203000957,
+      "loss": 0.7291,
+      "step": 3934
+    },
+    {
+      "epoch": 0.701989117830702,
+      "grad_norm": 0.46290135383605957,
+      "learning_rate": 0.00036277913832344875,
+      "loss": 0.8746,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7021675140487021,
+      "grad_norm": 0.4666912257671356,
+      "learning_rate": 0.0003627166057592629,
+      "loss": 0.8483,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7023459102667023,
+      "grad_norm": 0.44675248861312866,
+      "learning_rate": 0.0003626540643423634,
+      "loss": 0.8229,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7025243064847025,
+      "grad_norm": 0.46909114718437195,
+      "learning_rate": 0.0003625915140776621,
+      "loss": 0.7874,
+      "step": 3938
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 0.47172433137893677,
+      "learning_rate": 0.00036252895497007175,
+      "loss": 0.9997,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7028810989207028,
+      "grad_norm": 0.5105620622634888,
+      "learning_rate": 0.0003624663870245057,
+      "loss": 0.9346,
+      "step": 3940
+    },
+    {
+      "epoch": 0.703059495138703,
+      "grad_norm": 0.4702140688896179,
+      "learning_rate": 0.0003624038102458781,
+      "loss": 0.9306,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7032378913567032,
+      "grad_norm": 0.49693727493286133,
+      "learning_rate": 0.0003623412246391035,
+      "loss": 0.979,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7034162875747034,
+      "grad_norm": 0.47724950313568115,
+      "learning_rate": 0.00036227863020909753,
+      "loss": 0.9922,
+      "step": 3943
+    },
+    {
+      "epoch": 0.7035946837927036,
+      "grad_norm": 0.4332534968852997,
+      "learning_rate": 0.0003622160269607762,
+      "loss": 0.8226,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7037730800107038,
+      "grad_norm": 0.4521944522857666,
+      "learning_rate": 0.00036215341489905645,
+      "loss": 0.9031,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7039514762287039,
+      "grad_norm": 0.46443673968315125,
+      "learning_rate": 0.00036209079402885577,
+      "loss": 0.8799,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7041298724467041,
+      "grad_norm": 0.48729583621025085,
+      "learning_rate": 0.00036202816435509233,
+      "loss": 0.8665,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7043082686647043,
+      "grad_norm": 0.48398932814598083,
+      "learning_rate": 0.00036196552588268506,
+      "loss": 0.8229,
+      "step": 3948
+    },
+    {
+      "epoch": 0.7044866648827045,
+      "grad_norm": 0.47495603561401367,
+      "learning_rate": 0.0003619028786165536,
+      "loss": 0.8878,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7046650611007047,
+      "grad_norm": 0.5229374766349792,
+      "learning_rate": 0.0003618402225616182,
+      "loss": 0.9401,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7048434573187048,
+      "grad_norm": 0.4787358343601227,
+      "learning_rate": 0.00036177755772279983,
+      "loss": 0.9042,
+      "step": 3951
+    },
+    {
+      "epoch": 0.705021853536705,
+      "grad_norm": 0.4975121021270752,
+      "learning_rate": 0.00036171488410502016,
+      "loss": 1.0987,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7052002497547052,
+      "grad_norm": 0.44728225469589233,
+      "learning_rate": 0.00036165220171320166,
+      "loss": 0.774,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7053786459727054,
+      "grad_norm": 0.5024580955505371,
+      "learning_rate": 0.0003615895105522672,
+      "loss": 0.9369,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7055570421907056,
+      "grad_norm": 0.482075035572052,
+      "learning_rate": 0.00036152681062714064,
+      "loss": 0.9221,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7057354384087058,
+      "grad_norm": 0.4962591230869293,
+      "learning_rate": 0.0003614641019427464,
+      "loss": 0.8499,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7059138346267059,
+      "grad_norm": 0.5466794967651367,
+      "learning_rate": 0.0003614013845040095,
+      "loss": 0.8993,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7060922308447061,
+      "grad_norm": 0.5112261176109314,
+      "learning_rate": 0.00036133865831585577,
+      "loss": 0.8757,
+      "step": 3958
+    },
+    {
+      "epoch": 0.7062706270627063,
+      "grad_norm": 0.4203137159347534,
+      "learning_rate": 0.0003612759233832118,
+      "loss": 0.7496,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7064490232807065,
+      "grad_norm": 0.4720518887042999,
+      "learning_rate": 0.00036121317971100464,
+      "loss": 0.9851,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7066274194987067,
+      "grad_norm": 0.43325722217559814,
+      "learning_rate": 0.0003611504273041623,
+      "loss": 0.6659,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7068058157167068,
+      "grad_norm": 0.44432467222213745,
+      "learning_rate": 0.0003610876661676131,
+      "loss": 0.8509,
+      "step": 3962
+    },
+    {
+      "epoch": 0.706984211934707,
+      "grad_norm": 0.4786689281463623,
+      "learning_rate": 0.0003610248963062865,
+      "loss": 0.9216,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7071626081527072,
+      "grad_norm": 0.4814835488796234,
+      "learning_rate": 0.00036096211772511226,
+      "loss": 1.0138,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7073410043707073,
+      "grad_norm": 0.4865817725658417,
+      "learning_rate": 0.0003608993304290211,
+      "loss": 1.0312,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7075194005887075,
+      "grad_norm": 0.49532395601272583,
+      "learning_rate": 0.00036083653442294417,
+      "loss": 0.7395,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7076977968067077,
+      "grad_norm": 0.4562307298183441,
+      "learning_rate": 0.0003607737297118136,
+      "loss": 0.7416,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7078761930247078,
+      "grad_norm": 0.5441547632217407,
+      "learning_rate": 0.00036071091630056204,
+      "loss": 1.2372,
+      "step": 3968
+    },
+    {
+      "epoch": 0.708054589242708,
+      "grad_norm": 0.4169851839542389,
+      "learning_rate": 0.00036064809419412264,
+      "loss": 0.8411,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7082329854607082,
+      "grad_norm": 0.48741570115089417,
+      "learning_rate": 0.0003605852633974296,
+      "loss": 1.1458,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7084113816787084,
+      "grad_norm": 0.47009482979774475,
+      "learning_rate": 0.0003605224239154175,
+      "loss": 0.8117,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7085897778967086,
+      "grad_norm": 0.4695878028869629,
+      "learning_rate": 0.00036045957575302174,
+      "loss": 0.8465,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7087681741147087,
+      "grad_norm": 0.46119317412376404,
+      "learning_rate": 0.0003603967189151785,
+      "loss": 0.7996,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7089465703327089,
+      "grad_norm": 0.46736615896224976,
+      "learning_rate": 0.0003603338534068245,
+      "loss": 0.8513,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7091249665507091,
+      "grad_norm": 0.588634729385376,
+      "learning_rate": 0.00036027097923289707,
+      "loss": 1.0962,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7093033627687093,
+      "grad_norm": 0.4701951742172241,
+      "learning_rate": 0.00036020809639833446,
+      "loss": 0.8302,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7094817589867095,
+      "grad_norm": 0.4576779305934906,
+      "learning_rate": 0.00036014520490807535,
+      "loss": 0.8504,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7096601552047097,
+      "grad_norm": 0.5647838711738586,
+      "learning_rate": 0.00036008230476705915,
+      "loss": 0.8446,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7098385514227098,
+      "grad_norm": 0.48183631896972656,
+      "learning_rate": 0.00036001939598022625,
+      "loss": 0.9849,
+      "step": 3979
+    },
+    {
+      "epoch": 0.71001694764071,
+      "grad_norm": 0.4493800699710846,
+      "learning_rate": 0.00035995647855251726,
+      "loss": 0.8262,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7101953438587102,
+      "grad_norm": 0.5250908136367798,
+      "learning_rate": 0.00035989355248887384,
+      "loss": 0.7036,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7103737400767104,
+      "grad_norm": 0.4472370743751526,
+      "learning_rate": 0.000359830617794238,
+      "loss": 0.8584,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7105521362947106,
+      "grad_norm": 0.48600542545318604,
+      "learning_rate": 0.00035976767447355273,
+      "loss": 0.7812,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7107305325127107,
+      "grad_norm": 0.49188825488090515,
+      "learning_rate": 0.00035970472253176155,
+      "loss": 0.8518,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7109089287307109,
+      "grad_norm": 0.482270210981369,
+      "learning_rate": 0.0003596417619738087,
+      "loss": 0.767,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7110873249487111,
+      "grad_norm": 0.460519403219223,
+      "learning_rate": 0.000359578792804639,
+      "loss": 0.9311,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7112657211667113,
+      "grad_norm": 0.5443057417869568,
+      "learning_rate": 0.00035951581502919813,
+      "loss": 0.8974,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7114441173847115,
+      "grad_norm": 0.44756433367729187,
+      "learning_rate": 0.0003594528286524322,
+      "loss": 0.8031,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7116225136027117,
+      "grad_norm": 0.7050688862800598,
+      "learning_rate": 0.0003593898336792883,
+      "loss": 1.1276,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7118009098207118,
+      "grad_norm": 0.5873562097549438,
+      "learning_rate": 0.0003593268301147139,
+      "loss": 1.1023,
+      "step": 3990
+    },
+    {
+      "epoch": 0.711979306038712,
+      "grad_norm": 0.4501679241657257,
+      "learning_rate": 0.0003592638179636573,
+      "loss": 0.8911,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7121577022567122,
+      "grad_norm": 0.48664426803588867,
+      "learning_rate": 0.0003592007972310674,
+      "loss": 0.8171,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7123360984747124,
+      "grad_norm": 0.45436549186706543,
+      "learning_rate": 0.000359137767921894,
+      "loss": 0.7589,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7125144946927126,
+      "grad_norm": 0.5350733995437622,
+      "learning_rate": 0.0003590747300410871,
+      "loss": 1.1801,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7126928909107126,
+      "grad_norm": 0.49254482984542847,
+      "learning_rate": 0.00035901168359359797,
+      "loss": 0.9704,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7128712871287128,
+      "grad_norm": 0.4413221776485443,
+      "learning_rate": 0.0003589486285843781,
+      "loss": 0.9475,
+      "step": 3996
+    },
+    {
+      "epoch": 0.713049683346713,
+      "grad_norm": 0.4686570465564728,
+      "learning_rate": 0.0003588855650183798,
+      "loss": 0.8909,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7132280795647132,
+      "grad_norm": 0.5192868113517761,
+      "learning_rate": 0.0003588224929005561,
+      "loss": 0.9769,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7134064757827134,
+      "grad_norm": 0.4699283242225647,
+      "learning_rate": 0.0003587594122358607,
+      "loss": 0.8673,
+      "step": 3999
+    },
+    {
+      "epoch": 0.7135848720007136,
+      "grad_norm": 0.44109442830085754,
+      "learning_rate": 0.00035869632302924776,
+      "loss": 0.8594,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7137632682187137,
+      "grad_norm": 0.44390279054641724,
+      "learning_rate": 0.00035863322528567246,
+      "loss": 0.8449,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7139416644367139,
+      "grad_norm": 0.5160883069038391,
+      "learning_rate": 0.00035857011901009036,
+      "loss": 0.9412,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7141200606547141,
+      "grad_norm": 0.49634233117103577,
+      "learning_rate": 0.00035850700420745783,
+      "loss": 1.0827,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7142984568727143,
+      "grad_norm": 0.45757240056991577,
+      "learning_rate": 0.0003584438808827319,
+      "loss": 0.8453,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7144768530907145,
+      "grad_norm": 2.338979959487915,
+      "learning_rate": 0.0003583807490408702,
+      "loss": 1.1377,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7146552493087146,
+      "grad_norm": 0.46473217010498047,
+      "learning_rate": 0.00035831760868683117,
+      "loss": 0.9048,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7148336455267148,
+      "grad_norm": 0.47779056429862976,
+      "learning_rate": 0.0003582544598255737,
+      "loss": 1.0637,
+      "step": 4007
+    },
+    {
+      "epoch": 0.715012041744715,
+      "grad_norm": 0.4706021845340729,
+      "learning_rate": 0.0003581913024620577,
+      "loss": 0.7752,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7151904379627152,
+      "grad_norm": 0.4808463752269745,
+      "learning_rate": 0.0003581281366012431,
+      "loss": 0.8006,
+      "step": 4009
+    },
+    {
+      "epoch": 0.7153688341807154,
+      "grad_norm": 0.5095382928848267,
+      "learning_rate": 0.0003580649622480914,
+      "loss": 0.9802,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7155472303987156,
+      "grad_norm": 0.5496415495872498,
+      "learning_rate": 0.000358001779407564,
+      "loss": 1.2067,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7157256266167157,
+      "grad_norm": 0.47136303782463074,
+      "learning_rate": 0.0003579385880846232,
+      "loss": 0.8652,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7159040228347159,
+      "grad_norm": 0.7725960612297058,
+      "learning_rate": 0.00035787538828423225,
+      "loss": 0.7905,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7160824190527161,
+      "grad_norm": 0.5106984376907349,
+      "learning_rate": 0.0003578121800113548,
+      "loss": 0.8218,
+      "step": 4014
+    },
+    {
+      "epoch": 0.7162608152707163,
+      "grad_norm": 0.4280400276184082,
+      "learning_rate": 0.0003577489632709551,
+      "loss": 0.7157,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7164392114887165,
+      "grad_norm": 0.439709335565567,
+      "learning_rate": 0.0003576857380679981,
+      "loss": 0.9857,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7166176077067166,
+      "grad_norm": 0.4796721041202545,
+      "learning_rate": 0.0003576225044074496,
+      "loss": 0.7725,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7167960039247168,
+      "grad_norm": 0.5130892992019653,
+      "learning_rate": 0.00035755926229427595,
+      "loss": 1.048,
+      "step": 4018
+    },
+    {
+      "epoch": 0.716974400142717,
+      "grad_norm": 0.6783162951469421,
+      "learning_rate": 0.0003574960117334441,
+      "loss": 0.9083,
+      "step": 4019
+    },
+    {
+      "epoch": 0.7171527963607172,
+      "grad_norm": 0.551443874835968,
+      "learning_rate": 0.00035743275272992177,
+      "loss": 0.9846,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7173311925787174,
+      "grad_norm": 0.5131356716156006,
+      "learning_rate": 0.0003573694852886773,
+      "loss": 0.8985,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7175095887967176,
+      "grad_norm": 0.5424959659576416,
+      "learning_rate": 0.0003573062094146796,
+      "loss": 1.0965,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7176879850147176,
+      "grad_norm": 0.5547580122947693,
+      "learning_rate": 0.0003572429251128984,
+      "loss": 0.9774,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7178663812327178,
+      "grad_norm": 0.4463725686073303,
+      "learning_rate": 0.000357179632388304,
+      "loss": 0.7453,
+      "step": 4024
+    },
+    {
+      "epoch": 0.718044777450718,
+      "grad_norm": 0.4756935238838196,
+      "learning_rate": 0.0003571163312458674,
+      "loss": 0.8413,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7182231736687182,
+      "grad_norm": 0.47553977370262146,
+      "learning_rate": 0.0003570530216905603,
+      "loss": 0.7793,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7184015698867184,
+      "grad_norm": 0.48220735788345337,
+      "learning_rate": 0.000356989703727355,
+      "loss": 0.858,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7185799661047185,
+      "grad_norm": 1.1826428174972534,
+      "learning_rate": 0.0003569263773612242,
+      "loss": 1.007,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7187583623227187,
+      "grad_norm": 0.5101688504219055,
+      "learning_rate": 0.0003568630425971419,
+      "loss": 1.0152,
+      "step": 4029
+    },
+    {
+      "epoch": 0.7189367585407189,
+      "grad_norm": 0.6145039200782776,
+      "learning_rate": 0.00035679969944008217,
+      "loss": 0.8944,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7191151547587191,
+      "grad_norm": 0.4975384771823883,
+      "learning_rate": 0.00035673634789502,
+      "loss": 0.9689,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7192935509767193,
+      "grad_norm": 0.5056141018867493,
+      "learning_rate": 0.00035667298796693097,
+      "loss": 0.8644,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7194719471947195,
+      "grad_norm": 0.51641845703125,
+      "learning_rate": 0.00035660961966079146,
+      "loss": 1.1757,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7196503434127196,
+      "grad_norm": 0.6307793259620667,
+      "learning_rate": 0.00035654624298157823,
+      "loss": 0.9336,
+      "step": 4034
+    },
+    {
+      "epoch": 0.7198287396307198,
+      "grad_norm": 0.5330437421798706,
+      "learning_rate": 0.0003564828579342689,
+      "loss": 0.9126,
+      "step": 4035
+    },
+    {
+      "epoch": 0.72000713584872,
+      "grad_norm": 0.4748086631298065,
+      "learning_rate": 0.00035641946452384183,
+      "loss": 0.9391,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7201855320667202,
+      "grad_norm": 0.4811098873615265,
+      "learning_rate": 0.00035635606275527575,
+      "loss": 0.9427,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7203639282847204,
+      "grad_norm": 0.551956832408905,
+      "learning_rate": 0.00035629265263355025,
+      "loss": 1.0388,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7205423245027205,
+      "grad_norm": 0.5036234855651855,
+      "learning_rate": 0.0003562292341636456,
+      "loss": 0.9119,
+      "step": 4039
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 0.5007266402244568,
+      "learning_rate": 0.0003561658073505426,
+      "loss": 1.0323,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7208991169387209,
+      "grad_norm": 0.5283923745155334,
+      "learning_rate": 0.0003561023721992228,
+      "loss": 0.8999,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7210775131567211,
+      "grad_norm": 0.885420024394989,
+      "learning_rate": 0.0003560389287146683,
+      "loss": 0.9517,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7212559093747213,
+      "grad_norm": 0.4583801329135895,
+      "learning_rate": 0.000355975476901862,
+      "loss": 0.8263,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7214343055927215,
+      "grad_norm": 0.4987337291240692,
+      "learning_rate": 0.00035591201676578733,
+      "loss": 0.8681,
+      "step": 4044
+    },
+    {
+      "epoch": 0.7216127018107216,
+      "grad_norm": 0.5699477195739746,
+      "learning_rate": 0.00035584854831142846,
+      "loss": 0.9642,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7217910980287218,
+      "grad_norm": 0.5087426900863647,
+      "learning_rate": 0.00035578507154377016,
+      "loss": 0.9291,
+      "step": 4046
+    },
+    {
+      "epoch": 0.721969494246722,
+      "grad_norm": 0.520293653011322,
+      "learning_rate": 0.00035572158646779786,
+      "loss": 0.988,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7221478904647222,
+      "grad_norm": 0.511569082736969,
+      "learning_rate": 0.0003556580930884976,
+      "loss": 0.9026,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7223262866827224,
+      "grad_norm": 0.4560140371322632,
+      "learning_rate": 0.0003555945914108562,
+      "loss": 0.8434,
+      "step": 4049
+    },
+    {
+      "epoch": 0.7225046829007225,
+      "grad_norm": 0.49847978353500366,
+      "learning_rate": 0.00035553108143986106,
+      "loss": 1.0567,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7226830791187226,
+      "grad_norm": 0.4683671295642853,
+      "learning_rate": 0.0003554675631805002,
+      "loss": 0.7456,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7228614753367228,
+      "grad_norm": 0.5200768113136292,
+      "learning_rate": 0.0003554040366377623,
+      "loss": 0.9804,
+      "step": 4052
+    },
+    {
+      "epoch": 0.723039871554723,
+      "grad_norm": 0.5331013202667236,
+      "learning_rate": 0.0003553405018166367,
+      "loss": 1.0793,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7232182677727232,
+      "grad_norm": 0.48200780153274536,
+      "learning_rate": 0.00035527695872211354,
+      "loss": 0.9059,
+      "step": 4054
+    },
+    {
+      "epoch": 0.7233966639907234,
+      "grad_norm": 0.5711750388145447,
+      "learning_rate": 0.00035521340735918317,
+      "loss": 1.0259,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7235750602087235,
+      "grad_norm": 0.5371613502502441,
+      "learning_rate": 0.00035514984773283713,
+      "loss": 0.9537,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7237534564267237,
+      "grad_norm": 0.47757935523986816,
+      "learning_rate": 0.0003550862798480673,
+      "loss": 0.9782,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7239318526447239,
+      "grad_norm": 0.5152607560157776,
+      "learning_rate": 0.0003550227037098663,
+      "loss": 0.9834,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7241102488627241,
+      "grad_norm": 0.44710201025009155,
+      "learning_rate": 0.0003549591193232273,
+      "loss": 0.9038,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7242886450807243,
+      "grad_norm": 0.4328174591064453,
+      "learning_rate": 0.00035489552669314427,
+      "loss": 0.5936,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7244670412987244,
+      "grad_norm": 0.5450094938278198,
+      "learning_rate": 0.00035483192582461175,
+      "loss": 0.9184,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7246454375167246,
+      "grad_norm": 0.4884642958641052,
+      "learning_rate": 0.0003547683167226249,
+      "loss": 0.982,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7248238337347248,
+      "grad_norm": 0.4127320647239685,
+      "learning_rate": 0.00035470469939217944,
+      "loss": 0.6299,
+      "step": 4063
+    },
+    {
+      "epoch": 0.725002229952725,
+      "grad_norm": 0.5126365423202515,
+      "learning_rate": 0.0003546410738382719,
+      "loss": 0.7333,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7251806261707252,
+      "grad_norm": 1.068241000175476,
+      "learning_rate": 0.0003545774400658996,
+      "loss": 0.7118,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7253590223887254,
+      "grad_norm": 0.5835636258125305,
+      "learning_rate": 0.00035451379808006014,
+      "loss": 0.9209,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7255374186067255,
+      "grad_norm": 0.4724158048629761,
+      "learning_rate": 0.0003544501478857519,
+      "loss": 0.7558,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7257158148247257,
+      "grad_norm": 0.536562442779541,
+      "learning_rate": 0.000354386489487974,
+      "loss": 0.9178,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7258942110427259,
+      "grad_norm": 1.3931955099105835,
+      "learning_rate": 0.0003543228228917262,
+      "loss": 1.0701,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7260726072607261,
+      "grad_norm": 0.6029762029647827,
+      "learning_rate": 0.0003542591481020087,
+      "loss": 0.7614,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7262510034787263,
+      "grad_norm": 2.0984432697296143,
+      "learning_rate": 0.0003541954651238226,
+      "loss": 0.9829,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7264293996967264,
+      "grad_norm": 1.1561270952224731,
+      "learning_rate": 0.0003541317739621695,
+      "loss": 1.1042,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7266077959147266,
+      "grad_norm": 0.9806638360023499,
+      "learning_rate": 0.0003540680746220518,
+      "loss": 0.7303,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7267861921327268,
+      "grad_norm": 0.5632558465003967,
+      "learning_rate": 0.0003540043671084722,
+      "loss": 1.0854,
+      "step": 4074
+    },
+    {
+      "epoch": 0.726964588350727,
+      "grad_norm": 0.5446335077285767,
+      "learning_rate": 0.00035394065142643435,
+      "loss": 1.0717,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7271429845687272,
+      "grad_norm": 1.9541873931884766,
+      "learning_rate": 0.0003538769275809425,
+      "loss": 0.9602,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7273213807867274,
+      "grad_norm": 0.5483476519584656,
+      "learning_rate": 0.00035381319557700145,
+      "loss": 1.0379,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7274997770047275,
+      "grad_norm": 0.5389299392700195,
+      "learning_rate": 0.00035374945541961656,
+      "loss": 0.9395,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7276781732227277,
+      "grad_norm": 0.5572075247764587,
+      "learning_rate": 0.00035368570711379423,
+      "loss": 1.0089,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7278565694407279,
+      "grad_norm": 0.4937756359577179,
+      "learning_rate": 0.00035362195066454116,
+      "loss": 0.9257,
+      "step": 4080
+    },
+    {
+      "epoch": 0.728034965658728,
+      "grad_norm": 0.5375781059265137,
+      "learning_rate": 0.00035355818607686455,
+      "loss": 0.9355,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7282133618767282,
+      "grad_norm": 0.4889019727706909,
+      "learning_rate": 0.0003534944133557726,
+      "loss": 0.8256,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7283917580947283,
+      "grad_norm": 0.5278054475784302,
+      "learning_rate": 0.000353430632506274,
+      "loss": 1.029,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7285701543127285,
+      "grad_norm": 0.5437285900115967,
+      "learning_rate": 0.000353366843533378,
+      "loss": 1.1467,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7287485505307287,
+      "grad_norm": 0.5061957240104675,
+      "learning_rate": 0.00035330304644209454,
+      "loss": 1.0266,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7289269467487289,
+      "grad_norm": 0.5521642565727234,
+      "learning_rate": 0.00035323924123743436,
+      "loss": 0.9845,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7291053429667291,
+      "grad_norm": 0.5419567823410034,
+      "learning_rate": 0.00035317542792440853,
+      "loss": 0.7985,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7292837391847293,
+      "grad_norm": 0.510177731513977,
+      "learning_rate": 0.00035311160650802905,
+      "loss": 0.9469,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7294621354027294,
+      "grad_norm": 0.5522062182426453,
+      "learning_rate": 0.0003530477769933083,
+      "loss": 0.9675,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7296405316207296,
+      "grad_norm": 0.5909835696220398,
+      "learning_rate": 0.00035298393938525954,
+      "loss": 1.0044,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7298189278387298,
+      "grad_norm": 0.5589690804481506,
+      "learning_rate": 0.0003529200936888965,
+      "loss": 1.0368,
+      "step": 4091
+    },
+    {
+      "epoch": 0.72999732405673,
+      "grad_norm": 0.5542118549346924,
+      "learning_rate": 0.00035285623990923356,
+      "loss": 0.7456,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7301757202747302,
+      "grad_norm": 2.88665509223938,
+      "learning_rate": 0.00035279237805128585,
+      "loss": 0.7459,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7303541164927303,
+      "grad_norm": 0.5373327732086182,
+      "learning_rate": 0.000352728508120069,
+      "loss": 1.2578,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7305325127107305,
+      "grad_norm": 0.48414483666419983,
+      "learning_rate": 0.00035266463012059924,
+      "loss": 0.7918,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7307109089287307,
+      "grad_norm": 0.45521920919418335,
+      "learning_rate": 0.00035260074405789365,
+      "loss": 0.8527,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7308893051467309,
+      "grad_norm": 0.44525134563446045,
+      "learning_rate": 0.00035253684993696984,
+      "loss": 0.7479,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7310677013647311,
+      "grad_norm": 0.4846455454826355,
+      "learning_rate": 0.0003524729477628459,
+      "loss": 0.7777,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7312460975827313,
+      "grad_norm": 0.5616981983184814,
+      "learning_rate": 0.00035240903754054075,
+      "loss": 1.0553,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7314244938007314,
+      "grad_norm": 0.5334004163742065,
+      "learning_rate": 0.00035234511927507387,
+      "loss": 0.8418,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7316028900187316,
+      "grad_norm": 0.5169308185577393,
+      "learning_rate": 0.00035228119297146533,
+      "loss": 0.7405,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7317812862367318,
+      "grad_norm": 0.7203896045684814,
+      "learning_rate": 0.00035221725863473596,
+      "loss": 0.8365,
+      "step": 4102
+    },
+    {
+      "epoch": 0.731959682454732,
+      "grad_norm": 0.4874943196773529,
+      "learning_rate": 0.0003521533162699071,
+      "loss": 0.8711,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7321380786727322,
+      "grad_norm": 0.5255458950996399,
+      "learning_rate": 0.0003520893658820007,
+      "loss": 1.1175,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7323164748907324,
+      "grad_norm": 0.4852457642555237,
+      "learning_rate": 0.0003520254074760394,
+      "loss": 0.9222,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7324948711087325,
+      "grad_norm": 0.49769327044487,
+      "learning_rate": 0.00035196144105704654,
+      "loss": 0.8612,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7326732673267327,
+      "grad_norm": 0.528090238571167,
+      "learning_rate": 0.00035189746663004607,
+      "loss": 0.8552,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7328516635447329,
+      "grad_norm": 0.5445149540901184,
+      "learning_rate": 0.00035183348420006233,
+      "loss": 0.9377,
+      "step": 4108
+    },
+    {
+      "epoch": 0.733030059762733,
+      "grad_norm": 0.49806612730026245,
+      "learning_rate": 0.00035176949377212045,
+      "loss": 0.8447,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7332084559807333,
+      "grad_norm": 0.5145043134689331,
+      "learning_rate": 0.00035170549535124647,
+      "loss": 0.8682,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7333868521987333,
+      "grad_norm": 0.45539185404777527,
+      "learning_rate": 0.0003516414889424666,
+      "loss": 0.7736,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7335652484167335,
+      "grad_norm": 0.5117284059524536,
+      "learning_rate": 0.00035157747455080794,
+      "loss": 0.8903,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7337436446347337,
+      "grad_norm": 0.5072416663169861,
+      "learning_rate": 0.0003515134521812983,
+      "loss": 0.8481,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7339220408527339,
+      "grad_norm": 0.46144771575927734,
+      "learning_rate": 0.0003514494218389656,
+      "loss": 0.7497,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7341004370707341,
+      "grad_norm": 0.4636537432670593,
+      "learning_rate": 0.000351385383528839,
+      "loss": 0.9987,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7342788332887343,
+      "grad_norm": 1.1357630491256714,
+      "learning_rate": 0.00035132133725594803,
+      "loss": 1.2211,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7344572295067344,
+      "grad_norm": 0.48777201771736145,
+      "learning_rate": 0.0003512572830253228,
+      "loss": 0.9007,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7346356257247346,
+      "grad_norm": 0.5612165331840515,
+      "learning_rate": 0.0003511932208419942,
+      "loss": 0.9013,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7348140219427348,
+      "grad_norm": 0.6152064204216003,
+      "learning_rate": 0.00035112915071099354,
+      "loss": 1.0285,
+      "step": 4119
+    },
+    {
+      "epoch": 0.734992418160735,
+      "grad_norm": 0.6013907790184021,
+      "learning_rate": 0.000351065072637353,
+      "loss": 0.8736,
+      "step": 4120
+    },
+    {
+      "epoch": 0.7351708143787352,
+      "grad_norm": 0.5271583199501038,
+      "learning_rate": 0.000351000986626105,
+      "loss": 1.1314,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7353492105967353,
+      "grad_norm": 0.583275556564331,
+      "learning_rate": 0.00035093689268228306,
+      "loss": 0.8167,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7355276068147355,
+      "grad_norm": 0.49837273359298706,
+      "learning_rate": 0.000350872790810921,
+      "loss": 0.9489,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7357060030327357,
+      "grad_norm": 0.4844464957714081,
+      "learning_rate": 0.0003508086810170533,
+      "loss": 0.9135,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7358843992507359,
+      "grad_norm": 0.5695939660072327,
+      "learning_rate": 0.00035074456330571517,
+      "loss": 0.9493,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7360627954687361,
+      "grad_norm": 0.5401460528373718,
+      "learning_rate": 0.0003506804376819425,
+      "loss": 0.9202,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7362411916867363,
+      "grad_norm": 0.4903864860534668,
+      "learning_rate": 0.0003506163041507715,
+      "loss": 1.0455,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7364195879047364,
+      "grad_norm": 0.7425463795661926,
+      "learning_rate": 0.00035055216271723933,
+      "loss": 0.9451,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7365979841227366,
+      "grad_norm": 0.5137443542480469,
+      "learning_rate": 0.0003504880133863835,
+      "loss": 1.0321,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7367763803407368,
+      "grad_norm": 0.5113863348960876,
+      "learning_rate": 0.0003504238561632424,
+      "loss": 1.0488,
+      "step": 4130
+    },
+    {
+      "epoch": 0.736954776558737,
+      "grad_norm": 0.469612717628479,
+      "learning_rate": 0.0003503596910528548,
+      "loss": 0.897,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7371331727767372,
+      "grad_norm": 0.5417755246162415,
+      "learning_rate": 0.00035029551806026025,
+      "loss": 0.9615,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7373115689947373,
+      "grad_norm": 2.0405690670013428,
+      "learning_rate": 0.00035023133719049894,
+      "loss": 0.8717,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7374899652127375,
+      "grad_norm": 0.4939228594303131,
+      "learning_rate": 0.00035016714844861155,
+      "loss": 0.7748,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7376683614307377,
+      "grad_norm": 0.5170210003852844,
+      "learning_rate": 0.00035010295183963936,
+      "loss": 0.9805,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7378467576487379,
+      "grad_norm": 0.5120583772659302,
+      "learning_rate": 0.0003500387473686244,
+      "loss": 0.7109,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7380251538667381,
+      "grad_norm": 0.5130133628845215,
+      "learning_rate": 0.0003499745350406093,
+      "loss": 0.9658,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7382035500847383,
+      "grad_norm": 0.4344574511051178,
+      "learning_rate": 0.0003499103148606372,
+      "loss": 0.783,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7383819463027383,
+      "grad_norm": 0.5518962144851685,
+      "learning_rate": 0.00034984608683375206,
+      "loss": 1.0517,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7385603425207385,
+      "grad_norm": 0.46893310546875,
+      "learning_rate": 0.00034978185096499814,
+      "loss": 1.0466,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7387387387387387,
+      "grad_norm": 0.566422164440155,
+      "learning_rate": 0.00034971760725942063,
+      "loss": 1.1278,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7389171349567389,
+      "grad_norm": 0.47913649678230286,
+      "learning_rate": 0.00034965335572206515,
+      "loss": 0.965,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7390955311747391,
+      "grad_norm": 0.4865665137767792,
+      "learning_rate": 0.000349589096357978,
+      "loss": 0.832,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7392739273927392,
+      "grad_norm": 0.7697198987007141,
+      "learning_rate": 0.0003495248291722061,
+      "loss": 0.9949,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7394523236107394,
+      "grad_norm": 0.49303874373435974,
+      "learning_rate": 0.00034946055416979686,
+      "loss": 1.1002,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7396307198287396,
+      "grad_norm": 0.4551609456539154,
+      "learning_rate": 0.00034939627135579854,
+      "loss": 0.826,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7398091160467398,
+      "grad_norm": 0.5011882781982422,
+      "learning_rate": 0.00034933198073525986,
+      "loss": 0.9986,
+      "step": 4147
+    },
+    {
+      "epoch": 0.73998751226474,
+      "grad_norm": 0.4582299590110779,
+      "learning_rate": 0.0003492676823132301,
+      "loss": 0.8219,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7401659084827402,
+      "grad_norm": 0.4924689531326294,
+      "learning_rate": 0.00034920337609475936,
+      "loss": 1.0327,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7403443047007403,
+      "grad_norm": 0.5348846912384033,
+      "learning_rate": 0.00034913906208489814,
+      "loss": 1.0617,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7405227009187405,
+      "grad_norm": 0.4941270649433136,
+      "learning_rate": 0.0003490747402886977,
+      "loss": 0.8433,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7407010971367407,
+      "grad_norm": 0.5572370290756226,
+      "learning_rate": 0.0003490104107112097,
+      "loss": 0.9576,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7408794933547409,
+      "grad_norm": 0.5385303497314453,
+      "learning_rate": 0.00034894607335748674,
+      "loss": 0.9523,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7410578895727411,
+      "grad_norm": 0.5019118189811707,
+      "learning_rate": 0.00034888172823258165,
+      "loss": 0.6984,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7412362857907412,
+      "grad_norm": 0.4633517861366272,
+      "learning_rate": 0.0003488173753415482,
+      "loss": 0.8653,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7414146820087414,
+      "grad_norm": 0.5317572355270386,
+      "learning_rate": 0.0003487530146894407,
+      "loss": 1.1626,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7415930782267416,
+      "grad_norm": 0.45128071308135986,
+      "learning_rate": 0.0003486886462813138,
+      "loss": 0.8697,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7417714744447418,
+      "grad_norm": 0.4344140589237213,
+      "learning_rate": 0.0003486242701222232,
+      "loss": 0.7773,
+      "step": 4158
+    },
+    {
+      "epoch": 0.741949870662742,
+      "grad_norm": 0.443740576505661,
+      "learning_rate": 0.0003485598862172248,
+      "loss": 0.8047,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7421282668807422,
+      "grad_norm": 0.4513658583164215,
+      "learning_rate": 0.00034849549457137543,
+      "loss": 0.8605,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7423066630987423,
+      "grad_norm": 0.4627962112426758,
+      "learning_rate": 0.00034843109518973225,
+      "loss": 0.768,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7424850593167425,
+      "grad_norm": 0.5590542554855347,
+      "learning_rate": 0.00034836668807735314,
+      "loss": 0.9791,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7426634555347427,
+      "grad_norm": 0.45700347423553467,
+      "learning_rate": 0.00034830227323929674,
+      "loss": 0.8522,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7428418517527429,
+      "grad_norm": 2.536508798599243,
+      "learning_rate": 0.00034823785068062213,
+      "loss": 0.9045,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7430202479707431,
+      "grad_norm": 0.5341298580169678,
+      "learning_rate": 0.00034817342040638897,
+      "loss": 0.9195,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7431986441887432,
+      "grad_norm": 0.45887595415115356,
+      "learning_rate": 0.00034810898242165766,
+      "loss": 0.8731,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7433770404067434,
+      "grad_norm": 0.4361880421638489,
+      "learning_rate": 0.00034804453673148905,
+      "loss": 0.8234,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7435554366247435,
+      "grad_norm": 0.4275573790073395,
+      "learning_rate": 0.0003479800833409448,
+      "loss": 0.7071,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7437338328427437,
+      "grad_norm": 0.5099348425865173,
+      "learning_rate": 0.0003479156222550869,
+      "loss": 0.8876,
+      "step": 4169
+    },
+    {
+      "epoch": 0.743912229060744,
+      "grad_norm": 0.4776555895805359,
+      "learning_rate": 0.00034785115347897805,
+      "loss": 0.7928,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7440906252787441,
+      "grad_norm": 63.1339225769043,
+      "learning_rate": 0.00034778667701768187,
+      "loss": 1.9367,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7442690214967442,
+      "grad_norm": 0.5373305678367615,
+      "learning_rate": 0.00034772219287626207,
+      "loss": 0.9005,
+      "step": 4172
+    },
+    {
+      "epoch": 0.7444474177147444,
+      "grad_norm": 0.5590469837188721,
+      "learning_rate": 0.0003476577010597834,
+      "loss": 0.8572,
+      "step": 4173
+    },
+    {
+      "epoch": 0.7446258139327446,
+      "grad_norm": 0.4844629466533661,
+      "learning_rate": 0.0003475932015733109,
+      "loss": 0.9525,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7448042101507448,
+      "grad_norm": 0.41167715191841125,
+      "learning_rate": 0.00034752869442191027,
+      "loss": 0.8344,
+      "step": 4175
+    },
+    {
+      "epoch": 0.744982606368745,
+      "grad_norm": 0.5152734518051147,
+      "learning_rate": 0.00034746417961064793,
+      "loss": 1.0936,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7451610025867451,
+      "grad_norm": 0.48306646943092346,
+      "learning_rate": 0.0003473996571445909,
+      "loss": 0.7804,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7453393988047453,
+      "grad_norm": 0.5196247100830078,
+      "learning_rate": 0.0003473351270288067,
+      "loss": 0.9043,
+      "step": 4178
+    },
+    {
+      "epoch": 0.7455177950227455,
+      "grad_norm": 0.5509787201881409,
+      "learning_rate": 0.0003472705892683636,
+      "loss": 0.8701,
+      "step": 4179
+    },
+    {
+      "epoch": 0.7456961912407457,
+      "grad_norm": 0.7097774744033813,
+      "learning_rate": 0.0003472060438683302,
+      "loss": 0.7195,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7458745874587459,
+      "grad_norm": 0.49308526515960693,
+      "learning_rate": 0.00034714149083377594,
+      "loss": 1.0288,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7460529836767461,
+      "grad_norm": 0.6980340480804443,
+      "learning_rate": 0.00034707693016977083,
+      "loss": 0.858,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7462313798947462,
+      "grad_norm": 0.4881076216697693,
+      "learning_rate": 0.0003470123618813854,
+      "loss": 0.9786,
+      "step": 4183
+    },
+    {
+      "epoch": 0.7464097761127464,
+      "grad_norm": 0.5367513298988342,
+      "learning_rate": 0.00034694778597369076,
+      "loss": 0.8505,
+      "step": 4184
+    },
+    {
+      "epoch": 0.7465881723307466,
+      "grad_norm": 0.47334301471710205,
+      "learning_rate": 0.00034688320245175873,
+      "loss": 0.8669,
+      "step": 4185
+    },
+    {
+      "epoch": 0.7467665685487468,
+      "grad_norm": 0.6059718728065491,
+      "learning_rate": 0.0003468186113206617,
+      "loss": 1.1616,
+      "step": 4186
+    },
+    {
+      "epoch": 0.746944964766747,
+      "grad_norm": 0.5258612036705017,
+      "learning_rate": 0.00034675401258547266,
+      "loss": 0.8902,
+      "step": 4187
+    },
+    {
+      "epoch": 0.7471233609847471,
+      "grad_norm": 0.5637246370315552,
+      "learning_rate": 0.00034668940625126506,
+      "loss": 1.2349,
+      "step": 4188
+    },
+    {
+      "epoch": 0.7473017572027473,
+      "grad_norm": 78.5971450805664,
+      "learning_rate": 0.00034662479232311306,
+      "loss": 1.0211,
+      "step": 4189
+    },
+    {
+      "epoch": 0.7474801534207475,
+      "grad_norm": 0.6106373071670532,
+      "learning_rate": 0.00034656017080609154,
+      "loss": 1.0691,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7476585496387477,
+      "grad_norm": 0.5662621259689331,
+      "learning_rate": 0.0003464955417052757,
+      "loss": 0.9236,
+      "step": 4191
+    },
+    {
+      "epoch": 0.7478369458567479,
+      "grad_norm": 0.5042510032653809,
+      "learning_rate": 0.0003464309050257415,
+      "loss": 0.8522,
+      "step": 4192
+    },
+    {
+      "epoch": 0.7480153420747481,
+      "grad_norm": 0.5739179849624634,
+      "learning_rate": 0.0003463662607725656,
+      "loss": 0.9128,
+      "step": 4193
+    },
+    {
+      "epoch": 0.7481937382927482,
+      "grad_norm": 0.4624471962451935,
+      "learning_rate": 0.000346301608950825,
+      "loss": 0.7043,
+      "step": 4194
+    },
+    {
+      "epoch": 0.7483721345107484,
+      "grad_norm": 0.5266431570053101,
+      "learning_rate": 0.00034623694956559747,
+      "loss": 0.7484,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7485505307287486,
+      "grad_norm": 0.5885065197944641,
+      "learning_rate": 0.0003461722826219614,
+      "loss": 1.179,
+      "step": 4196
+    },
+    {
+      "epoch": 0.7487289269467488,
+      "grad_norm": 0.5153014659881592,
+      "learning_rate": 0.0003461076081249956,
+      "loss": 0.7813,
+      "step": 4197
+    },
+    {
+      "epoch": 0.748907323164749,
+      "grad_norm": 0.5379844307899475,
+      "learning_rate": 0.0003460429260797796,
+      "loss": 0.7575,
+      "step": 4198
+    },
+    {
+      "epoch": 0.749085719382749,
+      "grad_norm": 0.502864420413971,
+      "learning_rate": 0.00034597823649139346,
+      "loss": 0.995,
+      "step": 4199
+    },
+    {
+      "epoch": 0.7492641156007492,
+      "grad_norm": 0.4639131724834442,
+      "learning_rate": 0.00034591353936491806,
+      "loss": 0.8029,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7494425118187494,
+      "grad_norm": 0.5059810876846313,
+      "learning_rate": 0.0003458488347054345,
+      "loss": 0.8779,
+      "step": 4201
+    },
+    {
+      "epoch": 0.7496209080367496,
+      "grad_norm": 0.4426543712615967,
+      "learning_rate": 0.00034578412251802466,
+      "loss": 0.7159,
+      "step": 4202
+    },
+    {
+      "epoch": 0.7497993042547498,
+      "grad_norm": 0.571412205696106,
+      "learning_rate": 0.0003457194028077711,
+      "loss": 0.9628,
+      "step": 4203
+    },
+    {
+      "epoch": 0.74997770047275,
+      "grad_norm": 0.5668926239013672,
+      "learning_rate": 0.00034565467557975683,
+      "loss": 1.0421,
+      "step": 4204
+    },
+    {
+      "epoch": 0.7501560966907501,
+      "grad_norm": 0.48620158433914185,
+      "learning_rate": 0.0003455899408390655,
+      "loss": 0.8639,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7503344929087503,
+      "grad_norm": 0.5328919887542725,
+      "learning_rate": 0.0003455251985907814,
+      "loss": 0.9498,
+      "step": 4206
+    },
+    {
+      "epoch": 0.7505128891267505,
+      "grad_norm": 0.5057373642921448,
+      "learning_rate": 0.0003454604488399893,
+      "loss": 0.9308,
+      "step": 4207
+    },
+    {
+      "epoch": 0.7506912853447507,
+      "grad_norm": 0.43592992424964905,
+      "learning_rate": 0.0003453956915917745,
+      "loss": 0.8826,
+      "step": 4208
+    },
+    {
+      "epoch": 0.7508696815627509,
+      "grad_norm": 0.46315765380859375,
+      "learning_rate": 0.00034533092685122324,
+      "loss": 0.9092,
+      "step": 4209
+    },
+    {
+      "epoch": 0.751048077780751,
+      "grad_norm": 0.48972928524017334,
+      "learning_rate": 0.000345266154623422,
+      "loss": 0.9986,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7512264739987512,
+      "grad_norm": 0.49682357907295227,
+      "learning_rate": 0.0003452013749134579,
+      "loss": 0.9315,
+      "step": 4211
+    },
+    {
+      "epoch": 0.7514048702167514,
+      "grad_norm": 0.5060596466064453,
+      "learning_rate": 0.00034513658772641887,
+      "loss": 1.0385,
+      "step": 4212
+    },
+    {
+      "epoch": 0.7515832664347516,
+      "grad_norm": 0.652193546295166,
+      "learning_rate": 0.00034507179306739324,
+      "loss": 0.9741,
+      "step": 4213
+    },
+    {
+      "epoch": 0.7517616626527518,
+      "grad_norm": 0.5387786030769348,
+      "learning_rate": 0.0003450069909414698,
+      "loss": 0.8605,
+      "step": 4214
+    },
+    {
+      "epoch": 0.751940058870752,
+      "grad_norm": 0.5394373536109924,
+      "learning_rate": 0.00034494218135373817,
+      "loss": 0.8344,
+      "step": 4215
+    },
+    {
+      "epoch": 0.7521184550887521,
+      "grad_norm": 0.501558780670166,
+      "learning_rate": 0.00034487736430928846,
+      "loss": 1.0367,
+      "step": 4216
+    },
+    {
+      "epoch": 0.7522968513067523,
+      "grad_norm": 0.49226245284080505,
+      "learning_rate": 0.00034481253981321144,
+      "loss": 0.8282,
+      "step": 4217
+    },
+    {
+      "epoch": 0.7524752475247525,
+      "grad_norm": 0.5037868022918701,
+      "learning_rate": 0.0003447477078705983,
+      "loss": 1.0139,
+      "step": 4218
+    },
+    {
+      "epoch": 0.7526536437427527,
+      "grad_norm": 0.4862633943557739,
+      "learning_rate": 0.00034468286848654106,
+      "loss": 0.8863,
+      "step": 4219
+    },
+    {
+      "epoch": 0.7528320399607529,
+      "grad_norm": 0.4658646881580353,
+      "learning_rate": 0.000344618021666132,
+      "loss": 0.8069,
+      "step": 4220
+    },
+    {
+      "epoch": 0.753010436178753,
+      "grad_norm": 0.44514206051826477,
+      "learning_rate": 0.0003445531674144642,
+      "loss": 1.0605,
+      "step": 4221
+    },
+    {
+      "epoch": 0.7531888323967532,
+      "grad_norm": 0.5069167613983154,
+      "learning_rate": 0.0003444883057366314,
+      "loss": 0.8131,
+      "step": 4222
+    },
+    {
+      "epoch": 0.7533672286147534,
+      "grad_norm": 0.48138898611068726,
+      "learning_rate": 0.00034442343663772755,
+      "loss": 0.8838,
+      "step": 4223
+    },
+    {
+      "epoch": 0.7535456248327536,
+      "grad_norm": 0.46832725405693054,
+      "learning_rate": 0.0003443585601228478,
+      "loss": 1.0179,
+      "step": 4224
+    },
+    {
+      "epoch": 0.7537240210507538,
+      "grad_norm": 0.5185703635215759,
+      "learning_rate": 0.00034429367619708733,
+      "loss": 0.8979,
+      "step": 4225
+    },
+    {
+      "epoch": 0.753902417268754,
+      "grad_norm": 0.5143477916717529,
+      "learning_rate": 0.0003442287848655421,
+      "loss": 0.8142,
+      "step": 4226
+    },
+    {
+      "epoch": 0.754080813486754,
+      "grad_norm": 0.5070206522941589,
+      "learning_rate": 0.00034416388613330864,
+      "loss": 1.0616,
+      "step": 4227
+    },
+    {
+      "epoch": 0.7542592097047542,
+      "grad_norm": 0.5132788419723511,
+      "learning_rate": 0.00034409898000548403,
+      "loss": 1.0537,
+      "step": 4228
+    },
+    {
+      "epoch": 0.7544376059227544,
+      "grad_norm": 0.4766051173210144,
+      "learning_rate": 0.00034403406648716604,
+      "loss": 0.9183,
+      "step": 4229
+    },
+    {
+      "epoch": 0.7546160021407546,
+      "grad_norm": 0.6825383305549622,
+      "learning_rate": 0.00034396914558345297,
+      "loss": 1.0334,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7547943983587548,
+      "grad_norm": 0.4257564842700958,
+      "learning_rate": 0.0003439042172994436,
+      "loss": 0.75,
+      "step": 4231
+    },
+    {
+      "epoch": 0.7549727945767549,
+      "grad_norm": 0.5030829906463623,
+      "learning_rate": 0.0003438392816402375,
+      "loss": 1.0223,
+      "step": 4232
+    },
+    {
+      "epoch": 0.7551511907947551,
+      "grad_norm": 0.4708796441555023,
+      "learning_rate": 0.00034377433861093457,
+      "loss": 0.7668,
+      "step": 4233
+    },
+    {
+      "epoch": 0.7553295870127553,
+      "grad_norm": 0.49385902285575867,
+      "learning_rate": 0.0003437093882166354,
+      "loss": 0.7773,
+      "step": 4234
+    },
+    {
+      "epoch": 0.7555079832307555,
+      "grad_norm": 0.5581505298614502,
+      "learning_rate": 0.00034364443046244124,
+      "loss": 0.9124,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7556863794487557,
+      "grad_norm": 0.4790385067462921,
+      "learning_rate": 0.0003435794653534538,
+      "loss": 0.8304,
+      "step": 4236
+    },
+    {
+      "epoch": 0.7558647756667559,
+      "grad_norm": 0.43401768803596497,
+      "learning_rate": 0.00034351449289477545,
+      "loss": 0.77,
+      "step": 4237
+    },
+    {
+      "epoch": 0.756043171884756,
+      "grad_norm": 0.5333930253982544,
+      "learning_rate": 0.000343449513091509,
+      "loss": 0.8807,
+      "step": 4238
+    },
+    {
+      "epoch": 0.7562215681027562,
+      "grad_norm": 0.4971325993537903,
+      "learning_rate": 0.0003433845259487581,
+      "loss": 0.8195,
+      "step": 4239
+    },
+    {
+      "epoch": 0.7563999643207564,
+      "grad_norm": 0.4987589716911316,
+      "learning_rate": 0.00034331953147162666,
+      "loss": 0.9902,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7565783605387566,
+      "grad_norm": 0.45375093817710876,
+      "learning_rate": 0.0003432545296652194,
+      "loss": 0.9725,
+      "step": 4241
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 1.4552748203277588,
+      "learning_rate": 0.00034318952053464147,
+      "loss": 0.8451,
+      "step": 4242
+    },
+    {
+      "epoch": 0.7569351529747569,
+      "grad_norm": 0.44463515281677246,
+      "learning_rate": 0.0003431245040849987,
+      "loss": 0.7589,
+      "step": 4243
+    },
+    {
+      "epoch": 0.7571135491927571,
+      "grad_norm": 0.5243865251541138,
+      "learning_rate": 0.00034305948032139745,
+      "loss": 0.776,
+      "step": 4244
+    },
+    {
+      "epoch": 0.7572919454107573,
+      "grad_norm": 14.643073081970215,
+      "learning_rate": 0.00034299444924894474,
+      "loss": 0.9523,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7574703416287575,
+      "grad_norm": 0.5148363709449768,
+      "learning_rate": 0.00034292941087274794,
+      "loss": 0.8977,
+      "step": 4246
+    },
+    {
+      "epoch": 0.7576487378467577,
+      "grad_norm": 0.5748273134231567,
+      "learning_rate": 0.0003428643651979152,
+      "loss": 0.9339,
+      "step": 4247
+    },
+    {
+      "epoch": 0.7578271340647579,
+      "grad_norm": 0.5037097930908203,
+      "learning_rate": 0.00034279931222955517,
+      "loss": 0.8882,
+      "step": 4248
+    },
+    {
+      "epoch": 0.758005530282758,
+      "grad_norm": 0.5515426397323608,
+      "learning_rate": 0.00034273425197277715,
+      "loss": 1.113,
+      "step": 4249
+    },
+    {
+      "epoch": 0.7581839265007582,
+      "grad_norm": 0.49536243081092834,
+      "learning_rate": 0.00034266918443269083,
+      "loss": 0.7182,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7583623227187584,
+      "grad_norm": 0.49721240997314453,
+      "learning_rate": 0.0003426041096144067,
+      "loss": 1.0869,
+      "step": 4251
+    },
+    {
+      "epoch": 0.7585407189367586,
+      "grad_norm": 0.48414304852485657,
+      "learning_rate": 0.0003425390275230356,
+      "loss": 0.8055,
+      "step": 4252
+    },
+    {
+      "epoch": 0.7587191151547588,
+      "grad_norm": 0.48049938678741455,
+      "learning_rate": 0.00034247393816368914,
+      "loss": 0.7842,
+      "step": 4253
+    },
+    {
+      "epoch": 0.7588975113727588,
+      "grad_norm": 0.5838775634765625,
+      "learning_rate": 0.00034240884154147934,
+      "loss": 0.9039,
+      "step": 4254
+    },
+    {
+      "epoch": 0.759075907590759,
+      "grad_norm": 0.5820823907852173,
+      "learning_rate": 0.0003423437376615189,
+      "loss": 1.0101,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7592543038087592,
+      "grad_norm": 0.6763784289360046,
+      "learning_rate": 0.00034227862652892103,
+      "loss": 1.065,
+      "step": 4256
+    },
+    {
+      "epoch": 0.7594327000267594,
+      "grad_norm": 0.5086767077445984,
+      "learning_rate": 0.0003422135081487996,
+      "loss": 1.1122,
+      "step": 4257
+    },
+    {
+      "epoch": 0.7596110962447596,
+      "grad_norm": 0.4516467750072479,
+      "learning_rate": 0.0003421483825262688,
+      "loss": 0.785,
+      "step": 4258
+    },
+    {
+      "epoch": 0.7597894924627598,
+      "grad_norm": 0.6000534892082214,
+      "learning_rate": 0.0003420832496664439,
+      "loss": 1.013,
+      "step": 4259
+    },
+    {
+      "epoch": 0.7599678886807599,
+      "grad_norm": 0.4857349395751953,
+      "learning_rate": 0.00034201810957444,
+      "loss": 1.0669,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7601462848987601,
+      "grad_norm": 0.5024265646934509,
+      "learning_rate": 0.0003419529622553735,
+      "loss": 0.9337,
+      "step": 4261
+    },
+    {
+      "epoch": 0.7603246811167603,
+      "grad_norm": 0.4981759190559387,
+      "learning_rate": 0.0003418878077143608,
+      "loss": 1.0209,
+      "step": 4262
+    },
+    {
+      "epoch": 0.7605030773347605,
+      "grad_norm": 0.4379788637161255,
+      "learning_rate": 0.00034182264595651927,
+      "loss": 0.668,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7606814735527607,
+      "grad_norm": 0.5320781469345093,
+      "learning_rate": 0.0003417574769869666,
+      "loss": 0.8506,
+      "step": 4264
+    },
+    {
+      "epoch": 0.7608598697707608,
+      "grad_norm": 0.5261650085449219,
+      "learning_rate": 0.0003416923008108213,
+      "loss": 1.0954,
+      "step": 4265
+    },
+    {
+      "epoch": 0.761038265988761,
+      "grad_norm": 0.47511598467826843,
+      "learning_rate": 0.00034162711743320205,
+      "loss": 0.9106,
+      "step": 4266
+    },
+    {
+      "epoch": 0.7612166622067612,
+      "grad_norm": 0.4774247109889984,
+      "learning_rate": 0.00034156192685922846,
+      "loss": 0.7263,
+      "step": 4267
+    },
+    {
+      "epoch": 0.7613950584247614,
+      "grad_norm": 0.46322304010391235,
+      "learning_rate": 0.00034149672909402056,
+      "loss": 0.7982,
+      "step": 4268
+    },
+    {
+      "epoch": 0.7615734546427616,
+      "grad_norm": 0.48991408944129944,
+      "learning_rate": 0.00034143152414269887,
+      "loss": 0.8453,
+      "step": 4269
+    },
+    {
+      "epoch": 0.7617518508607618,
+      "grad_norm": 0.47439679503440857,
+      "learning_rate": 0.00034136631201038466,
+      "loss": 0.8049,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7619302470787619,
+      "grad_norm": 0.4984125792980194,
+      "learning_rate": 0.0003413010927021996,
+      "loss": 0.8084,
+      "step": 4271
+    },
+    {
+      "epoch": 0.7621086432967621,
+      "grad_norm": 0.547827959060669,
+      "learning_rate": 0.0003412358662232661,
+      "loss": 0.9195,
+      "step": 4272
+    },
+    {
+      "epoch": 0.7622870395147623,
+      "grad_norm": 0.4370548725128174,
+      "learning_rate": 0.0003411706325787068,
+      "loss": 0.7599,
+      "step": 4273
+    },
+    {
+      "epoch": 0.7624654357327625,
+      "grad_norm": 0.47955775260925293,
+      "learning_rate": 0.00034110539177364534,
+      "loss": 0.8049,
+      "step": 4274
+    },
+    {
+      "epoch": 0.7626438319507627,
+      "grad_norm": 0.5160841941833496,
+      "learning_rate": 0.00034104014381320557,
+      "loss": 0.8973,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7628222281687628,
+      "grad_norm": 0.506690502166748,
+      "learning_rate": 0.0003409748887025121,
+      "loss": 1.0149,
+      "step": 4276
+    },
+    {
+      "epoch": 0.763000624386763,
+      "grad_norm": 0.5552113652229309,
+      "learning_rate": 0.00034090962644669,
+      "loss": 1.028,
+      "step": 4277
+    },
+    {
+      "epoch": 0.7631790206047632,
+      "grad_norm": 0.45480722188949585,
+      "learning_rate": 0.000340844357050865,
+      "loss": 0.7717,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7633574168227634,
+      "grad_norm": 0.5435570478439331,
+      "learning_rate": 0.0003407790805201633,
+      "loss": 1.2671,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7635358130407636,
+      "grad_norm": 0.43579912185668945,
+      "learning_rate": 0.0003407137968597116,
+      "loss": 0.8684,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7637142092587638,
+      "grad_norm": 0.4706798791885376,
+      "learning_rate": 0.00034064850607463736,
+      "loss": 0.9247,
+      "step": 4281
+    },
+    {
+      "epoch": 0.7638926054767639,
+      "grad_norm": 0.4617820084095001,
+      "learning_rate": 0.0003405832081700685,
+      "loss": 0.8273,
+      "step": 4282
+    },
+    {
+      "epoch": 0.764071001694764,
+      "grad_norm": 0.49580156803131104,
+      "learning_rate": 0.0003405179031511334,
+      "loss": 0.9895,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7642493979127643,
+      "grad_norm": 0.4556731879711151,
+      "learning_rate": 0.00034045259102296124,
+      "loss": 0.8607,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7644277941307644,
+      "grad_norm": 0.461270272731781,
+      "learning_rate": 0.0003403872717906814,
+      "loss": 0.8178,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7646061903487646,
+      "grad_norm": 0.4924364686012268,
+      "learning_rate": 0.00034032194545942417,
+      "loss": 0.8912,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7647845865667647,
+      "grad_norm": 0.5016777515411377,
+      "learning_rate": 0.00034025661203432024,
+      "loss": 1.1841,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7649629827847649,
+      "grad_norm": 0.48156607151031494,
+      "learning_rate": 0.0003401912715205008,
+      "loss": 0.7984,
+      "step": 4288
+    },
+    {
+      "epoch": 0.7651413790027651,
+      "grad_norm": 0.4831927716732025,
+      "learning_rate": 0.00034012592392309774,
+      "loss": 0.8091,
+      "step": 4289
+    },
+    {
+      "epoch": 0.7653197752207653,
+      "grad_norm": 0.48947733640670776,
+      "learning_rate": 0.0003400605692472433,
+      "loss": 0.8708,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7654981714387655,
+      "grad_norm": 0.5063187479972839,
+      "learning_rate": 0.0003399952074980706,
+      "loss": 0.8559,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7656765676567657,
+      "grad_norm": 0.5040058493614197,
+      "learning_rate": 0.00033992983868071303,
+      "loss": 0.8416,
+      "step": 4292
+    },
+    {
+      "epoch": 0.7658549638747658,
+      "grad_norm": 0.5163808465003967,
+      "learning_rate": 0.0003398644628003046,
+      "loss": 0.8919,
+      "step": 4293
+    },
+    {
+      "epoch": 0.766033360092766,
+      "grad_norm": 0.5070242881774902,
+      "learning_rate": 0.00033979907986197993,
+      "loss": 0.8687,
+      "step": 4294
+    },
+    {
+      "epoch": 0.7662117563107662,
+      "grad_norm": 0.5390428900718689,
+      "learning_rate": 0.00033973368987087423,
+      "loss": 0.9808,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7663901525287664,
+      "grad_norm": 0.4794482886791229,
+      "learning_rate": 0.0003396682928321231,
+      "loss": 0.9719,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7665685487467666,
+      "grad_norm": 0.482950896024704,
+      "learning_rate": 0.0003396028887508628,
+      "loss": 0.7459,
+      "step": 4297
+    },
+    {
+      "epoch": 0.7667469449647667,
+      "grad_norm": 0.5277696251869202,
+      "learning_rate": 0.00033953747763223026,
+      "loss": 1.1243,
+      "step": 4298
+    },
+    {
+      "epoch": 0.7669253411827669,
+      "grad_norm": 0.4380790889263153,
+      "learning_rate": 0.0003394720594813627,
+      "loss": 0.8427,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7671037374007671,
+      "grad_norm": 0.5023663640022278,
+      "learning_rate": 0.0003394066343033981,
+      "loss": 0.8703,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7672821336187673,
+      "grad_norm": 0.48462873697280884,
+      "learning_rate": 0.00033934120210347496,
+      "loss": 0.8443,
+      "step": 4301
+    },
+    {
+      "epoch": 0.7674605298367675,
+      "grad_norm": 1.0551936626434326,
+      "learning_rate": 0.0003392757628867322,
+      "loss": 0.7749,
+      "step": 4302
+    },
+    {
+      "epoch": 0.7676389260547677,
+      "grad_norm": 0.6180291771888733,
+      "learning_rate": 0.0003392103166583095,
+      "loss": 1.1561,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7678173222727678,
+      "grad_norm": 0.5031440854072571,
+      "learning_rate": 0.0003391448634233468,
+      "loss": 0.8814,
+      "step": 4304
+    },
+    {
+      "epoch": 0.767995718490768,
+      "grad_norm": 0.4589061737060547,
+      "learning_rate": 0.00033907940318698504,
+      "loss": 0.9982,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7681741147087682,
+      "grad_norm": 0.46637728810310364,
+      "learning_rate": 0.00033901393595436527,
+      "loss": 0.8032,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7683525109267684,
+      "grad_norm": 0.5185346007347107,
+      "learning_rate": 0.00033894846173062915,
+      "loss": 1.0561,
+      "step": 4307
+    },
+    {
+      "epoch": 0.7685309071447686,
+      "grad_norm": 0.43067285418510437,
+      "learning_rate": 0.00033888298052091916,
+      "loss": 0.7942,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7687093033627687,
+      "grad_norm": 0.6915571689605713,
+      "learning_rate": 0.00033881749233037817,
+      "loss": 1.0712,
+      "step": 4309
+    },
+    {
+      "epoch": 0.7688876995807689,
+      "grad_norm": 0.4629500210285187,
+      "learning_rate": 0.0003387519971641495,
+      "loss": 0.7828,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7690660957987691,
+      "grad_norm": 0.65446937084198,
+      "learning_rate": 0.00033868649502737726,
+      "loss": 0.8419,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7692444920167693,
+      "grad_norm": 0.4821755588054657,
+      "learning_rate": 0.0003386209859252058,
+      "loss": 0.9422,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7694228882347695,
+      "grad_norm": 3.9539215564727783,
+      "learning_rate": 0.0003385554698627803,
+      "loss": 0.7106,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7696012844527697,
+      "grad_norm": 0.5639221668243408,
+      "learning_rate": 0.00033848994684524623,
+      "loss": 0.8801,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7697796806707697,
+      "grad_norm": 0.4910728931427002,
+      "learning_rate": 0.0003384244168777498,
+      "loss": 0.8116,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7699580768887699,
+      "grad_norm": 0.5648394823074341,
+      "learning_rate": 0.0003383588799654378,
+      "loss": 0.9331,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7701364731067701,
+      "grad_norm": 0.5196200609207153,
+      "learning_rate": 0.00033829333611345736,
+      "loss": 0.8242,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7703148693247703,
+      "grad_norm": 0.4403199553489685,
+      "learning_rate": 0.0003382277853269564,
+      "loss": 0.8163,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7704932655427705,
+      "grad_norm": 0.7118186950683594,
+      "learning_rate": 0.000338162227611083,
+      "loss": 0.8899,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7706716617607706,
+      "grad_norm": 0.8601351976394653,
+      "learning_rate": 0.00033809666297098624,
+      "loss": 1.0249,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7708500579787708,
+      "grad_norm": 0.8977810144424438,
+      "learning_rate": 0.0003380310914118155,
+      "loss": 0.771,
+      "step": 4321
+    },
+    {
+      "epoch": 0.771028454196771,
+      "grad_norm": 0.5474297404289246,
+      "learning_rate": 0.0003379655129387207,
+      "loss": 0.9783,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7712068504147712,
+      "grad_norm": 6.160559177398682,
+      "learning_rate": 0.0003378999275568523,
+      "loss": 1.1071,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7713852466327714,
+      "grad_norm": 0.6195570230484009,
+      "learning_rate": 0.0003378343352713614,
+      "loss": 0.8238,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7715636428507716,
+      "grad_norm": 0.5602040886878967,
+      "learning_rate": 0.00033776873608739976,
+      "loss": 0.7919,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7717420390687717,
+      "grad_norm": 0.5972548127174377,
+      "learning_rate": 0.00033770313001011933,
+      "loss": 0.9286,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7719204352867719,
+      "grad_norm": 0.5935573577880859,
+      "learning_rate": 0.0003376375170446727,
+      "loss": 0.9524,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7720988315047721,
+      "grad_norm": 0.7750356197357178,
+      "learning_rate": 0.00033757189719621326,
+      "loss": 0.8514,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7722772277227723,
+      "grad_norm": 0.5310060381889343,
+      "learning_rate": 0.00033750627046989475,
+      "loss": 0.8609,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7724556239407725,
+      "grad_norm": 0.5345288515090942,
+      "learning_rate": 0.00033744063687087136,
+      "loss": 0.9829,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7726340201587726,
+      "grad_norm": 0.50868159532547,
+      "learning_rate": 0.000337374996404298,
+      "loss": 0.9928,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7728124163767728,
+      "grad_norm": 0.46974533796310425,
+      "learning_rate": 0.00033730934907532994,
+      "loss": 0.832,
+      "step": 4332
+    },
+    {
+      "epoch": 0.772990812594773,
+      "grad_norm": 0.4263937771320343,
+      "learning_rate": 0.0003372436948891233,
+      "loss": 0.7312,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7731692088127732,
+      "grad_norm": 0.5079303979873657,
+      "learning_rate": 0.0003371780338508343,
+      "loss": 0.9124,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7733476050307734,
+      "grad_norm": 0.5083035826683044,
+      "learning_rate": 0.00033711236596562004,
+      "loss": 0.8987,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7735260012487736,
+      "grad_norm": 0.4942117929458618,
+      "learning_rate": 0.00033704669123863813,
+      "loss": 1.0008,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7737043974667737,
+      "grad_norm": 0.5014617443084717,
+      "learning_rate": 0.00033698100967504655,
+      "loss": 0.6739,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7738827936847739,
+      "grad_norm": 0.4389418065547943,
+      "learning_rate": 0.0003369153212800038,
+      "loss": 0.6625,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7740611899027741,
+      "grad_norm": 0.4890193045139313,
+      "learning_rate": 0.0003368496260586692,
+      "loss": 0.7249,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7742395861207743,
+      "grad_norm": 0.49702176451683044,
+      "learning_rate": 0.00033678392401620226,
+      "loss": 0.8034,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7744179823387745,
+      "grad_norm": 0.5406886339187622,
+      "learning_rate": 0.00033671821515776336,
+      "loss": 0.9704,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7745963785567745,
+      "grad_norm": 0.460330992937088,
+      "learning_rate": 0.00033665249948851316,
+      "loss": 0.6877,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7747747747747747,
+      "grad_norm": 0.4634692370891571,
+      "learning_rate": 0.0003365867770136129,
+      "loss": 0.7982,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7749531709927749,
+      "grad_norm": 0.5319910049438477,
+      "learning_rate": 0.00033652104773822445,
+      "loss": 1.0181,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7751315672107751,
+      "grad_norm": 0.48047569394111633,
+      "learning_rate": 0.00033645531166751015,
+      "loss": 0.9439,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7753099634287753,
+      "grad_norm": 0.4725799858570099,
+      "learning_rate": 0.00033638956880663285,
+      "loss": 0.8295,
+      "step": 4346
+    },
+    {
+      "epoch": 0.7754883596467755,
+      "grad_norm": 0.4927878975868225,
+      "learning_rate": 0.000336323819160756,
+      "loss": 0.8699,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7756667558647756,
+      "grad_norm": 0.5209059715270996,
+      "learning_rate": 0.00033625806273504354,
+      "loss": 0.8208,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7758451520827758,
+      "grad_norm": 0.4795133173465729,
+      "learning_rate": 0.00033619229953465996,
+      "loss": 0.8273,
+      "step": 4349
+    },
+    {
+      "epoch": 0.776023548300776,
+      "grad_norm": 0.5821531414985657,
+      "learning_rate": 0.0003361265295647703,
+      "loss": 0.8388,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7762019445187762,
+      "grad_norm": 0.5282803177833557,
+      "learning_rate": 0.00033606075283054005,
+      "loss": 0.6619,
+      "step": 4351
+    },
+    {
+      "epoch": 0.7763803407367764,
+      "grad_norm": 0.6072514653205872,
+      "learning_rate": 0.00033599496933713535,
+      "loss": 1.0466,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7765587369547765,
+      "grad_norm": 0.4529389441013336,
+      "learning_rate": 0.0003359291790897227,
+      "loss": 0.8325,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7767371331727767,
+      "grad_norm": 0.47834986448287964,
+      "learning_rate": 0.0003358633820934692,
+      "loss": 0.9251,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7769155293907769,
+      "grad_norm": 0.5303228497505188,
+      "learning_rate": 0.0003357975783535428,
+      "loss": 1.1421,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7770939256087771,
+      "grad_norm": 0.5310399532318115,
+      "learning_rate": 0.00033573176787511145,
+      "loss": 0.8054,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7772723218267773,
+      "grad_norm": 0.46372997760772705,
+      "learning_rate": 0.0003356659506633439,
+      "loss": 1.0177,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7774507180447775,
+      "grad_norm": 0.45827627182006836,
+      "learning_rate": 0.00033560012672340957,
+      "loss": 0.7673,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7776291142627776,
+      "grad_norm": 0.5642030835151672,
+      "learning_rate": 0.000335534296060478,
+      "loss": 0.8029,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7778075104807778,
+      "grad_norm": 0.5288789868354797,
+      "learning_rate": 0.00033546845867971976,
+      "loss": 1.0097,
+      "step": 4360
+    },
+    {
+      "epoch": 0.777985906698778,
+      "grad_norm": 0.5099871158599854,
+      "learning_rate": 0.0003354026145863054,
+      "loss": 0.8843,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7781643029167782,
+      "grad_norm": 0.4645063579082489,
+      "learning_rate": 0.0003353367637854065,
+      "loss": 0.8116,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7783426991347784,
+      "grad_norm": 0.48817193508148193,
+      "learning_rate": 0.00033527090628219494,
+      "loss": 0.7076,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7785210953527785,
+      "grad_norm": 0.4992578625679016,
+      "learning_rate": 0.00033520504208184304,
+      "loss": 0.9631,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7786994915707787,
+      "grad_norm": 0.4915018081665039,
+      "learning_rate": 0.00033513917118952385,
+      "loss": 0.8473,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7788778877887789,
+      "grad_norm": 1.4712707996368408,
+      "learning_rate": 0.0003350732936104108,
+      "loss": 1.0,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7790562840067791,
+      "grad_norm": 0.5193944573402405,
+      "learning_rate": 0.0003350074093496778,
+      "loss": 1.0957,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7792346802247793,
+      "grad_norm": 0.4992116689682007,
+      "learning_rate": 0.0003349415184124995,
+      "loss": 0.9401,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7794130764427795,
+      "grad_norm": 0.512789785861969,
+      "learning_rate": 0.00033487562080405085,
+      "loss": 0.915,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7795914726607795,
+      "grad_norm": 1.0886067152023315,
+      "learning_rate": 0.00033480971652950753,
+      "loss": 0.9546,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7797698688787797,
+      "grad_norm": 0.49638882279396057,
+      "learning_rate": 0.0003347438055940456,
+      "loss": 0.6824,
+      "step": 4371
+    },
+    {
+      "epoch": 0.77994826509678,
+      "grad_norm": 0.5366002917289734,
+      "learning_rate": 0.0003346778880028416,
+      "loss": 1.0622,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7801266613147801,
+      "grad_norm": 0.45976123213768005,
+      "learning_rate": 0.00033461196376107275,
+      "loss": 0.7995,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7803050575327803,
+      "grad_norm": 0.48608720302581787,
+      "learning_rate": 0.00033454603287391666,
+      "loss": 1.0287,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7804834537507804,
+      "grad_norm": 3.942063570022583,
+      "learning_rate": 0.0003344800953465515,
+      "loss": 1.1003,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7806618499687806,
+      "grad_norm": 0.4614802598953247,
+      "learning_rate": 0.000334414151184156,
+      "loss": 0.8365,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7808402461867808,
+      "grad_norm": 0.4670039117336273,
+      "learning_rate": 0.00033434820039190944,
+      "loss": 0.9256,
+      "step": 4377
+    },
+    {
+      "epoch": 0.781018642404781,
+      "grad_norm": 0.5666428208351135,
+      "learning_rate": 0.0003342822429749915,
+      "loss": 0.8964,
+      "step": 4378
+    },
+    {
+      "epoch": 0.7811970386227812,
+      "grad_norm": 0.5543097257614136,
+      "learning_rate": 0.00033421627893858244,
+      "loss": 0.9927,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7813754348407814,
+      "grad_norm": 0.5561456680297852,
+      "learning_rate": 0.0003341503082878631,
+      "loss": 0.9066,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7815538310587815,
+      "grad_norm": 0.5354291796684265,
+      "learning_rate": 0.00033408433102801474,
+      "loss": 1.0043,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7817322272767817,
+      "grad_norm": 0.45063936710357666,
+      "learning_rate": 0.00033401834716421925,
+      "loss": 0.6419,
+      "step": 4382
+    },
+    {
+      "epoch": 0.7819106234947819,
+      "grad_norm": 0.5044664144515991,
+      "learning_rate": 0.0003339523567016589,
+      "loss": 0.9381,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7820890197127821,
+      "grad_norm": 0.4677816331386566,
+      "learning_rate": 0.0003338863596455166,
+      "loss": 0.879,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7822674159307823,
+      "grad_norm": 0.4861041009426117,
+      "learning_rate": 0.00033382035600097563,
+      "loss": 0.8487,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7824458121487824,
+      "grad_norm": 0.5526744723320007,
+      "learning_rate": 0.0003337543457732201,
+      "loss": 0.9411,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7826242083667826,
+      "grad_norm": 4.370226860046387,
+      "learning_rate": 0.0003336883289674342,
+      "loss": 0.9109,
+      "step": 4387
+    },
+    {
+      "epoch": 0.7828026045847828,
+      "grad_norm": 0.4913937449455261,
+      "learning_rate": 0.00033362230558880296,
+      "loss": 0.8254,
+      "step": 4388
+    },
+    {
+      "epoch": 0.782981000802783,
+      "grad_norm": 0.5903784036636353,
+      "learning_rate": 0.00033355627564251184,
+      "loss": 1.2674,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7831593970207832,
+      "grad_norm": 0.48795902729034424,
+      "learning_rate": 0.0003334902391337468,
+      "loss": 0.7577,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7833377932387834,
+      "grad_norm": 0.4892320930957794,
+      "learning_rate": 0.00033342419606769433,
+      "loss": 0.9373,
+      "step": 4391
+    },
+    {
+      "epoch": 0.7835161894567835,
+      "grad_norm": 0.4674488604068756,
+      "learning_rate": 0.00033335814644954137,
+      "loss": 0.7285,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7836945856747837,
+      "grad_norm": 0.4890602231025696,
+      "learning_rate": 0.00033329209028447543,
+      "loss": 0.9521,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7838729818927839,
+      "grad_norm": 0.4913036823272705,
+      "learning_rate": 0.0003332260275776846,
+      "loss": 0.8138,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7840513781107841,
+      "grad_norm": 0.4752870500087738,
+      "learning_rate": 0.0003331599583343574,
+      "loss": 1.065,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7842297743287843,
+      "grad_norm": 0.542586624622345,
+      "learning_rate": 0.0003330938825596828,
+      "loss": 0.9805,
+      "step": 4396
+    },
+    {
+      "epoch": 0.7844081705467844,
+      "grad_norm": 0.4948160946369171,
+      "learning_rate": 0.0003330278002588505,
+      "loss": 0.7871,
+      "step": 4397
+    },
+    {
+      "epoch": 0.7845865667647846,
+      "grad_norm": 0.5883384943008423,
+      "learning_rate": 0.0003329617114370505,
+      "loss": 1.1591,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7847649629827848,
+      "grad_norm": 0.484783798456192,
+      "learning_rate": 0.00033289561609947335,
+      "loss": 1.0079,
+      "step": 4399
+    },
+    {
+      "epoch": 0.784943359200785,
+      "grad_norm": 0.5016215443611145,
+      "learning_rate": 0.00033282951425131014,
+      "loss": 0.7401,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7851217554187851,
+      "grad_norm": 0.45084336400032043,
+      "learning_rate": 0.00033276340589775255,
+      "loss": 0.7707,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7853001516367853,
+      "grad_norm": 0.4217044413089752,
+      "learning_rate": 0.00033269729104399263,
+      "loss": 0.614,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7854785478547854,
+      "grad_norm": 0.5516197085380554,
+      "learning_rate": 0.00033263116969522316,
+      "loss": 0.9417,
+      "step": 4403
+    },
+    {
+      "epoch": 0.7856569440727856,
+      "grad_norm": 0.4416463375091553,
+      "learning_rate": 0.00033256504185663713,
+      "loss": 0.7047,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7858353402907858,
+      "grad_norm": 2.543560743331909,
+      "learning_rate": 0.00033249890753342826,
+      "loss": 0.981,
+      "step": 4405
+    },
+    {
+      "epoch": 0.786013736508786,
+      "grad_norm": 0.44152694940567017,
+      "learning_rate": 0.0003324327667307907,
+      "loss": 0.8247,
+      "step": 4406
+    },
+    {
+      "epoch": 0.7861921327267862,
+      "grad_norm": 0.7885439395904541,
+      "learning_rate": 0.00033236661945391905,
+      "loss": 0.9313,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7863705289447863,
+      "grad_norm": 0.6830782294273376,
+      "learning_rate": 0.00033230046570800867,
+      "loss": 0.713,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7865489251627865,
+      "grad_norm": 0.5507636666297913,
+      "learning_rate": 0.000332234305498255,
+      "loss": 1.0907,
+      "step": 4409
+    },
+    {
+      "epoch": 0.7867273213807867,
+      "grad_norm": 0.4962003231048584,
+      "learning_rate": 0.00033216813882985444,
+      "loss": 0.8278,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7869057175987869,
+      "grad_norm": 0.5796182155609131,
+      "learning_rate": 0.00033210196570800365,
+      "loss": 0.881,
+      "step": 4411
+    },
+    {
+      "epoch": 0.7870841138167871,
+      "grad_norm": 0.48952221870422363,
+      "learning_rate": 0.00033203578613789973,
+      "loss": 0.9298,
+      "step": 4412
+    },
+    {
+      "epoch": 0.7872625100347873,
+      "grad_norm": 0.49012526869773865,
+      "learning_rate": 0.00033196960012474053,
+      "loss": 0.8109,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7874409062527874,
+      "grad_norm": 0.49443262815475464,
+      "learning_rate": 0.0003319034076737242,
+      "loss": 0.9402,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7876193024707876,
+      "grad_norm": 0.4793679714202881,
+      "learning_rate": 0.00033183720879004935,
+      "loss": 0.7636,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7877976986887878,
+      "grad_norm": 0.49106279015541077,
+      "learning_rate": 0.0003317710034789154,
+      "loss": 0.8781,
+      "step": 4416
+    },
+    {
+      "epoch": 0.787976094906788,
+      "grad_norm": 0.46945035457611084,
+      "learning_rate": 0.00033170479174552217,
+      "loss": 0.7358,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7881544911247882,
+      "grad_norm": 0.48162609338760376,
+      "learning_rate": 0.00033163857359506965,
+      "loss": 0.8671,
+      "step": 4418
+    },
+    {
+      "epoch": 0.7883328873427883,
+      "grad_norm": 0.4425477385520935,
+      "learning_rate": 0.00033157234903275867,
+      "loss": 0.8101,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7885112835607885,
+      "grad_norm": 0.499623566865921,
+      "learning_rate": 0.00033150611806379054,
+      "loss": 0.8189,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7886896797787887,
+      "grad_norm": 0.44681599736213684,
+      "learning_rate": 0.000331439880693367,
+      "loss": 0.7808,
+      "step": 4421
+    },
+    {
+      "epoch": 0.7888680759967889,
+      "grad_norm": 0.4531422555446625,
+      "learning_rate": 0.00033137363692669014,
+      "loss": 0.7667,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7890464722147891,
+      "grad_norm": 0.43248048424720764,
+      "learning_rate": 0.00033130738676896297,
+      "loss": 0.7906,
+      "step": 4423
+    },
+    {
+      "epoch": 0.7892248684327893,
+      "grad_norm": 0.5049116611480713,
+      "learning_rate": 0.00033124113022538865,
+      "loss": 0.9246,
+      "step": 4424
+    },
+    {
+      "epoch": 0.7894032646507894,
+      "grad_norm": 0.48414015769958496,
+      "learning_rate": 0.00033117486730117093,
+      "loss": 0.8807,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7895816608687896,
+      "grad_norm": 0.5417115688323975,
+      "learning_rate": 0.000331108598001514,
+      "loss": 0.9549,
+      "step": 4426
+    },
+    {
+      "epoch": 0.7897600570867898,
+      "grad_norm": 0.4763809144496918,
+      "learning_rate": 0.0003310423223316227,
+      "loss": 0.8302,
+      "step": 4427
+    },
+    {
+      "epoch": 0.78993845330479,
+      "grad_norm": 0.4397316873073578,
+      "learning_rate": 0.00033097604029670225,
+      "loss": 0.8886,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7901168495227902,
+      "grad_norm": 0.451040118932724,
+      "learning_rate": 0.0003309097519019585,
+      "loss": 0.6568,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7902952457407902,
+      "grad_norm": 0.4390012323856354,
+      "learning_rate": 0.0003308434571525976,
+      "loss": 0.7361,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7904736419587904,
+      "grad_norm": 0.48856833577156067,
+      "learning_rate": 0.0003307771560538264,
+      "loss": 0.7811,
+      "step": 4431
+    },
+    {
+      "epoch": 0.7906520381767906,
+      "grad_norm": 0.49298450350761414,
+      "learning_rate": 0.00033071084861085217,
+      "loss": 0.8156,
+      "step": 4432
+    },
+    {
+      "epoch": 0.7908304343947908,
+      "grad_norm": 0.424003541469574,
+      "learning_rate": 0.00033064453482888257,
+      "loss": 0.6857,
+      "step": 4433
+    },
+    {
+      "epoch": 0.791008830612791,
+      "grad_norm": 0.7829723358154297,
+      "learning_rate": 0.0003305782147131259,
+      "loss": 0.9136,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7911872268307912,
+      "grad_norm": 0.549010157585144,
+      "learning_rate": 0.0003305118882687909,
+      "loss": 0.772,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7913656230487913,
+      "grad_norm": 0.4902380704879761,
+      "learning_rate": 0.00033044555550108693,
+      "loss": 0.7763,
+      "step": 4436
+    },
+    {
+      "epoch": 0.7915440192667915,
+      "grad_norm": 0.4383619725704193,
+      "learning_rate": 0.0003303792164152236,
+      "loss": 0.9407,
+      "step": 4437
+    },
+    {
+      "epoch": 0.7917224154847917,
+      "grad_norm": 0.486372709274292,
+      "learning_rate": 0.00033031287101641116,
+      "loss": 0.997,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7919008117027919,
+      "grad_norm": 0.5471330285072327,
+      "learning_rate": 0.00033024651930986044,
+      "loss": 0.8833,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7920792079207921,
+      "grad_norm": 0.47940507531166077,
+      "learning_rate": 0.0003301801613007826,
+      "loss": 1.008,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7922576041387922,
+      "grad_norm": 0.4837360382080078,
+      "learning_rate": 0.0003301137969943894,
+      "loss": 0.9464,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7924360003567924,
+      "grad_norm": 0.48073476552963257,
+      "learning_rate": 0.0003300474263958931,
+      "loss": 0.8527,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7926143965747926,
+      "grad_norm": 0.47916659712791443,
+      "learning_rate": 0.00032998104951050634,
+      "loss": 0.8098,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 0.5394938588142395,
+      "learning_rate": 0.00032991466634344234,
+      "loss": 0.9535,
+      "step": 4444
+    },
+    {
+      "epoch": 0.792971189010793,
+      "grad_norm": 0.44547122716903687,
+      "learning_rate": 0.00032984827689991493,
+      "loss": 0.7269,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7931495852287932,
+      "grad_norm": 0.5474995374679565,
+      "learning_rate": 0.0003297818811851381,
+      "loss": 1.0227,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7933279814467933,
+      "grad_norm": 0.5473585724830627,
+      "learning_rate": 0.0003297154792043268,
+      "loss": 1.043,
+      "step": 4447
+    },
+    {
+      "epoch": 0.7935063776647935,
+      "grad_norm": 0.4513511061668396,
+      "learning_rate": 0.0003296490709626959,
+      "loss": 0.7503,
+      "step": 4448
+    },
+    {
+      "epoch": 0.7936847738827937,
+      "grad_norm": 0.47088685631752014,
+      "learning_rate": 0.0003295826564654614,
+      "loss": 0.8366,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7938631701007939,
+      "grad_norm": 0.4641473889350891,
+      "learning_rate": 0.00032951623571783915,
+      "loss": 0.8643,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7940415663187941,
+      "grad_norm": 0.43831127882003784,
+      "learning_rate": 0.00032944980872504613,
+      "loss": 0.7894,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7942199625367942,
+      "grad_norm": 0.4277407228946686,
+      "learning_rate": 0.00032938337549229924,
+      "loss": 0.6743,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7943983587547944,
+      "grad_norm": 0.454748272895813,
+      "learning_rate": 0.00032931693602481616,
+      "loss": 0.829,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7945767549727946,
+      "grad_norm": 0.5081732273101807,
+      "learning_rate": 0.0003292504903278151,
+      "loss": 0.9585,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7947551511907948,
+      "grad_norm": 0.4319037199020386,
+      "learning_rate": 0.0003291840384065146,
+      "loss": 0.7298,
+      "step": 4455
+    },
+    {
+      "epoch": 0.794933547408795,
+      "grad_norm": 0.5093657374382019,
+      "learning_rate": 0.00032911758026613384,
+      "loss": 1.0046,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7951119436267952,
+      "grad_norm": 0.4518407881259918,
+      "learning_rate": 0.0003290511159118924,
+      "loss": 0.7589,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7952903398447952,
+      "grad_norm": 0.4718686044216156,
+      "learning_rate": 0.0003289846453490103,
+      "loss": 0.7775,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7954687360627954,
+      "grad_norm": 0.520844578742981,
+      "learning_rate": 0.00032891816858270816,
+      "loss": 1.2267,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7956471322807956,
+      "grad_norm": 0.4900132119655609,
+      "learning_rate": 0.000328851685618207,
+      "loss": 0.941,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7958255284987958,
+      "grad_norm": 0.4234750270843506,
+      "learning_rate": 0.00032878519646072833,
+      "loss": 0.8223,
+      "step": 4461
+    },
+    {
+      "epoch": 0.796003924716796,
+      "grad_norm": 0.42578887939453125,
+      "learning_rate": 0.0003287187011154943,
+      "loss": 0.6952,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7961823209347961,
+      "grad_norm": 0.4930354356765747,
+      "learning_rate": 0.00032865219958772734,
+      "loss": 0.8822,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7963607171527963,
+      "grad_norm": 0.5005689263343811,
+      "learning_rate": 0.0003285856918826505,
+      "loss": 0.8978,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7965391133707965,
+      "grad_norm": 0.46632346510887146,
+      "learning_rate": 0.00032851917800548724,
+      "loss": 0.9696,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7967175095887967,
+      "grad_norm": 0.4460965692996979,
+      "learning_rate": 0.0003284526579614615,
+      "loss": 0.8667,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7968959058067969,
+      "grad_norm": 0.45909878611564636,
+      "learning_rate": 0.0003283861317557978,
+      "loss": 0.9574,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7970743020247971,
+      "grad_norm": 0.48101097345352173,
+      "learning_rate": 0.0003283195993937209,
+      "loss": 0.8039,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7972526982427972,
+      "grad_norm": 0.5066887140274048,
+      "learning_rate": 0.0003282530608804565,
+      "loss": 0.8576,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7974310944607974,
+      "grad_norm": 0.44910162687301636,
+      "learning_rate": 0.0003281865162212304,
+      "loss": 0.846,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7976094906787976,
+      "grad_norm": 0.46548551321029663,
+      "learning_rate": 0.000328119965421269,
+      "loss": 0.7717,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7977878868967978,
+      "grad_norm": 0.603074312210083,
+      "learning_rate": 0.00032805340848579903,
+      "loss": 0.9443,
+      "step": 4472
+    },
+    {
+      "epoch": 0.797966283114798,
+      "grad_norm": 0.4617455005645752,
+      "learning_rate": 0.00032798684542004793,
+      "loss": 0.6775,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7981446793327981,
+      "grad_norm": 0.5626710653305054,
+      "learning_rate": 0.00032792027622924357,
+      "loss": 1.0429,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7983230755507983,
+      "grad_norm": 0.500339925289154,
+      "learning_rate": 0.00032785370091861435,
+      "loss": 1.0465,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7985014717687985,
+      "grad_norm": 0.44158151745796204,
+      "learning_rate": 0.0003277871194933888,
+      "loss": 0.6398,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7986798679867987,
+      "grad_norm": 0.47629523277282715,
+      "learning_rate": 0.0003277205319587965,
+      "loss": 0.8574,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7988582642047989,
+      "grad_norm": 0.6063137650489807,
+      "learning_rate": 0.00032765393832006695,
+      "loss": 0.928,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7990366604227991,
+      "grad_norm": 0.41350358724594116,
+      "learning_rate": 0.00032758733858243054,
+      "loss": 0.7943,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7992150566407992,
+      "grad_norm": 0.4761079251766205,
+      "learning_rate": 0.000327520732751118,
+      "loss": 0.9293,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7993934528587994,
+      "grad_norm": 0.46137312054634094,
+      "learning_rate": 0.0003274541208313604,
+      "loss": 0.6324,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7995718490767996,
+      "grad_norm": 0.46572351455688477,
+      "learning_rate": 0.00032738750282838955,
+      "loss": 0.796,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7997502452947998,
+      "grad_norm": 1.1689634323120117,
+      "learning_rate": 0.0003273208787474375,
+      "loss": 0.8631,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7999286415128,
+      "grad_norm": 0.5234804749488831,
+      "learning_rate": 0.00032725424859373687,
+      "loss": 1.022,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8001070377308,
+      "grad_norm": 0.46076735854148865,
+      "learning_rate": 0.0003271876123725208,
+      "loss": 0.8363,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8002854339488003,
+      "grad_norm": 0.5089244246482849,
+      "learning_rate": 0.0003271209700890229,
+      "loss": 1.0501,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8004638301668004,
+      "grad_norm": 0.49238088726997375,
+      "learning_rate": 0.0003270543217484772,
+      "loss": 0.7696,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8006422263848006,
+      "grad_norm": 0.5077065229415894,
+      "learning_rate": 0.0003269876673561183,
+      "loss": 0.7879,
+      "step": 4488
+    },
+    {
+      "epoch": 0.8008206226028008,
+      "grad_norm": 0.4703711271286011,
+      "learning_rate": 0.000326921006917181,
+      "loss": 0.7882,
+      "step": 4489
+    },
+    {
+      "epoch": 0.800999018820801,
+      "grad_norm": 0.4334394633769989,
+      "learning_rate": 0.000326854340436901,
+      "loss": 0.6552,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8011774150388011,
+      "grad_norm": 0.45196327567100525,
+      "learning_rate": 0.0003267876679205142,
+      "loss": 0.674,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8013558112568013,
+      "grad_norm": 0.629697859287262,
+      "learning_rate": 0.0003267209893732569,
+      "loss": 1.0846,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8015342074748015,
+      "grad_norm": 0.4871639907360077,
+      "learning_rate": 0.00032665430480036616,
+      "loss": 0.8173,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8017126036928017,
+      "grad_norm": 2.4100522994995117,
+      "learning_rate": 0.0003265876142070794,
+      "loss": 0.8096,
+      "step": 4494
+    },
+    {
+      "epoch": 0.8018909999108019,
+      "grad_norm": 0.567620038986206,
+      "learning_rate": 0.00032652091759863424,
+      "loss": 1.2446,
+      "step": 4495
+    },
+    {
+      "epoch": 0.802069396128802,
+      "grad_norm": 0.440104603767395,
+      "learning_rate": 0.0003264542149802692,
+      "loss": 0.7097,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8022477923468022,
+      "grad_norm": 0.49658969044685364,
+      "learning_rate": 0.0003263875063572231,
+      "loss": 1.0596,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8024261885648024,
+      "grad_norm": 0.5105962157249451,
+      "learning_rate": 0.000326320791734735,
+      "loss": 0.9645,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8026045847828026,
+      "grad_norm": 0.5017589926719666,
+      "learning_rate": 0.00032625407111804477,
+      "loss": 0.8783,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8027829810008028,
+      "grad_norm": 0.4669455885887146,
+      "learning_rate": 0.0003261873445123926,
+      "loss": 0.7848,
+      "step": 4500
+    },
+    {
+      "epoch": 0.802961377218803,
+      "grad_norm": 0.5145031809806824,
+      "learning_rate": 0.0003261206119230192,
+      "loss": 0.9043,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8031397734368031,
+      "grad_norm": 0.8292232155799866,
+      "learning_rate": 0.0003260538733551658,
+      "loss": 0.8444,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8033181696548033,
+      "grad_norm": 0.4397391974925995,
+      "learning_rate": 0.0003259871288140738,
+      "loss": 0.8279,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8034965658728035,
+      "grad_norm": 0.48555225133895874,
+      "learning_rate": 0.0003259203783049854,
+      "loss": 1.0585,
+      "step": 4504
+    },
+    {
+      "epoch": 0.8036749620908037,
+      "grad_norm": 0.4733032286167145,
+      "learning_rate": 0.00032585362183314327,
+      "loss": 0.899,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8038533583088039,
+      "grad_norm": 0.6270719766616821,
+      "learning_rate": 0.00032578685940379027,
+      "loss": 0.8007,
+      "step": 4506
+    },
+    {
+      "epoch": 0.804031754526804,
+      "grad_norm": 1.0983858108520508,
+      "learning_rate": 0.00032572009102216983,
+      "loss": 0.8323,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8042101507448042,
+      "grad_norm": 0.5489131212234497,
+      "learning_rate": 0.00032565331669352613,
+      "loss": 0.78,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8043885469628044,
+      "grad_norm": 0.4885414242744446,
+      "learning_rate": 0.00032558653642310347,
+      "loss": 0.9142,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8045669431808046,
+      "grad_norm": 0.5155977010726929,
+      "learning_rate": 0.0003255197502161468,
+      "loss": 0.9124,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8047453393988048,
+      "grad_norm": 0.44706472754478455,
+      "learning_rate": 0.0003254529580779014,
+      "loss": 0.7907,
+      "step": 4511
+    },
+    {
+      "epoch": 0.804923735616805,
+      "grad_norm": 0.6217315793037415,
+      "learning_rate": 0.0003253861600136132,
+      "loss": 0.7259,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8051021318348051,
+      "grad_norm": 0.5625154972076416,
+      "learning_rate": 0.0003253193560285284,
+      "loss": 0.9467,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8052805280528053,
+      "grad_norm": 0.46225252747535706,
+      "learning_rate": 0.00032525254612789377,
+      "loss": 0.8054,
+      "step": 4514
+    },
+    {
+      "epoch": 0.8054589242708055,
+      "grad_norm": 0.4399617612361908,
+      "learning_rate": 0.0003251857303169565,
+      "loss": 0.7312,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8056373204888057,
+      "grad_norm": 0.4609040319919586,
+      "learning_rate": 0.00032511890860096443,
+      "loss": 0.8695,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8058157167068059,
+      "grad_norm": 0.5766742825508118,
+      "learning_rate": 0.00032505208098516567,
+      "loss": 0.9441,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8059941129248059,
+      "grad_norm": 0.47482702136039734,
+      "learning_rate": 0.0003249852474748086,
+      "loss": 0.7759,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8061725091428061,
+      "grad_norm": 0.4535680115222931,
+      "learning_rate": 0.0003249184080751426,
+      "loss": 0.9628,
+      "step": 4519
+    },
+    {
+      "epoch": 0.8063509053608063,
+      "grad_norm": 0.44655969738960266,
+      "learning_rate": 0.00032485156279141695,
+      "loss": 0.8394,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8065293015788065,
+      "grad_norm": 0.5331109762191772,
+      "learning_rate": 0.00032478471162888185,
+      "loss": 1.0328,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8067076977968067,
+      "grad_norm": 0.5180974006652832,
+      "learning_rate": 0.00032471785459278757,
+      "loss": 0.914,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8068860940148069,
+      "grad_norm": 0.4896256923675537,
+      "learning_rate": 0.0003246509916883853,
+      "loss": 0.6507,
+      "step": 4523
+    },
+    {
+      "epoch": 0.807064490232807,
+      "grad_norm": 0.5117029547691345,
+      "learning_rate": 0.0003245841229209262,
+      "loss": 1.0229,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8072428864508072,
+      "grad_norm": 0.4755803346633911,
+      "learning_rate": 0.00032451724829566216,
+      "loss": 0.8733,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8074212826688074,
+      "grad_norm": 0.642292320728302,
+      "learning_rate": 0.0003244503678178455,
+      "loss": 1.0649,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8075996788868076,
+      "grad_norm": 0.5174018740653992,
+      "learning_rate": 0.000324383481492729,
+      "loss": 1.0805,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8077780751048078,
+      "grad_norm": 0.439130961894989,
+      "learning_rate": 0.0003243165893255659,
+      "loss": 0.7667,
+      "step": 4528
+    },
+    {
+      "epoch": 0.807956471322808,
+      "grad_norm": 0.5517845153808594,
+      "learning_rate": 0.00032424969132160985,
+      "loss": 1.0015,
+      "step": 4529
+    },
+    {
+      "epoch": 0.8081348675408081,
+      "grad_norm": 0.5080962777137756,
+      "learning_rate": 0.00032418278748611495,
+      "loss": 0.8835,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8083132637588083,
+      "grad_norm": 0.5714918971061707,
+      "learning_rate": 0.00032411587782433594,
+      "loss": 1.2275,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8084916599768085,
+      "grad_norm": 0.5037386417388916,
+      "learning_rate": 0.0003240489623415277,
+      "loss": 0.8592,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8086700561948087,
+      "grad_norm": 0.5023531913757324,
+      "learning_rate": 0.00032398204104294585,
+      "loss": 0.9501,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8088484524128089,
+      "grad_norm": 0.4968222677707672,
+      "learning_rate": 0.00032391511393384633,
+      "loss": 0.8091,
+      "step": 4534
+    },
+    {
+      "epoch": 0.809026848630809,
+      "grad_norm": 0.5576387643814087,
+      "learning_rate": 0.00032384818101948554,
+      "loss": 0.8792,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8092052448488092,
+      "grad_norm": 0.4761325716972351,
+      "learning_rate": 0.0003237812423051204,
+      "loss": 0.8114,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8093836410668094,
+      "grad_norm": 0.393627792596817,
+      "learning_rate": 0.00032371429779600824,
+      "loss": 0.5444,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8095620372848096,
+      "grad_norm": 0.5613236427307129,
+      "learning_rate": 0.00032364734749740687,
+      "loss": 0.855,
+      "step": 4538
+    },
+    {
+      "epoch": 0.8097404335028098,
+      "grad_norm": 0.46514925360679626,
+      "learning_rate": 0.00032358039141457454,
+      "loss": 0.7157,
+      "step": 4539
+    },
+    {
+      "epoch": 0.80991882972081,
+      "grad_norm": 0.4529775083065033,
+      "learning_rate": 0.0003235134295527699,
+      "loss": 0.8365,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8100972259388101,
+      "grad_norm": 0.47237035632133484,
+      "learning_rate": 0.0003234464619172522,
+      "loss": 0.852,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8102756221568103,
+      "grad_norm": 0.4663850963115692,
+      "learning_rate": 0.00032337948851328093,
+      "loss": 0.7608,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8104540183748105,
+      "grad_norm": 0.5153512954711914,
+      "learning_rate": 0.0003233125093461162,
+      "loss": 0.8918,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8106324145928107,
+      "grad_norm": 0.40413349866867065,
+      "learning_rate": 0.0003232455244210186,
+      "loss": 0.6576,
+      "step": 4544
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 0.4308657944202423,
+      "learning_rate": 0.0003231785337432489,
+      "loss": 0.9653,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8109892070288109,
+      "grad_norm": 0.4188125729560852,
+      "learning_rate": 0.00032311153731806873,
+      "loss": 0.7579,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8111676032468111,
+      "grad_norm": 0.5259095430374146,
+      "learning_rate": 0.00032304453515073994,
+      "loss": 1.1485,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8113459994648113,
+      "grad_norm": 0.4392109215259552,
+      "learning_rate": 0.0003229775272465247,
+      "loss": 0.8834,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8115243956828115,
+      "grad_norm": 0.4974541962146759,
+      "learning_rate": 0.0003229105136106859,
+      "loss": 0.865,
+      "step": 4549
+    },
+    {
+      "epoch": 0.8117027919008117,
+      "grad_norm": 1.2180891036987305,
+      "learning_rate": 0.0003228434942484869,
+      "loss": 0.9709,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8118811881188119,
+      "grad_norm": 0.4866902828216553,
+      "learning_rate": 0.000322776469165191,
+      "loss": 0.9429,
+      "step": 4551
+    },
+    {
+      "epoch": 0.812059584336812,
+      "grad_norm": 0.5774986147880554,
+      "learning_rate": 0.0003227094383660626,
+      "loss": 0.8982,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8122379805548122,
+      "grad_norm": 0.4660821259021759,
+      "learning_rate": 0.0003226424018563661,
+      "loss": 0.8265,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8124163767728124,
+      "grad_norm": 0.4910152554512024,
+      "learning_rate": 0.00032257535964136673,
+      "loss": 0.868,
+      "step": 4554
+    },
+    {
+      "epoch": 0.8125947729908126,
+      "grad_norm": 0.4666675627231598,
+      "learning_rate": 0.0003225083117263298,
+      "loss": 0.849,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8127731692088128,
+      "grad_norm": 0.5173287391662598,
+      "learning_rate": 0.00032244125811652135,
+      "loss": 0.9152,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8129515654268129,
+      "grad_norm": 0.503570020198822,
+      "learning_rate": 0.00032237419881720765,
+      "loss": 1.0545,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8131299616448131,
+      "grad_norm": 0.4771646559238434,
+      "learning_rate": 0.00032230713383365545,
+      "loss": 0.8746,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8133083578628133,
+      "grad_norm": 0.47005799412727356,
+      "learning_rate": 0.0003222400631711321,
+      "loss": 0.8811,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8134867540808135,
+      "grad_norm": 0.48531290888786316,
+      "learning_rate": 0.00032217298683490526,
+      "loss": 0.7489,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8136651502988137,
+      "grad_norm": 0.4748542010784149,
+      "learning_rate": 0.0003221059048302431,
+      "loss": 1.0024,
+      "step": 4561
+    },
+    {
+      "epoch": 0.8138435465168139,
+      "grad_norm": 0.46616560220718384,
+      "learning_rate": 0.00032203881716241426,
+      "loss": 0.8083,
+      "step": 4562
+    },
+    {
+      "epoch": 0.814021942734814,
+      "grad_norm": 1.1073871850967407,
+      "learning_rate": 0.00032197172383668763,
+      "loss": 0.6794,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8142003389528142,
+      "grad_norm": 0.5361044406890869,
+      "learning_rate": 0.0003219046248583329,
+      "loss": 1.0918,
+      "step": 4564
+    },
+    {
+      "epoch": 0.8143787351708144,
+      "grad_norm": 0.44288361072540283,
+      "learning_rate": 0.00032183752023261973,
+      "loss": 0.8201,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8145571313888146,
+      "grad_norm": 0.4622916579246521,
+      "learning_rate": 0.00032177040996481874,
+      "loss": 0.9029,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8147355276068148,
+      "grad_norm": 0.48599693179130554,
+      "learning_rate": 0.0003217032940602006,
+      "loss": 1.0285,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8149139238248149,
+      "grad_norm": 0.46129268407821655,
+      "learning_rate": 0.00032163617252403654,
+      "loss": 0.8161,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8150923200428151,
+      "grad_norm": 0.5313239097595215,
+      "learning_rate": 0.0003215690453615985,
+      "loss": 1.0821,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8152707162608153,
+      "grad_norm": 0.42274999618530273,
+      "learning_rate": 0.0003215019125781583,
+      "loss": 0.7682,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8154491124788155,
+      "grad_norm": 0.46743467450141907,
+      "learning_rate": 0.00032143477417898866,
+      "loss": 0.8906,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8156275086968157,
+      "grad_norm": 0.4919176697731018,
+      "learning_rate": 0.0003213676301693626,
+      "loss": 0.8782,
+      "step": 4572
+    },
+    {
+      "epoch": 0.8158059049148159,
+      "grad_norm": 0.479407399892807,
+      "learning_rate": 0.00032130048055455356,
+      "loss": 0.6938,
+      "step": 4573
+    },
+    {
+      "epoch": 0.815984301132816,
+      "grad_norm": 0.4484141767024994,
+      "learning_rate": 0.0003212333253398355,
+      "loss": 0.6904,
+      "step": 4574
+    },
+    {
+      "epoch": 0.8161626973508161,
+      "grad_norm": 0.5200289487838745,
+      "learning_rate": 0.0003211661645304827,
+      "loss": 0.9434,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8163410935688163,
+      "grad_norm": 0.5177778601646423,
+      "learning_rate": 0.00032109899813177,
+      "loss": 0.8227,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8165194897868165,
+      "grad_norm": 0.5002527236938477,
+      "learning_rate": 0.0003210318261489725,
+      "loss": 0.9783,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8166978860048167,
+      "grad_norm": 0.47500887513160706,
+      "learning_rate": 0.0003209646485873661,
+      "loss": 0.7346,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8168762822228168,
+      "grad_norm": 0.49525851011276245,
+      "learning_rate": 0.00032089746545222657,
+      "loss": 0.8249,
+      "step": 4579
+    },
+    {
+      "epoch": 0.817054678440817,
+      "grad_norm": 0.5161625742912292,
+      "learning_rate": 0.0003208302767488307,
+      "loss": 0.9479,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8172330746588172,
+      "grad_norm": 0.47351545095443726,
+      "learning_rate": 0.00032076308248245533,
+      "loss": 0.8557,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8174114708768174,
+      "grad_norm": 0.37471267580986023,
+      "learning_rate": 0.00032069588265837794,
+      "loss": 0.4742,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8175898670948176,
+      "grad_norm": 0.5405718088150024,
+      "learning_rate": 0.0003206286772818764,
+      "loss": 1.0018,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8177682633128178,
+      "grad_norm": 0.4767707884311676,
+      "learning_rate": 0.00032056146635822886,
+      "loss": 0.8955,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8179466595308179,
+      "grad_norm": 0.48326289653778076,
+      "learning_rate": 0.00032049424989271416,
+      "loss": 0.8068,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8181250557488181,
+      "grad_norm": 0.44311580061912537,
+      "learning_rate": 0.0003204270278906114,
+      "loss": 0.7806,
+      "step": 4586
+    },
+    {
+      "epoch": 0.8183034519668183,
+      "grad_norm": 0.46994298696517944,
+      "learning_rate": 0.00032035980035720015,
+      "loss": 0.9937,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8184818481848185,
+      "grad_norm": 1.0689364671707153,
+      "learning_rate": 0.0003202925672977605,
+      "loss": 0.9123,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8186602444028187,
+      "grad_norm": 0.5591593980789185,
+      "learning_rate": 0.0003202253287175728,
+      "loss": 1.0514,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8188386406208188,
+      "grad_norm": 5.0126953125,
+      "learning_rate": 0.00032015808462191816,
+      "loss": 0.8152,
+      "step": 4590
+    },
+    {
+      "epoch": 0.819017036838819,
+      "grad_norm": 1.514142394065857,
+      "learning_rate": 0.00032009083501607753,
+      "loss": 0.7788,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8191954330568192,
+      "grad_norm": 0.5694434642791748,
+      "learning_rate": 0.00032002357990533296,
+      "loss": 0.9116,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8193738292748194,
+      "grad_norm": 1.0806207656860352,
+      "learning_rate": 0.0003199563192949666,
+      "loss": 1.0898,
+      "step": 4593
+    },
+    {
+      "epoch": 0.8195522254928196,
+      "grad_norm": 0.490933895111084,
+      "learning_rate": 0.000319889053190261,
+      "loss": 0.7681,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8197306217108198,
+      "grad_norm": 0.5183424949645996,
+      "learning_rate": 0.00031982178159649925,
+      "loss": 0.7874,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8199090179288199,
+      "grad_norm": 0.5265378355979919,
+      "learning_rate": 0.0003197545045189648,
+      "loss": 0.8632,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8200874141468201,
+      "grad_norm": 0.4929676353931427,
+      "learning_rate": 0.0003196872219629417,
+      "loss": 0.833,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8202658103648203,
+      "grad_norm": 0.48382702469825745,
+      "learning_rate": 0.00031961993393371405,
+      "loss": 0.6996,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8204442065828205,
+      "grad_norm": 0.7561694383621216,
+      "learning_rate": 0.00031955264043656675,
+      "loss": 0.8474,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8206226028008207,
+      "grad_norm": 0.6017073392868042,
+      "learning_rate": 0.000319485341476785,
+      "loss": 0.9392,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8208009990188208,
+      "grad_norm": 0.5361237525939941,
+      "learning_rate": 0.00031941803705965447,
+      "loss": 0.8068,
+      "step": 4601
+    },
+    {
+      "epoch": 0.820979395236821,
+      "grad_norm": 0.5034250617027283,
+      "learning_rate": 0.00031935072719046115,
+      "loss": 1.0488,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8211577914548212,
+      "grad_norm": 0.48589175939559937,
+      "learning_rate": 0.0003192834118744916,
+      "loss": 1.0141,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8213361876728213,
+      "grad_norm": 0.43965524435043335,
+      "learning_rate": 0.0003192160911170327,
+      "loss": 0.6931,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8215145838908215,
+      "grad_norm": 0.6355204582214355,
+      "learning_rate": 0.00031914876492337177,
+      "loss": 0.8362,
+      "step": 4605
+    },
+    {
+      "epoch": 0.8216929801088217,
+      "grad_norm": 0.502838671207428,
+      "learning_rate": 0.0003190814332987965,
+      "loss": 0.9824,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8218713763268218,
+      "grad_norm": 0.48031777143478394,
+      "learning_rate": 0.00031901409624859536,
+      "loss": 0.8348,
+      "step": 4607
+    },
+    {
+      "epoch": 0.822049772544822,
+      "grad_norm": 0.47550857067108154,
+      "learning_rate": 0.00031894675377805665,
+      "loss": 0.8757,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8222281687628222,
+      "grad_norm": 1.1182494163513184,
+      "learning_rate": 0.0003188794058924697,
+      "loss": 0.9129,
+      "step": 4609
+    },
+    {
+      "epoch": 0.8224065649808224,
+      "grad_norm": 0.5143523812294006,
+      "learning_rate": 0.00031881205259712384,
+      "loss": 0.9542,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8225849611988226,
+      "grad_norm": 0.5969739556312561,
+      "learning_rate": 0.00031874469389730884,
+      "loss": 0.9868,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8227633574168227,
+      "grad_norm": 0.5791580677032471,
+      "learning_rate": 0.0003186773297983153,
+      "loss": 0.8447,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8229417536348229,
+      "grad_norm": 0.4807104766368866,
+      "learning_rate": 0.00031860996030543383,
+      "loss": 0.9425,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8231201498528231,
+      "grad_norm": 0.48759788274765015,
+      "learning_rate": 0.00031854258542395546,
+      "loss": 0.7504,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8232985460708233,
+      "grad_norm": 0.5369093418121338,
+      "learning_rate": 0.00031847520515917207,
+      "loss": 0.8581,
+      "step": 4615
+    },
+    {
+      "epoch": 0.8234769422888235,
+      "grad_norm": 0.5256065130233765,
+      "learning_rate": 0.00031840781951637554,
+      "loss": 0.8261,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8236553385068237,
+      "grad_norm": 0.47315025329589844,
+      "learning_rate": 0.0003183404285008582,
+      "loss": 0.786,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8238337347248238,
+      "grad_norm": 0.4942631125450134,
+      "learning_rate": 0.00031827303211791314,
+      "loss": 0.8627,
+      "step": 4618
+    },
+    {
+      "epoch": 0.824012130942824,
+      "grad_norm": 0.4402255713939667,
+      "learning_rate": 0.0003182056303728334,
+      "loss": 0.7299,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8241905271608242,
+      "grad_norm": 0.5043277740478516,
+      "learning_rate": 0.00031813822327091286,
+      "loss": 0.8561,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8243689233788244,
+      "grad_norm": 0.49375712871551514,
+      "learning_rate": 0.0003180708108174456,
+      "loss": 0.8253,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8245473195968246,
+      "grad_norm": 1.158111572265625,
+      "learning_rate": 0.00031800339301772614,
+      "loss": 0.844,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8247257158148247,
+      "grad_norm": 0.5060315728187561,
+      "learning_rate": 0.0003179359698770494,
+      "loss": 0.8552,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8249041120328249,
+      "grad_norm": 0.5320479273796082,
+      "learning_rate": 0.00031786854140071084,
+      "loss": 0.9473,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8250825082508251,
+      "grad_norm": 0.5282738208770752,
+      "learning_rate": 0.00031780110759400634,
+      "loss": 0.9197,
+      "step": 4625
+    },
+    {
+      "epoch": 0.8252609044688253,
+      "grad_norm": 0.4730012118816376,
+      "learning_rate": 0.00031773366846223197,
+      "loss": 0.7853,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8254393006868255,
+      "grad_norm": 0.5142650008201599,
+      "learning_rate": 0.00031766622401068433,
+      "loss": 0.8487,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8256176969048257,
+      "grad_norm": 0.4720161259174347,
+      "learning_rate": 0.0003175987742446607,
+      "loss": 0.8038,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8257960931228258,
+      "grad_norm": 0.5759044885635376,
+      "learning_rate": 0.00031753131916945835,
+      "loss": 0.712,
+      "step": 4629
+    },
+    {
+      "epoch": 0.825974489340826,
+      "grad_norm": 0.6872929930686951,
+      "learning_rate": 0.0003174638587903753,
+      "loss": 0.9687,
+      "step": 4630
+    },
+    {
+      "epoch": 0.8261528855588262,
+      "grad_norm": 0.6308576464653015,
+      "learning_rate": 0.0003173963931127099,
+      "loss": 0.8172,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8263312817768264,
+      "grad_norm": 0.5117534399032593,
+      "learning_rate": 0.0003173289221417606,
+      "loss": 0.9426,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8265096779948266,
+      "grad_norm": 0.5493487119674683,
+      "learning_rate": 0.00031726144588282686,
+      "loss": 0.9628,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8266880742128266,
+      "grad_norm": 3.2146034240722656,
+      "learning_rate": 0.0003171939643412081,
+      "loss": 0.9362,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8268664704308268,
+      "grad_norm": 0.49968186020851135,
+      "learning_rate": 0.00031712647752220427,
+      "loss": 0.6978,
+      "step": 4635
+    },
+    {
+      "epoch": 0.827044866648827,
+      "grad_norm": 0.5384466052055359,
+      "learning_rate": 0.0003170589854311157,
+      "loss": 1.0049,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8272232628668272,
+      "grad_norm": 0.547645628452301,
+      "learning_rate": 0.0003169914880732434,
+      "loss": 1.1843,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8274016590848274,
+      "grad_norm": 0.4683157503604889,
+      "learning_rate": 0.0003169239854538884,
+      "loss": 0.7661,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8275800553028276,
+      "grad_norm": 0.5019885301589966,
+      "learning_rate": 0.0003168564775783523,
+      "loss": 0.8682,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8277584515208277,
+      "grad_norm": 0.6580917835235596,
+      "learning_rate": 0.0003167889644519374,
+      "loss": 0.8636,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8279368477388279,
+      "grad_norm": 0.5439825654029846,
+      "learning_rate": 0.00031672144607994583,
+      "loss": 0.8918,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8281152439568281,
+      "grad_norm": 0.5139893889427185,
+      "learning_rate": 0.00031665392246768066,
+      "loss": 0.9374,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8282936401748283,
+      "grad_norm": 0.4646841883659363,
+      "learning_rate": 0.00031658639362044515,
+      "loss": 0.744,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8284720363928285,
+      "grad_norm": 0.4877662658691406,
+      "learning_rate": 0.00031651885954354285,
+      "loss": 0.635,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8286504326108286,
+      "grad_norm": 0.5172538757324219,
+      "learning_rate": 0.00031645132024227794,
+      "loss": 0.9473,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 0.8445831537246704,
+      "learning_rate": 0.000316383775721955,
+      "loss": 0.7455,
+      "step": 4646
+    },
+    {
+      "epoch": 0.829007225046829,
+      "grad_norm": 0.5219873785972595,
+      "learning_rate": 0.0003163162259878788,
+      "loss": 0.9679,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8291856212648292,
+      "grad_norm": 0.5205322504043579,
+      "learning_rate": 0.0003162486710453548,
+      "loss": 0.7856,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8293640174828294,
+      "grad_norm": 0.4920894503593445,
+      "learning_rate": 0.0003161811108996888,
+      "loss": 0.7642,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8295424137008296,
+      "grad_norm": 0.4882299602031708,
+      "learning_rate": 0.00031611354555618673,
+      "loss": 0.7915,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8297208099188297,
+      "grad_norm": 0.5416523814201355,
+      "learning_rate": 0.0003160459750201552,
+      "loss": 1.1522,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8298992061368299,
+      "grad_norm": 0.4422816038131714,
+      "learning_rate": 0.0003159783992969012,
+      "loss": 0.7744,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8300776023548301,
+      "grad_norm": 0.5312891006469727,
+      "learning_rate": 0.0003159108183917321,
+      "loss": 1.0687,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8302559985728303,
+      "grad_norm": 0.4771096706390381,
+      "learning_rate": 0.00031584323230995584,
+      "loss": 0.8311,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8304343947908305,
+      "grad_norm": 0.5976044535636902,
+      "learning_rate": 0.0003157756410568803,
+      "loss": 0.6884,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8306127910088306,
+      "grad_norm": 0.6168167591094971,
+      "learning_rate": 0.0003157080446378143,
+      "loss": 0.8282,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8307911872268308,
+      "grad_norm": 0.5120362639427185,
+      "learning_rate": 0.0003156404430580667,
+      "loss": 1.0674,
+      "step": 4657
+    },
+    {
+      "epoch": 0.830969583444831,
+      "grad_norm": 0.5644657015800476,
+      "learning_rate": 0.000315572836322947,
+      "loss": 1.0913,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8311479796628312,
+      "grad_norm": 0.4254976511001587,
+      "learning_rate": 0.00031550522443776497,
+      "loss": 0.8372,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8313263758808314,
+      "grad_norm": 0.4404276907444,
+      "learning_rate": 0.0003154376074078307,
+      "loss": 0.8619,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8315047720988316,
+      "grad_norm": 0.4840773046016693,
+      "learning_rate": 0.00031536998523845497,
+      "loss": 0.87,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8316831683168316,
+      "grad_norm": 0.4883440136909485,
+      "learning_rate": 0.0003153023579349487,
+      "loss": 1.0516,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8318615645348318,
+      "grad_norm": 0.39201605319976807,
+      "learning_rate": 0.0003152347255026234,
+      "loss": 0.6851,
+      "step": 4663
+    },
+    {
+      "epoch": 0.832039960752832,
+      "grad_norm": 0.4473479688167572,
+      "learning_rate": 0.0003151670879467908,
+      "loss": 0.6324,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8322183569708322,
+      "grad_norm": 1.3589231967926025,
+      "learning_rate": 0.0003150994452727631,
+      "loss": 0.954,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8323967531888324,
+      "grad_norm": 0.44502779841423035,
+      "learning_rate": 0.00031503179748585303,
+      "loss": 0.6668,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8325751494068325,
+      "grad_norm": 0.49202197790145874,
+      "learning_rate": 0.0003149641445913736,
+      "loss": 0.8664,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8327535456248327,
+      "grad_norm": 0.6371123790740967,
+      "learning_rate": 0.0003148964865946381,
+      "loss": 0.7495,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8329319418428329,
+      "grad_norm": 0.5421488881111145,
+      "learning_rate": 0.00031482882350096063,
+      "loss": 1.1947,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8331103380608331,
+      "grad_norm": 0.49121710658073425,
+      "learning_rate": 0.0003147611553156552,
+      "loss": 0.8075,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8332887342788333,
+      "grad_norm": 0.5611662268638611,
+      "learning_rate": 0.00031469348204403647,
+      "loss": 0.9551,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8334671304968335,
+      "grad_norm": 0.4589937627315521,
+      "learning_rate": 0.0003146258036914195,
+      "loss": 0.8752,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8336455267148336,
+      "grad_norm": 0.49224385619163513,
+      "learning_rate": 0.0003145581202631197,
+      "loss": 0.7851,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8338239229328338,
+      "grad_norm": 0.4641133248806,
+      "learning_rate": 0.00031449043176445297,
+      "loss": 0.8309,
+      "step": 4674
+    },
+    {
+      "epoch": 0.834002319150834,
+      "grad_norm": 0.46971380710601807,
+      "learning_rate": 0.0003144227382007355,
+      "loss": 0.5713,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8341807153688342,
+      "grad_norm": 0.6246667504310608,
+      "learning_rate": 0.00031435503957728383,
+      "loss": 0.9682,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8343591115868344,
+      "grad_norm": 0.5531001091003418,
+      "learning_rate": 0.00031428733589941506,
+      "loss": 0.9917,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8345375078048345,
+      "grad_norm": 0.5050147175788879,
+      "learning_rate": 0.00031421962717244654,
+      "loss": 0.8575,
+      "step": 4678
+    },
+    {
+      "epoch": 0.8347159040228347,
+      "grad_norm": 0.5118961930274963,
+      "learning_rate": 0.0003141519134016962,
+      "loss": 0.8731,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8348943002408349,
+      "grad_norm": 0.4847746789455414,
+      "learning_rate": 0.0003140841945924822,
+      "loss": 0.8582,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8350726964588351,
+      "grad_norm": 0.4473601281642914,
+      "learning_rate": 0.0003140164707501232,
+      "loss": 0.768,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8352510926768353,
+      "grad_norm": 0.46980252861976624,
+      "learning_rate": 0.00031394874187993805,
+      "loss": 0.9835,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8354294888948355,
+      "grad_norm": 0.42925623059272766,
+      "learning_rate": 0.00031388100798724624,
+      "loss": 0.7166,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8356078851128356,
+      "grad_norm": 0.4442448019981384,
+      "learning_rate": 0.0003138132690773675,
+      "loss": 0.8631,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8357862813308358,
+      "grad_norm": 0.5010333061218262,
+      "learning_rate": 0.00031374552515562215,
+      "loss": 0.7902,
+      "step": 4685
+    },
+    {
+      "epoch": 0.835964677548836,
+      "grad_norm": 0.473787784576416,
+      "learning_rate": 0.0003136777762273306,
+      "loss": 0.9016,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8361430737668362,
+      "grad_norm": 0.4981488287448883,
+      "learning_rate": 0.000313610022297814,
+      "loss": 0.8935,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8363214699848364,
+      "grad_norm": 0.5715591907501221,
+      "learning_rate": 0.0003135422633723936,
+      "loss": 0.9939,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8364998662028365,
+      "grad_norm": 0.587060809135437,
+      "learning_rate": 0.0003134744994563912,
+      "loss": 1.0045,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8366782624208366,
+      "grad_norm": 0.4668729901313782,
+      "learning_rate": 0.0003134067305551289,
+      "loss": 0.9922,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8368566586388368,
+      "grad_norm": 0.44394639134407043,
+      "learning_rate": 0.0003133389566739292,
+      "loss": 0.8563,
+      "step": 4691
+    },
+    {
+      "epoch": 0.837035054856837,
+      "grad_norm": 0.41054239869117737,
+      "learning_rate": 0.0003132711778181152,
+      "loss": 0.6808,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8372134510748372,
+      "grad_norm": 0.45655539631843567,
+      "learning_rate": 0.00031320339399301005,
+      "loss": 0.7885,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8373918472928374,
+      "grad_norm": 0.528401255607605,
+      "learning_rate": 0.00031313560520393756,
+      "loss": 1.3023,
+      "step": 4694
+    },
+    {
+      "epoch": 0.8375702435108375,
+      "grad_norm": 0.5124956369400024,
+      "learning_rate": 0.0003130678114562218,
+      "loss": 0.9328,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8377486397288377,
+      "grad_norm": 0.5178223252296448,
+      "learning_rate": 0.00031300001275518733,
+      "loss": 1.0182,
+      "step": 4696
+    },
+    {
+      "epoch": 0.8379270359468379,
+      "grad_norm": 0.49288439750671387,
+      "learning_rate": 0.00031293220910615896,
+      "loss": 1.0634,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8381054321648381,
+      "grad_norm": 0.47642451524734497,
+      "learning_rate": 0.00031286440051446187,
+      "loss": 0.9013,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8382838283828383,
+      "grad_norm": 0.44043177366256714,
+      "learning_rate": 0.0003127965869854219,
+      "loss": 0.7628,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8384622246008384,
+      "grad_norm": 0.4588578939437866,
+      "learning_rate": 0.00031272876852436493,
+      "loss": 0.7756,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8386406208188386,
+      "grad_norm": 0.4768531620502472,
+      "learning_rate": 0.0003126609451366176,
+      "loss": 0.7546,
+      "step": 4701
+    },
+    {
+      "epoch": 0.8388190170368388,
+      "grad_norm": 0.4645833671092987,
+      "learning_rate": 0.00031259311682750655,
+      "loss": 0.7292,
+      "step": 4702
+    },
+    {
+      "epoch": 0.838997413254839,
+      "grad_norm": 0.49236834049224854,
+      "learning_rate": 0.00031252528360235907,
+      "loss": 0.884,
+      "step": 4703
+    },
+    {
+      "epoch": 0.8391758094728392,
+      "grad_norm": 0.5257160067558289,
+      "learning_rate": 0.0003124574454665027,
+      "loss": 0.9921,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8393542056908394,
+      "grad_norm": 2.2589757442474365,
+      "learning_rate": 0.0003123896024252654,
+      "loss": 0.7997,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8395326019088395,
+      "grad_norm": 0.4541998505592346,
+      "learning_rate": 0.00031232175448397547,
+      "loss": 0.7263,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8397109981268397,
+      "grad_norm": 0.8749123215675354,
+      "learning_rate": 0.00031225390164796193,
+      "loss": 0.7226,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8398893943448399,
+      "grad_norm": 0.46198785305023193,
+      "learning_rate": 0.0003121860439225537,
+      "loss": 0.7642,
+      "step": 4708
+    },
+    {
+      "epoch": 0.8400677905628401,
+      "grad_norm": 0.4428096115589142,
+      "learning_rate": 0.0003121181813130804,
+      "loss": 0.7683,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8402461867808403,
+      "grad_norm": 0.4557209014892578,
+      "learning_rate": 0.0003120503138248718,
+      "loss": 0.6965,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8404245829988404,
+      "grad_norm": 0.498859167098999,
+      "learning_rate": 0.0003119824414632583,
+      "loss": 0.7745,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8406029792168406,
+      "grad_norm": 0.44227057695388794,
+      "learning_rate": 0.00031191456423357045,
+      "loss": 0.8255,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8407813754348408,
+      "grad_norm": 0.5205174684524536,
+      "learning_rate": 0.0003118466821411394,
+      "loss": 1.0533,
+      "step": 4713
+    },
+    {
+      "epoch": 0.840959771652841,
+      "grad_norm": 0.48816418647766113,
+      "learning_rate": 0.0003117787951912965,
+      "loss": 0.8577,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8411381678708412,
+      "grad_norm": 0.5432656407356262,
+      "learning_rate": 0.00031171090338937376,
+      "loss": 1.0221,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8413165640888414,
+      "grad_norm": 0.4946669340133667,
+      "learning_rate": 0.0003116430067407031,
+      "loss": 0.8704,
+      "step": 4716
+    },
+    {
+      "epoch": 0.8414949603068415,
+      "grad_norm": 0.4849919378757477,
+      "learning_rate": 0.0003115751052506173,
+      "loss": 0.8248,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8416733565248417,
+      "grad_norm": 0.5005682110786438,
+      "learning_rate": 0.0003115071989244491,
+      "loss": 0.9195,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8418517527428419,
+      "grad_norm": 0.5993044972419739,
+      "learning_rate": 0.00031143928776753213,
+      "loss": 0.8419,
+      "step": 4719
+    },
+    {
+      "epoch": 0.842030148960842,
+      "grad_norm": 0.704413115978241,
+      "learning_rate": 0.0003113713717851998,
+      "loss": 0.9343,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8422085451788422,
+      "grad_norm": 0.5452966094017029,
+      "learning_rate": 0.0003113034509827864,
+      "loss": 0.7155,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8423869413968423,
+      "grad_norm": 0.5383533835411072,
+      "learning_rate": 0.0003112355253656263,
+      "loss": 0.7432,
+      "step": 4722
+    },
+    {
+      "epoch": 0.8425653376148425,
+      "grad_norm": 0.624819278717041,
+      "learning_rate": 0.00031116759493905445,
+      "loss": 0.8583,
+      "step": 4723
+    },
+    {
+      "epoch": 0.8427437338328427,
+      "grad_norm": 0.982383668422699,
+      "learning_rate": 0.000311099659708406,
+      "loss": 0.6117,
+      "step": 4724
+    },
+    {
+      "epoch": 0.8429221300508429,
+      "grad_norm": 0.6751213669776917,
+      "learning_rate": 0.00031103171967901655,
+      "loss": 1.022,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8431005262688431,
+      "grad_norm": 1.0229430198669434,
+      "learning_rate": 0.00031096377485622214,
+      "loss": 0.7461,
+      "step": 4726
+    },
+    {
+      "epoch": 0.8432789224868433,
+      "grad_norm": 0.5392288565635681,
+      "learning_rate": 0.000310895825245359,
+      "loss": 0.7769,
+      "step": 4727
+    },
+    {
+      "epoch": 0.8434573187048434,
+      "grad_norm": 1.207713007926941,
+      "learning_rate": 0.0003108278708517641,
+      "loss": 0.5888,
+      "step": 4728
+    },
+    {
+      "epoch": 0.8436357149228436,
+      "grad_norm": 1.1378090381622314,
+      "learning_rate": 0.0003107599116807743,
+      "loss": 0.7526,
+      "step": 4729
+    },
+    {
+      "epoch": 0.8438141111408438,
+      "grad_norm": 0.7071771621704102,
+      "learning_rate": 0.00031069194773772715,
+      "loss": 1.0347,
+      "step": 4730
+    },
+    {
+      "epoch": 0.843992507358844,
+      "grad_norm": 0.7279558181762695,
+      "learning_rate": 0.0003106239790279606,
+      "loss": 0.8911,
+      "step": 4731
+    },
+    {
+      "epoch": 0.8441709035768442,
+      "grad_norm": 0.49013325572013855,
+      "learning_rate": 0.0003105560055568128,
+      "loss": 0.6986,
+      "step": 4732
+    },
+    {
+      "epoch": 0.8443492997948443,
+      "grad_norm": 0.5166245698928833,
+      "learning_rate": 0.0003104880273296224,
+      "loss": 1.0091,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8445276960128445,
+      "grad_norm": 0.5410045385360718,
+      "learning_rate": 0.00031042004435172834,
+      "loss": 1.0074,
+      "step": 4734
+    },
+    {
+      "epoch": 0.8447060922308447,
+      "grad_norm": 0.40799999237060547,
+      "learning_rate": 0.00031035205662847005,
+      "loss": 0.7189,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8448844884488449,
+      "grad_norm": 0.468019038438797,
+      "learning_rate": 0.0003102840641651872,
+      "loss": 0.9389,
+      "step": 4736
+    },
+    {
+      "epoch": 0.8450628846668451,
+      "grad_norm": 0.4619864225387573,
+      "learning_rate": 0.00031021606696721984,
+      "loss": 0.7924,
+      "step": 4737
+    },
+    {
+      "epoch": 0.8452412808848453,
+      "grad_norm": 1.7745176553726196,
+      "learning_rate": 0.0003101480650399085,
+      "loss": 0.8259,
+      "step": 4738
+    },
+    {
+      "epoch": 0.8454196771028454,
+      "grad_norm": 0.5147114396095276,
+      "learning_rate": 0.000310080058388594,
+      "loss": 0.7269,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8455980733208456,
+      "grad_norm": 1.0041489601135254,
+      "learning_rate": 0.00031001204701861765,
+      "loss": 0.8016,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8457764695388458,
+      "grad_norm": 0.5194323062896729,
+      "learning_rate": 0.00030994403093532086,
+      "loss": 0.8482,
+      "step": 4741
+    },
+    {
+      "epoch": 0.845954865756846,
+      "grad_norm": 1.3192564249038696,
+      "learning_rate": 0.00030987601014404576,
+      "loss": 0.9586,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8461332619748462,
+      "grad_norm": 0.5231749415397644,
+      "learning_rate": 0.00030980798465013454,
+      "loss": 0.9005,
+      "step": 4743
+    },
+    {
+      "epoch": 0.8463116581928463,
+      "grad_norm": 0.52949458360672,
+      "learning_rate": 0.00030973995445892987,
+      "loss": 0.9448,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8464900544108465,
+      "grad_norm": 0.509884238243103,
+      "learning_rate": 0.00030967191957577503,
+      "loss": 0.7253,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8466684506288467,
+      "grad_norm": 0.4490432143211365,
+      "learning_rate": 0.00030960388000601325,
+      "loss": 0.7781,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8468468468468469,
+      "grad_norm": 0.5485023856163025,
+      "learning_rate": 0.0003095358357549883,
+      "loss": 0.8696,
+      "step": 4747
+    },
+    {
+      "epoch": 0.847025243064847,
+      "grad_norm": 0.46540600061416626,
+      "learning_rate": 0.00030946778682804457,
+      "loss": 0.837,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8472036392828473,
+      "grad_norm": 0.46531352400779724,
+      "learning_rate": 0.0003093997332305264,
+      "loss": 0.879,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8473820355008473,
+      "grad_norm": 0.5620687007904053,
+      "learning_rate": 0.00030933167496777875,
+      "loss": 1.2353,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8475604317188475,
+      "grad_norm": 0.4240880310535431,
+      "learning_rate": 0.0003092636120451469,
+      "loss": 0.6647,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8477388279368477,
+      "grad_norm": 0.5051021575927734,
+      "learning_rate": 0.0003091955444679763,
+      "loss": 0.9274,
+      "step": 4752
+    },
+    {
+      "epoch": 0.8479172241548479,
+      "grad_norm": 0.47416186332702637,
+      "learning_rate": 0.00030912747224161324,
+      "loss": 0.7746,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8480956203728481,
+      "grad_norm": 0.4895223081111908,
+      "learning_rate": 0.000309059395371404,
+      "loss": 0.7968,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8482740165908482,
+      "grad_norm": 0.5074788331985474,
+      "learning_rate": 0.00030899131386269527,
+      "loss": 1.1851,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8484524128088484,
+      "grad_norm": 0.45386070013046265,
+      "learning_rate": 0.000308923227720834,
+      "loss": 0.8569,
+      "step": 4756
+    },
+    {
+      "epoch": 0.8486308090268486,
+      "grad_norm": 0.5263837575912476,
+      "learning_rate": 0.0003088551369511679,
+      "loss": 0.9021,
+      "step": 4757
+    },
+    {
+      "epoch": 0.8488092052448488,
+      "grad_norm": 0.6763511896133423,
+      "learning_rate": 0.00030878704155904465,
+      "loss": 0.8721,
+      "step": 4758
+    },
+    {
+      "epoch": 0.848987601462849,
+      "grad_norm": 0.48090678453445435,
+      "learning_rate": 0.0003087189415498124,
+      "loss": 0.8649,
+      "step": 4759
+    },
+    {
+      "epoch": 0.8491659976808492,
+      "grad_norm": 0.44986963272094727,
+      "learning_rate": 0.0003086508369288198,
+      "loss": 0.9613,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8493443938988493,
+      "grad_norm": 4.108095169067383,
+      "learning_rate": 0.00030858272770141574,
+      "loss": 0.7813,
+      "step": 4761
+    },
+    {
+      "epoch": 0.8495227901168495,
+      "grad_norm": 0.4738391935825348,
+      "learning_rate": 0.0003085146138729494,
+      "loss": 0.8247,
+      "step": 4762
+    },
+    {
+      "epoch": 0.8497011863348497,
+      "grad_norm": 0.45482027530670166,
+      "learning_rate": 0.0003084464954487705,
+      "loss": 0.7193,
+      "step": 4763
+    },
+    {
+      "epoch": 0.8498795825528499,
+      "grad_norm": 0.6398509740829468,
+      "learning_rate": 0.00030837837243422896,
+      "loss": 0.7891,
+      "step": 4764
+    },
+    {
+      "epoch": 0.8500579787708501,
+      "grad_norm": 0.45158737897872925,
+      "learning_rate": 0.00030831024483467517,
+      "loss": 0.7787,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8502363749888502,
+      "grad_norm": 0.46626511216163635,
+      "learning_rate": 0.00030824211265545985,
+      "loss": 0.8941,
+      "step": 4766
+    },
+    {
+      "epoch": 0.8504147712068504,
+      "grad_norm": 0.4571463167667389,
+      "learning_rate": 0.00030817397590193404,
+      "loss": 0.9276,
+      "step": 4767
+    },
+    {
+      "epoch": 0.8505931674248506,
+      "grad_norm": 0.4970678687095642,
+      "learning_rate": 0.0003081058345794493,
+      "loss": 0.8267,
+      "step": 4768
+    },
+    {
+      "epoch": 0.8507715636428508,
+      "grad_norm": 0.5197835564613342,
+      "learning_rate": 0.0003080376886933572,
+      "loss": 0.8656,
+      "step": 4769
+    },
+    {
+      "epoch": 0.850949959860851,
+      "grad_norm": 0.48982325196266174,
+      "learning_rate": 0.0003079695382490101,
+      "loss": 0.7117,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8511283560788512,
+      "grad_norm": 0.5008729696273804,
+      "learning_rate": 0.0003079013832517603,
+      "loss": 0.8803,
+      "step": 4771
+    },
+    {
+      "epoch": 0.8513067522968513,
+      "grad_norm": 0.4859359562397003,
+      "learning_rate": 0.00030783322370696087,
+      "loss": 0.8494,
+      "step": 4772
+    },
+    {
+      "epoch": 0.8514851485148515,
+      "grad_norm": 0.4618992209434509,
+      "learning_rate": 0.00030776505961996494,
+      "loss": 0.6753,
+      "step": 4773
+    },
+    {
+      "epoch": 0.8516635447328517,
+      "grad_norm": 0.6842380166053772,
+      "learning_rate": 0.00030769689099612604,
+      "loss": 0.8521,
+      "step": 4774
+    },
+    {
+      "epoch": 0.8518419409508519,
+      "grad_norm": 0.46096381545066833,
+      "learning_rate": 0.00030762871784079815,
+      "loss": 0.6602,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8520203371688521,
+      "grad_norm": 0.5396759510040283,
+      "learning_rate": 0.0003075605401593356,
+      "loss": 1.053,
+      "step": 4776
+    },
+    {
+      "epoch": 0.8521987333868521,
+      "grad_norm": 0.4654025435447693,
+      "learning_rate": 0.000307492357957093,
+      "loss": 0.6658,
+      "step": 4777
+    },
+    {
+      "epoch": 0.8523771296048523,
+      "grad_norm": 0.46096524596214294,
+      "learning_rate": 0.0003074241712394253,
+      "loss": 0.7662,
+      "step": 4778
+    },
+    {
+      "epoch": 0.8525555258228525,
+      "grad_norm": 0.5289954543113708,
+      "learning_rate": 0.0003073559800116879,
+      "loss": 0.7916,
+      "step": 4779
+    },
+    {
+      "epoch": 0.8527339220408527,
+      "grad_norm": 0.4729297459125519,
+      "learning_rate": 0.00030728778427923655,
+      "loss": 0.8108,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8529123182588529,
+      "grad_norm": 1.2165110111236572,
+      "learning_rate": 0.0003072195840474273,
+      "loss": 1.0334,
+      "step": 4781
+    },
+    {
+      "epoch": 0.8530907144768531,
+      "grad_norm": 0.6247953176498413,
+      "learning_rate": 0.00030715137932161646,
+      "loss": 1.1616,
+      "step": 4782
+    },
+    {
+      "epoch": 0.8532691106948532,
+      "grad_norm": 0.5182445645332336,
+      "learning_rate": 0.00030708317010716093,
+      "loss": 1.0503,
+      "step": 4783
+    },
+    {
+      "epoch": 0.8534475069128534,
+      "grad_norm": 0.5359323620796204,
+      "learning_rate": 0.0003070149564094178,
+      "loss": 1.0459,
+      "step": 4784
+    },
+    {
+      "epoch": 0.8536259031308536,
+      "grad_norm": 0.4704750180244446,
+      "learning_rate": 0.0003069467382337445,
+      "loss": 0.8514,
+      "step": 4785
+    },
+    {
+      "epoch": 0.8538042993488538,
+      "grad_norm": 0.5520228743553162,
+      "learning_rate": 0.0003068785155854989,
+      "loss": 0.8987,
+      "step": 4786
+    },
+    {
+      "epoch": 0.853982695566854,
+      "grad_norm": 0.4699079394340515,
+      "learning_rate": 0.0003068102884700391,
+      "loss": 0.736,
+      "step": 4787
+    },
+    {
+      "epoch": 0.8541610917848541,
+      "grad_norm": 0.5631700754165649,
+      "learning_rate": 0.00030674205689272375,
+      "loss": 0.9386,
+      "step": 4788
+    },
+    {
+      "epoch": 0.8543394880028543,
+      "grad_norm": 0.4554060101509094,
+      "learning_rate": 0.00030667382085891175,
+      "loss": 0.7084,
+      "step": 4789
+    },
+    {
+      "epoch": 0.8545178842208545,
+      "grad_norm": 0.46370020508766174,
+      "learning_rate": 0.00030660558037396216,
+      "loss": 0.9635,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8546962804388547,
+      "grad_norm": 0.46454691886901855,
+      "learning_rate": 0.0003065373354432346,
+      "loss": 0.8481,
+      "step": 4791
+    },
+    {
+      "epoch": 0.8548746766568549,
+      "grad_norm": 0.4899047017097473,
+      "learning_rate": 0.0003064690860720891,
+      "loss": 0.8236,
+      "step": 4792
+    },
+    {
+      "epoch": 0.8550530728748551,
+      "grad_norm": 0.9299049377441406,
+      "learning_rate": 0.0003064008322658859,
+      "loss": 0.7952,
+      "step": 4793
+    },
+    {
+      "epoch": 0.8552314690928552,
+      "grad_norm": 1.665650725364685,
+      "learning_rate": 0.0003063325740299855,
+      "loss": 1.0987,
+      "step": 4794
+    },
+    {
+      "epoch": 0.8554098653108554,
+      "grad_norm": 0.4786631464958191,
+      "learning_rate": 0.0003062643113697492,
+      "loss": 0.8729,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8555882615288556,
+      "grad_norm": 1.025985836982727,
+      "learning_rate": 0.00030619604429053793,
+      "loss": 0.891,
+      "step": 4796
+    },
+    {
+      "epoch": 0.8557666577468558,
+      "grad_norm": 0.5616780519485474,
+      "learning_rate": 0.0003061277727977135,
+      "loss": 1.0594,
+      "step": 4797
+    },
+    {
+      "epoch": 0.855945053964856,
+      "grad_norm": 0.4509812593460083,
+      "learning_rate": 0.000306059496896638,
+      "loss": 0.6073,
+      "step": 4798
+    },
+    {
+      "epoch": 0.8561234501828561,
+      "grad_norm": 0.4322461783885956,
+      "learning_rate": 0.0003059912165926738,
+      "loss": 0.772,
+      "step": 4799
+    },
+    {
+      "epoch": 0.8563018464008563,
+      "grad_norm": 0.40975263714790344,
+      "learning_rate": 0.00030592293189118344,
+      "loss": 0.7681,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8564802426188565,
+      "grad_norm": 0.5190335512161255,
+      "learning_rate": 0.00030585464279753015,
+      "loss": 0.8196,
+      "step": 4801
+    },
+    {
+      "epoch": 0.8566586388368567,
+      "grad_norm": 0.5089114904403687,
+      "learning_rate": 0.0003057863493170772,
+      "loss": 0.7172,
+      "step": 4802
+    },
+    {
+      "epoch": 0.8568370350548569,
+      "grad_norm": 0.548192024230957,
+      "learning_rate": 0.0003057180514551884,
+      "loss": 0.8321,
+      "step": 4803
+    },
+    {
+      "epoch": 0.8570154312728571,
+      "grad_norm": 0.4767385721206665,
+      "learning_rate": 0.0003056497492172278,
+      "loss": 0.8879,
+      "step": 4804
+    },
+    {
+      "epoch": 0.8571938274908572,
+      "grad_norm": 0.488613098859787,
+      "learning_rate": 0.00030558144260855986,
+      "loss": 0.9423,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8573722237088574,
+      "grad_norm": 0.48882797360420227,
+      "learning_rate": 0.0003055131316345493,
+      "loss": 0.7837,
+      "step": 4806
+    },
+    {
+      "epoch": 0.8575506199268575,
+      "grad_norm": 0.5603160262107849,
+      "learning_rate": 0.0003054448163005613,
+      "loss": 0.8026,
+      "step": 4807
+    },
+    {
+      "epoch": 0.8577290161448577,
+      "grad_norm": 0.46545833349227905,
+      "learning_rate": 0.00030537649661196135,
+      "loss": 0.8075,
+      "step": 4808
+    },
+    {
+      "epoch": 0.857907412362858,
+      "grad_norm": 0.4614555537700653,
+      "learning_rate": 0.00030530817257411517,
+      "loss": 0.7662,
+      "step": 4809
+    },
+    {
+      "epoch": 0.858085808580858,
+      "grad_norm": 0.5005578398704529,
+      "learning_rate": 0.0003052398441923888,
+      "loss": 0.9496,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8582642047988582,
+      "grad_norm": 0.45669299364089966,
+      "learning_rate": 0.00030517151147214895,
+      "loss": 0.7207,
+      "step": 4811
+    },
+    {
+      "epoch": 0.8584426010168584,
+      "grad_norm": 0.4442518949508667,
+      "learning_rate": 0.0003051031744187623,
+      "loss": 0.7858,
+      "step": 4812
+    },
+    {
+      "epoch": 0.8586209972348586,
+      "grad_norm": 0.4947311580181122,
+      "learning_rate": 0.00030503483303759597,
+      "loss": 1.0889,
+      "step": 4813
+    },
+    {
+      "epoch": 0.8587993934528588,
+      "grad_norm": 0.4674645960330963,
+      "learning_rate": 0.00030496648733401764,
+      "loss": 0.7326,
+      "step": 4814
+    },
+    {
+      "epoch": 0.858977789670859,
+      "grad_norm": 0.49321508407592773,
+      "learning_rate": 0.00030489813731339504,
+      "loss": 0.8671,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8591561858888591,
+      "grad_norm": 0.48469820618629456,
+      "learning_rate": 0.00030482978298109636,
+      "loss": 1.0099,
+      "step": 4816
+    },
+    {
+      "epoch": 0.8593345821068593,
+      "grad_norm": 0.5168629884719849,
+      "learning_rate": 0.0003047614243424901,
+      "loss": 1.0685,
+      "step": 4817
+    },
+    {
+      "epoch": 0.8595129783248595,
+      "grad_norm": 0.5016402006149292,
+      "learning_rate": 0.0003046930614029451,
+      "loss": 0.9678,
+      "step": 4818
+    },
+    {
+      "epoch": 0.8596913745428597,
+      "grad_norm": 0.458768367767334,
+      "learning_rate": 0.00030462469416783067,
+      "loss": 0.8762,
+      "step": 4819
+    },
+    {
+      "epoch": 0.8598697707608599,
+      "grad_norm": 0.6688326597213745,
+      "learning_rate": 0.0003045563226425162,
+      "loss": 0.9441,
+      "step": 4820
+    },
+    {
+      "epoch": 0.86004816697886,
+      "grad_norm": 0.5008928179740906,
+      "learning_rate": 0.0003044879468323716,
+      "loss": 0.7504,
+      "step": 4821
+    },
+    {
+      "epoch": 0.8602265631968602,
+      "grad_norm": 2.919682502746582,
+      "learning_rate": 0.0003044195667427672,
+      "loss": 1.1196,
+      "step": 4822
+    },
+    {
+      "epoch": 0.8604049594148604,
+      "grad_norm": 1.0007497072219849,
+      "learning_rate": 0.0003043511823790734,
+      "loss": 0.638,
+      "step": 4823
+    },
+    {
+      "epoch": 0.8605833556328606,
+      "grad_norm": 0.5225825905799866,
+      "learning_rate": 0.00030428279374666113,
+      "loss": 0.8362,
+      "step": 4824
+    },
+    {
+      "epoch": 0.8607617518508608,
+      "grad_norm": 0.533996045589447,
+      "learning_rate": 0.0003042144008509016,
+      "loss": 0.8759,
+      "step": 4825
+    },
+    {
+      "epoch": 0.860940148068861,
+      "grad_norm": 0.45570921897888184,
+      "learning_rate": 0.00030414600369716636,
+      "loss": 0.7521,
+      "step": 4826
+    },
+    {
+      "epoch": 0.8611185442868611,
+      "grad_norm": 2.225616216659546,
+      "learning_rate": 0.0003040776022908273,
+      "loss": 0.8446,
+      "step": 4827
+    },
+    {
+      "epoch": 0.8612969405048613,
+      "grad_norm": 0.5153058767318726,
+      "learning_rate": 0.00030400919663725655,
+      "loss": 1.0429,
+      "step": 4828
+    },
+    {
+      "epoch": 0.8614753367228615,
+      "grad_norm": 0.44439637660980225,
+      "learning_rate": 0.00030394078674182684,
+      "loss": 0.8713,
+      "step": 4829
+    },
+    {
+      "epoch": 0.8616537329408617,
+      "grad_norm": 0.5092588663101196,
+      "learning_rate": 0.0003038723726099109,
+      "loss": 0.7983,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8618321291588619,
+      "grad_norm": 0.5052730441093445,
+      "learning_rate": 0.000303803954246882,
+      "loss": 0.9045,
+      "step": 4831
+    },
+    {
+      "epoch": 0.862010525376862,
+      "grad_norm": 0.5321127772331238,
+      "learning_rate": 0.00030373553165811377,
+      "loss": 0.9417,
+      "step": 4832
+    },
+    {
+      "epoch": 0.8621889215948622,
+      "grad_norm": 0.4712226688861847,
+      "learning_rate": 0.00030366710484897984,
+      "loss": 0.8645,
+      "step": 4833
+    },
+    {
+      "epoch": 0.8623673178128624,
+      "grad_norm": 0.5059930682182312,
+      "learning_rate": 0.0003035986738248547,
+      "loss": 0.9814,
+      "step": 4834
+    },
+    {
+      "epoch": 0.8625457140308626,
+      "grad_norm": 0.4837188124656677,
+      "learning_rate": 0.00030353023859111284,
+      "loss": 0.8914,
+      "step": 4835
+    },
+    {
+      "epoch": 0.8627241102488628,
+      "grad_norm": 0.4856887757778168,
+      "learning_rate": 0.0003034617991531289,
+      "loss": 0.9607,
+      "step": 4836
+    },
+    {
+      "epoch": 0.862902506466863,
+      "grad_norm": 0.5175963044166565,
+      "learning_rate": 0.0003033933555162784,
+      "loss": 0.9011,
+      "step": 4837
+    },
+    {
+      "epoch": 0.863080902684863,
+      "grad_norm": 0.4169839024543762,
+      "learning_rate": 0.0003033249076859367,
+      "loss": 0.7326,
+      "step": 4838
+    },
+    {
+      "epoch": 0.8632592989028632,
+      "grad_norm": 0.4630539119243622,
+      "learning_rate": 0.0003032564556674797,
+      "loss": 0.7955,
+      "step": 4839
+    },
+    {
+      "epoch": 0.8634376951208634,
+      "grad_norm": 0.47739243507385254,
+      "learning_rate": 0.0003031879994662836,
+      "loss": 0.8976,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8636160913388636,
+      "grad_norm": 0.5732921957969666,
+      "learning_rate": 0.00030311953908772495,
+      "loss": 0.9195,
+      "step": 4841
+    },
+    {
+      "epoch": 0.8637944875568638,
+      "grad_norm": 0.45784977078437805,
+      "learning_rate": 0.0003030510745371805,
+      "loss": 0.8676,
+      "step": 4842
+    },
+    {
+      "epoch": 0.8639728837748639,
+      "grad_norm": 0.40974244475364685,
+      "learning_rate": 0.00030298260582002753,
+      "loss": 0.6207,
+      "step": 4843
+    },
+    {
+      "epoch": 0.8641512799928641,
+      "grad_norm": 0.5293233394622803,
+      "learning_rate": 0.00030291413294164336,
+      "loss": 0.8833,
+      "step": 4844
+    },
+    {
+      "epoch": 0.8643296762108643,
+      "grad_norm": 0.5298712253570557,
+      "learning_rate": 0.00030284565590740607,
+      "loss": 0.8459,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8645080724288645,
+      "grad_norm": 0.5469498038291931,
+      "learning_rate": 0.00030277717472269373,
+      "loss": 0.8662,
+      "step": 4846
+    },
+    {
+      "epoch": 0.8646864686468647,
+      "grad_norm": 0.5502684116363525,
+      "learning_rate": 0.00030270868939288474,
+      "loss": 0.7796,
+      "step": 4847
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.5410329699516296,
+      "learning_rate": 0.00030264019992335805,
+      "loss": 0.6923,
+      "step": 4848
+    },
+    {
+      "epoch": 0.865043261082865,
+      "grad_norm": 0.5078060030937195,
+      "learning_rate": 0.00030257170631949265,
+      "loss": 0.8679,
+      "step": 4849
+    },
+    {
+      "epoch": 0.8652216573008652,
+      "grad_norm": 0.4987890422344208,
+      "learning_rate": 0.0003025032085866681,
+      "loss": 0.8628,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8654000535188654,
+      "grad_norm": 0.5421337485313416,
+      "learning_rate": 0.000302434706730264,
+      "loss": 0.8687,
+      "step": 4851
+    },
+    {
+      "epoch": 0.8655784497368656,
+      "grad_norm": 0.4919837415218353,
+      "learning_rate": 0.0003023662007556607,
+      "loss": 0.9298,
+      "step": 4852
+    },
+    {
+      "epoch": 0.8657568459548658,
+      "grad_norm": 0.4684261381626129,
+      "learning_rate": 0.0003022976906682385,
+      "loss": 0.7224,
+      "step": 4853
+    },
+    {
+      "epoch": 0.8659352421728659,
+      "grad_norm": 0.5145940780639648,
+      "learning_rate": 0.00030222917647337833,
+      "loss": 0.9305,
+      "step": 4854
+    },
+    {
+      "epoch": 0.8661136383908661,
+      "grad_norm": 0.5013492107391357,
+      "learning_rate": 0.00030216065817646097,
+      "loss": 0.9581,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8662920346088663,
+      "grad_norm": 0.4868033528327942,
+      "learning_rate": 0.000302092135782868,
+      "loss": 0.9121,
+      "step": 4856
+    },
+    {
+      "epoch": 0.8664704308268665,
+      "grad_norm": 0.5112666487693787,
+      "learning_rate": 0.0003020236092979811,
+      "loss": 0.9431,
+      "step": 4857
+    },
+    {
+      "epoch": 0.8666488270448667,
+      "grad_norm": 0.588441789150238,
+      "learning_rate": 0.0003019550787271823,
+      "loss": 0.9898,
+      "step": 4858
+    },
+    {
+      "epoch": 0.8668272232628669,
+      "grad_norm": 0.42368215322494507,
+      "learning_rate": 0.00030188654407585394,
+      "loss": 0.6658,
+      "step": 4859
+    },
+    {
+      "epoch": 0.867005619480867,
+      "grad_norm": 0.4669845700263977,
+      "learning_rate": 0.00030181800534937874,
+      "loss": 0.9835,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8671840156988672,
+      "grad_norm": 0.8458566665649414,
+      "learning_rate": 0.00030174946255313986,
+      "loss": 0.8751,
+      "step": 4861
+    },
+    {
+      "epoch": 0.8673624119168674,
+      "grad_norm": 0.43754637241363525,
+      "learning_rate": 0.0003016809156925203,
+      "loss": 0.7733,
+      "step": 4862
+    },
+    {
+      "epoch": 0.8675408081348676,
+      "grad_norm": 0.539728045463562,
+      "learning_rate": 0.00030161236477290387,
+      "loss": 1.0081,
+      "step": 4863
+    },
+    {
+      "epoch": 0.8677192043528678,
+      "grad_norm": 0.6148440837860107,
+      "learning_rate": 0.00030154380979967457,
+      "loss": 1.0732,
+      "step": 4864
+    },
+    {
+      "epoch": 0.8678976005708678,
+      "grad_norm": 0.5512346625328064,
+      "learning_rate": 0.0003014752507782166,
+      "loss": 0.9659,
+      "step": 4865
+    },
+    {
+      "epoch": 0.868075996788868,
+      "grad_norm": 0.4736751616001129,
+      "learning_rate": 0.0003014066877139146,
+      "loss": 0.567,
+      "step": 4866
+    },
+    {
+      "epoch": 0.8682543930068682,
+      "grad_norm": 0.9616851806640625,
+      "learning_rate": 0.00030133812061215346,
+      "loss": 0.7497,
+      "step": 4867
+    },
+    {
+      "epoch": 0.8684327892248684,
+      "grad_norm": 0.5108407139778137,
+      "learning_rate": 0.00030126954947831843,
+      "loss": 0.954,
+      "step": 4868
+    },
+    {
+      "epoch": 0.8686111854428686,
+      "grad_norm": 0.5192012786865234,
+      "learning_rate": 0.000301200974317795,
+      "loss": 0.8119,
+      "step": 4869
+    },
+    {
+      "epoch": 0.8687895816608688,
+      "grad_norm": 0.5100924372673035,
+      "learning_rate": 0.0003011323951359692,
+      "loss": 0.9276,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8689679778788689,
+      "grad_norm": 0.46075180172920227,
+      "learning_rate": 0.00030106381193822695,
+      "loss": 0.764,
+      "step": 4871
+    },
+    {
+      "epoch": 0.8691463740968691,
+      "grad_norm": 0.45364946126937866,
+      "learning_rate": 0.000300995224729955,
+      "loss": 0.7632,
+      "step": 4872
+    },
+    {
+      "epoch": 0.8693247703148693,
+      "grad_norm": 0.5815824270248413,
+      "learning_rate": 0.00030092663351654,
+      "loss": 0.9709,
+      "step": 4873
+    },
+    {
+      "epoch": 0.8695031665328695,
+      "grad_norm": 0.4578329920768738,
+      "learning_rate": 0.00030085803830336903,
+      "loss": 0.6986,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8696815627508697,
+      "grad_norm": 0.9438912272453308,
+      "learning_rate": 0.0003007894390958297,
+      "loss": 0.9468,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8698599589688698,
+      "grad_norm": 1.1074215173721313,
+      "learning_rate": 0.0003007208358993097,
+      "loss": 0.9185,
+      "step": 4876
+    },
+    {
+      "epoch": 0.87003835518687,
+      "grad_norm": 2.9237184524536133,
+      "learning_rate": 0.00030065222871919706,
+      "loss": 1.0407,
+      "step": 4877
+    },
+    {
+      "epoch": 0.8702167514048702,
+      "grad_norm": 0.5479034185409546,
+      "learning_rate": 0.00030058361756088014,
+      "loss": 1.0267,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8703951476228704,
+      "grad_norm": 0.5139360427856445,
+      "learning_rate": 0.0003005150024297477,
+      "loss": 0.8951,
+      "step": 4879
+    },
+    {
+      "epoch": 0.8705735438408706,
+      "grad_norm": 0.48596107959747314,
+      "learning_rate": 0.00030044638333118873,
+      "loss": 0.8362,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8707519400588708,
+      "grad_norm": 0.5204296708106995,
+      "learning_rate": 0.00030037776027059247,
+      "loss": 0.7555,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8709303362768709,
+      "grad_norm": 0.49391311407089233,
+      "learning_rate": 0.00030030913325334864,
+      "loss": 0.9305,
+      "step": 4882
+    },
+    {
+      "epoch": 0.8711087324948711,
+      "grad_norm": 0.4564768970012665,
+      "learning_rate": 0.00030024050228484714,
+      "loss": 0.7117,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8712871287128713,
+      "grad_norm": 0.434636652469635,
+      "learning_rate": 0.00030017186737047813,
+      "loss": 0.7125,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8714655249308715,
+      "grad_norm": 0.542251706123352,
+      "learning_rate": 0.00030010322851563233,
+      "loss": 1.061,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8716439211488717,
+      "grad_norm": 0.5167363882064819,
+      "learning_rate": 0.0003000345857257005,
+      "loss": 0.7555,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8718223173668718,
+      "grad_norm": 0.4324584901332855,
+      "learning_rate": 0.0002999659390060738,
+      "loss": 0.6161,
+      "step": 4887
+    },
+    {
+      "epoch": 0.872000713584872,
+      "grad_norm": 0.49977007508277893,
+      "learning_rate": 0.0002998972883621439,
+      "loss": 0.7964,
+      "step": 4888
+    },
+    {
+      "epoch": 0.8721791098028722,
+      "grad_norm": 0.5305691361427307,
+      "learning_rate": 0.00029982863379930224,
+      "loss": 1.1202,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8723575060208724,
+      "grad_norm": 0.5107898712158203,
+      "learning_rate": 0.0002997599753229412,
+      "loss": 1.0624,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8725359022388726,
+      "grad_norm": 0.5440460443496704,
+      "learning_rate": 0.00029969131293845313,
+      "loss": 1.0675,
+      "step": 4891
+    },
+    {
+      "epoch": 0.8727142984568728,
+      "grad_norm": 0.48299506306648254,
+      "learning_rate": 0.00029962264665123076,
+      "loss": 0.8612,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8728926946748728,
+      "grad_norm": 0.49671709537506104,
+      "learning_rate": 0.000299553976466667,
+      "loss": 0.9027,
+      "step": 4893
+    },
+    {
+      "epoch": 0.873071090892873,
+      "grad_norm": 0.5757230520248413,
+      "learning_rate": 0.00029948530239015534,
+      "loss": 0.9273,
+      "step": 4894
+    },
+    {
+      "epoch": 0.8732494871108732,
+      "grad_norm": 0.5014898180961609,
+      "learning_rate": 0.0002994166244270893,
+      "loss": 0.9075,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8734278833288734,
+      "grad_norm": 0.5563370585441589,
+      "learning_rate": 0.0002993479425828628,
+      "loss": 1.0369,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8736062795468736,
+      "grad_norm": 0.48165515065193176,
+      "learning_rate": 0.00029927925686287006,
+      "loss": 0.7797,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8737846757648737,
+      "grad_norm": 0.5211619734764099,
+      "learning_rate": 0.0002992105672725058,
+      "loss": 0.8904,
+      "step": 4898
+    },
+    {
+      "epoch": 0.8739630719828739,
+      "grad_norm": 0.5099582672119141,
+      "learning_rate": 0.00029914187381716473,
+      "loss": 0.7874,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8741414682008741,
+      "grad_norm": 0.48764798045158386,
+      "learning_rate": 0.00029907317650224204,
+      "loss": 0.6851,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8743198644188743,
+      "grad_norm": 0.5231783986091614,
+      "learning_rate": 0.0002990044753331332,
+      "loss": 0.7969,
+      "step": 4901
+    },
+    {
+      "epoch": 0.8744982606368745,
+      "grad_norm": 0.5043377876281738,
+      "learning_rate": 0.00029893577031523403,
+      "loss": 0.7843,
+      "step": 4902
+    },
+    {
+      "epoch": 0.8746766568548747,
+      "grad_norm": 0.4754558801651001,
+      "learning_rate": 0.0002988670614539404,
+      "loss": 0.7068,
+      "step": 4903
+    },
+    {
+      "epoch": 0.8748550530728748,
+      "grad_norm": 0.4813019037246704,
+      "learning_rate": 0.0002987983487546488,
+      "loss": 0.8119,
+      "step": 4904
+    },
+    {
+      "epoch": 0.875033449290875,
+      "grad_norm": 0.5080868601799011,
+      "learning_rate": 0.0002987296322227559,
+      "loss": 0.993,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8752118455088752,
+      "grad_norm": 0.5094708800315857,
+      "learning_rate": 0.00029866091186365865,
+      "loss": 1.0395,
+      "step": 4906
+    },
+    {
+      "epoch": 0.8753902417268754,
+      "grad_norm": 0.5148309469223022,
+      "learning_rate": 0.0002985921876827544,
+      "loss": 0.8391,
+      "step": 4907
+    },
+    {
+      "epoch": 0.8755686379448756,
+      "grad_norm": 0.47196096181869507,
+      "learning_rate": 0.00029852345968544057,
+      "loss": 0.7106,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8757470341628757,
+      "grad_norm": 0.5898258090019226,
+      "learning_rate": 0.00029845472787711516,
+      "loss": 0.8076,
+      "step": 4909
+    },
+    {
+      "epoch": 0.8759254303808759,
+      "grad_norm": 0.5204131007194519,
+      "learning_rate": 0.0002983859922631762,
+      "loss": 0.7482,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8761038265988761,
+      "grad_norm": 0.4837625026702881,
+      "learning_rate": 0.0002983172528490223,
+      "loss": 0.8764,
+      "step": 4911
+    },
+    {
+      "epoch": 0.8762822228168763,
+      "grad_norm": 0.46985456347465515,
+      "learning_rate": 0.00029824850964005215,
+      "loss": 0.6421,
+      "step": 4912
+    },
+    {
+      "epoch": 0.8764606190348765,
+      "grad_norm": 0.8495760560035706,
+      "learning_rate": 0.00029817976264166475,
+      "loss": 0.9803,
+      "step": 4913
+    },
+    {
+      "epoch": 0.8766390152528767,
+      "grad_norm": 0.5043582916259766,
+      "learning_rate": 0.00029811101185925955,
+      "loss": 1.0006,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8768174114708768,
+      "grad_norm": 13.705134391784668,
+      "learning_rate": 0.00029804225729823615,
+      "loss": 0.9303,
+      "step": 4915
+    },
+    {
+      "epoch": 0.876995807688877,
+      "grad_norm": 0.5199925899505615,
+      "learning_rate": 0.00029797349896399457,
+      "loss": 0.8513,
+      "step": 4916
+    },
+    {
+      "epoch": 0.8771742039068772,
+      "grad_norm": 0.5620130300521851,
+      "learning_rate": 0.000297904736861935,
+      "loss": 0.7335,
+      "step": 4917
+    },
+    {
+      "epoch": 0.8773526001248774,
+      "grad_norm": 1.1893454790115356,
+      "learning_rate": 0.0002978359709974581,
+      "loss": 0.9265,
+      "step": 4918
+    },
+    {
+      "epoch": 0.8775309963428776,
+      "grad_norm": 0.9579685926437378,
+      "learning_rate": 0.0002977672013759645,
+      "loss": 0.7829,
+      "step": 4919
+    },
+    {
+      "epoch": 0.8777093925608777,
+      "grad_norm": 0.8897736668586731,
+      "learning_rate": 0.0002976984280028556,
+      "loss": 0.8619,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8778877887788779,
+      "grad_norm": 0.6339399814605713,
+      "learning_rate": 0.00029762965088353256,
+      "loss": 0.9676,
+      "step": 4921
+    },
+    {
+      "epoch": 0.878066184996878,
+      "grad_norm": 0.5451995134353638,
+      "learning_rate": 0.00029756087002339734,
+      "loss": 1.0137,
+      "step": 4922
+    },
+    {
+      "epoch": 0.8782445812148783,
+      "grad_norm": 0.4819452166557312,
+      "learning_rate": 0.00029749208542785175,
+      "loss": 0.8346,
+      "step": 4923
+    },
+    {
+      "epoch": 0.8784229774328784,
+      "grad_norm": 0.500859260559082,
+      "learning_rate": 0.0002974232971022983,
+      "loss": 0.7695,
+      "step": 4924
+    },
+    {
+      "epoch": 0.8786013736508786,
+      "grad_norm": 0.7881564497947693,
+      "learning_rate": 0.00029735450505213943,
+      "loss": 1.0873,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8787797698688787,
+      "grad_norm": 1.3103078603744507,
+      "learning_rate": 0.0002972857092827781,
+      "loss": 0.889,
+      "step": 4926
+    },
+    {
+      "epoch": 0.8789581660868789,
+      "grad_norm": 0.5103085041046143,
+      "learning_rate": 0.00029721690979961764,
+      "loss": 0.7738,
+      "step": 4927
+    },
+    {
+      "epoch": 0.8791365623048791,
+      "grad_norm": 0.7508847117424011,
+      "learning_rate": 0.0002971481066080613,
+      "loss": 0.7429,
+      "step": 4928
+    },
+    {
+      "epoch": 0.8793149585228793,
+      "grad_norm": 0.5857597589492798,
+      "learning_rate": 0.000297079299713513,
+      "loss": 0.9272,
+      "step": 4929
+    },
+    {
+      "epoch": 0.8794933547408795,
+      "grad_norm": 0.5782235264778137,
+      "learning_rate": 0.00029701048912137676,
+      "loss": 0.9134,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8796717509588796,
+      "grad_norm": 0.5405158400535583,
+      "learning_rate": 0.00029694167483705684,
+      "loss": 0.8666,
+      "step": 4931
+    },
+    {
+      "epoch": 0.8798501471768798,
+      "grad_norm": 0.5647017359733582,
+      "learning_rate": 0.0002968728568659581,
+      "loss": 1.007,
+      "step": 4932
+    },
+    {
+      "epoch": 0.88002854339488,
+      "grad_norm": 0.537854015827179,
+      "learning_rate": 0.0002968040352134853,
+      "loss": 0.9957,
+      "step": 4933
+    },
+    {
+      "epoch": 0.8802069396128802,
+      "grad_norm": 0.5491927862167358,
+      "learning_rate": 0.00029673520988504376,
+      "loss": 0.9025,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8803853358308804,
+      "grad_norm": 0.5295958518981934,
+      "learning_rate": 0.0002966663808860389,
+      "loss": 0.9705,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8805637320488806,
+      "grad_norm": 0.5388423800468445,
+      "learning_rate": 0.0002965975482218766,
+      "loss": 0.7576,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8807421282668807,
+      "grad_norm": 0.5519341230392456,
+      "learning_rate": 0.00029652871189796284,
+      "loss": 0.8774,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8809205244848809,
+      "grad_norm": 0.5609647035598755,
+      "learning_rate": 0.00029645987191970414,
+      "loss": 0.9987,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8810989207028811,
+      "grad_norm": 0.47703075408935547,
+      "learning_rate": 0.000296391028292507,
+      "loss": 0.7438,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8812773169208813,
+      "grad_norm": 0.5322765111923218,
+      "learning_rate": 0.0002963221810217786,
+      "loss": 1.1033,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8814557131388815,
+      "grad_norm": 0.4712493121623993,
+      "learning_rate": 0.000296253330112926,
+      "loss": 0.7569,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8816341093568816,
+      "grad_norm": 0.5673714876174927,
+      "learning_rate": 0.00029618447557135677,
+      "loss": 0.7299,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8818125055748818,
+      "grad_norm": 0.5213398337364197,
+      "learning_rate": 0.00029611561740247854,
+      "loss": 0.828,
+      "step": 4943
+    },
+    {
+      "epoch": 0.881990901792882,
+      "grad_norm": 0.4355573058128357,
+      "learning_rate": 0.0002960467556116997,
+      "loss": 0.6324,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8821692980108822,
+      "grad_norm": 0.53075110912323,
+      "learning_rate": 0.0002959778902044285,
+      "loss": 0.852,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8823476942288824,
+      "grad_norm": 0.5229327082633972,
+      "learning_rate": 0.00029590902118607353,
+      "loss": 0.8341,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8825260904468826,
+      "grad_norm": 0.5234019160270691,
+      "learning_rate": 0.00029584014856204387,
+      "loss": 0.9364,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8827044866648827,
+      "grad_norm": 0.5076524019241333,
+      "learning_rate": 0.0002957712723377487,
+      "loss": 0.8151,
+      "step": 4948
+    },
+    {
+      "epoch": 0.8828828828828829,
+      "grad_norm": 0.4554321765899658,
+      "learning_rate": 0.00029570239251859744,
+      "loss": 0.8135,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8830612791008831,
+      "grad_norm": 0.47473978996276855,
+      "learning_rate": 0.0002956335091099999,
+      "loss": 0.7501,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8832396753188833,
+      "grad_norm": 0.4464404284954071,
+      "learning_rate": 0.00029556462211736614,
+      "loss": 0.7138,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8834180715368835,
+      "grad_norm": 0.5105646848678589,
+      "learning_rate": 0.0002954957315461066,
+      "loss": 0.9071,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8835964677548835,
+      "grad_norm": 0.49081435799598694,
+      "learning_rate": 0.00029542683740163203,
+      "loss": 0.9046,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8837748639728837,
+      "grad_norm": 0.45001259446144104,
+      "learning_rate": 0.0002953579396893531,
+      "loss": 0.7903,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8839532601908839,
+      "grad_norm": 0.512221097946167,
+      "learning_rate": 0.00029528903841468106,
+      "loss": 0.7811,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8841316564088841,
+      "grad_norm": 0.9528411030769348,
+      "learning_rate": 0.0002952201335830275,
+      "loss": 0.9261,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8843100526268843,
+      "grad_norm": 0.5219822525978088,
+      "learning_rate": 0.00029515122519980407,
+      "loss": 0.7963,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8844884488448845,
+      "grad_norm": 0.4960407614707947,
+      "learning_rate": 0.0002950823132704228,
+      "loss": 0.8326,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8846668450628846,
+      "grad_norm": 0.6182901263237,
+      "learning_rate": 0.0002950133978002961,
+      "loss": 0.8034,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8848452412808848,
+      "grad_norm": 0.5232715010643005,
+      "learning_rate": 0.00029494447879483657,
+      "loss": 0.9416,
+      "step": 4960
+    },
+    {
+      "epoch": 0.885023637498885,
+      "grad_norm": 0.514710009098053,
+      "learning_rate": 0.00029487555625945695,
+      "loss": 0.8282,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8852020337168852,
+      "grad_norm": 0.49220260977745056,
+      "learning_rate": 0.0002948066301995704,
+      "loss": 0.7967,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8853804299348854,
+      "grad_norm": 0.5210900902748108,
+      "learning_rate": 0.0002947377006205905,
+      "loss": 0.8346,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8855588261528856,
+      "grad_norm": 0.5323128700256348,
+      "learning_rate": 0.0002946687675279308,
+      "loss": 0.7406,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8857372223708857,
+      "grad_norm": 0.4670845866203308,
+      "learning_rate": 0.0002945998309270053,
+      "loss": 0.6944,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8859156185888859,
+      "grad_norm": 0.4710376262664795,
+      "learning_rate": 0.0002945308908232283,
+      "loss": 0.6828,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8860940148068861,
+      "grad_norm": 0.4978684186935425,
+      "learning_rate": 0.0002944619472220143,
+      "loss": 0.8463,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8862724110248863,
+      "grad_norm": 0.5263177156448364,
+      "learning_rate": 0.0002943930001287781,
+      "loss": 0.8929,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8864508072428865,
+      "grad_norm": 0.5331815481185913,
+      "learning_rate": 0.0002943240495489348,
+      "loss": 1.123,
+      "step": 4969
+    },
+    {
+      "epoch": 0.8866292034608866,
+      "grad_norm": 0.5104199051856995,
+      "learning_rate": 0.00029425509548789965,
+      "loss": 0.7605,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8868075996788868,
+      "grad_norm": 0.5087189674377441,
+      "learning_rate": 0.00029418613795108837,
+      "loss": 0.8925,
+      "step": 4971
+    },
+    {
+      "epoch": 0.886985995896887,
+      "grad_norm": 0.4605477750301361,
+      "learning_rate": 0.0002941171769439168,
+      "loss": 0.8883,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8871643921148872,
+      "grad_norm": 0.476589173078537,
+      "learning_rate": 0.0002940482124718012,
+      "loss": 0.8115,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8873427883328874,
+      "grad_norm": 0.5065289735794067,
+      "learning_rate": 0.00029397924454015797,
+      "loss": 0.9306,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8875211845508876,
+      "grad_norm": 0.5235819220542908,
+      "learning_rate": 0.0002939102731544037,
+      "loss": 1.0201,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8876995807688877,
+      "grad_norm": 0.6401598453521729,
+      "learning_rate": 0.0002938412983199555,
+      "loss": 0.9119,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8878779769868879,
+      "grad_norm": 0.5362576246261597,
+      "learning_rate": 0.00029377232004223065,
+      "loss": 1.006,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8880563732048881,
+      "grad_norm": 0.4627399146556854,
+      "learning_rate": 0.00029370333832664657,
+      "loss": 0.7415,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8882347694228883,
+      "grad_norm": 0.46277347207069397,
+      "learning_rate": 0.0002936343531786212,
+      "loss": 0.8446,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8884131656408885,
+      "grad_norm": 0.48642510175704956,
+      "learning_rate": 0.0002935653646035724,
+      "loss": 0.9397,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8885915618588885,
+      "grad_norm": 0.4110255837440491,
+      "learning_rate": 0.00029349637260691865,
+      "loss": 0.7168,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8887699580768887,
+      "grad_norm": 0.47086673974990845,
+      "learning_rate": 0.0002934273771940785,
+      "loss": 0.7632,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8889483542948889,
+      "grad_norm": 0.48255836963653564,
+      "learning_rate": 0.0002933583783704709,
+      "loss": 0.937,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8891267505128891,
+      "grad_norm": 0.44749948382377625,
+      "learning_rate": 0.00029328937614151487,
+      "loss": 0.7854,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8893051467308893,
+      "grad_norm": 0.5080204010009766,
+      "learning_rate": 0.0002932203705126298,
+      "loss": 0.7553,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8894835429488895,
+      "grad_norm": 0.47025632858276367,
+      "learning_rate": 0.0002931513614892355,
+      "loss": 0.8807,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8896619391668896,
+      "grad_norm": 0.46297869086265564,
+      "learning_rate": 0.0002930823490767519,
+      "loss": 0.8596,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8898403353848898,
+      "grad_norm": 0.413412868976593,
+      "learning_rate": 0.0002930133332805991,
+      "loss": 0.7766,
+      "step": 4988
+    },
+    {
+      "epoch": 0.89001873160289,
+      "grad_norm": 0.4820116460323334,
+      "learning_rate": 0.0002929443141061975,
+      "loss": 0.8289,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8901971278208902,
+      "grad_norm": 0.48540276288986206,
+      "learning_rate": 0.00029287529155896805,
+      "loss": 0.9301,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8903755240388904,
+      "grad_norm": 1.949074149131775,
+      "learning_rate": 0.0002928062656443317,
+      "loss": 0.6227,
+      "step": 4991
+    },
+    {
+      "epoch": 0.8905539202568905,
+      "grad_norm": 2.410193920135498,
+      "learning_rate": 0.00029273723636770953,
+      "loss": 0.699,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8907323164748907,
+      "grad_norm": 2.254347562789917,
+      "learning_rate": 0.00029266820373452334,
+      "loss": 0.7326,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8909107126928909,
+      "grad_norm": 0.8029910326004028,
+      "learning_rate": 0.00029259916775019475,
+      "loss": 0.8962,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8910891089108911,
+      "grad_norm": 0.6749674081802368,
+      "learning_rate": 0.0002925301284201458,
+      "loss": 0.9521,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8912675051288913,
+      "grad_norm": 0.700823962688446,
+      "learning_rate": 0.00029246108574979896,
+      "loss": 0.8662,
+      "step": 4996
+    },
+    {
+      "epoch": 0.8914459013468915,
+      "grad_norm": 0.7377640604972839,
+      "learning_rate": 0.0002923920397445766,
+      "loss": 0.9585,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8916242975648916,
+      "grad_norm": 0.5655547380447388,
+      "learning_rate": 0.00029232299040990174,
+      "loss": 0.8825,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8918026937828918,
+      "grad_norm": 0.4939744472503662,
+      "learning_rate": 0.0002922539377511974,
+      "loss": 0.7917,
+      "step": 4999
+    },
+    {
+      "epoch": 0.891981090000892,
+      "grad_norm": 0.5311207175254822,
+      "learning_rate": 0.00029218488177388705,
+      "loss": 1.0183,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8921594862188922,
+      "grad_norm": 0.5403327345848083,
+      "learning_rate": 0.00029211582248339424,
+      "loss": 0.7361,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8923378824368924,
+      "grad_norm": 0.47330448031425476,
+      "learning_rate": 0.0002920467598851428,
+      "loss": 0.6321,
+      "step": 5002
+    },
+    {
+      "epoch": 0.8925162786548925,
+      "grad_norm": 0.5282453298568726,
+      "learning_rate": 0.0002919776939845569,
+      "loss": 0.8446,
+      "step": 5003
+    },
+    {
+      "epoch": 0.8926946748728927,
+      "grad_norm": 0.48607364296913147,
+      "learning_rate": 0.000291908624787061,
+      "loss": 0.8299,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8928730710908929,
+      "grad_norm": 0.43826302886009216,
+      "learning_rate": 0.0002918395522980798,
+      "loss": 0.8356,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8930514673088931,
+      "grad_norm": 0.4682931900024414,
+      "learning_rate": 0.00029177047652303816,
+      "loss": 0.835,
+      "step": 5006
+    },
+    {
+      "epoch": 0.8932298635268933,
+      "grad_norm": 0.6613882780075073,
+      "learning_rate": 0.0002917013974673612,
+      "loss": 0.7867,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8934082597448935,
+      "grad_norm": 0.41586834192276,
+      "learning_rate": 0.00029163231513647454,
+      "loss": 0.772,
+      "step": 5008
+    },
+    {
+      "epoch": 0.8935866559628936,
+      "grad_norm": 0.5220624804496765,
+      "learning_rate": 0.00029156322953580367,
+      "loss": 0.8566,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8937650521808937,
+      "grad_norm": 0.48346883058547974,
+      "learning_rate": 0.00029149414067077467,
+      "loss": 0.9499,
+      "step": 5010
+    },
+    {
+      "epoch": 0.893943448398894,
+      "grad_norm": 0.4970376789569855,
+      "learning_rate": 0.00029142504854681375,
+      "loss": 1.0051,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8941218446168941,
+      "grad_norm": 0.532818615436554,
+      "learning_rate": 0.0002913559531693472,
+      "loss": 0.8754,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8943002408348943,
+      "grad_norm": 0.4270954430103302,
+      "learning_rate": 0.00029128685454380207,
+      "loss": 0.6813,
+      "step": 5013
+    },
+    {
+      "epoch": 0.8944786370528944,
+      "grad_norm": 0.5017000436782837,
+      "learning_rate": 0.0002912177526756051,
+      "loss": 0.9326,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8946570332708946,
+      "grad_norm": 0.5018951892852783,
+      "learning_rate": 0.0002911486475701835,
+      "loss": 0.867,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8948354294888948,
+      "grad_norm": 0.5230339765548706,
+      "learning_rate": 0.0002910795392329649,
+      "loss": 0.8305,
+      "step": 5016
+    },
+    {
+      "epoch": 0.895013825706895,
+      "grad_norm": 0.4868851900100708,
+      "learning_rate": 0.00029101042766937693,
+      "loss": 0.8635,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8951922219248952,
+      "grad_norm": 0.5056290626525879,
+      "learning_rate": 0.0002909413128848476,
+      "loss": 0.861,
+      "step": 5018
+    },
+    {
+      "epoch": 0.8953706181428954,
+      "grad_norm": 0.5414198040962219,
+      "learning_rate": 0.0002908721948848052,
+      "loss": 0.749,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8955490143608955,
+      "grad_norm": 0.4794885814189911,
+      "learning_rate": 0.00029080307367467824,
+      "loss": 0.7712,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8957274105788957,
+      "grad_norm": 0.454140841960907,
+      "learning_rate": 0.0002907339492598954,
+      "loss": 0.7231,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8959058067968959,
+      "grad_norm": 0.5277546644210815,
+      "learning_rate": 0.0002906648216458857,
+      "loss": 0.8037,
+      "step": 5022
+    },
+    {
+      "epoch": 0.8960842030148961,
+      "grad_norm": 0.5510913729667664,
+      "learning_rate": 0.0002905956908380784,
+      "loss": 0.972,
+      "step": 5023
+    },
+    {
+      "epoch": 0.8962625992328963,
+      "grad_norm": 0.4881914556026459,
+      "learning_rate": 0.00029052655684190304,
+      "loss": 0.8542,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8964409954508964,
+      "grad_norm": 0.5819581151008606,
+      "learning_rate": 0.0002904574196627893,
+      "loss": 0.6731,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8966193916688966,
+      "grad_norm": 0.48652219772338867,
+      "learning_rate": 0.0002903882793061673,
+      "loss": 0.7228,
+      "step": 5026
+    },
+    {
+      "epoch": 0.8967977878868968,
+      "grad_norm": 0.49010154604911804,
+      "learning_rate": 0.00029031913577746716,
+      "loss": 0.7869,
+      "step": 5027
+    },
+    {
+      "epoch": 0.896976184104897,
+      "grad_norm": 0.4732781648635864,
+      "learning_rate": 0.00029024998908211945,
+      "loss": 0.8617,
+      "step": 5028
+    },
+    {
+      "epoch": 0.8971545803228972,
+      "grad_norm": 0.47294124960899353,
+      "learning_rate": 0.000290180839225555,
+      "loss": 1.0253,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8973329765408974,
+      "grad_norm": 0.4338665306568146,
+      "learning_rate": 0.00029011168621320466,
+      "loss": 0.6808,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8975113727588975,
+      "grad_norm": 0.5743736028671265,
+      "learning_rate": 0.00029004253005049976,
+      "loss": 0.6331,
+      "step": 5031
+    },
+    {
+      "epoch": 0.8976897689768977,
+      "grad_norm": 0.4324184060096741,
+      "learning_rate": 0.0002899733707428718,
+      "loss": 0.7521,
+      "step": 5032
+    },
+    {
+      "epoch": 0.8978681651948979,
+      "grad_norm": 0.95427405834198,
+      "learning_rate": 0.0002899042082957525,
+      "loss": 0.8488,
+      "step": 5033
+    },
+    {
+      "epoch": 0.8980465614128981,
+      "grad_norm": 0.4829496741294861,
+      "learning_rate": 0.00028983504271457385,
+      "loss": 0.8179,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8982249576308983,
+      "grad_norm": 0.47819754481315613,
+      "learning_rate": 0.000289765874004768,
+      "loss": 0.8563,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8984033538488984,
+      "grad_norm": 0.5096835494041443,
+      "learning_rate": 0.0002896967021717676,
+      "loss": 0.7857,
+      "step": 5036
+    },
+    {
+      "epoch": 0.8985817500668986,
+      "grad_norm": 0.47968176007270813,
+      "learning_rate": 0.0002896275272210053,
+      "loss": 0.7974,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8987601462848988,
+      "grad_norm": 0.38809946179389954,
+      "learning_rate": 0.00028955834915791404,
+      "loss": 0.5523,
+      "step": 5038
+    },
+    {
+      "epoch": 0.898938542502899,
+      "grad_norm": 0.47436532378196716,
+      "learning_rate": 0.0002894891679879271,
+      "loss": 0.6803,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8991169387208992,
+      "grad_norm": 0.4444058835506439,
+      "learning_rate": 0.0002894199837164779,
+      "loss": 0.6817,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8992953349388993,
+      "grad_norm": 0.4691022038459778,
+      "learning_rate": 0.00028935079634900016,
+      "loss": 0.8635,
+      "step": 5041
+    },
+    {
+      "epoch": 0.8994737311568994,
+      "grad_norm": 0.4624834358692169,
+      "learning_rate": 0.0002892816058909277,
+      "loss": 0.8735,
+      "step": 5042
+    },
+    {
+      "epoch": 0.8996521273748996,
+      "grad_norm": 0.4934956431388855,
+      "learning_rate": 0.00028921241234769484,
+      "loss": 1.0094,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8998305235928998,
+      "grad_norm": 0.481513649225235,
+      "learning_rate": 0.000289143215724736,
+      "loss": 0.8711,
+      "step": 5044
+    },
+    {
+      "epoch": 0.9000089198109,
+      "grad_norm": 0.47422125935554504,
+      "learning_rate": 0.0002890740160274859,
+      "loss": 0.8002,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9001873160289002,
+      "grad_norm": 0.5127121806144714,
+      "learning_rate": 0.00028900481326137945,
+      "loss": 0.7709,
+      "step": 5046
+    },
+    {
+      "epoch": 0.9003657122469003,
+      "grad_norm": 0.4381260275840759,
+      "learning_rate": 0.00028893560743185166,
+      "loss": 0.693,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9005441084649005,
+      "grad_norm": 0.49497178196907043,
+      "learning_rate": 0.000288866398544338,
+      "loss": 0.7078,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9007225046829007,
+      "grad_norm": 0.511747419834137,
+      "learning_rate": 0.00028879718660427417,
+      "loss": 0.8916,
+      "step": 5049
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 0.48092395067214966,
+      "learning_rate": 0.00028872797161709593,
+      "loss": 0.8367,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9010792971189011,
+      "grad_norm": 0.4676063060760498,
+      "learning_rate": 0.0002886587535882395,
+      "loss": 0.7425,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9012576933369013,
+      "grad_norm": 0.45306262373924255,
+      "learning_rate": 0.00028858953252314126,
+      "loss": 0.7591,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9014360895549014,
+      "grad_norm": 0.5242826342582703,
+      "learning_rate": 0.0002885203084272377,
+      "loss": 0.922,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9016144857729016,
+      "grad_norm": 0.4822898209095001,
+      "learning_rate": 0.0002884510813059657,
+      "loss": 0.9461,
+      "step": 5054
+    },
+    {
+      "epoch": 0.9017928819909018,
+      "grad_norm": 0.46535515785217285,
+      "learning_rate": 0.0002883818511647623,
+      "loss": 0.8912,
+      "step": 5055
+    },
+    {
+      "epoch": 0.901971278208902,
+      "grad_norm": 0.4802592694759369,
+      "learning_rate": 0.0002883126180090648,
+      "loss": 0.9853,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9021496744269022,
+      "grad_norm": 0.497173547744751,
+      "learning_rate": 0.0002882433818443109,
+      "loss": 0.9567,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9023280706449023,
+      "grad_norm": 0.5876064896583557,
+      "learning_rate": 0.00028817414267593805,
+      "loss": 0.9769,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9025064668629025,
+      "grad_norm": 0.4283727705478668,
+      "learning_rate": 0.0002881049005093846,
+      "loss": 0.7212,
+      "step": 5059
+    },
+    {
+      "epoch": 0.9026848630809027,
+      "grad_norm": 0.5137040019035339,
+      "learning_rate": 0.0002880356553500886,
+      "loss": 0.7794,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9028632592989029,
+      "grad_norm": 0.6085144281387329,
+      "learning_rate": 0.00028796640720348866,
+      "loss": 0.8848,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9030416555169031,
+      "grad_norm": 0.5343783497810364,
+      "learning_rate": 0.0002878971560750234,
+      "loss": 0.7601,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9032200517349033,
+      "grad_norm": 0.5269213914871216,
+      "learning_rate": 0.0002878279019701318,
+      "loss": 0.8196,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9033984479529034,
+      "grad_norm": 0.4757898449897766,
+      "learning_rate": 0.00028775864489425306,
+      "loss": 0.8186,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9035768441709036,
+      "grad_norm": 0.5592185258865356,
+      "learning_rate": 0.0002876893848528266,
+      "loss": 1.1946,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9037552403889038,
+      "grad_norm": 0.43977245688438416,
+      "learning_rate": 0.0002876201218512921,
+      "loss": 0.7215,
+      "step": 5066
+    },
+    {
+      "epoch": 0.903933636606904,
+      "grad_norm": 0.49869504570961,
+      "learning_rate": 0.0002875508558950894,
+      "loss": 0.8863,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9041120328249042,
+      "grad_norm": 0.47307130694389343,
+      "learning_rate": 0.00028748158698965867,
+      "loss": 0.7705,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9042904290429042,
+      "grad_norm": 0.43691718578338623,
+      "learning_rate": 0.00028741231514044013,
+      "loss": 0.6947,
+      "step": 5069
+    },
+    {
+      "epoch": 0.9044688252609044,
+      "grad_norm": 0.4944881200790405,
+      "learning_rate": 0.00028734304035287454,
+      "loss": 0.7556,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9046472214789046,
+      "grad_norm": 0.48639604449272156,
+      "learning_rate": 0.00028727376263240265,
+      "loss": 0.7911,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9048256176969048,
+      "grad_norm": 0.4833623468875885,
+      "learning_rate": 0.0002872044819844654,
+      "loss": 0.802,
+      "step": 5072
+    },
+    {
+      "epoch": 0.905004013914905,
+      "grad_norm": 0.47726356983184814,
+      "learning_rate": 0.0002871351984145042,
+      "loss": 0.8817,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9051824101329052,
+      "grad_norm": 0.5050178170204163,
+      "learning_rate": 0.0002870659119279605,
+      "loss": 1.0378,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9053608063509053,
+      "grad_norm": 0.4867967367172241,
+      "learning_rate": 0.00028699662253027606,
+      "loss": 0.7799,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9055392025689055,
+      "grad_norm": 0.4266904890537262,
+      "learning_rate": 0.00028692733022689273,
+      "loss": 0.753,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9057175987869057,
+      "grad_norm": 0.4467776417732239,
+      "learning_rate": 0.0002868580350232528,
+      "loss": 0.6849,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9058959950049059,
+      "grad_norm": 0.46932464838027954,
+      "learning_rate": 0.0002867887369247987,
+      "loss": 1.0174,
+      "step": 5078
+    },
+    {
+      "epoch": 0.9060743912229061,
+      "grad_norm": 0.5115159153938293,
+      "learning_rate": 0.000286719435936973,
+      "loss": 0.8624,
+      "step": 5079
+    },
+    {
+      "epoch": 0.9062527874409062,
+      "grad_norm": 0.4789755046367645,
+      "learning_rate": 0.0002866501320652186,
+      "loss": 0.9705,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9064311836589064,
+      "grad_norm": 0.4835361838340759,
+      "learning_rate": 0.0002865808253149786,
+      "loss": 0.7777,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9066095798769066,
+      "grad_norm": 0.4759388566017151,
+      "learning_rate": 0.0002865115156916963,
+      "loss": 0.7886,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9067879760949068,
+      "grad_norm": 0.5399003028869629,
+      "learning_rate": 0.0002864422032008153,
+      "loss": 0.9909,
+      "step": 5083
+    },
+    {
+      "epoch": 0.906966372312907,
+      "grad_norm": 0.5101630687713623,
+      "learning_rate": 0.0002863728878477793,
+      "loss": 0.932,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9071447685309072,
+      "grad_norm": 0.5102178454399109,
+      "learning_rate": 0.0002863035696380324,
+      "loss": 0.9908,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9073231647489073,
+      "grad_norm": 0.5420662760734558,
+      "learning_rate": 0.0002862342485770188,
+      "loss": 1.0111,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9075015609669075,
+      "grad_norm": 0.4109286665916443,
+      "learning_rate": 0.00028616492467018286,
+      "loss": 0.8073,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9076799571849077,
+      "grad_norm": 0.489951491355896,
+      "learning_rate": 0.0002860955979229693,
+      "loss": 0.8515,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9078583534029079,
+      "grad_norm": 0.45059677958488464,
+      "learning_rate": 0.00028602626834082297,
+      "loss": 0.7568,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9080367496209081,
+      "grad_norm": 0.4857634902000427,
+      "learning_rate": 0.00028595693592918905,
+      "loss": 0.8072,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9082151458389082,
+      "grad_norm": 0.4536861479282379,
+      "learning_rate": 0.00028588760069351286,
+      "loss": 0.7684,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9083935420569084,
+      "grad_norm": 0.4914097487926483,
+      "learning_rate": 0.0002858182626392399,
+      "loss": 0.7621,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9085719382749086,
+      "grad_norm": 0.4842384159564972,
+      "learning_rate": 0.0002857489217718162,
+      "loss": 0.8194,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9087503344929088,
+      "grad_norm": 0.6547600030899048,
+      "learning_rate": 0.00028567957809668744,
+      "loss": 0.9461,
+      "step": 5094
+    },
+    {
+      "epoch": 0.908928730710909,
+      "grad_norm": 0.7207086682319641,
+      "learning_rate": 0.00028561023161929996,
+      "loss": 0.7638,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9091071269289092,
+      "grad_norm": 0.4441107511520386,
+      "learning_rate": 0.0002855408823451002,
+      "loss": 0.7315,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9092855231469092,
+      "grad_norm": 0.42800113558769226,
+      "learning_rate": 0.00028547153027953483,
+      "loss": 0.6776,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9094639193649094,
+      "grad_norm": 0.46136319637298584,
+      "learning_rate": 0.00028540217542805075,
+      "loss": 0.9207,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9096423155829096,
+      "grad_norm": 0.4710569977760315,
+      "learning_rate": 0.0002853328177960951,
+      "loss": 0.8327,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9098207118009098,
+      "grad_norm": 0.48937293887138367,
+      "learning_rate": 0.00028526345738911514,
+      "loss": 0.8316,
+      "step": 5100
+    },
+    {
+      "epoch": 0.90999910801891,
+      "grad_norm": 0.49789008498191833,
+      "learning_rate": 0.0002851940942125584,
+      "loss": 0.8759,
+      "step": 5101
+    },
+    {
+      "epoch": 0.9101775042369101,
+      "grad_norm": 0.4753686487674713,
+      "learning_rate": 0.0002851247282718726,
+      "loss": 0.8414,
+      "step": 5102
+    },
+    {
+      "epoch": 0.9103559004549103,
+      "grad_norm": 0.43923425674438477,
+      "learning_rate": 0.0002850553595725057,
+      "loss": 0.6973,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9105342966729105,
+      "grad_norm": 0.4735580086708069,
+      "learning_rate": 0.000284985988119906,
+      "loss": 0.9447,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9107126928909107,
+      "grad_norm": 0.543528139591217,
+      "learning_rate": 0.00028491661391952196,
+      "loss": 0.9476,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9108910891089109,
+      "grad_norm": 0.48239144682884216,
+      "learning_rate": 0.00028484723697680197,
+      "loss": 0.9228,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9110694853269111,
+      "grad_norm": 0.49379172921180725,
+      "learning_rate": 0.00028477785729719504,
+      "loss": 0.8531,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9112478815449112,
+      "grad_norm": 0.4407452344894409,
+      "learning_rate": 0.00028470847488615015,
+      "loss": 0.6475,
+      "step": 5108
+    },
+    {
+      "epoch": 0.9114262777629114,
+      "grad_norm": 0.4452020227909088,
+      "learning_rate": 0.00028463908974911656,
+      "loss": 0.6565,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9116046739809116,
+      "grad_norm": 0.48613160848617554,
+      "learning_rate": 0.0002845697018915437,
+      "loss": 0.6659,
+      "step": 5110
+    },
+    {
+      "epoch": 0.9117830701989118,
+      "grad_norm": 0.43221384286880493,
+      "learning_rate": 0.00028450031131888146,
+      "loss": 0.7194,
+      "step": 5111
+    },
+    {
+      "epoch": 0.911961466416912,
+      "grad_norm": 0.46625545620918274,
+      "learning_rate": 0.00028443091803657955,
+      "loss": 0.8089,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9121398626349121,
+      "grad_norm": 0.4777173399925232,
+      "learning_rate": 0.0002843615220500881,
+      "loss": 0.8622,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9123182588529123,
+      "grad_norm": 0.47781699895858765,
+      "learning_rate": 0.0002842921233648576,
+      "loss": 0.9369,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9124966550709125,
+      "grad_norm": 0.5193250775337219,
+      "learning_rate": 0.0002842227219863385,
+      "loss": 0.967,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9126750512889127,
+      "grad_norm": 0.46008825302124023,
+      "learning_rate": 0.00028415331791998145,
+      "loss": 0.9416,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9128534475069129,
+      "grad_norm": 0.4777912199497223,
+      "learning_rate": 0.00028408391117123755,
+      "loss": 0.8793,
+      "step": 5117
+    },
+    {
+      "epoch": 0.9130318437249131,
+      "grad_norm": 0.45857903361320496,
+      "learning_rate": 0.00028401450174555794,
+      "loss": 0.908,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9132102399429132,
+      "grad_norm": 0.4434804320335388,
+      "learning_rate": 0.000283945089648394,
+      "loss": 0.6586,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9133886361609134,
+      "grad_norm": 0.47657546401023865,
+      "learning_rate": 0.0002838756748851973,
+      "loss": 0.5956,
+      "step": 5120
+    },
+    {
+      "epoch": 0.9135670323789136,
+      "grad_norm": 0.4849657118320465,
+      "learning_rate": 0.0002838062574614197,
+      "loss": 0.9679,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9137454285969138,
+      "grad_norm": 0.5244645476341248,
+      "learning_rate": 0.00028373683738251314,
+      "loss": 1.0226,
+      "step": 5122
+    },
+    {
+      "epoch": 0.913923824814914,
+      "grad_norm": 0.4713945686817169,
+      "learning_rate": 0.00028366741465393,
+      "loss": 0.9147,
+      "step": 5123
+    },
+    {
+      "epoch": 0.914102221032914,
+      "grad_norm": 0.4651153087615967,
+      "learning_rate": 0.00028359798928112253,
+      "loss": 1.0022,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9142806172509143,
+      "grad_norm": 0.4441218972206116,
+      "learning_rate": 0.0002835285612695434,
+      "loss": 0.7423,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9144590134689145,
+      "grad_norm": 0.44443660974502563,
+      "learning_rate": 0.00028345913062464555,
+      "loss": 0.592,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9146374096869146,
+      "grad_norm": 0.431494802236557,
+      "learning_rate": 0.00028338969735188196,
+      "loss": 0.7291,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9148158059049148,
+      "grad_norm": 0.4383372366428375,
+      "learning_rate": 0.00028332026145670594,
+      "loss": 0.7891,
+      "step": 5128
+    },
+    {
+      "epoch": 0.914994202122915,
+      "grad_norm": 0.49089357256889343,
+      "learning_rate": 0.00028325082294457086,
+      "loss": 0.951,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9151725983409151,
+      "grad_norm": 0.4921533167362213,
+      "learning_rate": 0.00028318138182093053,
+      "loss": 0.9203,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9153509945589153,
+      "grad_norm": 0.48249009251594543,
+      "learning_rate": 0.0002831119380912387,
+      "loss": 0.7326,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9155293907769155,
+      "grad_norm": 0.48014095425605774,
+      "learning_rate": 0.00028304249176094946,
+      "loss": 0.7696,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9157077869949157,
+      "grad_norm": 0.5415335297584534,
+      "learning_rate": 0.00028297304283551725,
+      "loss": 0.9026,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9158861832129159,
+      "grad_norm": 0.4653064012527466,
+      "learning_rate": 0.00028290359132039644,
+      "loss": 1.0105,
+      "step": 5134
+    },
+    {
+      "epoch": 0.916064579430916,
+      "grad_norm": 0.482500821352005,
+      "learning_rate": 0.00028283413722104164,
+      "loss": 0.8853,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9162429756489162,
+      "grad_norm": 0.44477859139442444,
+      "learning_rate": 0.00028276468054290785,
+      "loss": 0.9191,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9164213718669164,
+      "grad_norm": 0.4712953567504883,
+      "learning_rate": 0.00028269522129145013,
+      "loss": 0.7727,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9165997680849166,
+      "grad_norm": 0.4651871621608734,
+      "learning_rate": 0.0002826257594721238,
+      "loss": 0.8326,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9167781643029168,
+      "grad_norm": 0.4735598564147949,
+      "learning_rate": 0.00028255629509038447,
+      "loss": 0.8486,
+      "step": 5139
+    },
+    {
+      "epoch": 0.916956560520917,
+      "grad_norm": 0.4831654727458954,
+      "learning_rate": 0.00028248682815168767,
+      "loss": 0.7643,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9171349567389171,
+      "grad_norm": 0.4459700584411621,
+      "learning_rate": 0.0002824173586614894,
+      "loss": 0.7463,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9173133529569173,
+      "grad_norm": 0.5149909853935242,
+      "learning_rate": 0.0002823478866252456,
+      "loss": 0.9273,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9174917491749175,
+      "grad_norm": 0.4949319362640381,
+      "learning_rate": 0.0002822784120484128,
+      "loss": 0.7418,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9176701453929177,
+      "grad_norm": 0.47984182834625244,
+      "learning_rate": 0.00028220893493644737,
+      "loss": 0.7807,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9178485416109179,
+      "grad_norm": 0.5391241908073425,
+      "learning_rate": 0.0002821394552948062,
+      "loss": 0.8166,
+      "step": 5145
+    },
+    {
+      "epoch": 0.918026937828918,
+      "grad_norm": 0.45041200518608093,
+      "learning_rate": 0.0002820699731289459,
+      "loss": 0.6369,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9182053340469182,
+      "grad_norm": 0.45535093545913696,
+      "learning_rate": 0.00028200048844432375,
+      "loss": 0.6831,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9183837302649184,
+      "grad_norm": 1.066908597946167,
+      "learning_rate": 0.000281931001246397,
+      "loss": 0.8285,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9185621264829186,
+      "grad_norm": 0.46443071961402893,
+      "learning_rate": 0.0002818615115406231,
+      "loss": 0.7574,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9187405227009188,
+      "grad_norm": 1.1375192403793335,
+      "learning_rate": 0.0002817920193324598,
+      "loss": 0.6816,
+      "step": 5150
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 0.49952948093414307,
+      "learning_rate": 0.000281722524627365,
+      "loss": 0.8653,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9190973151369191,
+      "grad_norm": 0.978867769241333,
+      "learning_rate": 0.00028165302743079693,
+      "loss": 0.8213,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9192757113549193,
+      "grad_norm": 0.4918247163295746,
+      "learning_rate": 0.0002815835277482135,
+      "loss": 0.8755,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9194541075729195,
+      "grad_norm": 0.4470139443874359,
+      "learning_rate": 0.0002815140255850735,
+      "loss": 0.6909,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9196325037909197,
+      "grad_norm": 0.47432026267051697,
+      "learning_rate": 0.0002814445209468354,
+      "loss": 0.8101,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9198109000089199,
+      "grad_norm": 0.5474990606307983,
+      "learning_rate": 0.00028137501383895824,
+      "loss": 1.0695,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9199892962269199,
+      "grad_norm": 0.47270458936691284,
+      "learning_rate": 0.00028130550426690095,
+      "loss": 0.7906,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9201676924449201,
+      "grad_norm": 0.4652954339981079,
+      "learning_rate": 0.0002812359922361228,
+      "loss": 0.6902,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9203460886629203,
+      "grad_norm": 0.4987577199935913,
+      "learning_rate": 0.00028116647775208335,
+      "loss": 0.7337,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9205244848809205,
+      "grad_norm": 0.530262291431427,
+      "learning_rate": 0.0002810969608202421,
+      "loss": 0.9222,
+      "step": 5160
+    },
+    {
+      "epoch": 0.9207028810989207,
+      "grad_norm": 0.48284605145454407,
+      "learning_rate": 0.00028102744144605895,
+      "loss": 0.7397,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9208812773169209,
+      "grad_norm": 0.4856660068035126,
+      "learning_rate": 0.00028095791963499384,
+      "loss": 0.7192,
+      "step": 5162
+    },
+    {
+      "epoch": 0.921059673534921,
+      "grad_norm": 0.7125374674797058,
+      "learning_rate": 0.0002808883953925071,
+      "loss": 1.0275,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9212380697529212,
+      "grad_norm": 0.4433026611804962,
+      "learning_rate": 0.0002808188687240591,
+      "loss": 0.7499,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9214164659709214,
+      "grad_norm": 0.44827622175216675,
+      "learning_rate": 0.00028074933963511035,
+      "loss": 0.8216,
+      "step": 5165
+    },
+    {
+      "epoch": 0.9215948621889216,
+      "grad_norm": 0.5151771306991577,
+      "learning_rate": 0.0002806798081311217,
+      "loss": 0.9786,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9217732584069218,
+      "grad_norm": 0.5016061663627625,
+      "learning_rate": 0.0002806102742175542,
+      "loss": 1.008,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9219516546249219,
+      "grad_norm": 0.4590628743171692,
+      "learning_rate": 0.00028054073789986883,
+      "loss": 0.7287,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9221300508429221,
+      "grad_norm": 0.47854331135749817,
+      "learning_rate": 0.00028047119918352717,
+      "loss": 1.1118,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9223084470609223,
+      "grad_norm": 0.49446046352386475,
+      "learning_rate": 0.00028040165807399054,
+      "loss": 1.0227,
+      "step": 5170
+    },
+    {
+      "epoch": 0.9224868432789225,
+      "grad_norm": 0.49778786301612854,
+      "learning_rate": 0.0002803321145767208,
+      "loss": 0.7794,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9226652394969227,
+      "grad_norm": 0.46323299407958984,
+      "learning_rate": 0.0002802625686971798,
+      "loss": 0.8109,
+      "step": 5172
+    },
+    {
+      "epoch": 0.9228436357149229,
+      "grad_norm": 0.4294353425502777,
+      "learning_rate": 0.0002801930204408297,
+      "loss": 0.7047,
+      "step": 5173
+    },
+    {
+      "epoch": 0.923022031932923,
+      "grad_norm": 0.4814031720161438,
+      "learning_rate": 0.0002801234698131328,
+      "loss": 0.6691,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9232004281509232,
+      "grad_norm": 0.5088090300559998,
+      "learning_rate": 0.0002800539168195515,
+      "loss": 0.7468,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9233788243689234,
+      "grad_norm": 0.486738383769989,
+      "learning_rate": 0.00027998436146554857,
+      "loss": 0.8424,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9235572205869236,
+      "grad_norm": 0.43971219658851624,
+      "learning_rate": 0.0002799148037565867,
+      "loss": 0.829,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9237356168049238,
+      "grad_norm": 0.47104689478874207,
+      "learning_rate": 0.0002798452436981291,
+      "loss": 0.608,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9239140130229239,
+      "grad_norm": 0.477444589138031,
+      "learning_rate": 0.0002797756812956389,
+      "loss": 0.7896,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9240924092409241,
+      "grad_norm": 0.7103281617164612,
+      "learning_rate": 0.00027970611655457953,
+      "loss": 0.7084,
+      "step": 5180
+    },
+    {
+      "epoch": 0.9242708054589243,
+      "grad_norm": 1.2961317300796509,
+      "learning_rate": 0.0002796365494804144,
+      "loss": 0.9616,
+      "step": 5181
+    },
+    {
+      "epoch": 0.9244492016769245,
+      "grad_norm": 0.540886402130127,
+      "learning_rate": 0.00027956698007860754,
+      "loss": 0.8563,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9246275978949247,
+      "grad_norm": 0.4906296133995056,
+      "learning_rate": 0.0002794974083546227,
+      "loss": 0.8361,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9248059941129249,
+      "grad_norm": 0.47894391417503357,
+      "learning_rate": 0.0002794278343139242,
+      "loss": 0.881,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9249843903309249,
+      "grad_norm": 0.5145757794380188,
+      "learning_rate": 0.0002793582579619762,
+      "loss": 0.8234,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9251627865489251,
+      "grad_norm": 0.43651270866394043,
+      "learning_rate": 0.0002792886793042434,
+      "loss": 0.625,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9253411827669253,
+      "grad_norm": 0.5134057998657227,
+      "learning_rate": 0.00027921909834619017,
+      "loss": 0.9787,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9255195789849255,
+      "grad_norm": 0.6394853591918945,
+      "learning_rate": 0.0002791495150932815,
+      "loss": 0.9377,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9256979752029257,
+      "grad_norm": 0.4257347583770752,
+      "learning_rate": 0.0002790799295509825,
+      "loss": 0.6986,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9258763714209258,
+      "grad_norm": 0.7512332797050476,
+      "learning_rate": 0.0002790103417247584,
+      "loss": 0.826,
+      "step": 5190
+    },
+    {
+      "epoch": 0.926054767638926,
+      "grad_norm": 0.49140146374702454,
+      "learning_rate": 0.0002789407516200746,
+      "loss": 0.7517,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9262331638569262,
+      "grad_norm": 0.4644830524921417,
+      "learning_rate": 0.0002788711592423966,
+      "loss": 0.7325,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9264115600749264,
+      "grad_norm": 0.466547429561615,
+      "learning_rate": 0.0002788015645971901,
+      "loss": 0.7623,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9265899562929266,
+      "grad_norm": 0.4740476906299591,
+      "learning_rate": 0.00027873196768992114,
+      "loss": 0.7261,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9267683525109268,
+      "grad_norm": 0.4969778060913086,
+      "learning_rate": 0.00027866236852605575,
+      "loss": 0.8977,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9269467487289269,
+      "grad_norm": 0.5204504132270813,
+      "learning_rate": 0.0002785927671110603,
+      "loss": 0.9751,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9271251449469271,
+      "grad_norm": 0.5292174220085144,
+      "learning_rate": 0.00027852316345040125,
+      "loss": 0.927,
+      "step": 5197
+    },
+    {
+      "epoch": 0.9273035411649273,
+      "grad_norm": 0.5514991283416748,
+      "learning_rate": 0.0002784535575495453,
+      "loss": 1.0084,
+      "step": 5198
+    },
+    {
+      "epoch": 0.9274819373829275,
+      "grad_norm": 0.5040212869644165,
+      "learning_rate": 0.00027838394941395907,
+      "loss": 0.7041,
+      "step": 5199
+    },
+    {
+      "epoch": 0.9276603336009277,
+      "grad_norm": 0.5497444272041321,
+      "learning_rate": 0.00027831433904910963,
+      "loss": 1.0214,
+      "step": 5200
+    },
+    {
+      "epoch": 0.9278387298189278,
+      "grad_norm": 0.4607004225254059,
+      "learning_rate": 0.0002782447264604643,
+      "loss": 0.7548,
+      "step": 5201
+    },
+    {
+      "epoch": 0.928017126036928,
+      "grad_norm": 0.4299633502960205,
+      "learning_rate": 0.00027817511165349024,
+      "loss": 0.7039,
+      "step": 5202
+    },
+    {
+      "epoch": 0.9281955222549282,
+      "grad_norm": 0.5070714950561523,
+      "learning_rate": 0.000278105494633655,
+      "loss": 0.9753,
+      "step": 5203
+    },
+    {
+      "epoch": 0.9283739184729284,
+      "grad_norm": 0.49922603368759155,
+      "learning_rate": 0.0002780358754064263,
+      "loss": 0.8345,
+      "step": 5204
+    },
+    {
+      "epoch": 0.9285523146909286,
+      "grad_norm": 0.6079359650611877,
+      "learning_rate": 0.00027796625397727214,
+      "loss": 0.694,
+      "step": 5205
+    },
+    {
+      "epoch": 0.9287307109089288,
+      "grad_norm": 0.45466673374176025,
+      "learning_rate": 0.0002778966303516603,
+      "loss": 0.8286,
+      "step": 5206
+    },
+    {
+      "epoch": 0.9289091071269289,
+      "grad_norm": 0.5105868577957153,
+      "learning_rate": 0.00027782700453505925,
+      "loss": 0.7927,
+      "step": 5207
+    },
+    {
+      "epoch": 0.9290875033449291,
+      "grad_norm": 0.5233384370803833,
+      "learning_rate": 0.00027775737653293716,
+      "loss": 0.8307,
+      "step": 5208
+    },
+    {
+      "epoch": 0.9292658995629293,
+      "grad_norm": 0.508243978023529,
+      "learning_rate": 0.00027768774635076265,
+      "loss": 0.8838,
+      "step": 5209
+    },
+    {
+      "epoch": 0.9294442957809295,
+      "grad_norm": 0.46551835536956787,
+      "learning_rate": 0.0002776181139940045,
+      "loss": 0.7637,
+      "step": 5210
+    },
+    {
+      "epoch": 0.9296226919989297,
+      "grad_norm": 0.44400593638420105,
+      "learning_rate": 0.0002775484794681315,
+      "loss": 0.7503,
+      "step": 5211
+    },
+    {
+      "epoch": 0.9298010882169297,
+      "grad_norm": 0.4329807460308075,
+      "learning_rate": 0.0002774788427786128,
+      "loss": 0.6834,
+      "step": 5212
+    },
+    {
+      "epoch": 0.92997948443493,
+      "grad_norm": 0.5895024538040161,
+      "learning_rate": 0.0002774092039309176,
+      "loss": 0.928,
+      "step": 5213
+    },
+    {
+      "epoch": 0.9301578806529301,
+      "grad_norm": 0.46719425916671753,
+      "learning_rate": 0.0002773395629305154,
+      "loss": 0.6546,
+      "step": 5214
+    },
+    {
+      "epoch": 0.9303362768709303,
+      "grad_norm": 0.5935095548629761,
+      "learning_rate": 0.0002772699197828756,
+      "loss": 0.7523,
+      "step": 5215
+    },
+    {
+      "epoch": 0.9305146730889305,
+      "grad_norm": 0.5028880834579468,
+      "learning_rate": 0.00027720027449346806,
+      "loss": 0.9039,
+      "step": 5216
+    },
+    {
+      "epoch": 0.9306930693069307,
+      "grad_norm": 0.4530572295188904,
+      "learning_rate": 0.00027713062706776273,
+      "loss": 0.7171,
+      "step": 5217
+    },
+    {
+      "epoch": 0.9308714655249308,
+      "grad_norm": 0.4633224606513977,
+      "learning_rate": 0.0002770609775112295,
+      "loss": 0.8019,
+      "step": 5218
+    },
+    {
+      "epoch": 0.931049861742931,
+      "grad_norm": 0.5543755292892456,
+      "learning_rate": 0.00027699132582933886,
+      "loss": 1.0682,
+      "step": 5219
+    },
+    {
+      "epoch": 0.9312282579609312,
+      "grad_norm": 0.4689428508281708,
+      "learning_rate": 0.000276921672027561,
+      "loss": 0.7497,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9314066541789314,
+      "grad_norm": 0.5170400142669678,
+      "learning_rate": 0.0002768520161113667,
+      "loss": 0.8323,
+      "step": 5221
+    },
+    {
+      "epoch": 0.9315850503969316,
+      "grad_norm": 0.4578782320022583,
+      "learning_rate": 0.0002767823580862265,
+      "loss": 1.0117,
+      "step": 5222
+    },
+    {
+      "epoch": 0.9317634466149317,
+      "grad_norm": 0.44367796182632446,
+      "learning_rate": 0.00027671269795761155,
+      "loss": 0.7195,
+      "step": 5223
+    },
+    {
+      "epoch": 0.9319418428329319,
+      "grad_norm": 0.48492223024368286,
+      "learning_rate": 0.00027664303573099274,
+      "loss": 0.9721,
+      "step": 5224
+    },
+    {
+      "epoch": 0.9321202390509321,
+      "grad_norm": 0.696104109287262,
+      "learning_rate": 0.00027657337141184134,
+      "loss": 1.0126,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9322986352689323,
+      "grad_norm": 0.7282087206840515,
+      "learning_rate": 0.00027650370500562885,
+      "loss": 0.9032,
+      "step": 5226
+    },
+    {
+      "epoch": 0.9324770314869325,
+      "grad_norm": 0.42898091673851013,
+      "learning_rate": 0.00027643403651782673,
+      "loss": 0.6429,
+      "step": 5227
+    },
+    {
+      "epoch": 0.9326554277049327,
+      "grad_norm": 0.4315216541290283,
+      "learning_rate": 0.00027636436595390674,
+      "loss": 0.8703,
+      "step": 5228
+    },
+    {
+      "epoch": 0.9328338239229328,
+      "grad_norm": 0.44752299785614014,
+      "learning_rate": 0.0002762946933193408,
+      "loss": 0.8215,
+      "step": 5229
+    },
+    {
+      "epoch": 0.933012220140933,
+      "grad_norm": 0.4484764635562897,
+      "learning_rate": 0.00027622501861960104,
+      "loss": 0.6912,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9331906163589332,
+      "grad_norm": 0.371706485748291,
+      "learning_rate": 0.0002761553418601595,
+      "loss": 0.5803,
+      "step": 5231
+    },
+    {
+      "epoch": 0.9333690125769334,
+      "grad_norm": 0.4724928140640259,
+      "learning_rate": 0.0002760856630464888,
+      "loss": 0.8421,
+      "step": 5232
+    },
+    {
+      "epoch": 0.9335474087949336,
+      "grad_norm": 0.49815791845321655,
+      "learning_rate": 0.0002760159821840612,
+      "loss": 0.6765,
+      "step": 5233
+    },
+    {
+      "epoch": 0.9337258050129337,
+      "grad_norm": 0.5424796342849731,
+      "learning_rate": 0.00027594629927834956,
+      "loss": 0.8134,
+      "step": 5234
+    },
+    {
+      "epoch": 0.9339042012309339,
+      "grad_norm": 0.5402620434761047,
+      "learning_rate": 0.0002758766143348268,
+      "loss": 0.8324,
+      "step": 5235
+    },
+    {
+      "epoch": 0.9340825974489341,
+      "grad_norm": 0.4638790190219879,
+      "learning_rate": 0.0002758069273589659,
+      "loss": 0.8668,
+      "step": 5236
+    },
+    {
+      "epoch": 0.9342609936669343,
+      "grad_norm": 0.5340316295623779,
+      "learning_rate": 0.00027573723835624004,
+      "loss": 0.8628,
+      "step": 5237
+    },
+    {
+      "epoch": 0.9344393898849345,
+      "grad_norm": 0.6846177577972412,
+      "learning_rate": 0.00027566754733212255,
+      "loss": 0.831,
+      "step": 5238
+    },
+    {
+      "epoch": 0.9346177861029347,
+      "grad_norm": 0.502400815486908,
+      "learning_rate": 0.0002755978542920869,
+      "loss": 0.8583,
+      "step": 5239
+    },
+    {
+      "epoch": 0.9347961823209348,
+      "grad_norm": 0.44814401865005493,
+      "learning_rate": 0.00027552815924160686,
+      "loss": 0.7743,
+      "step": 5240
+    },
+    {
+      "epoch": 0.934974578538935,
+      "grad_norm": 0.6513227224349976,
+      "learning_rate": 0.0002754584621861561,
+      "loss": 0.922,
+      "step": 5241
+    },
+    {
+      "epoch": 0.9351529747569352,
+      "grad_norm": 0.4567849040031433,
+      "learning_rate": 0.0002753887631312086,
+      "loss": 0.7956,
+      "step": 5242
+    },
+    {
+      "epoch": 0.9353313709749353,
+      "grad_norm": 0.4874018430709839,
+      "learning_rate": 0.00027531906208223865,
+      "loss": 0.9146,
+      "step": 5243
+    },
+    {
+      "epoch": 0.9355097671929355,
+      "grad_norm": 0.4673248827457428,
+      "learning_rate": 0.00027524935904472053,
+      "loss": 0.8524,
+      "step": 5244
+    },
+    {
+      "epoch": 0.9356881634109356,
+      "grad_norm": 0.5060569047927856,
+      "learning_rate": 0.0002751796540241286,
+      "loss": 1.0127,
+      "step": 5245
+    },
+    {
+      "epoch": 0.9358665596289358,
+      "grad_norm": 0.4823893904685974,
+      "learning_rate": 0.00027510994702593743,
+      "loss": 0.8641,
+      "step": 5246
+    },
+    {
+      "epoch": 0.936044955846936,
+      "grad_norm": 0.5004808902740479,
+      "learning_rate": 0.0002750402380556218,
+      "loss": 0.8461,
+      "step": 5247
+    },
+    {
+      "epoch": 0.9362233520649362,
+      "grad_norm": 0.4644457995891571,
+      "learning_rate": 0.0002749705271186567,
+      "loss": 0.7319,
+      "step": 5248
+    },
+    {
+      "epoch": 0.9364017482829364,
+      "grad_norm": 0.49020230770111084,
+      "learning_rate": 0.0002749008142205171,
+      "loss": 0.7813,
+      "step": 5249
+    },
+    {
+      "epoch": 0.9365801445009366,
+      "grad_norm": 0.568228006362915,
+      "learning_rate": 0.0002748310993666783,
+      "loss": 1.0021,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9367585407189367,
+      "grad_norm": 0.46288859844207764,
+      "learning_rate": 0.00027476138256261575,
+      "loss": 0.7229,
+      "step": 5251
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 0.5582467913627625,
+      "learning_rate": 0.00027469166381380474,
+      "loss": 0.729,
+      "step": 5252
+    },
+    {
+      "epoch": 0.9371153331549371,
+      "grad_norm": 0.6614149808883667,
+      "learning_rate": 0.0002746219431257211,
+      "loss": 0.8033,
+      "step": 5253
+    },
+    {
+      "epoch": 0.9372937293729373,
+      "grad_norm": 0.4389813244342804,
+      "learning_rate": 0.0002745522205038406,
+      "loss": 0.6155,
+      "step": 5254
+    },
+    {
+      "epoch": 0.9374721255909375,
+      "grad_norm": 0.564568042755127,
+      "learning_rate": 0.0002744824959536393,
+      "loss": 0.8861,
+      "step": 5255
+    },
+    {
+      "epoch": 0.9376505218089376,
+      "grad_norm": 0.6031481623649597,
+      "learning_rate": 0.00027441276948059337,
+      "loss": 0.8549,
+      "step": 5256
+    },
+    {
+      "epoch": 0.9378289180269378,
+      "grad_norm": 0.4482172429561615,
+      "learning_rate": 0.0002743430410901789,
+      "loss": 0.6852,
+      "step": 5257
+    },
+    {
+      "epoch": 0.938007314244938,
+      "grad_norm": 0.47618693113327026,
+      "learning_rate": 0.0002742733107878726,
+      "loss": 0.6698,
+      "step": 5258
+    },
+    {
+      "epoch": 0.9381857104629382,
+      "grad_norm": 0.4913199543952942,
+      "learning_rate": 0.00027420357857915083,
+      "loss": 0.9027,
+      "step": 5259
+    },
+    {
+      "epoch": 0.9383641066809384,
+      "grad_norm": 0.510835587978363,
+      "learning_rate": 0.0002741338444694904,
+      "loss": 0.8386,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9385425028989386,
+      "grad_norm": 1.6068651676177979,
+      "learning_rate": 0.00027406410846436826,
+      "loss": 0.9289,
+      "step": 5261
+    },
+    {
+      "epoch": 0.9387208991169387,
+      "grad_norm": 0.5296999216079712,
+      "learning_rate": 0.0002739943705692614,
+      "loss": 0.9303,
+      "step": 5262
+    },
+    {
+      "epoch": 0.9388992953349389,
+      "grad_norm": 0.5108866095542908,
+      "learning_rate": 0.00027392463078964696,
+      "loss": 0.8593,
+      "step": 5263
+    },
+    {
+      "epoch": 0.9390776915529391,
+      "grad_norm": 0.48683467507362366,
+      "learning_rate": 0.0002738548891310023,
+      "loss": 0.8258,
+      "step": 5264
+    },
+    {
+      "epoch": 0.9392560877709393,
+      "grad_norm": 0.48959386348724365,
+      "learning_rate": 0.00027378514559880495,
+      "loss": 0.802,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9394344839889395,
+      "grad_norm": 0.5760032534599304,
+      "learning_rate": 0.0002737154001985325,
+      "loss": 0.8428,
+      "step": 5266
+    },
+    {
+      "epoch": 0.9396128802069396,
+      "grad_norm": 0.47354596853256226,
+      "learning_rate": 0.0002736456529356627,
+      "loss": 0.7726,
+      "step": 5267
+    },
+    {
+      "epoch": 0.9397912764249398,
+      "grad_norm": 0.5111311078071594,
+      "learning_rate": 0.00027357590381567353,
+      "loss": 0.8006,
+      "step": 5268
+    },
+    {
+      "epoch": 0.93996967264294,
+      "grad_norm": 0.4919393062591553,
+      "learning_rate": 0.00027350615284404305,
+      "loss": 0.8726,
+      "step": 5269
+    },
+    {
+      "epoch": 0.9401480688609402,
+      "grad_norm": 0.43833592534065247,
+      "learning_rate": 0.0002734364000262494,
+      "loss": 0.8131,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9403264650789404,
+      "grad_norm": 0.4701617658138275,
+      "learning_rate": 0.00027336664536777093,
+      "loss": 0.8316,
+      "step": 5271
+    },
+    {
+      "epoch": 0.9405048612969406,
+      "grad_norm": 0.430813729763031,
+      "learning_rate": 0.0002732968888740863,
+      "loss": 0.7455,
+      "step": 5272
+    },
+    {
+      "epoch": 0.9406832575149406,
+      "grad_norm": 0.46511298418045044,
+      "learning_rate": 0.00027322713055067397,
+      "loss": 0.8424,
+      "step": 5273
+    },
+    {
+      "epoch": 0.9408616537329408,
+      "grad_norm": 0.5365491509437561,
+      "learning_rate": 0.0002731573704030128,
+      "loss": 0.902,
+      "step": 5274
+    },
+    {
+      "epoch": 0.941040049950941,
+      "grad_norm": 0.4637891948223114,
+      "learning_rate": 0.0002730876084365817,
+      "loss": 0.7835,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9412184461689412,
+      "grad_norm": 0.47008875012397766,
+      "learning_rate": 0.00027301784465685983,
+      "loss": 0.8082,
+      "step": 5276
+    },
+    {
+      "epoch": 0.9413968423869414,
+      "grad_norm": 0.42450207471847534,
+      "learning_rate": 0.0002729480790693263,
+      "loss": 0.7095,
+      "step": 5277
+    },
+    {
+      "epoch": 0.9415752386049415,
+      "grad_norm": 0.5233426690101624,
+      "learning_rate": 0.0002728783116794606,
+      "loss": 0.9605,
+      "step": 5278
+    },
+    {
+      "epoch": 0.9417536348229417,
+      "grad_norm": 1.0770610570907593,
+      "learning_rate": 0.00027280854249274206,
+      "loss": 0.6148,
+      "step": 5279
+    },
+    {
+      "epoch": 0.9419320310409419,
+      "grad_norm": 0.6112112402915955,
+      "learning_rate": 0.00027273877151465036,
+      "loss": 1.1989,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9421104272589421,
+      "grad_norm": 0.5142430663108826,
+      "learning_rate": 0.0002726689987506654,
+      "loss": 0.9081,
+      "step": 5281
+    },
+    {
+      "epoch": 0.9422888234769423,
+      "grad_norm": 0.5745399594306946,
+      "learning_rate": 0.00027259922420626705,
+      "loss": 0.8929,
+      "step": 5282
+    },
+    {
+      "epoch": 0.9424672196949425,
+      "grad_norm": 0.6627066731452942,
+      "learning_rate": 0.00027252944788693536,
+      "loss": 0.6532,
+      "step": 5283
+    },
+    {
+      "epoch": 0.9426456159129426,
+      "grad_norm": 0.584989607334137,
+      "learning_rate": 0.00027245966979815044,
+      "loss": 0.8091,
+      "step": 5284
+    },
+    {
+      "epoch": 0.9428240121309428,
+      "grad_norm": 0.4927927851676941,
+      "learning_rate": 0.0002723898899453929,
+      "loss": 0.7812,
+      "step": 5285
+    },
+    {
+      "epoch": 0.943002408348943,
+      "grad_norm": 0.5424949526786804,
+      "learning_rate": 0.00027232010833414287,
+      "loss": 0.9299,
+      "step": 5286
+    },
+    {
+      "epoch": 0.9431808045669432,
+      "grad_norm": 0.5809288024902344,
+      "learning_rate": 0.0002722503249698812,
+      "loss": 0.8294,
+      "step": 5287
+    },
+    {
+      "epoch": 0.9433592007849434,
+      "grad_norm": 0.601315438747406,
+      "learning_rate": 0.0002721805398580885,
+      "loss": 0.7725,
+      "step": 5288
+    },
+    {
+      "epoch": 0.9435375970029435,
+      "grad_norm": 0.5783292055130005,
+      "learning_rate": 0.0002721107530042458,
+      "loss": 0.7394,
+      "step": 5289
+    },
+    {
+      "epoch": 0.9437159932209437,
+      "grad_norm": 0.4998534321784973,
+      "learning_rate": 0.00027204096441383414,
+      "loss": 0.8316,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9438943894389439,
+      "grad_norm": 0.8708257079124451,
+      "learning_rate": 0.0002719711740923346,
+      "loss": 0.811,
+      "step": 5291
+    },
+    {
+      "epoch": 0.9440727856569441,
+      "grad_norm": 0.5783541798591614,
+      "learning_rate": 0.00027190138204522847,
+      "loss": 0.9316,
+      "step": 5292
+    },
+    {
+      "epoch": 0.9442511818749443,
+      "grad_norm": 1.6886039972305298,
+      "learning_rate": 0.0002718315882779972,
+      "loss": 0.6536,
+      "step": 5293
+    },
+    {
+      "epoch": 0.9444295780929445,
+      "grad_norm": 0.5459927320480347,
+      "learning_rate": 0.0002717617927961224,
+      "loss": 0.8067,
+      "step": 5294
+    },
+    {
+      "epoch": 0.9446079743109446,
+      "grad_norm": 0.46528294682502747,
+      "learning_rate": 0.00027169199560508574,
+      "loss": 0.5931,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9447863705289448,
+      "grad_norm": 0.5428159832954407,
+      "learning_rate": 0.0002716221967103691,
+      "loss": 0.8225,
+      "step": 5296
+    },
+    {
+      "epoch": 0.944964766746945,
+      "grad_norm": 0.4363666772842407,
+      "learning_rate": 0.0002715523961174545,
+      "loss": 0.6572,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9451431629649452,
+      "grad_norm": 0.6103348731994629,
+      "learning_rate": 0.0002714825938318239,
+      "loss": 1.0737,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9453215591829454,
+      "grad_norm": 0.4808749854564667,
+      "learning_rate": 0.0002714127898589596,
+      "loss": 0.8564,
+      "step": 5299
+    },
+    {
+      "epoch": 0.9454999554009454,
+      "grad_norm": 0.46589285135269165,
+      "learning_rate": 0.00027134298420434405,
+      "loss": 0.7157,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9456783516189456,
+      "grad_norm": 0.4251248836517334,
+      "learning_rate": 0.0002712731768734597,
+      "loss": 0.6488,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9458567478369458,
+      "grad_norm": 0.4899739623069763,
+      "learning_rate": 0.0002712033678717892,
+      "loss": 0.9113,
+      "step": 5302
+    },
+    {
+      "epoch": 0.946035144054946,
+      "grad_norm": 0.506604790687561,
+      "learning_rate": 0.00027113355720481523,
+      "loss": 0.8268,
+      "step": 5303
+    },
+    {
+      "epoch": 0.9462135402729462,
+      "grad_norm": 0.4809357225894928,
+      "learning_rate": 0.00027106374487802096,
+      "loss": 0.9743,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9463919364909464,
+      "grad_norm": 0.4477202296257019,
+      "learning_rate": 0.00027099393089688906,
+      "loss": 0.8365,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9465703327089465,
+      "grad_norm": 0.42474350333213806,
+      "learning_rate": 0.0002709241152669029,
+      "loss": 0.654,
+      "step": 5306
+    },
+    {
+      "epoch": 0.9467487289269467,
+      "grad_norm": 0.4127642810344696,
+      "learning_rate": 0.00027085429799354575,
+      "loss": 0.7342,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9469271251449469,
+      "grad_norm": 0.4678402543067932,
+      "learning_rate": 0.00027078447908230105,
+      "loss": 0.7684,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9471055213629471,
+      "grad_norm": 0.434625506401062,
+      "learning_rate": 0.00027071465853865224,
+      "loss": 0.6473,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9472839175809473,
+      "grad_norm": 0.48882344365119934,
+      "learning_rate": 0.00027064483636808314,
+      "loss": 0.8451,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9474623137989474,
+      "grad_norm": 0.506234347820282,
+      "learning_rate": 0.0002705750125760774,
+      "loss": 1.075,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9476407100169476,
+      "grad_norm": 0.47516417503356934,
+      "learning_rate": 0.00027050518716811904,
+      "loss": 0.8066,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9478191062349478,
+      "grad_norm": 0.45935487747192383,
+      "learning_rate": 0.0002704353601496921,
+      "loss": 0.8011,
+      "step": 5313
+    },
+    {
+      "epoch": 0.947997502452948,
+      "grad_norm": 0.4938845634460449,
+      "learning_rate": 0.0002703655315262808,
+      "loss": 0.9695,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9481758986709482,
+      "grad_norm": 0.4160667359828949,
+      "learning_rate": 0.00027029570130336937,
+      "loss": 0.6122,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9483542948889484,
+      "grad_norm": 0.5093261003494263,
+      "learning_rate": 0.00027022586948644234,
+      "loss": 1.0263,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9485326911069485,
+      "grad_norm": 0.4197608530521393,
+      "learning_rate": 0.0002701560360809842,
+      "loss": 0.8012,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9487110873249487,
+      "grad_norm": 0.47987252473831177,
+      "learning_rate": 0.0002700862010924797,
+      "loss": 0.7935,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9488894835429489,
+      "grad_norm": 0.4315970242023468,
+      "learning_rate": 0.00027001636452641354,
+      "loss": 0.6472,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9490678797609491,
+      "grad_norm": 0.4990069270133972,
+      "learning_rate": 0.0002699465263882708,
+      "loss": 0.7692,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9492462759789493,
+      "grad_norm": 0.4555619955062866,
+      "learning_rate": 0.00026987668668353637,
+      "loss": 0.7722,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9494246721969494,
+      "grad_norm": 0.5311444401741028,
+      "learning_rate": 0.00026980684541769563,
+      "loss": 1.0347,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9496030684149496,
+      "grad_norm": 0.4669055640697479,
+      "learning_rate": 0.0002697370025962337,
+      "loss": 0.6929,
+      "step": 5323
+    },
+    {
+      "epoch": 0.9497814646329498,
+      "grad_norm": 0.4881044924259186,
+      "learning_rate": 0.0002696671582246361,
+      "loss": 0.8754,
+      "step": 5324
+    },
+    {
+      "epoch": 0.94995986085095,
+      "grad_norm": 0.45897579193115234,
+      "learning_rate": 0.0002695973123083884,
+      "loss": 0.8245,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9501382570689502,
+      "grad_norm": 0.4809344410896301,
+      "learning_rate": 0.00026952746485297614,
+      "loss": 0.6617,
+      "step": 5326
+    },
+    {
+      "epoch": 0.9503166532869504,
+      "grad_norm": 0.4562990963459015,
+      "learning_rate": 0.00026945761586388524,
+      "loss": 0.8179,
+      "step": 5327
+    },
+    {
+      "epoch": 0.9504950495049505,
+      "grad_norm": 0.4504075050354004,
+      "learning_rate": 0.0002693877653466015,
+      "loss": 0.8011,
+      "step": 5328
+    },
+    {
+      "epoch": 0.9506734457229506,
+      "grad_norm": 0.44723081588745117,
+      "learning_rate": 0.0002693179133066111,
+      "loss": 0.7159,
+      "step": 5329
+    },
+    {
+      "epoch": 0.9508518419409508,
+      "grad_norm": 0.49957334995269775,
+      "learning_rate": 0.00026924805974940007,
+      "loss": 0.8277,
+      "step": 5330
+    },
+    {
+      "epoch": 0.951030238158951,
+      "grad_norm": 0.4783354699611664,
+      "learning_rate": 0.0002691782046804548,
+      "loss": 0.9436,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9512086343769512,
+      "grad_norm": 0.39161327481269836,
+      "learning_rate": 0.00026910834810526147,
+      "loss": 0.5881,
+      "step": 5332
+    },
+    {
+      "epoch": 0.9513870305949513,
+      "grad_norm": 0.46436673402786255,
+      "learning_rate": 0.00026903849002930677,
+      "loss": 0.7549,
+      "step": 5333
+    },
+    {
+      "epoch": 0.9515654268129515,
+      "grad_norm": 0.4534311592578888,
+      "learning_rate": 0.00026896863045807715,
+      "loss": 0.6577,
+      "step": 5334
+    },
+    {
+      "epoch": 0.9517438230309517,
+      "grad_norm": 0.4262526333332062,
+      "learning_rate": 0.00026889876939705946,
+      "loss": 0.7005,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9519222192489519,
+      "grad_norm": 0.49773791432380676,
+      "learning_rate": 0.00026882890685174065,
+      "loss": 0.8739,
+      "step": 5336
+    },
+    {
+      "epoch": 0.9521006154669521,
+      "grad_norm": 0.46525028347969055,
+      "learning_rate": 0.00026875904282760765,
+      "loss": 0.6493,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9522790116849523,
+      "grad_norm": 0.42473146319389343,
+      "learning_rate": 0.00026868917733014743,
+      "loss": 0.6203,
+      "step": 5338
+    },
+    {
+      "epoch": 0.9524574079029524,
+      "grad_norm": 0.49023476243019104,
+      "learning_rate": 0.0002686193103648472,
+      "loss": 0.9317,
+      "step": 5339
+    },
+    {
+      "epoch": 0.9526358041209526,
+      "grad_norm": 0.5018748641014099,
+      "learning_rate": 0.00026854944193719445,
+      "loss": 0.9483,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9528142003389528,
+      "grad_norm": 0.49157464504241943,
+      "learning_rate": 0.00026847957205267635,
+      "loss": 0.9226,
+      "step": 5341
+    },
+    {
+      "epoch": 0.952992596556953,
+      "grad_norm": 0.474185049533844,
+      "learning_rate": 0.0002684097007167807,
+      "loss": 0.8999,
+      "step": 5342
+    },
+    {
+      "epoch": 0.9531709927749532,
+      "grad_norm": 0.4604833126068115,
+      "learning_rate": 0.0002683398279349952,
+      "loss": 0.8502,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9533493889929533,
+      "grad_norm": 0.5128932595252991,
+      "learning_rate": 0.0002682699537128074,
+      "loss": 0.8059,
+      "step": 5344
+    },
+    {
+      "epoch": 0.9535277852109535,
+      "grad_norm": 0.42968082427978516,
+      "learning_rate": 0.00026820007805570536,
+      "loss": 0.6065,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9537061814289537,
+      "grad_norm": 0.5407286286354065,
+      "learning_rate": 0.00026813020096917695,
+      "loss": 0.9448,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9538845776469539,
+      "grad_norm": 1.0155121088027954,
+      "learning_rate": 0.0002680603224587104,
+      "loss": 1.2441,
+      "step": 5347
+    },
+    {
+      "epoch": 0.9540629738649541,
+      "grad_norm": 0.5304948687553406,
+      "learning_rate": 0.0002679904425297938,
+      "loss": 1.0268,
+      "step": 5348
+    },
+    {
+      "epoch": 0.9542413700829543,
+      "grad_norm": 0.4759924113750458,
+      "learning_rate": 0.00026792056118791563,
+      "loss": 0.9434,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9544197663009544,
+      "grad_norm": 0.3957015573978424,
+      "learning_rate": 0.00026785067843856437,
+      "loss": 0.6315,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9545981625189546,
+      "grad_norm": 0.46599963307380676,
+      "learning_rate": 0.00026778079428722845,
+      "loss": 0.7155,
+      "step": 5351
+    },
+    {
+      "epoch": 0.9547765587369548,
+      "grad_norm": 0.4720132648944855,
+      "learning_rate": 0.0002677109087393966,
+      "loss": 0.8976,
+      "step": 5352
+    },
+    {
+      "epoch": 0.954954954954955,
+      "grad_norm": 0.5006893277168274,
+      "learning_rate": 0.00026764102180055766,
+      "loss": 0.8828,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9551333511729552,
+      "grad_norm": 0.4419143795967102,
+      "learning_rate": 0.0002675711334762004,
+      "loss": 0.7284,
+      "step": 5354
+    },
+    {
+      "epoch": 0.9553117473909553,
+      "grad_norm": 0.4111958146095276,
+      "learning_rate": 0.0002675012437718139,
+      "loss": 0.6126,
+      "step": 5355
+    },
+    {
+      "epoch": 0.9554901436089555,
+      "grad_norm": 0.5170559287071228,
+      "learning_rate": 0.0002674313526928872,
+      "loss": 0.9313,
+      "step": 5356
+    },
+    {
+      "epoch": 0.9556685398269557,
+      "grad_norm": 0.4762997329235077,
+      "learning_rate": 0.0002673614602449096,
+      "loss": 0.8053,
+      "step": 5357
+    },
+    {
+      "epoch": 0.9558469360449559,
+      "grad_norm": 0.4590204358100891,
+      "learning_rate": 0.0002672915664333704,
+      "loss": 0.85,
+      "step": 5358
+    },
+    {
+      "epoch": 0.956025332262956,
+      "grad_norm": 0.5133612155914307,
+      "learning_rate": 0.000267221671263759,
+      "loss": 0.7038,
+      "step": 5359
+    },
+    {
+      "epoch": 0.9562037284809562,
+      "grad_norm": 0.4732131063938141,
+      "learning_rate": 0.0002671517747415649,
+      "loss": 0.8324,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9563821246989563,
+      "grad_norm": 0.4754936099052429,
+      "learning_rate": 0.0002670818768722778,
+      "loss": 0.9258,
+      "step": 5361
+    },
+    {
+      "epoch": 0.9565605209169565,
+      "grad_norm": 0.4571801722049713,
+      "learning_rate": 0.0002670119776613875,
+      "loss": 0.8035,
+      "step": 5362
+    },
+    {
+      "epoch": 0.9567389171349567,
+      "grad_norm": 0.48625341057777405,
+      "learning_rate": 0.0002669420771143838,
+      "loss": 0.8987,
+      "step": 5363
+    },
+    {
+      "epoch": 0.9569173133529569,
+      "grad_norm": 0.4703410267829895,
+      "learning_rate": 0.0002668721752367566,
+      "loss": 0.9292,
+      "step": 5364
+    },
+    {
+      "epoch": 0.9570957095709571,
+      "grad_norm": 0.458659291267395,
+      "learning_rate": 0.00026680227203399604,
+      "loss": 0.7997,
+      "step": 5365
+    },
+    {
+      "epoch": 0.9572741057889572,
+      "grad_norm": 0.4934099018573761,
+      "learning_rate": 0.0002667323675115922,
+      "loss": 0.7636,
+      "step": 5366
+    },
+    {
+      "epoch": 0.9574525020069574,
+      "grad_norm": 0.4184805154800415,
+      "learning_rate": 0.0002666624616750355,
+      "loss": 0.6587,
+      "step": 5367
+    },
+    {
+      "epoch": 0.9576308982249576,
+      "grad_norm": 0.485462486743927,
+      "learning_rate": 0.00026659255452981623,
+      "loss": 0.9463,
+      "step": 5368
+    },
+    {
+      "epoch": 0.9578092944429578,
+      "grad_norm": 0.6165233254432678,
+      "learning_rate": 0.00026652264608142484,
+      "loss": 0.6606,
+      "step": 5369
+    },
+    {
+      "epoch": 0.957987690660958,
+      "grad_norm": 0.44948717951774597,
+      "learning_rate": 0.000266452736335352,
+      "loss": 0.7658,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9581660868789582,
+      "grad_norm": 0.40302959084510803,
+      "learning_rate": 0.0002663828252970883,
+      "loss": 0.6658,
+      "step": 5371
+    },
+    {
+      "epoch": 0.9583444830969583,
+      "grad_norm": 0.4867773652076721,
+      "learning_rate": 0.00026631291297212444,
+      "loss": 0.7478,
+      "step": 5372
+    },
+    {
+      "epoch": 0.9585228793149585,
+      "grad_norm": 0.4726101756095886,
+      "learning_rate": 0.0002662429993659515,
+      "loss": 0.8439,
+      "step": 5373
+    },
+    {
+      "epoch": 0.9587012755329587,
+      "grad_norm": 0.5044912099838257,
+      "learning_rate": 0.0002661730844840604,
+      "loss": 0.9133,
+      "step": 5374
+    },
+    {
+      "epoch": 0.9588796717509589,
+      "grad_norm": 0.4630397856235504,
+      "learning_rate": 0.0002661031683319422,
+      "loss": 0.9888,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9590580679689591,
+      "grad_norm": 0.4974066913127899,
+      "learning_rate": 0.00026603325091508807,
+      "loss": 1.1051,
+      "step": 5376
+    },
+    {
+      "epoch": 0.9592364641869592,
+      "grad_norm": 0.42207884788513184,
+      "learning_rate": 0.00026596333223898933,
+      "loss": 0.7558,
+      "step": 5377
+    },
+    {
+      "epoch": 0.9594148604049594,
+      "grad_norm": 0.5264635682106018,
+      "learning_rate": 0.00026589341230913736,
+      "loss": 0.9417,
+      "step": 5378
+    },
+    {
+      "epoch": 0.9595932566229596,
+      "grad_norm": 0.467877596616745,
+      "learning_rate": 0.0002658234911310236,
+      "loss": 0.6177,
+      "step": 5379
+    },
+    {
+      "epoch": 0.9597716528409598,
+      "grad_norm": 0.5078222155570984,
+      "learning_rate": 0.0002657535687101396,
+      "loss": 0.9433,
+      "step": 5380
+    },
+    {
+      "epoch": 0.95995004905896,
+      "grad_norm": 0.5080622434616089,
+      "learning_rate": 0.0002656836450519772,
+      "loss": 0.8492,
+      "step": 5381
+    },
+    {
+      "epoch": 0.9601284452769602,
+      "grad_norm": 0.4674982726573944,
+      "learning_rate": 0.000265613720162028,
+      "loss": 0.7062,
+      "step": 5382
+    },
+    {
+      "epoch": 0.9603068414949603,
+      "grad_norm": 0.544535219669342,
+      "learning_rate": 0.00026554379404578396,
+      "loss": 0.8553,
+      "step": 5383
+    },
+    {
+      "epoch": 0.9604852377129605,
+      "grad_norm": 0.4691609740257263,
+      "learning_rate": 0.00026547386670873707,
+      "loss": 0.747,
+      "step": 5384
+    },
+    {
+      "epoch": 0.9606636339309607,
+      "grad_norm": 0.5239211916923523,
+      "learning_rate": 0.00026540393815637924,
+      "loss": 0.7682,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9608420301489609,
+      "grad_norm": 0.4878639280796051,
+      "learning_rate": 0.00026533400839420286,
+      "loss": 1.0158,
+      "step": 5386
+    },
+    {
+      "epoch": 0.9610204263669611,
+      "grad_norm": 0.45283398032188416,
+      "learning_rate": 0.0002652640774276999,
+      "loss": 0.9548,
+      "step": 5387
+    },
+    {
+      "epoch": 0.9611988225849611,
+      "grad_norm": 0.4687561094760895,
+      "learning_rate": 0.00026519414526236297,
+      "loss": 0.8433,
+      "step": 5388
+    },
+    {
+      "epoch": 0.9613772188029613,
+      "grad_norm": 0.515119731426239,
+      "learning_rate": 0.0002651242119036844,
+      "loss": 0.8019,
+      "step": 5389
+    },
+    {
+      "epoch": 0.9615556150209615,
+      "grad_norm": 0.42879781126976013,
+      "learning_rate": 0.00026505427735715675,
+      "loss": 0.7789,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9617340112389617,
+      "grad_norm": 0.49096372723579407,
+      "learning_rate": 0.00026498434162827266,
+      "loss": 0.9128,
+      "step": 5391
+    },
+    {
+      "epoch": 0.9619124074569619,
+      "grad_norm": 0.45153510570526123,
+      "learning_rate": 0.00026491440472252475,
+      "loss": 0.7555,
+      "step": 5392
+    },
+    {
+      "epoch": 0.9620908036749621,
+      "grad_norm": 0.43661659955978394,
+      "learning_rate": 0.00026484446664540594,
+      "loss": 0.6922,
+      "step": 5393
+    },
+    {
+      "epoch": 0.9622691998929622,
+      "grad_norm": 0.4993753731250763,
+      "learning_rate": 0.00026477452740240914,
+      "loss": 0.7653,
+      "step": 5394
+    },
+    {
+      "epoch": 0.9624475961109624,
+      "grad_norm": 0.4706138074398041,
+      "learning_rate": 0.00026470458699902723,
+      "loss": 0.9497,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9626259923289626,
+      "grad_norm": 0.4838847815990448,
+      "learning_rate": 0.00026463464544075344,
+      "loss": 0.8203,
+      "step": 5396
+    },
+    {
+      "epoch": 0.9628043885469628,
+      "grad_norm": 0.4903802275657654,
+      "learning_rate": 0.000264564702733081,
+      "loss": 0.8855,
+      "step": 5397
+    },
+    {
+      "epoch": 0.962982784764963,
+      "grad_norm": 0.5173416137695312,
+      "learning_rate": 0.00026449475888150293,
+      "loss": 0.872,
+      "step": 5398
+    },
+    {
+      "epoch": 0.9631611809829632,
+      "grad_norm": 0.4816221594810486,
+      "learning_rate": 0.0002644248138915128,
+      "loss": 0.8021,
+      "step": 5399
+    },
+    {
+      "epoch": 0.9633395772009633,
+      "grad_norm": 0.4750434160232544,
+      "learning_rate": 0.00026435486776860395,
+      "loss": 0.8306,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9635179734189635,
+      "grad_norm": 0.47214189171791077,
+      "learning_rate": 0.00026428492051827,
+      "loss": 0.8921,
+      "step": 5401
+    },
+    {
+      "epoch": 0.9636963696369637,
+      "grad_norm": 0.4591917097568512,
+      "learning_rate": 0.0002642149721460045,
+      "loss": 0.7029,
+      "step": 5402
+    },
+    {
+      "epoch": 0.9638747658549639,
+      "grad_norm": 0.5314555168151855,
+      "learning_rate": 0.00026414502265730125,
+      "loss": 0.9167,
+      "step": 5403
+    },
+    {
+      "epoch": 0.9640531620729641,
+      "grad_norm": 0.4804200232028961,
+      "learning_rate": 0.000264075072057654,
+      "loss": 0.9891,
+      "step": 5404
+    },
+    {
+      "epoch": 0.9642315582909642,
+      "grad_norm": 0.4692907929420471,
+      "learning_rate": 0.00026400512035255663,
+      "loss": 0.6985,
+      "step": 5405
+    },
+    {
+      "epoch": 0.9644099545089644,
+      "grad_norm": 0.4961127042770386,
+      "learning_rate": 0.00026393516754750313,
+      "loss": 0.8653,
+      "step": 5406
+    },
+    {
+      "epoch": 0.9645883507269646,
+      "grad_norm": 0.455837607383728,
+      "learning_rate": 0.0002638652136479876,
+      "loss": 0.6942,
+      "step": 5407
+    },
+    {
+      "epoch": 0.9647667469449648,
+      "grad_norm": 0.5186680555343628,
+      "learning_rate": 0.0002637952586595041,
+      "loss": 1.0216,
+      "step": 5408
+    },
+    {
+      "epoch": 0.964945143162965,
+      "grad_norm": 0.49785247445106506,
+      "learning_rate": 0.00026372530258754695,
+      "loss": 0.7818,
+      "step": 5409
+    },
+    {
+      "epoch": 0.9651235393809652,
+      "grad_norm": 0.4302063286304474,
+      "learning_rate": 0.0002636553454376105,
+      "loss": 0.6761,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9653019355989653,
+      "grad_norm": 0.4406273066997528,
+      "learning_rate": 0.00026358538721518905,
+      "loss": 0.7473,
+      "step": 5411
+    },
+    {
+      "epoch": 0.9654803318169655,
+      "grad_norm": 0.42952749133110046,
+      "learning_rate": 0.0002635154279257771,
+      "loss": 0.6809,
+      "step": 5412
+    },
+    {
+      "epoch": 0.9656587280349657,
+      "grad_norm": 0.502311110496521,
+      "learning_rate": 0.00026344546757486924,
+      "loss": 1.1199,
+      "step": 5413
+    },
+    {
+      "epoch": 0.9658371242529659,
+      "grad_norm": 0.4300912022590637,
+      "learning_rate": 0.00026337550616796024,
+      "loss": 0.9147,
+      "step": 5414
+    },
+    {
+      "epoch": 0.9660155204709661,
+      "grad_norm": 0.4525073766708374,
+      "learning_rate": 0.00026330554371054466,
+      "loss": 0.9604,
+      "step": 5415
+    },
+    {
+      "epoch": 0.9661939166889661,
+      "grad_norm": 0.41568025946617126,
+      "learning_rate": 0.00026323558020811745,
+      "loss": 0.7829,
+      "step": 5416
+    },
+    {
+      "epoch": 0.9663723129069663,
+      "grad_norm": 0.5461636781692505,
+      "learning_rate": 0.00026316561566617347,
+      "loss": 0.8039,
+      "step": 5417
+    },
+    {
+      "epoch": 0.9665507091249665,
+      "grad_norm": 0.4481653571128845,
+      "learning_rate": 0.00026309565009020766,
+      "loss": 0.8262,
+      "step": 5418
+    },
+    {
+      "epoch": 0.9667291053429667,
+      "grad_norm": 0.911888599395752,
+      "learning_rate": 0.00026302568348571514,
+      "loss": 0.8544,
+      "step": 5419
+    },
+    {
+      "epoch": 0.9669075015609669,
+      "grad_norm": 0.42414039373397827,
+      "learning_rate": 0.0002629557158581911,
+      "loss": 0.6467,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9670858977789671,
+      "grad_norm": 0.5355694890022278,
+      "learning_rate": 0.00026288574721313064,
+      "loss": 0.7265,
+      "step": 5421
+    },
+    {
+      "epoch": 0.9672642939969672,
+      "grad_norm": 0.8004781007766724,
+      "learning_rate": 0.0002628157775560291,
+      "loss": 0.8402,
+      "step": 5422
+    },
+    {
+      "epoch": 0.9674426902149674,
+      "grad_norm": 0.5186832547187805,
+      "learning_rate": 0.00026274580689238206,
+      "loss": 0.9926,
+      "step": 5423
+    },
+    {
+      "epoch": 0.9676210864329676,
+      "grad_norm": 0.5586687922477722,
+      "learning_rate": 0.00026267583522768473,
+      "loss": 1.045,
+      "step": 5424
+    },
+    {
+      "epoch": 0.9677994826509678,
+      "grad_norm": 0.4568381607532501,
+      "learning_rate": 0.0002626058625674328,
+      "loss": 0.6261,
+      "step": 5425
+    },
+    {
+      "epoch": 0.967977878868968,
+      "grad_norm": 0.4433128237724304,
+      "learning_rate": 0.0002625358889171217,
+      "loss": 0.9197,
+      "step": 5426
+    },
+    {
+      "epoch": 0.9681562750869681,
+      "grad_norm": 0.4375893175601959,
+      "learning_rate": 0.00026246591428224743,
+      "loss": 0.6837,
+      "step": 5427
+    },
+    {
+      "epoch": 0.9683346713049683,
+      "grad_norm": 0.48804354667663574,
+      "learning_rate": 0.0002623959386683056,
+      "loss": 0.9795,
+      "step": 5428
+    },
+    {
+      "epoch": 0.9685130675229685,
+      "grad_norm": 0.43355077505111694,
+      "learning_rate": 0.00026232596208079203,
+      "loss": 0.7921,
+      "step": 5429
+    },
+    {
+      "epoch": 0.9686914637409687,
+      "grad_norm": 0.4870491623878479,
+      "learning_rate": 0.00026225598452520277,
+      "loss": 0.9069,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9688698599589689,
+      "grad_norm": 0.463943749666214,
+      "learning_rate": 0.00026218600600703376,
+      "loss": 0.7366,
+      "step": 5431
+    },
+    {
+      "epoch": 0.9690482561769691,
+      "grad_norm": 0.46572330594062805,
+      "learning_rate": 0.000262116026531781,
+      "loss": 0.7924,
+      "step": 5432
+    },
+    {
+      "epoch": 0.9692266523949692,
+      "grad_norm": 0.42976751923561096,
+      "learning_rate": 0.00026204604610494077,
+      "loss": 0.638,
+      "step": 5433
+    },
+    {
+      "epoch": 0.9694050486129694,
+      "grad_norm": 0.5167571306228638,
+      "learning_rate": 0.0002619760647320092,
+      "loss": 0.7912,
+      "step": 5434
+    },
+    {
+      "epoch": 0.9695834448309696,
+      "grad_norm": 0.39845386147499084,
+      "learning_rate": 0.0002619060824184828,
+      "loss": 0.6637,
+      "step": 5435
+    },
+    {
+      "epoch": 0.9697618410489698,
+      "grad_norm": 0.4528100788593292,
+      "learning_rate": 0.00026183609916985776,
+      "loss": 0.6917,
+      "step": 5436
+    },
+    {
+      "epoch": 0.96994023726697,
+      "grad_norm": 0.4622432291507721,
+      "learning_rate": 0.00026176611499163056,
+      "loss": 0.721,
+      "step": 5437
+    },
+    {
+      "epoch": 0.9701186334849701,
+      "grad_norm": 0.447052538394928,
+      "learning_rate": 0.00026169612988929773,
+      "loss": 0.6908,
+      "step": 5438
+    },
+    {
+      "epoch": 0.9702970297029703,
+      "grad_norm": 0.4297953248023987,
+      "learning_rate": 0.00026162614386835597,
+      "loss": 0.7084,
+      "step": 5439
+    },
+    {
+      "epoch": 0.9704754259209705,
+      "grad_norm": 0.4557560682296753,
+      "learning_rate": 0.0002615561569343018,
+      "loss": 0.8465,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9706538221389707,
+      "grad_norm": 0.42266544699668884,
+      "learning_rate": 0.000261486169092632,
+      "loss": 0.6536,
+      "step": 5441
+    },
+    {
+      "epoch": 0.9708322183569709,
+      "grad_norm": 0.4929708242416382,
+      "learning_rate": 0.0002614161803488435,
+      "loss": 0.9094,
+      "step": 5442
+    },
+    {
+      "epoch": 0.9710106145749711,
+      "grad_norm": 0.45966464281082153,
+      "learning_rate": 0.0002613461907084331,
+      "loss": 0.7899,
+      "step": 5443
+    },
+    {
+      "epoch": 0.9711890107929712,
+      "grad_norm": 0.5083897113800049,
+      "learning_rate": 0.0002612762001768978,
+      "loss": 0.9044,
+      "step": 5444
+    },
+    {
+      "epoch": 0.9713674070109714,
+      "grad_norm": 0.4741891026496887,
+      "learning_rate": 0.00026120620875973453,
+      "loss": 0.9287,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9715458032289715,
+      "grad_norm": 1.0679816007614136,
+      "learning_rate": 0.00026113621646244045,
+      "loss": 0.999,
+      "step": 5446
+    },
+    {
+      "epoch": 0.9717241994469717,
+      "grad_norm": 0.4491601288318634,
+      "learning_rate": 0.0002610662232905127,
+      "loss": 0.8127,
+      "step": 5447
+    },
+    {
+      "epoch": 0.971902595664972,
+      "grad_norm": 0.538118302822113,
+      "learning_rate": 0.00026099622924944863,
+      "loss": 0.905,
+      "step": 5448
+    },
+    {
+      "epoch": 0.972080991882972,
+      "grad_norm": 0.4619334936141968,
+      "learning_rate": 0.0002609262343447454,
+      "loss": 0.6049,
+      "step": 5449
+    },
+    {
+      "epoch": 0.9722593881009722,
+      "grad_norm": 0.5011905431747437,
+      "learning_rate": 0.0002608562385819004,
+      "loss": 0.6995,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9724377843189724,
+      "grad_norm": 0.5354875326156616,
+      "learning_rate": 0.0002607862419664111,
+      "loss": 0.585,
+      "step": 5451
+    },
+    {
+      "epoch": 0.9726161805369726,
+      "grad_norm": 0.46661576628685,
+      "learning_rate": 0.00026071624450377495,
+      "loss": 0.7073,
+      "step": 5452
+    },
+    {
+      "epoch": 0.9727945767549728,
+      "grad_norm": 0.511663556098938,
+      "learning_rate": 0.0002606462461994896,
+      "loss": 0.8606,
+      "step": 5453
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.48708683252334595,
+      "learning_rate": 0.0002605762470590527,
+      "loss": 0.6398,
+      "step": 5454
+    },
+    {
+      "epoch": 0.9731513691909731,
+      "grad_norm": 0.48029178380966187,
+      "learning_rate": 0.0002605062470879619,
+      "loss": 0.7369,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9733297654089733,
+      "grad_norm": 0.481996089220047,
+      "learning_rate": 0.00026043624629171495,
+      "loss": 0.9269,
+      "step": 5456
+    },
+    {
+      "epoch": 0.9735081616269735,
+      "grad_norm": 0.5002886652946472,
+      "learning_rate": 0.0002603662446758097,
+      "loss": 0.8318,
+      "step": 5457
+    },
+    {
+      "epoch": 0.9736865578449737,
+      "grad_norm": 0.48849231004714966,
+      "learning_rate": 0.0002602962422457441,
+      "loss": 0.7965,
+      "step": 5458
+    },
+    {
+      "epoch": 0.9738649540629739,
+      "grad_norm": 0.4990347623825073,
+      "learning_rate": 0.000260226239007016,
+      "loss": 0.87,
+      "step": 5459
+    },
+    {
+      "epoch": 0.974043350280974,
+      "grad_norm": 0.5066533088684082,
+      "learning_rate": 0.0002601562349651235,
+      "loss": 0.7364,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9742217464989742,
+      "grad_norm": 0.5521520972251892,
+      "learning_rate": 0.0002600862301255647,
+      "loss": 0.9395,
+      "step": 5461
+    },
+    {
+      "epoch": 0.9744001427169744,
+      "grad_norm": 0.7457981109619141,
+      "learning_rate": 0.00026001622449383776,
+      "loss": 0.8701,
+      "step": 5462
+    },
+    {
+      "epoch": 0.9745785389349746,
+      "grad_norm": 0.5160167217254639,
+      "learning_rate": 0.00025994621807544084,
+      "loss": 0.9626,
+      "step": 5463
+    },
+    {
+      "epoch": 0.9747569351529748,
+      "grad_norm": 0.504137396812439,
+      "learning_rate": 0.0002598762108758722,
+      "loss": 0.7476,
+      "step": 5464
+    },
+    {
+      "epoch": 0.974935331370975,
+      "grad_norm": 0.4687594473361969,
+      "learning_rate": 0.00025980620290063023,
+      "loss": 0.7256,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9751137275889751,
+      "grad_norm": 0.44819939136505127,
+      "learning_rate": 0.0002597361941552133,
+      "loss": 0.6916,
+      "step": 5466
+    },
+    {
+      "epoch": 0.9752921238069753,
+      "grad_norm": 2.4401307106018066,
+      "learning_rate": 0.00025966618464511986,
+      "loss": 0.7513,
+      "step": 5467
+    },
+    {
+      "epoch": 0.9754705200249755,
+      "grad_norm": 0.48852846026420593,
+      "learning_rate": 0.0002595961743758484,
+      "loss": 0.9704,
+      "step": 5468
+    },
+    {
+      "epoch": 0.9756489162429757,
+      "grad_norm": 0.47070446610450745,
+      "learning_rate": 0.00025952616335289766,
+      "loss": 0.8079,
+      "step": 5469
+    },
+    {
+      "epoch": 0.9758273124609759,
+      "grad_norm": 0.49295809864997864,
+      "learning_rate": 0.00025945615158176605,
+      "loss": 1.0164,
+      "step": 5470
+    },
+    {
+      "epoch": 0.976005708678976,
+      "grad_norm": 0.5102391839027405,
+      "learning_rate": 0.00025938613906795237,
+      "loss": 1.1723,
+      "step": 5471
+    },
+    {
+      "epoch": 0.9761841048969762,
+      "grad_norm": 0.4273965656757355,
+      "learning_rate": 0.0002593161258169554,
+      "loss": 0.5914,
+      "step": 5472
+    },
+    {
+      "epoch": 0.9763625011149764,
+      "grad_norm": 0.4733925759792328,
+      "learning_rate": 0.00025924611183427386,
+      "loss": 0.8552,
+      "step": 5473
+    },
+    {
+      "epoch": 0.9765408973329766,
+      "grad_norm": 0.4754532277584076,
+      "learning_rate": 0.00025917609712540674,
+      "loss": 0.8698,
+      "step": 5474
+    },
+    {
+      "epoch": 0.9767192935509768,
+      "grad_norm": 0.5383017063140869,
+      "learning_rate": 0.0002591060816958529,
+      "loss": 1.0911,
+      "step": 5475
+    },
+    {
+      "epoch": 0.976897689768977,
+      "grad_norm": 0.50196772813797,
+      "learning_rate": 0.00025903606555111123,
+      "loss": 0.9058,
+      "step": 5476
+    },
+    {
+      "epoch": 0.977076085986977,
+      "grad_norm": 0.5124291181564331,
+      "learning_rate": 0.000258966048696681,
+      "loss": 0.9839,
+      "step": 5477
+    },
+    {
+      "epoch": 0.9772544822049772,
+      "grad_norm": 0.5031840801239014,
+      "learning_rate": 0.0002588960311380611,
+      "loss": 0.7175,
+      "step": 5478
+    },
+    {
+      "epoch": 0.9774328784229774,
+      "grad_norm": 0.46920791268348694,
+      "learning_rate": 0.0002588260128807507,
+      "loss": 0.8517,
+      "step": 5479
+    },
+    {
+      "epoch": 0.9776112746409776,
+      "grad_norm": 0.5168631672859192,
+      "learning_rate": 0.0002587559939302491,
+      "loss": 0.964,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9777896708589778,
+      "grad_norm": 0.44529473781585693,
+      "learning_rate": 0.00025868597429205543,
+      "loss": 0.778,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9779680670769779,
+      "grad_norm": 0.4642591178417206,
+      "learning_rate": 0.00025861595397166915,
+      "loss": 0.7513,
+      "step": 5482
+    },
+    {
+      "epoch": 0.9781464632949781,
+      "grad_norm": 0.47108370065689087,
+      "learning_rate": 0.00025854593297458956,
+      "loss": 0.6897,
+      "step": 5483
+    },
+    {
+      "epoch": 0.9783248595129783,
+      "grad_norm": 0.5786851048469543,
+      "learning_rate": 0.00025847591130631603,
+      "loss": 0.9704,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9785032557309785,
+      "grad_norm": 0.4903102219104767,
+      "learning_rate": 0.0002584058889723481,
+      "loss": 0.9588,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9786816519489787,
+      "grad_norm": 0.40852758288383484,
+      "learning_rate": 0.00025833586597818526,
+      "loss": 0.6194,
+      "step": 5486
+    },
+    {
+      "epoch": 0.9788600481669789,
+      "grad_norm": 0.42910122871398926,
+      "learning_rate": 0.00025826584232932704,
+      "loss": 0.7026,
+      "step": 5487
+    },
+    {
+      "epoch": 0.979038444384979,
+      "grad_norm": 0.6865221261978149,
+      "learning_rate": 0.00025819581803127316,
+      "loss": 0.932,
+      "step": 5488
+    },
+    {
+      "epoch": 0.9792168406029792,
+      "grad_norm": 0.44170790910720825,
+      "learning_rate": 0.0002581257930895233,
+      "loss": 0.7926,
+      "step": 5489
+    },
+    {
+      "epoch": 0.9793952368209794,
+      "grad_norm": 0.4942973256111145,
+      "learning_rate": 0.00025805576750957714,
+      "loss": 0.9616,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9795736330389796,
+      "grad_norm": 0.5086950063705444,
+      "learning_rate": 0.0002579857412969345,
+      "loss": 0.9139,
+      "step": 5491
+    },
+    {
+      "epoch": 0.9797520292569798,
+      "grad_norm": 0.5349062085151672,
+      "learning_rate": 0.00025791571445709505,
+      "loss": 0.9468,
+      "step": 5492
+    },
+    {
+      "epoch": 0.9799304254749799,
+      "grad_norm": 0.5083227157592773,
+      "learning_rate": 0.0002578456869955589,
+      "loss": 0.9298,
+      "step": 5493
+    },
+    {
+      "epoch": 0.9801088216929801,
+      "grad_norm": 0.4308398962020874,
+      "learning_rate": 0.0002577756589178258,
+      "loss": 0.6871,
+      "step": 5494
+    },
+    {
+      "epoch": 0.9802872179109803,
+      "grad_norm": 0.3946779668331146,
+      "learning_rate": 0.0002577056302293958,
+      "loss": 0.7342,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9804656141289805,
+      "grad_norm": 0.4864563047885895,
+      "learning_rate": 0.000257635600935769,
+      "loss": 0.9185,
+      "step": 5496
+    },
+    {
+      "epoch": 0.9806440103469807,
+      "grad_norm": 0.427357941865921,
+      "learning_rate": 0.00025756557104244534,
+      "loss": 0.7518,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9808224065649809,
+      "grad_norm": 0.4161073565483093,
+      "learning_rate": 0.000257495540554925,
+      "loss": 0.5618,
+      "step": 5498
+    },
+    {
+      "epoch": 0.981000802782981,
+      "grad_norm": 0.437459796667099,
+      "learning_rate": 0.00025742550947870806,
+      "loss": 0.638,
+      "step": 5499
+    },
+    {
+      "epoch": 0.9811791990009812,
+      "grad_norm": 0.4389030635356903,
+      "learning_rate": 0.00025735547781929484,
+      "loss": 0.6333,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9813575952189814,
+      "grad_norm": 0.47596055269241333,
+      "learning_rate": 0.00025728544558218557,
+      "loss": 0.7331,
+      "step": 5501
+    },
+    {
+      "epoch": 0.9815359914369816,
+      "grad_norm": 0.49048876762390137,
+      "learning_rate": 0.00025721541277288053,
+      "loss": 0.7336,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9817143876549818,
+      "grad_norm": 0.4905683398246765,
+      "learning_rate": 0.0002571453793968801,
+      "loss": 0.8261,
+      "step": 5503
+    },
+    {
+      "epoch": 0.9818927838729818,
+      "grad_norm": 0.46471673250198364,
+      "learning_rate": 0.0002570753454596846,
+      "loss": 0.8081,
+      "step": 5504
+    },
+    {
+      "epoch": 0.982071180090982,
+      "grad_norm": 0.49744272232055664,
+      "learning_rate": 0.00025700531096679456,
+      "loss": 0.8207,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9822495763089822,
+      "grad_norm": 0.4972292184829712,
+      "learning_rate": 0.0002569352759237104,
+      "loss": 1.0645,
+      "step": 5506
+    },
+    {
+      "epoch": 0.9824279725269824,
+      "grad_norm": 0.47991102933883667,
+      "learning_rate": 0.00025686524033593263,
+      "loss": 0.8118,
+      "step": 5507
+    },
+    {
+      "epoch": 0.9826063687449826,
+      "grad_norm": 0.41838109493255615,
+      "learning_rate": 0.00025679520420896184,
+      "loss": 0.7066,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9827847649629828,
+      "grad_norm": 0.40462827682495117,
+      "learning_rate": 0.00025672516754829866,
+      "loss": 0.6267,
+      "step": 5509
+    },
+    {
+      "epoch": 0.9829631611809829,
+      "grad_norm": 0.4698152542114258,
+      "learning_rate": 0.00025665513035944373,
+      "loss": 0.9033,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9831415573989831,
+      "grad_norm": 0.4284512400627136,
+      "learning_rate": 0.0002565850926478977,
+      "loss": 0.7329,
+      "step": 5511
+    },
+    {
+      "epoch": 0.9833199536169833,
+      "grad_norm": 0.4176551401615143,
+      "learning_rate": 0.0002565150544191613,
+      "loss": 0.7597,
+      "step": 5512
+    },
+    {
+      "epoch": 0.9834983498349835,
+      "grad_norm": 0.46959561109542847,
+      "learning_rate": 0.00025644501567873533,
+      "loss": 0.7277,
+      "step": 5513
+    },
+    {
+      "epoch": 0.9836767460529837,
+      "grad_norm": 0.4930439591407776,
+      "learning_rate": 0.0002563749764321207,
+      "loss": 0.8034,
+      "step": 5514
+    },
+    {
+      "epoch": 0.9838551422709838,
+      "grad_norm": 0.4476320743560791,
+      "learning_rate": 0.0002563049366848181,
+      "loss": 0.7959,
+      "step": 5515
+    },
+    {
+      "epoch": 0.984033538488984,
+      "grad_norm": 0.7321682572364807,
+      "learning_rate": 0.00025623489644232845,
+      "loss": 0.7308,
+      "step": 5516
+    },
+    {
+      "epoch": 0.9842119347069842,
+      "grad_norm": 0.447443425655365,
+      "learning_rate": 0.00025616485571015277,
+      "loss": 0.7,
+      "step": 5517
+    },
+    {
+      "epoch": 0.9843903309249844,
+      "grad_norm": 0.4954073131084442,
+      "learning_rate": 0.0002560948144937919,
+      "loss": 0.8115,
+      "step": 5518
+    },
+    {
+      "epoch": 0.9845687271429846,
+      "grad_norm": 0.5299807190895081,
+      "learning_rate": 0.00025602477279874697,
+      "loss": 0.9499,
+      "step": 5519
+    },
+    {
+      "epoch": 0.9847471233609848,
+      "grad_norm": 0.4475019574165344,
+      "learning_rate": 0.000255954730630519,
+      "loss": 0.7138,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9849255195789849,
+      "grad_norm": 0.4217219650745392,
+      "learning_rate": 0.000255884687994609,
+      "loss": 0.697,
+      "step": 5521
+    },
+    {
+      "epoch": 0.9851039157969851,
+      "grad_norm": 0.4400602877140045,
+      "learning_rate": 0.0002558146448965182,
+      "loss": 0.7563,
+      "step": 5522
+    },
+    {
+      "epoch": 0.9852823120149853,
+      "grad_norm": 0.42823049426078796,
+      "learning_rate": 0.0002557446013417477,
+      "loss": 0.6568,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9854607082329855,
+      "grad_norm": 0.4775795638561249,
+      "learning_rate": 0.00025567455733579867,
+      "loss": 0.6586,
+      "step": 5524
+    },
+    {
+      "epoch": 0.9856391044509857,
+      "grad_norm": 0.4584738314151764,
+      "learning_rate": 0.00025560451288417224,
+      "loss": 0.7584,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9858175006689858,
+      "grad_norm": 0.6482552289962769,
+      "learning_rate": 0.00025553446799236987,
+      "loss": 0.9836,
+      "step": 5526
+    },
+    {
+      "epoch": 0.985995896886986,
+      "grad_norm": 2.181297779083252,
+      "learning_rate": 0.00025546442266589274,
+      "loss": 0.9366,
+      "step": 5527
+    },
+    {
+      "epoch": 0.9861742931049862,
+      "grad_norm": 0.4901970326900482,
+      "learning_rate": 0.0002553943769102422,
+      "loss": 0.7734,
+      "step": 5528
+    },
+    {
+      "epoch": 0.9863526893229864,
+      "grad_norm": 0.5219231843948364,
+      "learning_rate": 0.00025532433073091967,
+      "loss": 0.7347,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9865310855409866,
+      "grad_norm": 0.5067287087440491,
+      "learning_rate": 0.0002552542841334265,
+      "loss": 0.8365,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9867094817589868,
+      "grad_norm": 3.1279547214508057,
+      "learning_rate": 0.0002551842371232641,
+      "loss": 0.8209,
+      "step": 5531
+    },
+    {
+      "epoch": 0.9868878779769868,
+      "grad_norm": 0.48862382769584656,
+      "learning_rate": 0.00025511418970593393,
+      "loss": 0.8062,
+      "step": 5532
+    },
+    {
+      "epoch": 0.987066274194987,
+      "grad_norm": 1.2135387659072876,
+      "learning_rate": 0.0002550441418869374,
+      "loss": 0.8549,
+      "step": 5533
+    },
+    {
+      "epoch": 0.9872446704129872,
+      "grad_norm": 0.8009121417999268,
+      "learning_rate": 0.00025497409367177627,
+      "loss": 0.8502,
+      "step": 5534
+    },
+    {
+      "epoch": 0.9874230666309874,
+      "grad_norm": 0.6405108571052551,
+      "learning_rate": 0.0002549040450659519,
+      "loss": 0.9773,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9876014628489876,
+      "grad_norm": 3.273838758468628,
+      "learning_rate": 0.00025483399607496604,
+      "loss": 0.9752,
+      "step": 5536
+    },
+    {
+      "epoch": 0.9877798590669877,
+      "grad_norm": 0.4358234703540802,
+      "learning_rate": 0.0002547639467043201,
+      "loss": 0.6485,
+      "step": 5537
+    },
+    {
+      "epoch": 0.9879582552849879,
+      "grad_norm": 0.5305948853492737,
+      "learning_rate": 0.00025469389695951595,
+      "loss": 0.8997,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9881366515029881,
+      "grad_norm": 0.7307850122451782,
+      "learning_rate": 0.0002546238468460551,
+      "loss": 0.7562,
+      "step": 5539
+    },
+    {
+      "epoch": 0.9883150477209883,
+      "grad_norm": 0.47485119104385376,
+      "learning_rate": 0.0002545537963694392,
+      "loss": 0.7355,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9884934439389885,
+      "grad_norm": 0.5570200085639954,
+      "learning_rate": 0.0002544837455351702,
+      "loss": 0.8597,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9886718401569887,
+      "grad_norm": 0.592291533946991,
+      "learning_rate": 0.00025441369434874977,
+      "loss": 1.0744,
+      "step": 5542
+    },
+    {
+      "epoch": 0.9888502363749888,
+      "grad_norm": 0.487498939037323,
+      "learning_rate": 0.0002543436428156796,
+      "loss": 0.8186,
+      "step": 5543
+    },
+    {
+      "epoch": 0.989028632592989,
+      "grad_norm": 0.475384920835495,
+      "learning_rate": 0.0002542735909414617,
+      "loss": 0.6558,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9892070288109892,
+      "grad_norm": 0.44877302646636963,
+      "learning_rate": 0.00025420353873159774,
+      "loss": 0.7219,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9893854250289894,
+      "grad_norm": 0.49043965339660645,
+      "learning_rate": 0.0002541334861915897,
+      "loss": 0.8794,
+      "step": 5546
+    },
+    {
+      "epoch": 0.9895638212469896,
+      "grad_norm": 0.47729647159576416,
+      "learning_rate": 0.00025406343332693934,
+      "loss": 0.8159,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9897422174649897,
+      "grad_norm": 0.48762455582618713,
+      "learning_rate": 0.0002539933801431487,
+      "loss": 0.7083,
+      "step": 5548
+    },
+    {
+      "epoch": 0.9899206136829899,
+      "grad_norm": 0.4554971158504486,
+      "learning_rate": 0.0002539233266457198,
+      "loss": 0.6679,
+      "step": 5549
+    },
+    {
+      "epoch": 0.9900990099009901,
+      "grad_norm": 0.5924381613731384,
+      "learning_rate": 0.0002538532728401544,
+      "loss": 0.9812,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9902774061189903,
+      "grad_norm": 0.5297411680221558,
+      "learning_rate": 0.0002537832187319547,
+      "loss": 0.7363,
+      "step": 5551
+    },
+    {
+      "epoch": 0.9904558023369905,
+      "grad_norm": 0.5259188413619995,
+      "learning_rate": 0.00025371316432662254,
+      "loss": 0.924,
+      "step": 5552
+    },
+    {
+      "epoch": 0.9906341985549907,
+      "grad_norm": 0.5142114162445068,
+      "learning_rate": 0.0002536431096296601,
+      "loss": 0.94,
+      "step": 5553
+    },
+    {
+      "epoch": 0.9908125947729908,
+      "grad_norm": 0.4876602590084076,
+      "learning_rate": 0.00025357305464656943,
+      "loss": 0.9249,
+      "step": 5554
+    },
+    {
+      "epoch": 0.990990990990991,
+      "grad_norm": 0.4915853440761566,
+      "learning_rate": 0.00025350299938285253,
+      "loss": 1.1248,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9911693872089912,
+      "grad_norm": 0.5007473826408386,
+      "learning_rate": 0.0002534329438440116,
+      "loss": 1.0163,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9913477834269914,
+      "grad_norm": 0.4649515151977539,
+      "learning_rate": 0.0002533628880355487,
+      "loss": 0.8398,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9915261796449916,
+      "grad_norm": 0.4449246823787689,
+      "learning_rate": 0.0002532928319629661,
+      "loss": 0.8213,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9917045758629917,
+      "grad_norm": 0.47130802273750305,
+      "learning_rate": 0.00025322277563176584,
+      "loss": 0.8082,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9918829720809919,
+      "grad_norm": 0.5442295670509338,
+      "learning_rate": 0.00025315271904745014,
+      "loss": 0.9156,
+      "step": 5560
+    },
+    {
+      "epoch": 0.992061368298992,
+      "grad_norm": 0.5047063827514648,
+      "learning_rate": 0.0002530826622155213,
+      "loss": 0.8635,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9922397645169923,
+      "grad_norm": 0.504047155380249,
+      "learning_rate": 0.00025301260514148146,
+      "loss": 0.9712,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9924181607349924,
+      "grad_norm": 0.46802255511283875,
+      "learning_rate": 0.0002529425478308329,
+      "loss": 0.8758,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9925965569529926,
+      "grad_norm": 0.5220746397972107,
+      "learning_rate": 0.00025287249028907796,
+      "loss": 1.001,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9927749531709927,
+      "grad_norm": 0.43903014063835144,
+      "learning_rate": 0.0002528024325217188,
+      "loss": 0.6244,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9929533493889929,
+      "grad_norm": 0.4588782787322998,
+      "learning_rate": 0.0002527323745342578,
+      "loss": 0.6223,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9931317456069931,
+      "grad_norm": 0.501520574092865,
+      "learning_rate": 0.0002526623163321973,
+      "loss": 0.9619,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9933101418249933,
+      "grad_norm": 0.4871865510940552,
+      "learning_rate": 0.0002525922579210396,
+      "loss": 0.9316,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9934885380429935,
+      "grad_norm": 0.4477481544017792,
+      "learning_rate": 0.0002525221993062871,
+      "loss": 0.7816,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9936669342609936,
+      "grad_norm": 0.4530073404312134,
+      "learning_rate": 0.00025245214049344225,
+      "loss": 0.7685,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9938453304789938,
+      "grad_norm": 0.4744289815425873,
+      "learning_rate": 0.0002523820814880072,
+      "loss": 0.7734,
+      "step": 5571
+    },
+    {
+      "epoch": 0.994023726696994,
+      "grad_norm": 0.4457191526889801,
+      "learning_rate": 0.0002523120222954845,
+      "loss": 0.895,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9942021229149942,
+      "grad_norm": 0.4728289842605591,
+      "learning_rate": 0.00025224196292137664,
+      "loss": 0.7997,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9943805191329944,
+      "grad_norm": 0.4260111451148987,
+      "learning_rate": 0.00025217190337118594,
+      "loss": 0.7461,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9945589153509946,
+      "grad_norm": 0.4607539176940918,
+      "learning_rate": 0.0002521018436504149,
+      "loss": 0.7959,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9947373115689947,
+      "grad_norm": 0.6200941801071167,
+      "learning_rate": 0.000252031783764566,
+      "loss": 0.8551,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9949157077869949,
+      "grad_norm": 0.4465862810611725,
+      "learning_rate": 0.0002519617237191416,
+      "loss": 0.7163,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9950941040049951,
+      "grad_norm": 0.4562668800354004,
+      "learning_rate": 0.00025189166351964425,
+      "loss": 0.7838,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9952725002229953,
+      "grad_norm": 0.45672228932380676,
+      "learning_rate": 0.0002518216031715765,
+      "loss": 0.7227,
+      "step": 5579
+    },
+    {
+      "epoch": 0.9954508964409955,
+      "grad_norm": 0.4691329002380371,
+      "learning_rate": 0.0002517515426804408,
+      "loss": 0.9586,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9956292926589956,
+      "grad_norm": 1.6988149881362915,
+      "learning_rate": 0.00025168148205173974,
+      "loss": 0.7888,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9958076888769958,
+      "grad_norm": 0.46691417694091797,
+      "learning_rate": 0.0002516114212909758,
+      "loss": 0.6838,
+      "step": 5582
+    },
+    {
+      "epoch": 0.995986085094996,
+      "grad_norm": 0.45711347460746765,
+      "learning_rate": 0.0002515413604036515,
+      "loss": 0.7862,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9961644813129962,
+      "grad_norm": 0.49638915061950684,
+      "learning_rate": 0.0002514712993952694,
+      "loss": 0.7689,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9963428775309964,
+      "grad_norm": 0.4389818608760834,
+      "learning_rate": 0.0002514012382713321,
+      "loss": 0.6992,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9965212737489966,
+      "grad_norm": 0.4768664538860321,
+      "learning_rate": 0.00025133117703734207,
+      "loss": 0.6965,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9966996699669967,
+      "grad_norm": 0.5143462419509888,
+      "learning_rate": 0.0002512611156988021,
+      "loss": 1.0382,
+      "step": 5587
+    },
+    {
+      "epoch": 0.9968780661849969,
+      "grad_norm": 0.47159501910209656,
+      "learning_rate": 0.00025119105426121455,
+      "loss": 0.7476,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9970564624029971,
+      "grad_norm": 0.4691692292690277,
+      "learning_rate": 0.0002511209927300822,
+      "loss": 0.8535,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9972348586209973,
+      "grad_norm": 0.44228291511535645,
+      "learning_rate": 0.00025105093111090756,
+      "loss": 0.777,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9974132548389975,
+      "grad_norm": 0.43507784605026245,
+      "learning_rate": 0.00025098086940919317,
+      "loss": 0.6588,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9975916510569975,
+      "grad_norm": 0.4751743972301483,
+      "learning_rate": 0.00025091080763044177,
+      "loss": 0.6748,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9977700472749977,
+      "grad_norm": 4.461918354034424,
+      "learning_rate": 0.0002508407457801559,
+      "loss": 0.669,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9979484434929979,
+      "grad_norm": 0.8328418135643005,
+      "learning_rate": 0.00025077068386383816,
+      "loss": 0.8627,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9981268397109981,
+      "grad_norm": 1.091975212097168,
+      "learning_rate": 0.00025070062188699136,
+      "loss": 0.8833,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9983052359289983,
+      "grad_norm": 0.4713873565196991,
+      "learning_rate": 0.00025063055985511794,
+      "loss": 0.6602,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9984836321469985,
+      "grad_norm": 0.5721749663352966,
+      "learning_rate": 0.0002505604977737207,
+      "loss": 0.8543,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9986620283649986,
+      "grad_norm": 0.5203986167907715,
+      "learning_rate": 0.00025049043564830207,
+      "loss": 0.7544,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9988404245829988,
+      "grad_norm": 0.7062872052192688,
+      "learning_rate": 0.00025042037348436497,
+      "loss": 0.8659,
+      "step": 5599
+    },
+    {
+      "epoch": 0.999018820800999,
+      "grad_norm": 0.476829469203949,
+      "learning_rate": 0.00025035031128741185,
+      "loss": 0.7615,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9991972170189992,
+      "grad_norm": 0.42391034960746765,
+      "learning_rate": 0.0002502802490629454,
+      "loss": 0.5286,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9993756132369994,
+      "grad_norm": 0.5202667713165283,
+      "learning_rate": 0.0002502101868164684,
+      "loss": 0.9911,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9995540094549995,
+      "grad_norm": 0.47505563497543335,
+      "learning_rate": 0.0002501401245534834,
+      "loss": 0.9305,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9997324056729997,
+      "grad_norm": 0.49714621901512146,
+      "learning_rate": 0.000250070062279493,
+      "loss": 0.8849,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9999108018909999,
+      "grad_norm": 0.5128735899925232,
+      "learning_rate": 0.00025,
+      "loss": 0.8105,
+      "step": 5605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8229463696479797,
+      "learning_rate": 0.000249929937720507,
+      "loss": 0.5342,
+      "step": 5606
+    },
+    {
+      "epoch": 1.0001783962180002,
+      "grad_norm": 0.48117566108703613,
+      "learning_rate": 0.00024985987544651667,
+      "loss": 0.6871,
+      "step": 5607
+    },
+    {
+      "epoch": 1.0003567924360004,
+      "grad_norm": 0.4833311438560486,
+      "learning_rate": 0.0002497898131835316,
+      "loss": 0.6285,
+      "step": 5608
+    },
+    {
+      "epoch": 1.0005351886540006,
+      "grad_norm": 0.5113769769668579,
+      "learning_rate": 0.0002497197509370546,
+      "loss": 0.7797,
+      "step": 5609
+    },
+    {
+      "epoch": 1.0007135848720008,
+      "grad_norm": 0.5292689204216003,
+      "learning_rate": 0.0002496496887125882,
+      "loss": 0.8098,
+      "step": 5610
+    },
+    {
+      "epoch": 1.000891981090001,
+      "grad_norm": 0.5784763097763062,
+      "learning_rate": 0.0002495796265156351,
+      "loss": 0.8951,
+      "step": 5611
+    },
+    {
+      "epoch": 1.001070377308001,
+      "grad_norm": 0.516196608543396,
+      "learning_rate": 0.000249509564351698,
+      "loss": 0.8567,
+      "step": 5612
+    },
+    {
+      "epoch": 1.0012487735260012,
+      "grad_norm": 0.49342212080955505,
+      "learning_rate": 0.0002494395022262793,
+      "loss": 0.6955,
+      "step": 5613
+    },
+    {
+      "epoch": 1.0014271697440014,
+      "grad_norm": 0.45294061303138733,
+      "learning_rate": 0.00024936944014488207,
+      "loss": 0.6438,
+      "step": 5614
+    },
+    {
+      "epoch": 1.0016055659620016,
+      "grad_norm": 0.5231368541717529,
+      "learning_rate": 0.0002492993781130087,
+      "loss": 0.7458,
+      "step": 5615
+    },
+    {
+      "epoch": 1.0017839621800018,
+      "grad_norm": 0.49573010206222534,
+      "learning_rate": 0.0002492293161361618,
+      "loss": 0.6514,
+      "step": 5616
+    },
+    {
+      "epoch": 1.001962358398002,
+      "grad_norm": 0.4060691297054291,
+      "learning_rate": 0.00024915925421984417,
+      "loss": 0.5311,
+      "step": 5617
+    },
+    {
+      "epoch": 1.0021407546160022,
+      "grad_norm": 0.5073151588439941,
+      "learning_rate": 0.0002490891923695583,
+      "loss": 0.7972,
+      "step": 5618
+    },
+    {
+      "epoch": 1.0023191508340024,
+      "grad_norm": 0.5470726490020752,
+      "learning_rate": 0.00024901913059080684,
+      "loss": 0.6951,
+      "step": 5619
+    },
+    {
+      "epoch": 1.0024975470520026,
+      "grad_norm": 0.4746871888637543,
+      "learning_rate": 0.0002489490688890925,
+      "loss": 0.678,
+      "step": 5620
+    },
+    {
+      "epoch": 1.0026759432700028,
+      "grad_norm": 0.6704702377319336,
+      "learning_rate": 0.0002488790072699178,
+      "loss": 0.9827,
+      "step": 5621
+    },
+    {
+      "epoch": 1.002854339488003,
+      "grad_norm": 0.5733181834220886,
+      "learning_rate": 0.00024880894573878546,
+      "loss": 0.6582,
+      "step": 5622
+    },
+    {
+      "epoch": 1.003032735706003,
+      "grad_norm": 0.4528675377368927,
+      "learning_rate": 0.00024873888430119794,
+      "loss": 0.6188,
+      "step": 5623
+    },
+    {
+      "epoch": 1.0032111319240031,
+      "grad_norm": 0.4826764166355133,
+      "learning_rate": 0.00024866882296265794,
+      "loss": 0.7185,
+      "step": 5624
+    },
+    {
+      "epoch": 1.0033895281420033,
+      "grad_norm": 0.4640424847602844,
+      "learning_rate": 0.000248598761728668,
+      "loss": 0.5546,
+      "step": 5625
+    },
+    {
+      "epoch": 1.0035679243600035,
+      "grad_norm": 0.57952880859375,
+      "learning_rate": 0.0002485287006047307,
+      "loss": 0.5323,
+      "step": 5626
+    },
+    {
+      "epoch": 1.0037463205780037,
+      "grad_norm": 0.4803033173084259,
+      "learning_rate": 0.0002484586395963486,
+      "loss": 0.7079,
+      "step": 5627
+    },
+    {
+      "epoch": 1.003924716796004,
+      "grad_norm": 0.47530657052993774,
+      "learning_rate": 0.0002483885787090242,
+      "loss": 0.7073,
+      "step": 5628
+    },
+    {
+      "epoch": 1.0041031130140041,
+      "grad_norm": 0.5128652453422546,
+      "learning_rate": 0.00024831851794826027,
+      "loss": 0.7797,
+      "step": 5629
+    },
+    {
+      "epoch": 1.0042815092320043,
+      "grad_norm": 0.4710089862346649,
+      "learning_rate": 0.0002482484573195592,
+      "loss": 0.7584,
+      "step": 5630
+    },
+    {
+      "epoch": 1.0044599054500045,
+      "grad_norm": 0.5013584494590759,
+      "learning_rate": 0.0002481783968284235,
+      "loss": 0.7168,
+      "step": 5631
+    },
+    {
+      "epoch": 1.0046383016680047,
+      "grad_norm": 0.48483264446258545,
+      "learning_rate": 0.00024810833648035576,
+      "loss": 0.7159,
+      "step": 5632
+    },
+    {
+      "epoch": 1.004816697886005,
+      "grad_norm": 0.43199649453163147,
+      "learning_rate": 0.00024803827628085845,
+      "loss": 0.6508,
+      "step": 5633
+    },
+    {
+      "epoch": 1.004995094104005,
+      "grad_norm": 0.48384973406791687,
+      "learning_rate": 0.00024796821623543407,
+      "loss": 0.8085,
+      "step": 5634
+    },
+    {
+      "epoch": 1.005173490322005,
+      "grad_norm": 0.5814610719680786,
+      "learning_rate": 0.00024789815634958517,
+      "loss": 0.7655,
+      "step": 5635
+    },
+    {
+      "epoch": 1.0053518865400053,
+      "grad_norm": 0.4795527160167694,
+      "learning_rate": 0.0002478280966288141,
+      "loss": 0.7431,
+      "step": 5636
+    },
+    {
+      "epoch": 1.0055302827580055,
+      "grad_norm": 0.4990082383155823,
+      "learning_rate": 0.0002477580370786234,
+      "loss": 0.7987,
+      "step": 5637
+    },
+    {
+      "epoch": 1.0057086789760057,
+      "grad_norm": 0.5206576585769653,
+      "learning_rate": 0.0002476879777045155,
+      "loss": 0.7974,
+      "step": 5638
+    },
+    {
+      "epoch": 1.0058870751940059,
+      "grad_norm": 0.38653305172920227,
+      "learning_rate": 0.00024761791851199286,
+      "loss": 0.5669,
+      "step": 5639
+    },
+    {
+      "epoch": 1.006065471412006,
+      "grad_norm": 0.49176260828971863,
+      "learning_rate": 0.0002475478595065578,
+      "loss": 0.7417,
+      "step": 5640
+    },
+    {
+      "epoch": 1.0062438676300063,
+      "grad_norm": 0.4860985279083252,
+      "learning_rate": 0.0002474778006937129,
+      "loss": 0.5978,
+      "step": 5641
+    },
+    {
+      "epoch": 1.0064222638480065,
+      "grad_norm": 0.4277432858943939,
+      "learning_rate": 0.0002474077420789604,
+      "loss": 0.4554,
+      "step": 5642
+    },
+    {
+      "epoch": 1.0066006600660067,
+      "grad_norm": 0.43579888343811035,
+      "learning_rate": 0.0002473376836678028,
+      "loss": 0.6738,
+      "step": 5643
+    },
+    {
+      "epoch": 1.0067790562840069,
+      "grad_norm": 0.5105220079421997,
+      "learning_rate": 0.00024726762546574215,
+      "loss": 0.9275,
+      "step": 5644
+    },
+    {
+      "epoch": 1.0069574525020069,
+      "grad_norm": 0.4739680886268616,
+      "learning_rate": 0.0002471975674782812,
+      "loss": 0.7133,
+      "step": 5645
+    },
+    {
+      "epoch": 1.007135848720007,
+      "grad_norm": 0.4873730540275574,
+      "learning_rate": 0.00024712750971092205,
+      "loss": 0.7338,
+      "step": 5646
+    },
+    {
+      "epoch": 1.0073142449380073,
+      "grad_norm": 0.45006611943244934,
+      "learning_rate": 0.0002470574521691671,
+      "loss": 0.6444,
+      "step": 5647
+    },
+    {
+      "epoch": 1.0074926411560075,
+      "grad_norm": 0.46951183676719666,
+      "learning_rate": 0.0002469873948585186,
+      "loss": 0.7305,
+      "step": 5648
+    },
+    {
+      "epoch": 1.0076710373740076,
+      "grad_norm": 0.4490543007850647,
+      "learning_rate": 0.00024691733778447875,
+      "loss": 0.6245,
+      "step": 5649
+    },
+    {
+      "epoch": 1.0078494335920078,
+      "grad_norm": 0.5618079900741577,
+      "learning_rate": 0.00024684728095254987,
+      "loss": 0.787,
+      "step": 5650
+    },
+    {
+      "epoch": 1.008027829810008,
+      "grad_norm": 7.213865756988525,
+      "learning_rate": 0.0002467772243682342,
+      "loss": 0.5768,
+      "step": 5651
+    },
+    {
+      "epoch": 1.0082062260280082,
+      "grad_norm": 0.5140239000320435,
+      "learning_rate": 0.000246707168037034,
+      "loss": 0.7213,
+      "step": 5652
+    },
+    {
+      "epoch": 1.0083846222460084,
+      "grad_norm": 0.5297546982765198,
+      "learning_rate": 0.00024663711196445135,
+      "loss": 0.7577,
+      "step": 5653
+    },
+    {
+      "epoch": 1.0085630184640086,
+      "grad_norm": 0.5135631561279297,
+      "learning_rate": 0.00024656705615598844,
+      "loss": 0.8142,
+      "step": 5654
+    },
+    {
+      "epoch": 1.0087414146820088,
+      "grad_norm": 0.4576781094074249,
+      "learning_rate": 0.0002464970006171475,
+      "loss": 0.6083,
+      "step": 5655
+    },
+    {
+      "epoch": 1.0089198109000088,
+      "grad_norm": 0.5894820690155029,
+      "learning_rate": 0.0002464269453534307,
+      "loss": 0.8443,
+      "step": 5656
+    },
+    {
+      "epoch": 1.009098207118009,
+      "grad_norm": 0.5274859070777893,
+      "learning_rate": 0.00024635689037034,
+      "loss": 0.9412,
+      "step": 5657
+    },
+    {
+      "epoch": 1.0092766033360092,
+      "grad_norm": 0.46422895789146423,
+      "learning_rate": 0.0002462868356733775,
+      "loss": 0.7357,
+      "step": 5658
+    },
+    {
+      "epoch": 1.0094549995540094,
+      "grad_norm": 0.49951404333114624,
+      "learning_rate": 0.0002462167812680453,
+      "loss": 0.8181,
+      "step": 5659
+    },
+    {
+      "epoch": 1.0096333957720096,
+      "grad_norm": 0.48407530784606934,
+      "learning_rate": 0.00024614672715984556,
+      "loss": 0.6384,
+      "step": 5660
+    },
+    {
+      "epoch": 1.0098117919900098,
+      "grad_norm": 1.810511827468872,
+      "learning_rate": 0.0002460766733542803,
+      "loss": 0.7134,
+      "step": 5661
+    },
+    {
+      "epoch": 1.00999018820801,
+      "grad_norm": 0.4726197421550751,
+      "learning_rate": 0.0002460066198568513,
+      "loss": 0.6611,
+      "step": 5662
+    },
+    {
+      "epoch": 1.0101685844260102,
+      "grad_norm": 0.512795090675354,
+      "learning_rate": 0.0002459365666730607,
+      "loss": 0.7829,
+      "step": 5663
+    },
+    {
+      "epoch": 1.0103469806440104,
+      "grad_norm": 0.6051098108291626,
+      "learning_rate": 0.0002458665138084104,
+      "loss": 0.7989,
+      "step": 5664
+    },
+    {
+      "epoch": 1.0105253768620106,
+      "grad_norm": 0.5258930921554565,
+      "learning_rate": 0.00024579646126840233,
+      "loss": 0.5785,
+      "step": 5665
+    },
+    {
+      "epoch": 1.0107037730800108,
+      "grad_norm": 0.41359806060791016,
+      "learning_rate": 0.0002457264090585384,
+      "loss": 0.5786,
+      "step": 5666
+    },
+    {
+      "epoch": 1.0108821692980108,
+      "grad_norm": 0.4931972622871399,
+      "learning_rate": 0.0002456563571843204,
+      "loss": 0.8471,
+      "step": 5667
+    },
+    {
+      "epoch": 1.011060565516011,
+      "grad_norm": 0.49193498492240906,
+      "learning_rate": 0.0002455863056512503,
+      "loss": 0.8302,
+      "step": 5668
+    },
+    {
+      "epoch": 1.0112389617340112,
+      "grad_norm": 0.5347012877464294,
+      "learning_rate": 0.0002455162544648299,
+      "loss": 1.1447,
+      "step": 5669
+    },
+    {
+      "epoch": 1.0114173579520114,
+      "grad_norm": 0.46643638610839844,
+      "learning_rate": 0.00024544620363056084,
+      "loss": 0.5934,
+      "step": 5670
+    },
+    {
+      "epoch": 1.0115957541700116,
+      "grad_norm": 0.6032198071479797,
+      "learning_rate": 0.00024537615315394504,
+      "loss": 0.5384,
+      "step": 5671
+    },
+    {
+      "epoch": 1.0117741503880118,
+      "grad_norm": 0.5089538097381592,
+      "learning_rate": 0.00024530610304048417,
+      "loss": 0.8041,
+      "step": 5672
+    },
+    {
+      "epoch": 1.011952546606012,
+      "grad_norm": 0.5048344731330872,
+      "learning_rate": 0.00024523605329567996,
+      "loss": 0.7037,
+      "step": 5673
+    },
+    {
+      "epoch": 1.0121309428240122,
+      "grad_norm": 0.40389716625213623,
+      "learning_rate": 0.00024516600392503397,
+      "loss": 0.5336,
+      "step": 5674
+    },
+    {
+      "epoch": 1.0123093390420124,
+      "grad_norm": 0.4871548116207123,
+      "learning_rate": 0.0002450959549340481,
+      "loss": 0.6761,
+      "step": 5675
+    },
+    {
+      "epoch": 1.0124877352600126,
+      "grad_norm": 0.49612149596214294,
+      "learning_rate": 0.00024502590632822374,
+      "loss": 0.7691,
+      "step": 5676
+    },
+    {
+      "epoch": 1.0126661314780128,
+      "grad_norm": 0.4323468506336212,
+      "learning_rate": 0.0002449558581130626,
+      "loss": 0.5754,
+      "step": 5677
+    },
+    {
+      "epoch": 1.0128445276960127,
+      "grad_norm": 0.4288893938064575,
+      "learning_rate": 0.00024488581029406614,
+      "loss": 0.6697,
+      "step": 5678
+    },
+    {
+      "epoch": 1.013022923914013,
+      "grad_norm": 0.4768436551094055,
+      "learning_rate": 0.00024481576287673596,
+      "loss": 0.657,
+      "step": 5679
+    },
+    {
+      "epoch": 1.0132013201320131,
+      "grad_norm": 0.4346489906311035,
+      "learning_rate": 0.00024474571586657353,
+      "loss": 0.663,
+      "step": 5680
+    },
+    {
+      "epoch": 1.0133797163500133,
+      "grad_norm": 0.5226327180862427,
+      "learning_rate": 0.0002446756692690804,
+      "loss": 0.9833,
+      "step": 5681
+    },
+    {
+      "epoch": 1.0135581125680135,
+      "grad_norm": 0.5677915811538696,
+      "learning_rate": 0.0002446056230897578,
+      "loss": 0.9305,
+      "step": 5682
+    },
+    {
+      "epoch": 1.0137365087860137,
+      "grad_norm": 0.5274161100387573,
+      "learning_rate": 0.0002445355773341073,
+      "loss": 0.7953,
+      "step": 5683
+    },
+    {
+      "epoch": 1.013914905004014,
+      "grad_norm": 0.557561457157135,
+      "learning_rate": 0.0002444655320076302,
+      "loss": 0.9001,
+      "step": 5684
+    },
+    {
+      "epoch": 1.0140933012220141,
+      "grad_norm": 0.3999897241592407,
+      "learning_rate": 0.0002443954871158278,
+      "loss": 0.6062,
+      "step": 5685
+    },
+    {
+      "epoch": 1.0142716974400143,
+      "grad_norm": 0.6423475742340088,
+      "learning_rate": 0.00024432544266420145,
+      "loss": 0.8672,
+      "step": 5686
+    },
+    {
+      "epoch": 1.0144500936580145,
+      "grad_norm": 0.5340455770492554,
+      "learning_rate": 0.0002442553986582524,
+      "loss": 0.7465,
+      "step": 5687
+    },
+    {
+      "epoch": 1.0146284898760147,
+      "grad_norm": 0.4833824634552002,
+      "learning_rate": 0.00024418535510348184,
+      "loss": 0.7415,
+      "step": 5688
+    },
+    {
+      "epoch": 1.0148068860940147,
+      "grad_norm": 0.452972948551178,
+      "learning_rate": 0.00024411531200539102,
+      "loss": 0.6568,
+      "step": 5689
+    },
+    {
+      "epoch": 1.014985282312015,
+      "grad_norm": 0.45635709166526794,
+      "learning_rate": 0.00024404526936948098,
+      "loss": 0.5435,
+      "step": 5690
+    },
+    {
+      "epoch": 1.015163678530015,
+      "grad_norm": 0.5237349271774292,
+      "learning_rate": 0.00024397522720125302,
+      "loss": 0.7549,
+      "step": 5691
+    },
+    {
+      "epoch": 1.0153420747480153,
+      "grad_norm": 0.627680778503418,
+      "learning_rate": 0.00024390518550620807,
+      "loss": 0.6618,
+      "step": 5692
+    },
+    {
+      "epoch": 1.0155204709660155,
+      "grad_norm": 0.5116132497787476,
+      "learning_rate": 0.00024383514428984727,
+      "loss": 0.7378,
+      "step": 5693
+    },
+    {
+      "epoch": 1.0156988671840157,
+      "grad_norm": 1.0455267429351807,
+      "learning_rate": 0.00024376510355767161,
+      "loss": 0.8657,
+      "step": 5694
+    },
+    {
+      "epoch": 1.015877263402016,
+      "grad_norm": 0.5515667200088501,
+      "learning_rate": 0.000243695063315182,
+      "loss": 0.7739,
+      "step": 5695
+    },
+    {
+      "epoch": 1.016055659620016,
+      "grad_norm": 0.5279601812362671,
+      "learning_rate": 0.0002436250235678794,
+      "loss": 0.7465,
+      "step": 5696
+    },
+    {
+      "epoch": 1.0162340558380163,
+      "grad_norm": 0.4827010929584503,
+      "learning_rate": 0.00024355498432126468,
+      "loss": 0.7674,
+      "step": 5697
+    },
+    {
+      "epoch": 1.0164124520560165,
+      "grad_norm": 0.5047492384910583,
+      "learning_rate": 0.00024348494558083873,
+      "loss": 0.7607,
+      "step": 5698
+    },
+    {
+      "epoch": 1.0165908482740167,
+      "grad_norm": 0.4985560476779938,
+      "learning_rate": 0.00024341490735210237,
+      "loss": 0.8236,
+      "step": 5699
+    },
+    {
+      "epoch": 1.0167692444920167,
+      "grad_norm": 0.48609042167663574,
+      "learning_rate": 0.00024334486964055634,
+      "loss": 0.7264,
+      "step": 5700
+    },
+    {
+      "epoch": 1.0169476407100169,
+      "grad_norm": 0.4970250129699707,
+      "learning_rate": 0.00024327483245170138,
+      "loss": 0.6943,
+      "step": 5701
+    },
+    {
+      "epoch": 1.017126036928017,
+      "grad_norm": 0.5171718001365662,
+      "learning_rate": 0.00024320479579103825,
+      "loss": 0.6666,
+      "step": 5702
+    },
+    {
+      "epoch": 1.0173044331460173,
+      "grad_norm": 0.48517653346061707,
+      "learning_rate": 0.00024313475966406746,
+      "loss": 0.7518,
+      "step": 5703
+    },
+    {
+      "epoch": 1.0174828293640175,
+      "grad_norm": 0.5177552700042725,
+      "learning_rate": 0.0002430647240762897,
+      "loss": 0.6988,
+      "step": 5704
+    },
+    {
+      "epoch": 1.0176612255820177,
+      "grad_norm": 0.4849720597267151,
+      "learning_rate": 0.00024299468903320542,
+      "loss": 0.6417,
+      "step": 5705
+    },
+    {
+      "epoch": 1.0178396218000179,
+      "grad_norm": 0.5243503451347351,
+      "learning_rate": 0.00024292465454031536,
+      "loss": 0.7603,
+      "step": 5706
+    },
+    {
+      "epoch": 1.018018018018018,
+      "grad_norm": 0.48335912823677063,
+      "learning_rate": 0.00024285462060311995,
+      "loss": 0.8571,
+      "step": 5707
+    },
+    {
+      "epoch": 1.0181964142360183,
+      "grad_norm": 0.48116248846054077,
+      "learning_rate": 0.00024278458722711948,
+      "loss": 0.717,
+      "step": 5708
+    },
+    {
+      "epoch": 1.0183748104540185,
+      "grad_norm": 0.5435453653335571,
+      "learning_rate": 0.0002427145544178145,
+      "loss": 0.7622,
+      "step": 5709
+    },
+    {
+      "epoch": 1.0185532066720187,
+      "grad_norm": 0.5478907227516174,
+      "learning_rate": 0.0002426445221807052,
+      "loss": 0.6225,
+      "step": 5710
+    },
+    {
+      "epoch": 1.0187316028900186,
+      "grad_norm": 0.44758474826812744,
+      "learning_rate": 0.000242574490521292,
+      "loss": 0.7621,
+      "step": 5711
+    },
+    {
+      "epoch": 1.0189099991080188,
+      "grad_norm": 0.4896325170993805,
+      "learning_rate": 0.0002425044594450751,
+      "loss": 0.803,
+      "step": 5712
+    },
+    {
+      "epoch": 1.019088395326019,
+      "grad_norm": 0.6267058253288269,
+      "learning_rate": 0.00024243442895755476,
+      "loss": 0.782,
+      "step": 5713
+    },
+    {
+      "epoch": 1.0192667915440192,
+      "grad_norm": 0.49481847882270813,
+      "learning_rate": 0.00024236439906423105,
+      "loss": 0.7271,
+      "step": 5714
+    },
+    {
+      "epoch": 1.0194451877620194,
+      "grad_norm": 0.42377015948295593,
+      "learning_rate": 0.00024229436977060427,
+      "loss": 0.529,
+      "step": 5715
+    },
+    {
+      "epoch": 1.0196235839800196,
+      "grad_norm": 0.4759003520011902,
+      "learning_rate": 0.00024222434108217428,
+      "loss": 0.7413,
+      "step": 5716
+    },
+    {
+      "epoch": 1.0198019801980198,
+      "grad_norm": 0.41128072142601013,
+      "learning_rate": 0.0002421543130044412,
+      "loss": 0.5624,
+      "step": 5717
+    },
+    {
+      "epoch": 1.01998037641602,
+      "grad_norm": 0.5285554528236389,
+      "learning_rate": 0.00024208428554290502,
+      "loss": 0.71,
+      "step": 5718
+    },
+    {
+      "epoch": 1.0201587726340202,
+      "grad_norm": 0.4953111410140991,
+      "learning_rate": 0.00024201425870306565,
+      "loss": 0.6101,
+      "step": 5719
+    },
+    {
+      "epoch": 1.0203371688520204,
+      "grad_norm": 0.5717734098434448,
+      "learning_rate": 0.0002419442324904229,
+      "loss": 0.8363,
+      "step": 5720
+    },
+    {
+      "epoch": 1.0205155650700206,
+      "grad_norm": 0.49335336685180664,
+      "learning_rate": 0.0002418742069104767,
+      "loss": 0.5884,
+      "step": 5721
+    },
+    {
+      "epoch": 1.0206939612880206,
+      "grad_norm": 0.7236892580986023,
+      "learning_rate": 0.0002418041819687268,
+      "loss": 0.7232,
+      "step": 5722
+    },
+    {
+      "epoch": 1.0208723575060208,
+      "grad_norm": 0.5189496278762817,
+      "learning_rate": 0.00024173415767067295,
+      "loss": 0.7079,
+      "step": 5723
+    },
+    {
+      "epoch": 1.021050753724021,
+      "grad_norm": 0.5228090882301331,
+      "learning_rate": 0.00024166413402181477,
+      "loss": 0.7657,
+      "step": 5724
+    },
+    {
+      "epoch": 1.0212291499420212,
+      "grad_norm": 0.4883708953857422,
+      "learning_rate": 0.00024159411102765195,
+      "loss": 0.6471,
+      "step": 5725
+    },
+    {
+      "epoch": 1.0214075461600214,
+      "grad_norm": 0.5337876677513123,
+      "learning_rate": 0.00024152408869368398,
+      "loss": 0.8151,
+      "step": 5726
+    },
+    {
+      "epoch": 1.0215859423780216,
+      "grad_norm": 0.6350103616714478,
+      "learning_rate": 0.00024145406702541047,
+      "loss": 0.694,
+      "step": 5727
+    },
+    {
+      "epoch": 1.0217643385960218,
+      "grad_norm": 0.49717575311660767,
+      "learning_rate": 0.00024138404602833092,
+      "loss": 0.6491,
+      "step": 5728
+    },
+    {
+      "epoch": 1.021942734814022,
+      "grad_norm": 0.4632187783718109,
+      "learning_rate": 0.0002413140257079446,
+      "loss": 0.6376,
+      "step": 5729
+    },
+    {
+      "epoch": 1.0221211310320222,
+      "grad_norm": 0.7505040764808655,
+      "learning_rate": 0.000241244006069751,
+      "loss": 0.7011,
+      "step": 5730
+    },
+    {
+      "epoch": 1.0222995272500224,
+      "grad_norm": 0.4758153557777405,
+      "learning_rate": 0.00024117398711924937,
+      "loss": 0.6989,
+      "step": 5731
+    },
+    {
+      "epoch": 1.0224779234680226,
+      "grad_norm": 0.43656665086746216,
+      "learning_rate": 0.000241103968861939,
+      "loss": 0.5964,
+      "step": 5732
+    },
+    {
+      "epoch": 1.0226563196860226,
+      "grad_norm": 0.5343319177627563,
+      "learning_rate": 0.00024103395130331909,
+      "loss": 0.855,
+      "step": 5733
+    },
+    {
+      "epoch": 1.0228347159040228,
+      "grad_norm": 0.6632677912712097,
+      "learning_rate": 0.00024096393444888878,
+      "loss": 0.5942,
+      "step": 5734
+    },
+    {
+      "epoch": 1.023013112122023,
+      "grad_norm": 0.40866273641586304,
+      "learning_rate": 0.0002408939183041472,
+      "loss": 0.5326,
+      "step": 5735
+    },
+    {
+      "epoch": 1.0231915083400231,
+      "grad_norm": 0.43526768684387207,
+      "learning_rate": 0.00024082390287459327,
+      "loss": 0.6601,
+      "step": 5736
+    },
+    {
+      "epoch": 1.0233699045580233,
+      "grad_norm": 0.5318346619606018,
+      "learning_rate": 0.00024075388816572612,
+      "loss": 0.709,
+      "step": 5737
+    },
+    {
+      "epoch": 1.0235483007760235,
+      "grad_norm": 0.5785126090049744,
+      "learning_rate": 0.0002406838741830446,
+      "loss": 0.6006,
+      "step": 5738
+    },
+    {
+      "epoch": 1.0237266969940237,
+      "grad_norm": 0.4882996380329132,
+      "learning_rate": 0.0002406138609320476,
+      "loss": 0.7155,
+      "step": 5739
+    },
+    {
+      "epoch": 1.023905093212024,
+      "grad_norm": 0.5886842012405396,
+      "learning_rate": 0.00024054384841823396,
+      "loss": 0.7278,
+      "step": 5740
+    },
+    {
+      "epoch": 1.0240834894300241,
+      "grad_norm": 0.5136778354644775,
+      "learning_rate": 0.00024047383664710243,
+      "loss": 0.7661,
+      "step": 5741
+    },
+    {
+      "epoch": 1.0242618856480243,
+      "grad_norm": 0.563352644443512,
+      "learning_rate": 0.0002404038256241516,
+      "loss": 0.9258,
+      "step": 5742
+    },
+    {
+      "epoch": 1.0244402818660245,
+      "grad_norm": 0.4579552710056305,
+      "learning_rate": 0.0002403338153548802,
+      "loss": 0.5869,
+      "step": 5743
+    },
+    {
+      "epoch": 1.0246186780840245,
+      "grad_norm": 0.5219502449035645,
+      "learning_rate": 0.00024026380584478676,
+      "loss": 0.7318,
+      "step": 5744
+    },
+    {
+      "epoch": 1.0247970743020247,
+      "grad_norm": 0.4938696622848511,
+      "learning_rate": 0.00024019379709936984,
+      "loss": 0.8765,
+      "step": 5745
+    },
+    {
+      "epoch": 1.024975470520025,
+      "grad_norm": 0.6994608640670776,
+      "learning_rate": 0.00024012378912412785,
+      "loss": 0.7989,
+      "step": 5746
+    },
+    {
+      "epoch": 1.025153866738025,
+      "grad_norm": 0.4768621027469635,
+      "learning_rate": 0.00024005378192455923,
+      "loss": 0.6248,
+      "step": 5747
+    },
+    {
+      "epoch": 1.0253322629560253,
+      "grad_norm": 1.2631269693374634,
+      "learning_rate": 0.00023998377550616228,
+      "loss": 0.6479,
+      "step": 5748
+    },
+    {
+      "epoch": 1.0255106591740255,
+      "grad_norm": 0.48576176166534424,
+      "learning_rate": 0.00023991376987443535,
+      "loss": 0.7885,
+      "step": 5749
+    },
+    {
+      "epoch": 1.0256890553920257,
+      "grad_norm": 0.5313836336135864,
+      "learning_rate": 0.00023984376503487657,
+      "loss": 0.8569,
+      "step": 5750
+    },
+    {
+      "epoch": 1.025867451610026,
+      "grad_norm": 1.306301236152649,
+      "learning_rate": 0.00023977376099298397,
+      "loss": 0.7603,
+      "step": 5751
+    },
+    {
+      "epoch": 1.026045847828026,
+      "grad_norm": 0.6186336278915405,
+      "learning_rate": 0.0002397037577542559,
+      "loss": 0.8546,
+      "step": 5752
+    },
+    {
+      "epoch": 1.0262242440460263,
+      "grad_norm": 0.4687788784503937,
+      "learning_rate": 0.00023963375532419032,
+      "loss": 0.7342,
+      "step": 5753
+    },
+    {
+      "epoch": 1.0264026402640265,
+      "grad_norm": 0.5840802192687988,
+      "learning_rate": 0.00023956375370828508,
+      "loss": 0.6639,
+      "step": 5754
+    },
+    {
+      "epoch": 1.0265810364820265,
+      "grad_norm": 0.5559926629066467,
+      "learning_rate": 0.00023949375291203815,
+      "loss": 0.8283,
+      "step": 5755
+    },
+    {
+      "epoch": 1.0267594327000267,
+      "grad_norm": 0.47672855854034424,
+      "learning_rate": 0.0002394237529409473,
+      "loss": 0.7104,
+      "step": 5756
+    },
+    {
+      "epoch": 1.0269378289180269,
+      "grad_norm": 0.49324122071266174,
+      "learning_rate": 0.00023935375380051038,
+      "loss": 0.7398,
+      "step": 5757
+    },
+    {
+      "epoch": 1.027116225136027,
+      "grad_norm": 0.547975480556488,
+      "learning_rate": 0.0002392837554962251,
+      "loss": 0.8606,
+      "step": 5758
+    },
+    {
+      "epoch": 1.0272946213540273,
+      "grad_norm": 0.48279041051864624,
+      "learning_rate": 0.00023921375803358897,
+      "loss": 0.7028,
+      "step": 5759
+    },
+    {
+      "epoch": 1.0274730175720275,
+      "grad_norm": 0.5213961601257324,
+      "learning_rate": 0.00023914376141809967,
+      "loss": 0.6459,
+      "step": 5760
+    },
+    {
+      "epoch": 1.0276514137900277,
+      "grad_norm": 0.46963804960250854,
+      "learning_rate": 0.00023907376565525469,
+      "loss": 0.6467,
+      "step": 5761
+    },
+    {
+      "epoch": 1.0278298100080279,
+      "grad_norm": 0.4742526113986969,
+      "learning_rate": 0.0002390037707505515,
+      "loss": 0.6972,
+      "step": 5762
+    },
+    {
+      "epoch": 1.028008206226028,
+      "grad_norm": 0.6087868213653564,
+      "learning_rate": 0.00023893377670948735,
+      "loss": 0.6751,
+      "step": 5763
+    },
+    {
+      "epoch": 1.0281866024440283,
+      "grad_norm": 0.5412265062332153,
+      "learning_rate": 0.00023886378353755964,
+      "loss": 0.7496,
+      "step": 5764
+    },
+    {
+      "epoch": 1.0283649986620285,
+      "grad_norm": 0.6296983957290649,
+      "learning_rate": 0.00023879379124026556,
+      "loss": 0.7903,
+      "step": 5765
+    },
+    {
+      "epoch": 1.0285433948800284,
+      "grad_norm": 0.5297571420669556,
+      "learning_rate": 0.00023872379982310224,
+      "loss": 0.7143,
+      "step": 5766
+    },
+    {
+      "epoch": 1.0287217910980286,
+      "grad_norm": 0.4944717586040497,
+      "learning_rate": 0.00023865380929156691,
+      "loss": 0.7,
+      "step": 5767
+    },
+    {
+      "epoch": 1.0289001873160288,
+      "grad_norm": 0.44193384051322937,
+      "learning_rate": 0.0002385838196511565,
+      "loss": 0.6133,
+      "step": 5768
+    },
+    {
+      "epoch": 1.029078583534029,
+      "grad_norm": 2.3271965980529785,
+      "learning_rate": 0.000238513830907368,
+      "loss": 0.712,
+      "step": 5769
+    },
+    {
+      "epoch": 1.0292569797520292,
+      "grad_norm": 0.45124170184135437,
+      "learning_rate": 0.00023844384306569825,
+      "loss": 0.6838,
+      "step": 5770
+    },
+    {
+      "epoch": 1.0294353759700294,
+      "grad_norm": 0.488295316696167,
+      "learning_rate": 0.0002383738561316441,
+      "loss": 0.6474,
+      "step": 5771
+    },
+    {
+      "epoch": 1.0296137721880296,
+      "grad_norm": 0.4569397270679474,
+      "learning_rate": 0.00023830387011070225,
+      "loss": 0.7513,
+      "step": 5772
+    },
+    {
+      "epoch": 1.0297921684060298,
+      "grad_norm": 0.5301244258880615,
+      "learning_rate": 0.00023823388500836945,
+      "loss": 0.7864,
+      "step": 5773
+    },
+    {
+      "epoch": 1.02997056462403,
+      "grad_norm": 0.4296034872531891,
+      "learning_rate": 0.00023816390083014234,
+      "loss": 0.5479,
+      "step": 5774
+    },
+    {
+      "epoch": 1.0301489608420302,
+      "grad_norm": 0.49546340107917786,
+      "learning_rate": 0.00023809391758151726,
+      "loss": 0.6903,
+      "step": 5775
+    },
+    {
+      "epoch": 1.0303273570600304,
+      "grad_norm": 0.47323548793792725,
+      "learning_rate": 0.0002380239352679908,
+      "loss": 0.6874,
+      "step": 5776
+    },
+    {
+      "epoch": 1.0305057532780304,
+      "grad_norm": 0.45178884267807007,
+      "learning_rate": 0.00023795395389505927,
+      "loss": 0.8751,
+      "step": 5777
+    },
+    {
+      "epoch": 1.0306841494960306,
+      "grad_norm": 0.5081565380096436,
+      "learning_rate": 0.00023788397346821905,
+      "loss": 0.7723,
+      "step": 5778
+    },
+    {
+      "epoch": 1.0308625457140308,
+      "grad_norm": 0.4901391565799713,
+      "learning_rate": 0.00023781399399296635,
+      "loss": 0.628,
+      "step": 5779
+    },
+    {
+      "epoch": 1.031040941932031,
+      "grad_norm": 0.4235868752002716,
+      "learning_rate": 0.0002377440154747973,
+      "loss": 0.5508,
+      "step": 5780
+    },
+    {
+      "epoch": 1.0312193381500312,
+      "grad_norm": 0.45451483130455017,
+      "learning_rate": 0.000237674037919208,
+      "loss": 0.6556,
+      "step": 5781
+    },
+    {
+      "epoch": 1.0313977343680314,
+      "grad_norm": 0.5195580720901489,
+      "learning_rate": 0.0002376040613316944,
+      "loss": 0.9728,
+      "step": 5782
+    },
+    {
+      "epoch": 1.0315761305860316,
+      "grad_norm": 0.49326059222221375,
+      "learning_rate": 0.00023753408571775255,
+      "loss": 0.6793,
+      "step": 5783
+    },
+    {
+      "epoch": 1.0317545268040318,
+      "grad_norm": 0.4820156395435333,
+      "learning_rate": 0.00023746411108287825,
+      "loss": 0.562,
+      "step": 5784
+    },
+    {
+      "epoch": 1.031932923022032,
+      "grad_norm": 0.48524320125579834,
+      "learning_rate": 0.00023739413743256726,
+      "loss": 0.6854,
+      "step": 5785
+    },
+    {
+      "epoch": 1.0321113192400322,
+      "grad_norm": 0.49960482120513916,
+      "learning_rate": 0.0002373241647723153,
+      "loss": 0.8011,
+      "step": 5786
+    },
+    {
+      "epoch": 1.0322897154580324,
+      "grad_norm": 0.5163941979408264,
+      "learning_rate": 0.00023725419310761803,
+      "loss": 0.7166,
+      "step": 5787
+    },
+    {
+      "epoch": 1.0324681116760324,
+      "grad_norm": 0.507465660572052,
+      "learning_rate": 0.0002371842224439709,
+      "loss": 0.8175,
+      "step": 5788
+    },
+    {
+      "epoch": 1.0326465078940326,
+      "grad_norm": 0.47466954588890076,
+      "learning_rate": 0.00023711425278686945,
+      "loss": 0.6952,
+      "step": 5789
+    },
+    {
+      "epoch": 1.0328249041120328,
+      "grad_norm": 0.4188692569732666,
+      "learning_rate": 0.000237044284141809,
+      "loss": 0.5577,
+      "step": 5790
+    },
+    {
+      "epoch": 1.033003300330033,
+      "grad_norm": 0.5226710438728333,
+      "learning_rate": 0.0002369743165142849,
+      "loss": 0.8422,
+      "step": 5791
+    },
+    {
+      "epoch": 1.0331816965480332,
+      "grad_norm": 0.49022915959358215,
+      "learning_rate": 0.00023690434990979238,
+      "loss": 0.8066,
+      "step": 5792
+    },
+    {
+      "epoch": 1.0333600927660334,
+      "grad_norm": 0.49656999111175537,
+      "learning_rate": 0.0002368343843338266,
+      "loss": 0.7251,
+      "step": 5793
+    },
+    {
+      "epoch": 1.0335384889840336,
+      "grad_norm": 0.40922585129737854,
+      "learning_rate": 0.00023676441979188258,
+      "loss": 0.5209,
+      "step": 5794
+    },
+    {
+      "epoch": 1.0337168852020338,
+      "grad_norm": 0.46607285737991333,
+      "learning_rate": 0.0002366944562894554,
+      "loss": 0.6608,
+      "step": 5795
+    },
+    {
+      "epoch": 1.033895281420034,
+      "grad_norm": 0.43443769216537476,
+      "learning_rate": 0.00023662449383203988,
+      "loss": 0.6759,
+      "step": 5796
+    },
+    {
+      "epoch": 1.0340736776380341,
+      "grad_norm": 0.5248075127601624,
+      "learning_rate": 0.0002365545324251307,
+      "loss": 0.8347,
+      "step": 5797
+    },
+    {
+      "epoch": 1.0342520738560343,
+      "grad_norm": 0.5106051564216614,
+      "learning_rate": 0.0002364845720742229,
+      "loss": 0.8222,
+      "step": 5798
+    },
+    {
+      "epoch": 1.0344304700740343,
+      "grad_norm": 0.4647291600704193,
+      "learning_rate": 0.00023641461278481096,
+      "loss": 0.5955,
+      "step": 5799
+    },
+    {
+      "epoch": 1.0346088662920345,
+      "grad_norm": 0.3780640661716461,
+      "learning_rate": 0.00023634465456238957,
+      "loss": 0.4401,
+      "step": 5800
+    },
+    {
+      "epoch": 1.0347872625100347,
+      "grad_norm": 0.4369570016860962,
+      "learning_rate": 0.00023627469741245306,
+      "loss": 0.5725,
+      "step": 5801
+    },
+    {
+      "epoch": 1.034965658728035,
+      "grad_norm": 0.4812033772468567,
+      "learning_rate": 0.0002362047413404959,
+      "loss": 0.7532,
+      "step": 5802
+    },
+    {
+      "epoch": 1.0351440549460351,
+      "grad_norm": 0.49813112616539,
+      "learning_rate": 0.00023613478635201246,
+      "loss": 0.7331,
+      "step": 5803
+    },
+    {
+      "epoch": 1.0353224511640353,
+      "grad_norm": 0.45200127363204956,
+      "learning_rate": 0.0002360648324524969,
+      "loss": 0.6147,
+      "step": 5804
+    },
+    {
+      "epoch": 1.0355008473820355,
+      "grad_norm": 0.5208128690719604,
+      "learning_rate": 0.0002359948796474434,
+      "loss": 0.6966,
+      "step": 5805
+    },
+    {
+      "epoch": 1.0356792436000357,
+      "grad_norm": 0.49845966696739197,
+      "learning_rate": 0.00023592492794234605,
+      "loss": 0.7405,
+      "step": 5806
+    },
+    {
+      "epoch": 1.035857639818036,
+      "grad_norm": 0.4411349594593048,
+      "learning_rate": 0.0002358549773426988,
+      "loss": 0.6018,
+      "step": 5807
+    },
+    {
+      "epoch": 1.0360360360360361,
+      "grad_norm": 0.5190544128417969,
+      "learning_rate": 0.00023578502785399558,
+      "loss": 0.9016,
+      "step": 5808
+    },
+    {
+      "epoch": 1.0362144322540363,
+      "grad_norm": 0.45174866914749146,
+      "learning_rate": 0.0002357150794817301,
+      "loss": 0.5387,
+      "step": 5809
+    },
+    {
+      "epoch": 1.0363928284720363,
+      "grad_norm": 0.56296706199646,
+      "learning_rate": 0.00023564513223139615,
+      "loss": 0.6236,
+      "step": 5810
+    },
+    {
+      "epoch": 1.0365712246900365,
+      "grad_norm": 0.4660092294216156,
+      "learning_rate": 0.0002355751861084873,
+      "loss": 0.6932,
+      "step": 5811
+    },
+    {
+      "epoch": 1.0367496209080367,
+      "grad_norm": 0.4947931170463562,
+      "learning_rate": 0.00023550524111849705,
+      "loss": 0.6975,
+      "step": 5812
+    },
+    {
+      "epoch": 1.0369280171260369,
+      "grad_norm": 0.49009427428245544,
+      "learning_rate": 0.0002354352972669191,
+      "loss": 0.6677,
+      "step": 5813
+    },
+    {
+      "epoch": 1.037106413344037,
+      "grad_norm": 0.5009847283363342,
+      "learning_rate": 0.00023536535455924654,
+      "loss": 0.7397,
+      "step": 5814
+    },
+    {
+      "epoch": 1.0372848095620373,
+      "grad_norm": 0.5054614543914795,
+      "learning_rate": 0.00023529541300097275,
+      "loss": 0.7261,
+      "step": 5815
+    },
+    {
+      "epoch": 1.0374632057800375,
+      "grad_norm": 0.46940669417381287,
+      "learning_rate": 0.0002352254725975909,
+      "loss": 0.7628,
+      "step": 5816
+    },
+    {
+      "epoch": 1.0376416019980377,
+      "grad_norm": 0.4963463842868805,
+      "learning_rate": 0.00023515553335459407,
+      "loss": 0.9402,
+      "step": 5817
+    },
+    {
+      "epoch": 1.0378199982160379,
+      "grad_norm": 0.46906179189682007,
+      "learning_rate": 0.00023508559527747527,
+      "loss": 0.6101,
+      "step": 5818
+    },
+    {
+      "epoch": 1.037998394434038,
+      "grad_norm": 0.5523307919502258,
+      "learning_rate": 0.0002350156583717274,
+      "loss": 0.6492,
+      "step": 5819
+    },
+    {
+      "epoch": 1.0381767906520383,
+      "grad_norm": 0.5532516837120056,
+      "learning_rate": 0.00023494572264284326,
+      "loss": 0.82,
+      "step": 5820
+    },
+    {
+      "epoch": 1.0383551868700382,
+      "grad_norm": 0.5024081468582153,
+      "learning_rate": 0.00023487578809631567,
+      "loss": 0.725,
+      "step": 5821
+    },
+    {
+      "epoch": 1.0385335830880384,
+      "grad_norm": 0.4690212905406952,
+      "learning_rate": 0.00023480585473763707,
+      "loss": 0.7607,
+      "step": 5822
+    },
+    {
+      "epoch": 1.0387119793060386,
+      "grad_norm": 0.39463046193122864,
+      "learning_rate": 0.00023473592257230015,
+      "loss": 0.5608,
+      "step": 5823
+    },
+    {
+      "epoch": 1.0388903755240388,
+      "grad_norm": 0.44406580924987793,
+      "learning_rate": 0.00023466599160579726,
+      "loss": 0.6521,
+      "step": 5824
+    },
+    {
+      "epoch": 1.039068771742039,
+      "grad_norm": 0.46876436471939087,
+      "learning_rate": 0.0002345960618436208,
+      "loss": 0.643,
+      "step": 5825
+    },
+    {
+      "epoch": 1.0392471679600392,
+      "grad_norm": 0.5072786211967468,
+      "learning_rate": 0.00023452613329126305,
+      "loss": 0.91,
+      "step": 5826
+    },
+    {
+      "epoch": 1.0394255641780394,
+      "grad_norm": 0.45098283886909485,
+      "learning_rate": 0.0002344562059542161,
+      "loss": 0.6326,
+      "step": 5827
+    },
+    {
+      "epoch": 1.0396039603960396,
+      "grad_norm": 0.507174551486969,
+      "learning_rate": 0.000234386279837972,
+      "loss": 0.8313,
+      "step": 5828
+    },
+    {
+      "epoch": 1.0397823566140398,
+      "grad_norm": 0.44841310381889343,
+      "learning_rate": 0.0002343163549480228,
+      "loss": 0.6486,
+      "step": 5829
+    },
+    {
+      "epoch": 1.03996075283204,
+      "grad_norm": 0.45851168036460876,
+      "learning_rate": 0.00023424643128986037,
+      "loss": 0.6491,
+      "step": 5830
+    },
+    {
+      "epoch": 1.0401391490500402,
+      "grad_norm": 0.4500889778137207,
+      "learning_rate": 0.0002341765088689764,
+      "loss": 0.6452,
+      "step": 5831
+    },
+    {
+      "epoch": 1.0403175452680402,
+      "grad_norm": 0.4633578658103943,
+      "learning_rate": 0.00023410658769086265,
+      "loss": 0.6371,
+      "step": 5832
+    },
+    {
+      "epoch": 1.0404959414860404,
+      "grad_norm": 0.46451932191848755,
+      "learning_rate": 0.0002340366677610107,
+      "loss": 0.6226,
+      "step": 5833
+    },
+    {
+      "epoch": 1.0406743377040406,
+      "grad_norm": 0.5377410054206848,
+      "learning_rate": 0.00023396674908491194,
+      "loss": 0.8014,
+      "step": 5834
+    },
+    {
+      "epoch": 1.0408527339220408,
+      "grad_norm": 0.46611928939819336,
+      "learning_rate": 0.00023389683166805784,
+      "loss": 0.7689,
+      "step": 5835
+    },
+    {
+      "epoch": 1.041031130140041,
+      "grad_norm": 0.43398740887641907,
+      "learning_rate": 0.00023382691551593964,
+      "loss": 0.5971,
+      "step": 5836
+    },
+    {
+      "epoch": 1.0412095263580412,
+      "grad_norm": 0.46025121212005615,
+      "learning_rate": 0.0002337570006340485,
+      "loss": 0.548,
+      "step": 5837
+    },
+    {
+      "epoch": 1.0413879225760414,
+      "grad_norm": 0.5142043828964233,
+      "learning_rate": 0.00023368708702787555,
+      "loss": 0.7064,
+      "step": 5838
+    },
+    {
+      "epoch": 1.0415663187940416,
+      "grad_norm": 0.4556623697280884,
+      "learning_rate": 0.00023361717470291176,
+      "loss": 0.7167,
+      "step": 5839
+    },
+    {
+      "epoch": 1.0417447150120418,
+      "grad_norm": 0.46669575572013855,
+      "learning_rate": 0.00023354726366464808,
+      "loss": 0.6693,
+      "step": 5840
+    },
+    {
+      "epoch": 1.041923111230042,
+      "grad_norm": 0.5870226621627808,
+      "learning_rate": 0.00023347735391857517,
+      "loss": 0.81,
+      "step": 5841
+    },
+    {
+      "epoch": 1.0421015074480422,
+      "grad_norm": 0.5324650406837463,
+      "learning_rate": 0.00023340744547018384,
+      "loss": 0.6455,
+      "step": 5842
+    },
+    {
+      "epoch": 1.0422799036660422,
+      "grad_norm": 0.4658534824848175,
+      "learning_rate": 0.00023333753832496443,
+      "loss": 0.716,
+      "step": 5843
+    },
+    {
+      "epoch": 1.0424582998840424,
+      "grad_norm": 0.6966456174850464,
+      "learning_rate": 0.0002332676324884077,
+      "loss": 0.4844,
+      "step": 5844
+    },
+    {
+      "epoch": 1.0426366961020426,
+      "grad_norm": 0.440407395362854,
+      "learning_rate": 0.00023319772796600395,
+      "loss": 0.5498,
+      "step": 5845
+    },
+    {
+      "epoch": 1.0428150923200428,
+      "grad_norm": 0.5414045453071594,
+      "learning_rate": 0.00023312782476324345,
+      "loss": 0.8414,
+      "step": 5846
+    },
+    {
+      "epoch": 1.042993488538043,
+      "grad_norm": 0.4795417785644531,
+      "learning_rate": 0.0002330579228856163,
+      "loss": 0.7409,
+      "step": 5847
+    },
+    {
+      "epoch": 1.0431718847560432,
+      "grad_norm": 0.47270727157592773,
+      "learning_rate": 0.00023298802233861254,
+      "loss": 0.5839,
+      "step": 5848
+    },
+    {
+      "epoch": 1.0433502809740434,
+      "grad_norm": 0.45291322469711304,
+      "learning_rate": 0.0002329181231277222,
+      "loss": 0.6999,
+      "step": 5849
+    },
+    {
+      "epoch": 1.0435286771920436,
+      "grad_norm": 0.49870064854621887,
+      "learning_rate": 0.00023284822525843513,
+      "loss": 0.6977,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0437070734100438,
+      "grad_norm": 0.500630259513855,
+      "learning_rate": 0.00023277832873624108,
+      "loss": 0.8385,
+      "step": 5851
+    },
+    {
+      "epoch": 1.043885469628044,
+      "grad_norm": 0.4953528642654419,
+      "learning_rate": 0.00023270843356662968,
+      "loss": 0.7302,
+      "step": 5852
+    },
+    {
+      "epoch": 1.0440638658460442,
+      "grad_norm": 0.4294586181640625,
+      "learning_rate": 0.00023263853975509044,
+      "loss": 0.529,
+      "step": 5853
+    },
+    {
+      "epoch": 1.0442422620640441,
+      "grad_norm": 0.5091565847396851,
+      "learning_rate": 0.00023256864730711289,
+      "loss": 0.7218,
+      "step": 5854
+    },
+    {
+      "epoch": 1.0444206582820443,
+      "grad_norm": 0.4561258554458618,
+      "learning_rate": 0.00023249875622818623,
+      "loss": 0.6372,
+      "step": 5855
+    },
+    {
+      "epoch": 1.0445990545000445,
+      "grad_norm": 0.46639683842658997,
+      "learning_rate": 0.00023242886652379973,
+      "loss": 0.6476,
+      "step": 5856
+    },
+    {
+      "epoch": 1.0447774507180447,
+      "grad_norm": 0.4989601671695709,
+      "learning_rate": 0.00023235897819944245,
+      "loss": 0.7827,
+      "step": 5857
+    },
+    {
+      "epoch": 1.044955846936045,
+      "grad_norm": 0.4990067780017853,
+      "learning_rate": 0.00023228909126060335,
+      "loss": 0.7969,
+      "step": 5858
+    },
+    {
+      "epoch": 1.0451342431540451,
+      "grad_norm": 0.4401039183139801,
+      "learning_rate": 0.00023221920571277159,
+      "loss": 0.6276,
+      "step": 5859
+    },
+    {
+      "epoch": 1.0453126393720453,
+      "grad_norm": 0.496707558631897,
+      "learning_rate": 0.00023214932156143564,
+      "loss": 0.7849,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0454910355900455,
+      "grad_norm": 0.5055187940597534,
+      "learning_rate": 0.00023207943881208435,
+      "loss": 0.75,
+      "step": 5861
+    },
+    {
+      "epoch": 1.0456694318080457,
+      "grad_norm": 0.4924319088459015,
+      "learning_rate": 0.0002320095574702062,
+      "loss": 0.741,
+      "step": 5862
+    },
+    {
+      "epoch": 1.045847828026046,
+      "grad_norm": 0.4404760003089905,
+      "learning_rate": 0.0002319396775412897,
+      "loss": 0.5559,
+      "step": 5863
+    },
+    {
+      "epoch": 1.0460262242440461,
+      "grad_norm": 0.4808800518512726,
+      "learning_rate": 0.0002318697990308231,
+      "loss": 0.776,
+      "step": 5864
+    },
+    {
+      "epoch": 1.046204620462046,
+      "grad_norm": 0.4353162348270416,
+      "learning_rate": 0.00023179992194429473,
+      "loss": 0.534,
+      "step": 5865
+    },
+    {
+      "epoch": 1.0463830166800463,
+      "grad_norm": 0.489243745803833,
+      "learning_rate": 0.00023173004628719262,
+      "loss": 0.72,
+      "step": 5866
+    },
+    {
+      "epoch": 1.0465614128980465,
+      "grad_norm": 0.5229580998420715,
+      "learning_rate": 0.0002316601720650049,
+      "loss": 0.8733,
+      "step": 5867
+    },
+    {
+      "epoch": 1.0467398091160467,
+      "grad_norm": 0.4031560719013214,
+      "learning_rate": 0.0002315902992832193,
+      "loss": 0.4158,
+      "step": 5868
+    },
+    {
+      "epoch": 1.046918205334047,
+      "grad_norm": 0.5013259053230286,
+      "learning_rate": 0.00023152042794732366,
+      "loss": 0.6999,
+      "step": 5869
+    },
+    {
+      "epoch": 1.047096601552047,
+      "grad_norm": 0.4599556624889374,
+      "learning_rate": 0.00023145055806280567,
+      "loss": 0.6991,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0472749977700473,
+      "grad_norm": 0.4679622948169708,
+      "learning_rate": 0.00023138068963515288,
+      "loss": 0.9515,
+      "step": 5871
+    },
+    {
+      "epoch": 1.0474533939880475,
+      "grad_norm": 0.4390574097633362,
+      "learning_rate": 0.0002313108226698527,
+      "loss": 0.5369,
+      "step": 5872
+    },
+    {
+      "epoch": 1.0476317902060477,
+      "grad_norm": 0.42867380380630493,
+      "learning_rate": 0.00023124095717239241,
+      "loss": 0.5489,
+      "step": 5873
+    },
+    {
+      "epoch": 1.0478101864240479,
+      "grad_norm": 0.43418964743614197,
+      "learning_rate": 0.00023117109314825933,
+      "loss": 0.7322,
+      "step": 5874
+    },
+    {
+      "epoch": 1.047988582642048,
+      "grad_norm": 0.4745190441608429,
+      "learning_rate": 0.00023110123060294047,
+      "loss": 0.682,
+      "step": 5875
+    },
+    {
+      "epoch": 1.048166978860048,
+      "grad_norm": 0.4581608176231384,
+      "learning_rate": 0.00023103136954192286,
+      "loss": 0.6654,
+      "step": 5876
+    },
+    {
+      "epoch": 1.0483453750780483,
+      "grad_norm": 0.46257057785987854,
+      "learning_rate": 0.0002309615099706933,
+      "loss": 0.6197,
+      "step": 5877
+    },
+    {
+      "epoch": 1.0485237712960485,
+      "grad_norm": 0.528059184551239,
+      "learning_rate": 0.00023089165189473857,
+      "loss": 0.7513,
+      "step": 5878
+    },
+    {
+      "epoch": 1.0487021675140487,
+      "grad_norm": 0.44776850938796997,
+      "learning_rate": 0.00023082179531954525,
+      "loss": 0.6703,
+      "step": 5879
+    },
+    {
+      "epoch": 1.0488805637320489,
+      "grad_norm": 0.46082332730293274,
+      "learning_rate": 0.00023075194025059994,
+      "loss": 0.5771,
+      "step": 5880
+    },
+    {
+      "epoch": 1.049058959950049,
+      "grad_norm": 0.49523085355758667,
+      "learning_rate": 0.00023068208669338894,
+      "loss": 0.5816,
+      "step": 5881
+    },
+    {
+      "epoch": 1.0492373561680493,
+      "grad_norm": 0.6181836128234863,
+      "learning_rate": 0.0002306122346533985,
+      "loss": 0.5963,
+      "step": 5882
+    },
+    {
+      "epoch": 1.0494157523860494,
+      "grad_norm": 0.8168792128562927,
+      "learning_rate": 0.00023054238413611482,
+      "loss": 0.609,
+      "step": 5883
+    },
+    {
+      "epoch": 1.0495941486040496,
+      "grad_norm": 0.5016876459121704,
+      "learning_rate": 0.0002304725351470239,
+      "loss": 0.8368,
+      "step": 5884
+    },
+    {
+      "epoch": 1.0497725448220498,
+      "grad_norm": 0.4856698215007782,
+      "learning_rate": 0.00023040268769161168,
+      "loss": 0.5799,
+      "step": 5885
+    },
+    {
+      "epoch": 1.04995094104005,
+      "grad_norm": 0.8010654449462891,
+      "learning_rate": 0.00023033284177536396,
+      "loss": 0.7327,
+      "step": 5886
+    },
+    {
+      "epoch": 1.05012933725805,
+      "grad_norm": 0.48454806208610535,
+      "learning_rate": 0.00023026299740376633,
+      "loss": 0.7485,
+      "step": 5887
+    },
+    {
+      "epoch": 1.0503077334760502,
+      "grad_norm": 0.47134482860565186,
+      "learning_rate": 0.00023019315458230449,
+      "loss": 0.7593,
+      "step": 5888
+    },
+    {
+      "epoch": 1.0504861296940504,
+      "grad_norm": 0.5328679084777832,
+      "learning_rate": 0.00023012331331646359,
+      "loss": 0.6926,
+      "step": 5889
+    },
+    {
+      "epoch": 1.0506645259120506,
+      "grad_norm": 0.46135085821151733,
+      "learning_rate": 0.0002300534736117292,
+      "loss": 0.6585,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0508429221300508,
+      "grad_norm": 0.7322617173194885,
+      "learning_rate": 0.00022998363547358641,
+      "loss": 0.7974,
+      "step": 5891
+    },
+    {
+      "epoch": 1.051021318348051,
+      "grad_norm": 0.46016210317611694,
+      "learning_rate": 0.0002299137989075203,
+      "loss": 0.7362,
+      "step": 5892
+    },
+    {
+      "epoch": 1.0511997145660512,
+      "grad_norm": 0.4320010244846344,
+      "learning_rate": 0.00022984396391901582,
+      "loss": 0.5613,
+      "step": 5893
+    },
+    {
+      "epoch": 1.0513781107840514,
+      "grad_norm": 0.45406126976013184,
+      "learning_rate": 0.0002297741305135577,
+      "loss": 0.5832,
+      "step": 5894
+    },
+    {
+      "epoch": 1.0515565070020516,
+      "grad_norm": 0.5657575130462646,
+      "learning_rate": 0.00022970429869663064,
+      "loss": 0.9168,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0517349032200518,
+      "grad_norm": 0.47483891248703003,
+      "learning_rate": 0.00022963446847371925,
+      "loss": 0.5917,
+      "step": 5896
+    },
+    {
+      "epoch": 1.051913299438052,
+      "grad_norm": 0.4458398222923279,
+      "learning_rate": 0.00022956463985030794,
+      "loss": 0.5259,
+      "step": 5897
+    },
+    {
+      "epoch": 1.0520916956560522,
+      "grad_norm": 0.4733836054801941,
+      "learning_rate": 0.000229494812831881,
+      "loss": 0.6835,
+      "step": 5898
+    },
+    {
+      "epoch": 1.0522700918740522,
+      "grad_norm": 0.49233517050743103,
+      "learning_rate": 0.00022942498742392265,
+      "loss": 0.8004,
+      "step": 5899
+    },
+    {
+      "epoch": 1.0524484880920524,
+      "grad_norm": 0.7832338809967041,
+      "learning_rate": 0.00022935516363191695,
+      "loss": 0.6962,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0526268843100526,
+      "grad_norm": 0.7512785792350769,
+      "learning_rate": 0.00022928534146134783,
+      "loss": 0.7948,
+      "step": 5901
+    },
+    {
+      "epoch": 1.0528052805280528,
+      "grad_norm": 0.6257414221763611,
+      "learning_rate": 0.00022921552091769907,
+      "loss": 0.5168,
+      "step": 5902
+    },
+    {
+      "epoch": 1.052983676746053,
+      "grad_norm": 0.5249746441841125,
+      "learning_rate": 0.0002291457020064543,
+      "loss": 0.9227,
+      "step": 5903
+    },
+    {
+      "epoch": 1.0531620729640532,
+      "grad_norm": 0.4546600878238678,
+      "learning_rate": 0.00022907588473309703,
+      "loss": 0.5357,
+      "step": 5904
+    },
+    {
+      "epoch": 1.0533404691820534,
+      "grad_norm": 0.5810390114784241,
+      "learning_rate": 0.00022900606910311098,
+      "loss": 0.6856,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0535188654000536,
+      "grad_norm": 0.45941051840782166,
+      "learning_rate": 0.00022893625512197913,
+      "loss": 0.6229,
+      "step": 5906
+    },
+    {
+      "epoch": 1.0536972616180538,
+      "grad_norm": 0.5021553039550781,
+      "learning_rate": 0.00022886644279518473,
+      "loss": 0.9207,
+      "step": 5907
+    },
+    {
+      "epoch": 1.053875657836054,
+      "grad_norm": 0.48117703199386597,
+      "learning_rate": 0.00022879663212821083,
+      "loss": 0.648,
+      "step": 5908
+    },
+    {
+      "epoch": 1.054054054054054,
+      "grad_norm": 0.5499297976493835,
+      "learning_rate": 0.00022872682312654032,
+      "loss": 0.7953,
+      "step": 5909
+    },
+    {
+      "epoch": 1.0542324502720541,
+      "grad_norm": 0.4166644513607025,
+      "learning_rate": 0.000228657015795656,
+      "loss": 0.506,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0544108464900543,
+      "grad_norm": 0.5143164992332458,
+      "learning_rate": 0.00022858721014104043,
+      "loss": 0.7587,
+      "step": 5911
+    },
+    {
+      "epoch": 1.0545892427080545,
+      "grad_norm": 0.5195590257644653,
+      "learning_rate": 0.00022851740616817615,
+      "loss": 0.8337,
+      "step": 5912
+    },
+    {
+      "epoch": 1.0547676389260547,
+      "grad_norm": 0.5148821473121643,
+      "learning_rate": 0.00022844760388254556,
+      "loss": 0.9393,
+      "step": 5913
+    },
+    {
+      "epoch": 1.054946035144055,
+      "grad_norm": 0.5613206028938293,
+      "learning_rate": 0.00022837780328963095,
+      "loss": 1.0002,
+      "step": 5914
+    },
+    {
+      "epoch": 1.0551244313620551,
+      "grad_norm": 0.4363964796066284,
+      "learning_rate": 0.00022830800439491435,
+      "loss": 0.6343,
+      "step": 5915
+    },
+    {
+      "epoch": 1.0553028275800553,
+      "grad_norm": 0.46682143211364746,
+      "learning_rate": 0.00022823820720387766,
+      "loss": 0.6258,
+      "step": 5916
+    },
+    {
+      "epoch": 1.0554812237980555,
+      "grad_norm": 0.45022645592689514,
+      "learning_rate": 0.00022816841172200287,
+      "loss": 0.7482,
+      "step": 5917
+    },
+    {
+      "epoch": 1.0556596200160557,
+      "grad_norm": 0.5296302437782288,
+      "learning_rate": 0.00022809861795477162,
+      "loss": 0.7556,
+      "step": 5918
+    },
+    {
+      "epoch": 1.055838016234056,
+      "grad_norm": 0.48178842663764954,
+      "learning_rate": 0.00022802882590766544,
+      "loss": 0.8764,
+      "step": 5919
+    },
+    {
+      "epoch": 1.0560164124520561,
+      "grad_norm": 0.5427073240280151,
+      "learning_rate": 0.00022795903558616587,
+      "loss": 0.9599,
+      "step": 5920
+    },
+    {
+      "epoch": 1.056194808670056,
+      "grad_norm": 0.5280442833900452,
+      "learning_rate": 0.00022788924699575417,
+      "loss": 0.6732,
+      "step": 5921
+    },
+    {
+      "epoch": 1.0563732048880563,
+      "grad_norm": 0.5916957259178162,
+      "learning_rate": 0.00022781946014191145,
+      "loss": 0.8213,
+      "step": 5922
+    },
+    {
+      "epoch": 1.0565516011060565,
+      "grad_norm": 0.566719651222229,
+      "learning_rate": 0.00022774967503011884,
+      "loss": 0.7634,
+      "step": 5923
+    },
+    {
+      "epoch": 1.0567299973240567,
+      "grad_norm": 0.5337679982185364,
+      "learning_rate": 0.00022767989166585717,
+      "loss": 0.8209,
+      "step": 5924
+    },
+    {
+      "epoch": 1.056908393542057,
+      "grad_norm": 1.3235185146331787,
+      "learning_rate": 0.0002276101100546072,
+      "loss": 0.706,
+      "step": 5925
+    },
+    {
+      "epoch": 1.057086789760057,
+      "grad_norm": 0.4909132719039917,
+      "learning_rate": 0.0002275403302018496,
+      "loss": 0.6735,
+      "step": 5926
+    },
+    {
+      "epoch": 1.0572651859780573,
+      "grad_norm": 0.4729708135128021,
+      "learning_rate": 0.00022747055211306473,
+      "loss": 0.6506,
+      "step": 5927
+    },
+    {
+      "epoch": 1.0574435821960575,
+      "grad_norm": 0.48028764128685,
+      "learning_rate": 0.000227400775793733,
+      "loss": 0.577,
+      "step": 5928
+    },
+    {
+      "epoch": 1.0576219784140577,
+      "grad_norm": 0.4471667408943176,
+      "learning_rate": 0.00022733100124933464,
+      "loss": 0.5678,
+      "step": 5929
+    },
+    {
+      "epoch": 1.057800374632058,
+      "grad_norm": 0.4753149449825287,
+      "learning_rate": 0.00022726122848534965,
+      "loss": 0.7441,
+      "step": 5930
+    },
+    {
+      "epoch": 1.0579787708500579,
+      "grad_norm": 0.5008271932601929,
+      "learning_rate": 0.00022719145750725803,
+      "loss": 0.7606,
+      "step": 5931
+    },
+    {
+      "epoch": 1.058157167068058,
+      "grad_norm": 0.48235058784484863,
+      "learning_rate": 0.0002271216883205395,
+      "loss": 0.6566,
+      "step": 5932
+    },
+    {
+      "epoch": 1.0583355632860583,
+      "grad_norm": 0.49578502774238586,
+      "learning_rate": 0.00022705192093067377,
+      "loss": 0.6123,
+      "step": 5933
+    },
+    {
+      "epoch": 1.0585139595040585,
+      "grad_norm": 0.4736134111881256,
+      "learning_rate": 0.0002269821553431403,
+      "loss": 0.5684,
+      "step": 5934
+    },
+    {
+      "epoch": 1.0586923557220587,
+      "grad_norm": 0.4547494351863861,
+      "learning_rate": 0.00022691239156341828,
+      "loss": 0.6157,
+      "step": 5935
+    },
+    {
+      "epoch": 1.0588707519400589,
+      "grad_norm": 0.45939305424690247,
+      "learning_rate": 0.0002268426295969872,
+      "loss": 0.6121,
+      "step": 5936
+    },
+    {
+      "epoch": 1.059049148158059,
+      "grad_norm": 0.5370938181877136,
+      "learning_rate": 0.00022677286944932604,
+      "loss": 0.8359,
+      "step": 5937
+    },
+    {
+      "epoch": 1.0592275443760593,
+      "grad_norm": 0.5468327403068542,
+      "learning_rate": 0.0002267031111259137,
+      "loss": 0.7847,
+      "step": 5938
+    },
+    {
+      "epoch": 1.0594059405940595,
+      "grad_norm": 0.472802996635437,
+      "learning_rate": 0.00022663335463222906,
+      "loss": 0.6689,
+      "step": 5939
+    },
+    {
+      "epoch": 1.0595843368120597,
+      "grad_norm": 0.5464509129524231,
+      "learning_rate": 0.00022656359997375063,
+      "loss": 0.9584,
+      "step": 5940
+    },
+    {
+      "epoch": 1.0597627330300599,
+      "grad_norm": 0.5138636827468872,
+      "learning_rate": 0.000226493847155957,
+      "loss": 0.7645,
+      "step": 5941
+    },
+    {
+      "epoch": 1.05994112924806,
+      "grad_norm": 0.4933434724807739,
+      "learning_rate": 0.00022642409618432648,
+      "loss": 0.7348,
+      "step": 5942
+    },
+    {
+      "epoch": 1.06011952546606,
+      "grad_norm": 0.4958341121673584,
+      "learning_rate": 0.00022635434706433727,
+      "loss": 0.7549,
+      "step": 5943
+    },
+    {
+      "epoch": 1.0602979216840602,
+      "grad_norm": 0.4594402015209198,
+      "learning_rate": 0.00022628459980146752,
+      "loss": 0.5529,
+      "step": 5944
+    },
+    {
+      "epoch": 1.0604763179020604,
+      "grad_norm": 0.5959436893463135,
+      "learning_rate": 0.00022621485440119506,
+      "loss": 0.65,
+      "step": 5945
+    },
+    {
+      "epoch": 1.0606547141200606,
+      "grad_norm": 0.5254977345466614,
+      "learning_rate": 0.00022614511086899768,
+      "loss": 0.7693,
+      "step": 5946
+    },
+    {
+      "epoch": 1.0608331103380608,
+      "grad_norm": 0.4639701843261719,
+      "learning_rate": 0.00022607536921035313,
+      "loss": 0.7103,
+      "step": 5947
+    },
+    {
+      "epoch": 1.061011506556061,
+      "grad_norm": 0.5139868855476379,
+      "learning_rate": 0.00022600562943073872,
+      "loss": 0.7358,
+      "step": 5948
+    },
+    {
+      "epoch": 1.0611899027740612,
+      "grad_norm": 0.4823780059814453,
+      "learning_rate": 0.00022593589153563183,
+      "loss": 0.6385,
+      "step": 5949
+    },
+    {
+      "epoch": 1.0613682989920614,
+      "grad_norm": 0.475702702999115,
+      "learning_rate": 0.00022586615553050958,
+      "loss": 0.7695,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0615466952100616,
+      "grad_norm": 0.4667466878890991,
+      "learning_rate": 0.00022579642142084918,
+      "loss": 0.7241,
+      "step": 5951
+    },
+    {
+      "epoch": 1.0617250914280618,
+      "grad_norm": 0.46129074692726135,
+      "learning_rate": 0.00022572668921212746,
+      "loss": 0.7244,
+      "step": 5952
+    },
+    {
+      "epoch": 1.0619034876460618,
+      "grad_norm": 0.45217975974082947,
+      "learning_rate": 0.0002256569589098211,
+      "loss": 0.5342,
+      "step": 5953
+    },
+    {
+      "epoch": 1.062081883864062,
+      "grad_norm": 0.4796280860900879,
+      "learning_rate": 0.0002255872305194067,
+      "loss": 0.7878,
+      "step": 5954
+    },
+    {
+      "epoch": 1.0622602800820622,
+      "grad_norm": 0.5046055912971497,
+      "learning_rate": 0.0002255175040463607,
+      "loss": 0.6931,
+      "step": 5955
+    },
+    {
+      "epoch": 1.0624386763000624,
+      "grad_norm": 0.5430681109428406,
+      "learning_rate": 0.00022544777949615942,
+      "loss": 0.6904,
+      "step": 5956
+    },
+    {
+      "epoch": 1.0626170725180626,
+      "grad_norm": 0.5339892506599426,
+      "learning_rate": 0.00022537805687427895,
+      "loss": 0.7239,
+      "step": 5957
+    },
+    {
+      "epoch": 1.0627954687360628,
+      "grad_norm": 0.510955274105072,
+      "learning_rate": 0.0002253083361861953,
+      "loss": 0.7252,
+      "step": 5958
+    },
+    {
+      "epoch": 1.062973864954063,
+      "grad_norm": 0.48307734727859497,
+      "learning_rate": 0.00022523861743738434,
+      "loss": 0.7571,
+      "step": 5959
+    },
+    {
+      "epoch": 1.0631522611720632,
+      "grad_norm": 0.4888105094432831,
+      "learning_rate": 0.00022516890063332173,
+      "loss": 0.6627,
+      "step": 5960
+    },
+    {
+      "epoch": 1.0633306573900634,
+      "grad_norm": 0.44385766983032227,
+      "learning_rate": 0.00022509918577948292,
+      "loss": 0.6508,
+      "step": 5961
+    },
+    {
+      "epoch": 1.0635090536080636,
+      "grad_norm": 0.4929356276988983,
+      "learning_rate": 0.00022502947288134334,
+      "loss": 0.6448,
+      "step": 5962
+    },
+    {
+      "epoch": 1.0636874498260638,
+      "grad_norm": 0.5029351711273193,
+      "learning_rate": 0.00022495976194437822,
+      "loss": 0.7504,
+      "step": 5963
+    },
+    {
+      "epoch": 1.063865846044064,
+      "grad_norm": 0.47953563928604126,
+      "learning_rate": 0.00022489005297406266,
+      "loss": 0.6591,
+      "step": 5964
+    },
+    {
+      "epoch": 1.064044242262064,
+      "grad_norm": 0.5105430483818054,
+      "learning_rate": 0.0002248203459758714,
+      "loss": 0.8449,
+      "step": 5965
+    },
+    {
+      "epoch": 1.0642226384800642,
+      "grad_norm": 0.5377670526504517,
+      "learning_rate": 0.00022475064095527948,
+      "loss": 0.8731,
+      "step": 5966
+    },
+    {
+      "epoch": 1.0644010346980644,
+      "grad_norm": 0.47288522124290466,
+      "learning_rate": 0.00022468093791776128,
+      "loss": 0.595,
+      "step": 5967
+    },
+    {
+      "epoch": 1.0645794309160646,
+      "grad_norm": 0.43630439043045044,
+      "learning_rate": 0.00022461123686879137,
+      "loss": 0.5941,
+      "step": 5968
+    },
+    {
+      "epoch": 1.0647578271340647,
+      "grad_norm": 0.4721398651599884,
+      "learning_rate": 0.00022454153781384395,
+      "loss": 0.6303,
+      "step": 5969
+    },
+    {
+      "epoch": 1.064936223352065,
+      "grad_norm": 0.5068725347518921,
+      "learning_rate": 0.00022447184075839323,
+      "loss": 0.7956,
+      "step": 5970
+    },
+    {
+      "epoch": 1.0651146195700651,
+      "grad_norm": 0.5235750675201416,
+      "learning_rate": 0.0002244021457079131,
+      "loss": 0.7384,
+      "step": 5971
+    },
+    {
+      "epoch": 1.0652930157880653,
+      "grad_norm": 0.44175130128860474,
+      "learning_rate": 0.00022433245266787749,
+      "loss": 0.6365,
+      "step": 5972
+    },
+    {
+      "epoch": 1.0654714120060655,
+      "grad_norm": 0.4824054539203644,
+      "learning_rate": 0.00022426276164376003,
+      "loss": 0.795,
+      "step": 5973
+    },
+    {
+      "epoch": 1.0656498082240657,
+      "grad_norm": 0.46271640062332153,
+      "learning_rate": 0.00022419307264103414,
+      "loss": 0.6397,
+      "step": 5974
+    },
+    {
+      "epoch": 1.0658282044420657,
+      "grad_norm": 0.4163155257701874,
+      "learning_rate": 0.0002241233856651732,
+      "loss": 0.4912,
+      "step": 5975
+    },
+    {
+      "epoch": 1.066006600660066,
+      "grad_norm": 0.46445098519325256,
+      "learning_rate": 0.00022405370072165043,
+      "loss": 0.587,
+      "step": 5976
+    },
+    {
+      "epoch": 1.0661849968780661,
+      "grad_norm": 0.4462120831012726,
+      "learning_rate": 0.00022398401781593884,
+      "loss": 0.6272,
+      "step": 5977
+    },
+    {
+      "epoch": 1.0663633930960663,
+      "grad_norm": 0.48081815242767334,
+      "learning_rate": 0.00022391433695351131,
+      "loss": 0.6237,
+      "step": 5978
+    },
+    {
+      "epoch": 1.0665417893140665,
+      "grad_norm": 0.5043575167655945,
+      "learning_rate": 0.00022384465813984054,
+      "loss": 0.6698,
+      "step": 5979
+    },
+    {
+      "epoch": 1.0667201855320667,
+      "grad_norm": 0.5385834574699402,
+      "learning_rate": 0.00022377498138039903,
+      "loss": 0.7248,
+      "step": 5980
+    },
+    {
+      "epoch": 1.066898581750067,
+      "grad_norm": 0.5373404622077942,
+      "learning_rate": 0.00022370530668065915,
+      "loss": 0.7383,
+      "step": 5981
+    },
+    {
+      "epoch": 1.067076977968067,
+      "grad_norm": 0.464138925075531,
+      "learning_rate": 0.0002236356340460932,
+      "loss": 0.5345,
+      "step": 5982
+    },
+    {
+      "epoch": 1.0672553741860673,
+      "grad_norm": 0.5200468301773071,
+      "learning_rate": 0.00022356596348217325,
+      "loss": 0.7235,
+      "step": 5983
+    },
+    {
+      "epoch": 1.0674337704040675,
+      "grad_norm": 0.5154780745506287,
+      "learning_rate": 0.00022349629499437116,
+      "loss": 0.8373,
+      "step": 5984
+    },
+    {
+      "epoch": 1.0676121666220677,
+      "grad_norm": 0.4483351707458496,
+      "learning_rate": 0.00022342662858815867,
+      "loss": 0.5755,
+      "step": 5985
+    },
+    {
+      "epoch": 1.067790562840068,
+      "grad_norm": 0.4439292848110199,
+      "learning_rate": 0.0002233569642690073,
+      "loss": 0.5772,
+      "step": 5986
+    },
+    {
+      "epoch": 1.0679689590580679,
+      "grad_norm": 0.49343034625053406,
+      "learning_rate": 0.00022328730204238852,
+      "loss": 0.7372,
+      "step": 5987
+    },
+    {
+      "epoch": 1.068147355276068,
+      "grad_norm": 0.4922356605529785,
+      "learning_rate": 0.00022321764191377347,
+      "loss": 0.7034,
+      "step": 5988
+    },
+    {
+      "epoch": 1.0683257514940683,
+      "grad_norm": 0.426301509141922,
+      "learning_rate": 0.00022314798388863336,
+      "loss": 0.6708,
+      "step": 5989
+    },
+    {
+      "epoch": 1.0685041477120685,
+      "grad_norm": 0.46274682879447937,
+      "learning_rate": 0.000223078327972439,
+      "loss": 0.5906,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0686825439300687,
+      "grad_norm": 0.43291884660720825,
+      "learning_rate": 0.0002230086741706612,
+      "loss": 0.6506,
+      "step": 5991
+    },
+    {
+      "epoch": 1.0688609401480689,
+      "grad_norm": 0.4762420952320099,
+      "learning_rate": 0.00022293902248877052,
+      "loss": 0.6862,
+      "step": 5992
+    },
+    {
+      "epoch": 1.069039336366069,
+      "grad_norm": 0.4900970160961151,
+      "learning_rate": 0.00022286937293223736,
+      "loss": 0.7371,
+      "step": 5993
+    },
+    {
+      "epoch": 1.0692177325840693,
+      "grad_norm": 0.44312936067581177,
+      "learning_rate": 0.00022279972550653203,
+      "loss": 0.7384,
+      "step": 5994
+    },
+    {
+      "epoch": 1.0693961288020695,
+      "grad_norm": 0.5585013628005981,
+      "learning_rate": 0.00022273008021712448,
+      "loss": 0.9225,
+      "step": 5995
+    },
+    {
+      "epoch": 1.0695745250200697,
+      "grad_norm": 0.4900633990764618,
+      "learning_rate": 0.00022266043706948462,
+      "loss": 0.6349,
+      "step": 5996
+    },
+    {
+      "epoch": 1.0697529212380696,
+      "grad_norm": 0.47767671942710876,
+      "learning_rate": 0.00022259079606908237,
+      "loss": 0.728,
+      "step": 5997
+    },
+    {
+      "epoch": 1.0699313174560698,
+      "grad_norm": 0.477839857339859,
+      "learning_rate": 0.00022252115722138724,
+      "loss": 0.626,
+      "step": 5998
+    },
+    {
+      "epoch": 1.07010971367407,
+      "grad_norm": 0.47772642970085144,
+      "learning_rate": 0.00022245152053186853,
+      "loss": 0.7116,
+      "step": 5999
+    },
+    {
+      "epoch": 1.0702881098920702,
+      "grad_norm": 0.45496633648872375,
+      "learning_rate": 0.00022238188600599558,
+      "loss": 0.6855,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0704665061100704,
+      "grad_norm": 0.5116882920265198,
+      "learning_rate": 0.00022231225364923736,
+      "loss": 0.7783,
+      "step": 6001
+    },
+    {
+      "epoch": 1.0706449023280706,
+      "grad_norm": 0.5786128044128418,
+      "learning_rate": 0.00022224262346706288,
+      "loss": 0.7753,
+      "step": 6002
+    },
+    {
+      "epoch": 1.0708232985460708,
+      "grad_norm": 0.4571937620639801,
+      "learning_rate": 0.00022217299546494078,
+      "loss": 0.6224,
+      "step": 6003
+    },
+    {
+      "epoch": 1.071001694764071,
+      "grad_norm": 0.4922640919685364,
+      "learning_rate": 0.00022210336964833966,
+      "loss": 0.7753,
+      "step": 6004
+    },
+    {
+      "epoch": 1.0711800909820712,
+      "grad_norm": 0.4930363595485687,
+      "learning_rate": 0.0002220337460227279,
+      "loss": 0.7542,
+      "step": 6005
+    },
+    {
+      "epoch": 1.0713584872000714,
+      "grad_norm": 0.4897131323814392,
+      "learning_rate": 0.00022196412459357372,
+      "loss": 0.7468,
+      "step": 6006
+    },
+    {
+      "epoch": 1.0715368834180716,
+      "grad_norm": 0.5091310739517212,
+      "learning_rate": 0.00022189450536634506,
+      "loss": 0.6646,
+      "step": 6007
+    },
+    {
+      "epoch": 1.0717152796360718,
+      "grad_norm": 0.5187907814979553,
+      "learning_rate": 0.00022182488834650987,
+      "loss": 0.816,
+      "step": 6008
+    },
+    {
+      "epoch": 1.0718936758540718,
+      "grad_norm": 0.5296580195426941,
+      "learning_rate": 0.00022175527353953585,
+      "loss": 0.8415,
+      "step": 6009
+    },
+    {
+      "epoch": 1.072072072072072,
+      "grad_norm": 0.4686925411224365,
+      "learning_rate": 0.00022168566095089043,
+      "loss": 0.5842,
+      "step": 6010
+    },
+    {
+      "epoch": 1.0722504682900722,
+      "grad_norm": 0.5479150414466858,
+      "learning_rate": 0.000221616050586041,
+      "loss": 0.7578,
+      "step": 6011
+    },
+    {
+      "epoch": 1.0724288645080724,
+      "grad_norm": 0.6131009459495544,
+      "learning_rate": 0.0002215464424504548,
+      "loss": 0.6768,
+      "step": 6012
+    },
+    {
+      "epoch": 1.0726072607260726,
+      "grad_norm": 0.47752901911735535,
+      "learning_rate": 0.00022147683654959876,
+      "loss": 0.6412,
+      "step": 6013
+    },
+    {
+      "epoch": 1.0727856569440728,
+      "grad_norm": 0.47805142402648926,
+      "learning_rate": 0.0002214072328889397,
+      "loss": 0.6528,
+      "step": 6014
+    },
+    {
+      "epoch": 1.072964053162073,
+      "grad_norm": 0.44289377331733704,
+      "learning_rate": 0.00022133763147394426,
+      "loss": 0.7201,
+      "step": 6015
+    },
+    {
+      "epoch": 1.0731424493800732,
+      "grad_norm": 0.5617145299911499,
+      "learning_rate": 0.00022126803231007893,
+      "loss": 0.7928,
+      "step": 6016
+    },
+    {
+      "epoch": 1.0733208455980734,
+      "grad_norm": 0.5184254050254822,
+      "learning_rate": 0.00022119843540280995,
+      "loss": 0.8179,
+      "step": 6017
+    },
+    {
+      "epoch": 1.0734992418160736,
+      "grad_norm": 0.4659225344657898,
+      "learning_rate": 0.00022112884075760347,
+      "loss": 0.5611,
+      "step": 6018
+    },
+    {
+      "epoch": 1.0736776380340736,
+      "grad_norm": 0.5036365389823914,
+      "learning_rate": 0.00022105924837992547,
+      "loss": 0.9739,
+      "step": 6019
+    },
+    {
+      "epoch": 1.0738560342520738,
+      "grad_norm": 0.4579788148403168,
+      "learning_rate": 0.0002209896582752416,
+      "loss": 0.6486,
+      "step": 6020
+    },
+    {
+      "epoch": 1.074034430470074,
+      "grad_norm": 0.4959608018398285,
+      "learning_rate": 0.00022092007044901746,
+      "loss": 0.9164,
+      "step": 6021
+    },
+    {
+      "epoch": 1.0742128266880742,
+      "grad_norm": 0.4846688508987427,
+      "learning_rate": 0.00022085048490671849,
+      "loss": 0.7229,
+      "step": 6022
+    },
+    {
+      "epoch": 1.0743912229060744,
+      "grad_norm": 0.4702696204185486,
+      "learning_rate": 0.0002207809016538099,
+      "loss": 0.5403,
+      "step": 6023
+    },
+    {
+      "epoch": 1.0745696191240746,
+      "grad_norm": 0.4317050576210022,
+      "learning_rate": 0.00022071132069575672,
+      "loss": 0.6298,
+      "step": 6024
+    },
+    {
+      "epoch": 1.0747480153420748,
+      "grad_norm": 0.4599125385284424,
+      "learning_rate": 0.00022064174203802382,
+      "loss": 0.6337,
+      "step": 6025
+    },
+    {
+      "epoch": 1.074926411560075,
+      "grad_norm": 0.5147960186004639,
+      "learning_rate": 0.00022057216568607582,
+      "loss": 0.6918,
+      "step": 6026
+    },
+    {
+      "epoch": 1.0751048077780752,
+      "grad_norm": 0.45055848360061646,
+      "learning_rate": 0.00022050259164537725,
+      "loss": 0.6336,
+      "step": 6027
+    },
+    {
+      "epoch": 1.0752832039960754,
+      "grad_norm": 0.5015688538551331,
+      "learning_rate": 0.00022043301992139247,
+      "loss": 0.7122,
+      "step": 6028
+    },
+    {
+      "epoch": 1.0754616002140756,
+      "grad_norm": 0.4941107928752899,
+      "learning_rate": 0.0002203634505195856,
+      "loss": 0.7284,
+      "step": 6029
+    },
+    {
+      "epoch": 1.0756399964320758,
+      "grad_norm": 0.4930790960788727,
+      "learning_rate": 0.00022029388344542056,
+      "loss": 0.6292,
+      "step": 6030
+    },
+    {
+      "epoch": 1.0758183926500757,
+      "grad_norm": 0.461085706949234,
+      "learning_rate": 0.00022022431870436114,
+      "loss": 0.5988,
+      "step": 6031
+    },
+    {
+      "epoch": 1.075996788868076,
+      "grad_norm": 0.5100793838500977,
+      "learning_rate": 0.00022015475630187095,
+      "loss": 0.773,
+      "step": 6032
+    },
+    {
+      "epoch": 1.0761751850860761,
+      "grad_norm": 0.45178812742233276,
+      "learning_rate": 0.00022008519624341333,
+      "loss": 0.6794,
+      "step": 6033
+    },
+    {
+      "epoch": 1.0763535813040763,
+      "grad_norm": 0.44591426849365234,
+      "learning_rate": 0.0002200156385344515,
+      "loss": 0.6673,
+      "step": 6034
+    },
+    {
+      "epoch": 1.0765319775220765,
+      "grad_norm": 0.47172197699546814,
+      "learning_rate": 0.00021994608318044853,
+      "loss": 0.7226,
+      "step": 6035
+    },
+    {
+      "epoch": 1.0767103737400767,
+      "grad_norm": 0.5064817070960999,
+      "learning_rate": 0.00021987653018686724,
+      "loss": 0.7875,
+      "step": 6036
+    },
+    {
+      "epoch": 1.076888769958077,
+      "grad_norm": 0.49598416686058044,
+      "learning_rate": 0.0002198069795591703,
+      "loss": 0.6392,
+      "step": 6037
+    },
+    {
+      "epoch": 1.0770671661760771,
+      "grad_norm": 0.5123727321624756,
+      "learning_rate": 0.00021973743130282024,
+      "loss": 0.7137,
+      "step": 6038
+    },
+    {
+      "epoch": 1.0772455623940773,
+      "grad_norm": 0.48429879546165466,
+      "learning_rate": 0.00021966788542327926,
+      "loss": 0.6153,
+      "step": 6039
+    },
+    {
+      "epoch": 1.0774239586120775,
+      "grad_norm": 0.5295233130455017,
+      "learning_rate": 0.00021959834192600958,
+      "loss": 0.7566,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0776023548300775,
+      "grad_norm": 0.5196706652641296,
+      "learning_rate": 0.00021952880081647298,
+      "loss": 0.7511,
+      "step": 6041
+    },
+    {
+      "epoch": 1.0777807510480777,
+      "grad_norm": 0.4832461476325989,
+      "learning_rate": 0.00021945926210013112,
+      "loss": 0.5969,
+      "step": 6042
+    },
+    {
+      "epoch": 1.0779591472660779,
+      "grad_norm": 0.46703991293907166,
+      "learning_rate": 0.00021938972578244582,
+      "loss": 0.5534,
+      "step": 6043
+    },
+    {
+      "epoch": 1.078137543484078,
+      "grad_norm": 0.5281887054443359,
+      "learning_rate": 0.00021932019186887824,
+      "loss": 0.8217,
+      "step": 6044
+    },
+    {
+      "epoch": 1.0783159397020783,
+      "grad_norm": 0.48991912603378296,
+      "learning_rate": 0.00021925066036488969,
+      "loss": 0.6933,
+      "step": 6045
+    },
+    {
+      "epoch": 1.0784943359200785,
+      "grad_norm": 0.49540868401527405,
+      "learning_rate": 0.00021918113127594098,
+      "loss": 0.8362,
+      "step": 6046
+    },
+    {
+      "epoch": 1.0786727321380787,
+      "grad_norm": 0.4319058358669281,
+      "learning_rate": 0.00021911160460749295,
+      "loss": 0.5308,
+      "step": 6047
+    },
+    {
+      "epoch": 1.0788511283560789,
+      "grad_norm": 0.4906058609485626,
+      "learning_rate": 0.00021904208036500618,
+      "loss": 0.6291,
+      "step": 6048
+    },
+    {
+      "epoch": 1.079029524574079,
+      "grad_norm": 0.46787381172180176,
+      "learning_rate": 0.0002189725585539411,
+      "loss": 0.7964,
+      "step": 6049
+    },
+    {
+      "epoch": 1.0792079207920793,
+      "grad_norm": 0.5059190988540649,
+      "learning_rate": 0.00021890303917975794,
+      "loss": 0.7308,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0793863170100795,
+      "grad_norm": 0.4896734356880188,
+      "learning_rate": 0.0002188335222479167,
+      "loss": 0.8345,
+      "step": 6051
+    },
+    {
+      "epoch": 1.0795647132280797,
+      "grad_norm": 0.48433777689933777,
+      "learning_rate": 0.0002187640077638772,
+      "loss": 0.5938,
+      "step": 6052
+    },
+    {
+      "epoch": 1.0797431094460797,
+      "grad_norm": 0.5422555804252625,
+      "learning_rate": 0.00021869449573309912,
+      "loss": 0.8484,
+      "step": 6053
+    },
+    {
+      "epoch": 1.0799215056640799,
+      "grad_norm": 0.4676606059074402,
+      "learning_rate": 0.00021862498616104188,
+      "loss": 0.6853,
+      "step": 6054
+    },
+    {
+      "epoch": 1.08009990188208,
+      "grad_norm": 0.5379725694656372,
+      "learning_rate": 0.00021855547905316467,
+      "loss": 0.8504,
+      "step": 6055
+    },
+    {
+      "epoch": 1.0802782981000802,
+      "grad_norm": 0.4376412034034729,
+      "learning_rate": 0.00021848597441492663,
+      "loss": 0.6913,
+      "step": 6056
+    },
+    {
+      "epoch": 1.0804566943180804,
+      "grad_norm": 0.48028087615966797,
+      "learning_rate": 0.0002184164722517865,
+      "loss": 0.7915,
+      "step": 6057
+    },
+    {
+      "epoch": 1.0806350905360806,
+      "grad_norm": 0.5001854300498962,
+      "learning_rate": 0.00021834697256920316,
+      "loss": 0.7001,
+      "step": 6058
+    },
+    {
+      "epoch": 1.0808134867540808,
+      "grad_norm": 0.4411584436893463,
+      "learning_rate": 0.00021827747537263496,
+      "loss": 0.5408,
+      "step": 6059
+    },
+    {
+      "epoch": 1.080991882972081,
+      "grad_norm": 0.4824419319629669,
+      "learning_rate": 0.0002182079806675402,
+      "loss": 0.6647,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0811702791900812,
+      "grad_norm": 0.4544886350631714,
+      "learning_rate": 0.00021813848845937691,
+      "loss": 0.6396,
+      "step": 6061
+    },
+    {
+      "epoch": 1.0813486754080814,
+      "grad_norm": 0.5271000266075134,
+      "learning_rate": 0.00021806899875360307,
+      "loss": 0.7279,
+      "step": 6062
+    },
+    {
+      "epoch": 1.0815270716260814,
+      "grad_norm": 0.5003283619880676,
+      "learning_rate": 0.00021799951155567632,
+      "loss": 0.7433,
+      "step": 6063
+    },
+    {
+      "epoch": 1.0817054678440816,
+      "grad_norm": 0.5777568221092224,
+      "learning_rate": 0.00021793002687105415,
+      "loss": 0.8419,
+      "step": 6064
+    },
+    {
+      "epoch": 1.0818838640620818,
+      "grad_norm": 0.4634368121623993,
+      "learning_rate": 0.00021786054470519388,
+      "loss": 0.6258,
+      "step": 6065
+    },
+    {
+      "epoch": 1.082062260280082,
+      "grad_norm": 0.4333091378211975,
+      "learning_rate": 0.00021779106506355264,
+      "loss": 0.6735,
+      "step": 6066
+    },
+    {
+      "epoch": 1.0822406564980822,
+      "grad_norm": 0.45657244324684143,
+      "learning_rate": 0.00021772158795158725,
+      "loss": 0.6708,
+      "step": 6067
+    },
+    {
+      "epoch": 1.0824190527160824,
+      "grad_norm": 0.5117029547691345,
+      "learning_rate": 0.00021765211337475445,
+      "loss": 0.7658,
+      "step": 6068
+    },
+    {
+      "epoch": 1.0825974489340826,
+      "grad_norm": 0.45610618591308594,
+      "learning_rate": 0.00021758264133851072,
+      "loss": 0.5174,
+      "step": 6069
+    },
+    {
+      "epoch": 1.0827758451520828,
+      "grad_norm": 0.4404855966567993,
+      "learning_rate": 0.0002175131718483124,
+      "loss": 0.5822,
+      "step": 6070
+    },
+    {
+      "epoch": 1.082954241370083,
+      "grad_norm": 0.5023370385169983,
+      "learning_rate": 0.0002174437049096156,
+      "loss": 0.7341,
+      "step": 6071
+    },
+    {
+      "epoch": 1.0831326375880832,
+      "grad_norm": 0.5234084725379944,
+      "learning_rate": 0.00021737424052787618,
+      "loss": 0.8447,
+      "step": 6072
+    },
+    {
+      "epoch": 1.0833110338060834,
+      "grad_norm": 0.49975350499153137,
+      "learning_rate": 0.00021730477870854985,
+      "loss": 0.6664,
+      "step": 6073
+    },
+    {
+      "epoch": 1.0834894300240836,
+      "grad_norm": 0.5313869118690491,
+      "learning_rate": 0.00021723531945709216,
+      "loss": 0.7133,
+      "step": 6074
+    },
+    {
+      "epoch": 1.0836678262420836,
+      "grad_norm": 0.513853907585144,
+      "learning_rate": 0.0002171658627789584,
+      "loss": 0.9083,
+      "step": 6075
+    },
+    {
+      "epoch": 1.0838462224600838,
+      "grad_norm": 0.48460695147514343,
+      "learning_rate": 0.00021709640867960362,
+      "loss": 0.8195,
+      "step": 6076
+    },
+    {
+      "epoch": 1.084024618678084,
+      "grad_norm": 0.49029091000556946,
+      "learning_rate": 0.00021702695716448276,
+      "loss": 0.6181,
+      "step": 6077
+    },
+    {
+      "epoch": 1.0842030148960842,
+      "grad_norm": 0.48168060183525085,
+      "learning_rate": 0.00021695750823905053,
+      "loss": 0.6585,
+      "step": 6078
+    },
+    {
+      "epoch": 1.0843814111140844,
+      "grad_norm": 0.5174769759178162,
+      "learning_rate": 0.00021688806190876136,
+      "loss": 0.684,
+      "step": 6079
+    },
+    {
+      "epoch": 1.0845598073320846,
+      "grad_norm": 0.47358494997024536,
+      "learning_rate": 0.00021681861817906954,
+      "loss": 0.6681,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0847382035500848,
+      "grad_norm": 0.5476037859916687,
+      "learning_rate": 0.00021674917705542918,
+      "loss": 0.8857,
+      "step": 6081
+    },
+    {
+      "epoch": 1.084916599768085,
+      "grad_norm": 0.47755712270736694,
+      "learning_rate": 0.00021667973854329415,
+      "loss": 0.6948,
+      "step": 6082
+    },
+    {
+      "epoch": 1.0850949959860852,
+      "grad_norm": 0.48054665327072144,
+      "learning_rate": 0.0002166103026481181,
+      "loss": 0.6114,
+      "step": 6083
+    },
+    {
+      "epoch": 1.0852733922040854,
+      "grad_norm": 0.5116613507270813,
+      "learning_rate": 0.00021654086937535449,
+      "loss": 0.7028,
+      "step": 6084
+    },
+    {
+      "epoch": 1.0854517884220853,
+      "grad_norm": 0.5386189222335815,
+      "learning_rate": 0.00021647143873045662,
+      "loss": 0.8417,
+      "step": 6085
+    },
+    {
+      "epoch": 1.0856301846400855,
+      "grad_norm": 0.770457923412323,
+      "learning_rate": 0.00021640201071887761,
+      "loss": 0.7823,
+      "step": 6086
+    },
+    {
+      "epoch": 1.0858085808580857,
+      "grad_norm": 0.45015949010849,
+      "learning_rate": 0.00021633258534607013,
+      "loss": 0.5068,
+      "step": 6087
+    },
+    {
+      "epoch": 1.085986977076086,
+      "grad_norm": 0.4800700843334198,
+      "learning_rate": 0.0002162631626174868,
+      "loss": 0.6321,
+      "step": 6088
+    },
+    {
+      "epoch": 1.0861653732940861,
+      "grad_norm": 0.5034077167510986,
+      "learning_rate": 0.0002161937425385803,
+      "loss": 0.7027,
+      "step": 6089
+    },
+    {
+      "epoch": 1.0863437695120863,
+      "grad_norm": 0.46986886858940125,
+      "learning_rate": 0.00021612432511480267,
+      "loss": 0.5984,
+      "step": 6090
+    },
+    {
+      "epoch": 1.0865221657300865,
+      "grad_norm": 1.3976149559020996,
+      "learning_rate": 0.00021605491035160603,
+      "loss": 0.81,
+      "step": 6091
+    },
+    {
+      "epoch": 1.0867005619480867,
+      "grad_norm": 0.4368259906768799,
+      "learning_rate": 0.0002159854982544421,
+      "loss": 0.5048,
+      "step": 6092
+    },
+    {
+      "epoch": 1.086878958166087,
+      "grad_norm": 0.45895346999168396,
+      "learning_rate": 0.00021591608882876249,
+      "loss": 0.6807,
+      "step": 6093
+    },
+    {
+      "epoch": 1.0870573543840871,
+      "grad_norm": 0.500730037689209,
+      "learning_rate": 0.00021584668208001856,
+      "loss": 0.7392,
+      "step": 6094
+    },
+    {
+      "epoch": 1.0872357506020873,
+      "grad_norm": 0.5281662344932556,
+      "learning_rate": 0.00021577727801366158,
+      "loss": 0.8316,
+      "step": 6095
+    },
+    {
+      "epoch": 1.0874141468200875,
+      "grad_norm": 0.5275529623031616,
+      "learning_rate": 0.00021570787663514242,
+      "loss": 0.7324,
+      "step": 6096
+    },
+    {
+      "epoch": 1.0875925430380875,
+      "grad_norm": 0.4927082061767578,
+      "learning_rate": 0.00021563847794991186,
+      "loss": 0.7293,
+      "step": 6097
+    },
+    {
+      "epoch": 1.0877709392560877,
+      "grad_norm": 0.49843451380729675,
+      "learning_rate": 0.0002155690819634205,
+      "loss": 0.7416,
+      "step": 6098
+    },
+    {
+      "epoch": 1.087949335474088,
+      "grad_norm": 0.46133577823638916,
+      "learning_rate": 0.00021549968868111863,
+      "loss": 0.5853,
+      "step": 6099
+    },
+    {
+      "epoch": 1.088127731692088,
+      "grad_norm": 0.45552465319633484,
+      "learning_rate": 0.00021543029810845634,
+      "loss": 0.6492,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0883061279100883,
+      "grad_norm": 0.40714430809020996,
+      "learning_rate": 0.00021536091025088356,
+      "loss": 0.5971,
+      "step": 6101
+    },
+    {
+      "epoch": 1.0884845241280885,
+      "grad_norm": 0.46682703495025635,
+      "learning_rate": 0.00021529152511384997,
+      "loss": 0.7605,
+      "step": 6102
+    },
+    {
+      "epoch": 1.0886629203460887,
+      "grad_norm": 0.45366188883781433,
+      "learning_rate": 0.00021522214270280497,
+      "loss": 0.6504,
+      "step": 6103
+    },
+    {
+      "epoch": 1.088841316564089,
+      "grad_norm": 0.44408997893333435,
+      "learning_rate": 0.00021515276302319807,
+      "loss": 0.6159,
+      "step": 6104
+    },
+    {
+      "epoch": 1.089019712782089,
+      "grad_norm": 0.4182736277580261,
+      "learning_rate": 0.0002150833860804781,
+      "loss": 0.6284,
+      "step": 6105
+    },
+    {
+      "epoch": 1.0891981090000893,
+      "grad_norm": 0.4445931911468506,
+      "learning_rate": 0.00021501401188009397,
+      "loss": 0.5778,
+      "step": 6106
+    },
+    {
+      "epoch": 1.0893765052180893,
+      "grad_norm": 0.46306318044662476,
+      "learning_rate": 0.00021494464042749427,
+      "loss": 0.6435,
+      "step": 6107
+    },
+    {
+      "epoch": 1.0895549014360895,
+      "grad_norm": 0.49822935461997986,
+      "learning_rate": 0.0002148752717281275,
+      "loss": 0.6868,
+      "step": 6108
+    },
+    {
+      "epoch": 1.0897332976540897,
+      "grad_norm": 0.4217550754547119,
+      "learning_rate": 0.0002148059057874417,
+      "loss": 0.5048,
+      "step": 6109
+    },
+    {
+      "epoch": 1.0899116938720899,
+      "grad_norm": 0.497555136680603,
+      "learning_rate": 0.00021473654261088492,
+      "loss": 0.741,
+      "step": 6110
+    },
+    {
+      "epoch": 1.09009009009009,
+      "grad_norm": 0.47492715716362,
+      "learning_rate": 0.00021466718220390494,
+      "loss": 0.6734,
+      "step": 6111
+    },
+    {
+      "epoch": 1.0902684863080903,
+      "grad_norm": 0.5085891485214233,
+      "learning_rate": 0.0002145978245719493,
+      "loss": 0.6618,
+      "step": 6112
+    },
+    {
+      "epoch": 1.0904468825260905,
+      "grad_norm": 0.45178478956222534,
+      "learning_rate": 0.00021452846972046523,
+      "loss": 0.5666,
+      "step": 6113
+    },
+    {
+      "epoch": 1.0906252787440907,
+      "grad_norm": 0.44206705689430237,
+      "learning_rate": 0.0002144591176548999,
+      "loss": 0.5961,
+      "step": 6114
+    },
+    {
+      "epoch": 1.0908036749620909,
+      "grad_norm": 0.4810873866081238,
+      "learning_rate": 0.00021438976838070016,
+      "loss": 0.7622,
+      "step": 6115
+    },
+    {
+      "epoch": 1.090982071180091,
+      "grad_norm": 0.45664024353027344,
+      "learning_rate": 0.00021432042190331266,
+      "loss": 0.7886,
+      "step": 6116
+    },
+    {
+      "epoch": 1.0911604673980912,
+      "grad_norm": 0.439494788646698,
+      "learning_rate": 0.0002142510782281839,
+      "loss": 0.7159,
+      "step": 6117
+    },
+    {
+      "epoch": 1.0913388636160914,
+      "grad_norm": 0.5126174688339233,
+      "learning_rate": 0.00021418173736076007,
+      "loss": 0.806,
+      "step": 6118
+    },
+    {
+      "epoch": 1.0915172598340914,
+      "grad_norm": 0.465787798166275,
+      "learning_rate": 0.00021411239930648713,
+      "loss": 0.5724,
+      "step": 6119
+    },
+    {
+      "epoch": 1.0916956560520916,
+      "grad_norm": 0.5169677138328552,
+      "learning_rate": 0.00021404306407081094,
+      "loss": 0.7997,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0918740522700918,
+      "grad_norm": 0.5252036452293396,
+      "learning_rate": 0.00021397373165917704,
+      "loss": 0.7794,
+      "step": 6121
+    },
+    {
+      "epoch": 1.092052448488092,
+      "grad_norm": 0.456243097782135,
+      "learning_rate": 0.00021390440207703075,
+      "loss": 0.6367,
+      "step": 6122
+    },
+    {
+      "epoch": 1.0922308447060922,
+      "grad_norm": 0.45093104243278503,
+      "learning_rate": 0.00021383507532981717,
+      "loss": 0.5247,
+      "step": 6123
+    },
+    {
+      "epoch": 1.0924092409240924,
+      "grad_norm": 0.40191787481307983,
+      "learning_rate": 0.00021376575142298122,
+      "loss": 0.4306,
+      "step": 6124
+    },
+    {
+      "epoch": 1.0925876371420926,
+      "grad_norm": 0.4974512755870819,
+      "learning_rate": 0.00021369643036196762,
+      "loss": 0.6029,
+      "step": 6125
+    },
+    {
+      "epoch": 1.0927660333600928,
+      "grad_norm": 0.4702337682247162,
+      "learning_rate": 0.0002136271121522207,
+      "loss": 0.6395,
+      "step": 6126
+    },
+    {
+      "epoch": 1.092944429578093,
+      "grad_norm": 0.48748883605003357,
+      "learning_rate": 0.00021355779679918475,
+      "loss": 0.6442,
+      "step": 6127
+    },
+    {
+      "epoch": 1.0931228257960932,
+      "grad_norm": 0.4701106548309326,
+      "learning_rate": 0.00021348848430830376,
+      "loss": 0.597,
+      "step": 6128
+    },
+    {
+      "epoch": 1.0933012220140932,
+      "grad_norm": 0.492400586605072,
+      "learning_rate": 0.00021341917468502148,
+      "loss": 0.6764,
+      "step": 6129
+    },
+    {
+      "epoch": 1.0934796182320934,
+      "grad_norm": 0.5202962756156921,
+      "learning_rate": 0.00021334986793478147,
+      "loss": 0.8682,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0936580144500936,
+      "grad_norm": 0.4661516547203064,
+      "learning_rate": 0.00021328056406302707,
+      "loss": 0.7784,
+      "step": 6131
+    },
+    {
+      "epoch": 1.0938364106680938,
+      "grad_norm": 0.5273587107658386,
+      "learning_rate": 0.0002132112630752014,
+      "loss": 0.6089,
+      "step": 6132
+    },
+    {
+      "epoch": 1.094014806886094,
+      "grad_norm": 0.44249820709228516,
+      "learning_rate": 0.0002131419649767473,
+      "loss": 0.539,
+      "step": 6133
+    },
+    {
+      "epoch": 1.0941932031040942,
+      "grad_norm": 0.4947591722011566,
+      "learning_rate": 0.00021307266977310728,
+      "loss": 0.8595,
+      "step": 6134
+    },
+    {
+      "epoch": 1.0943715993220944,
+      "grad_norm": 0.46888822317123413,
+      "learning_rate": 0.00021300337746972398,
+      "loss": 0.719,
+      "step": 6135
+    },
+    {
+      "epoch": 1.0945499955400946,
+      "grad_norm": 0.5104032158851624,
+      "learning_rate": 0.00021293408807203948,
+      "loss": 0.7808,
+      "step": 6136
+    },
+    {
+      "epoch": 1.0947283917580948,
+      "grad_norm": 0.5240311622619629,
+      "learning_rate": 0.00021286480158549582,
+      "loss": 0.7997,
+      "step": 6137
+    },
+    {
+      "epoch": 1.094906787976095,
+      "grad_norm": 0.5189229249954224,
+      "learning_rate": 0.00021279551801553463,
+      "loss": 0.8566,
+      "step": 6138
+    },
+    {
+      "epoch": 1.0950851841940952,
+      "grad_norm": 0.4285975396633148,
+      "learning_rate": 0.00021272623736759742,
+      "loss": 0.7213,
+      "step": 6139
+    },
+    {
+      "epoch": 1.0952635804120954,
+      "grad_norm": 0.5412182211875916,
+      "learning_rate": 0.0002126569596471255,
+      "loss": 0.8526,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0954419766300953,
+      "grad_norm": 0.5037285685539246,
+      "learning_rate": 0.00021258768485955988,
+      "loss": 0.7894,
+      "step": 6141
+    },
+    {
+      "epoch": 1.0956203728480955,
+      "grad_norm": 0.4721333980560303,
+      "learning_rate": 0.00021251841301034142,
+      "loss": 0.7696,
+      "step": 6142
+    },
+    {
+      "epoch": 1.0957987690660957,
+      "grad_norm": 0.5173086524009705,
+      "learning_rate": 0.00021244914410491062,
+      "loss": 0.8049,
+      "step": 6143
+    },
+    {
+      "epoch": 1.095977165284096,
+      "grad_norm": 0.47684210538864136,
+      "learning_rate": 0.00021237987814870795,
+      "loss": 0.6094,
+      "step": 6144
+    },
+    {
+      "epoch": 1.0961555615020961,
+      "grad_norm": 0.42279088497161865,
+      "learning_rate": 0.0002123106151471734,
+      "loss": 0.5532,
+      "step": 6145
+    },
+    {
+      "epoch": 1.0963339577200963,
+      "grad_norm": 0.5065193176269531,
+      "learning_rate": 0.00021224135510574703,
+      "loss": 0.6686,
+      "step": 6146
+    },
+    {
+      "epoch": 1.0965123539380965,
+      "grad_norm": 0.4805443584918976,
+      "learning_rate": 0.0002121720980298683,
+      "loss": 0.5822,
+      "step": 6147
+    },
+    {
+      "epoch": 1.0966907501560967,
+      "grad_norm": 0.4822993278503418,
+      "learning_rate": 0.0002121028439249767,
+      "loss": 0.7553,
+      "step": 6148
+    },
+    {
+      "epoch": 1.096869146374097,
+      "grad_norm": 0.4846903383731842,
+      "learning_rate": 0.00021203359279651132,
+      "loss": 0.6991,
+      "step": 6149
+    },
+    {
+      "epoch": 1.0970475425920971,
+      "grad_norm": 0.4852921962738037,
+      "learning_rate": 0.0002119643446499114,
+      "loss": 0.7108,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0972259388100971,
+      "grad_norm": 0.4798181354999542,
+      "learning_rate": 0.00021189509949061543,
+      "loss": 0.6163,
+      "step": 6151
+    },
+    {
+      "epoch": 1.0974043350280973,
+      "grad_norm": 0.5046129822731018,
+      "learning_rate": 0.00021182585732406196,
+      "loss": 0.8545,
+      "step": 6152
+    },
+    {
+      "epoch": 1.0975827312460975,
+      "grad_norm": 0.5011611580848694,
+      "learning_rate": 0.0002117566181556892,
+      "loss": 0.7396,
+      "step": 6153
+    },
+    {
+      "epoch": 1.0977611274640977,
+      "grad_norm": 0.4669513702392578,
+      "learning_rate": 0.0002116873819909352,
+      "loss": 0.6778,
+      "step": 6154
+    },
+    {
+      "epoch": 1.097939523682098,
+      "grad_norm": 0.4715968370437622,
+      "learning_rate": 0.00021161814883523773,
+      "loss": 0.7572,
+      "step": 6155
+    },
+    {
+      "epoch": 1.098117919900098,
+      "grad_norm": 0.9391183257102966,
+      "learning_rate": 0.00021154891869403433,
+      "loss": 0.7393,
+      "step": 6156
+    },
+    {
+      "epoch": 1.0982963161180983,
+      "grad_norm": 0.49894896149635315,
+      "learning_rate": 0.00021147969157276234,
+      "loss": 0.7028,
+      "step": 6157
+    },
+    {
+      "epoch": 1.0984747123360985,
+      "grad_norm": 0.4233601987361908,
+      "learning_rate": 0.00021141046747685883,
+      "loss": 0.6457,
+      "step": 6158
+    },
+    {
+      "epoch": 1.0986531085540987,
+      "grad_norm": 0.5224723219871521,
+      "learning_rate": 0.00021134124641176052,
+      "loss": 0.8284,
+      "step": 6159
+    },
+    {
+      "epoch": 1.098831504772099,
+      "grad_norm": 0.4575176537036896,
+      "learning_rate": 0.00021127202838290413,
+      "loss": 0.6213,
+      "step": 6160
+    },
+    {
+      "epoch": 1.099009900990099,
+      "grad_norm": 0.4861457347869873,
+      "learning_rate": 0.00021120281339572595,
+      "loss": 0.7269,
+      "step": 6161
+    },
+    {
+      "epoch": 1.0991882972080993,
+      "grad_norm": 0.49578866362571716,
+      "learning_rate": 0.00021113360145566206,
+      "loss": 0.7596,
+      "step": 6162
+    },
+    {
+      "epoch": 1.0993666934260993,
+      "grad_norm": 0.43700939416885376,
+      "learning_rate": 0.00021106439256814844,
+      "loss": 0.6501,
+      "step": 6163
+    },
+    {
+      "epoch": 1.0995450896440995,
+      "grad_norm": 0.5099201798439026,
+      "learning_rate": 0.00021099518673862061,
+      "loss": 0.9244,
+      "step": 6164
+    },
+    {
+      "epoch": 1.0997234858620997,
+      "grad_norm": 0.434171587228775,
+      "learning_rate": 0.00021092598397251408,
+      "loss": 0.543,
+      "step": 6165
+    },
+    {
+      "epoch": 1.0999018820800999,
+      "grad_norm": 0.45645880699157715,
+      "learning_rate": 0.00021085678427526394,
+      "loss": 0.661,
+      "step": 6166
+    },
+    {
+      "epoch": 1.1000802782981,
+      "grad_norm": 0.45062491297721863,
+      "learning_rate": 0.00021078758765230514,
+      "loss": 0.5422,
+      "step": 6167
+    },
+    {
+      "epoch": 1.1002586745161003,
+      "grad_norm": 0.4871162176132202,
+      "learning_rate": 0.00021071839410907232,
+      "loss": 0.7189,
+      "step": 6168
+    },
+    {
+      "epoch": 1.1004370707341005,
+      "grad_norm": 0.3932052254676819,
+      "learning_rate": 0.00021064920365099994,
+      "loss": 0.4697,
+      "step": 6169
+    },
+    {
+      "epoch": 1.1006154669521007,
+      "grad_norm": 0.46333831548690796,
+      "learning_rate": 0.00021058001628352214,
+      "loss": 0.6093,
+      "step": 6170
+    },
+    {
+      "epoch": 1.1007938631701009,
+      "grad_norm": 0.4353841543197632,
+      "learning_rate": 0.00021051083201207297,
+      "loss": 0.555,
+      "step": 6171
+    },
+    {
+      "epoch": 1.100972259388101,
+      "grad_norm": 0.4251698851585388,
+      "learning_rate": 0.000210441650842086,
+      "loss": 0.616,
+      "step": 6172
+    },
+    {
+      "epoch": 1.101150655606101,
+      "grad_norm": 0.5128315091133118,
+      "learning_rate": 0.00021037247277899473,
+      "loss": 0.8073,
+      "step": 6173
+    },
+    {
+      "epoch": 1.1013290518241012,
+      "grad_norm": 0.49934786558151245,
+      "learning_rate": 0.00021030329782823244,
+      "loss": 0.7532,
+      "step": 6174
+    },
+    {
+      "epoch": 1.1015074480421014,
+      "grad_norm": 0.45256903767585754,
+      "learning_rate": 0.00021023412599523202,
+      "loss": 0.6656,
+      "step": 6175
+    },
+    {
+      "epoch": 1.1016858442601016,
+      "grad_norm": 0.4254659414291382,
+      "learning_rate": 0.00021016495728542626,
+      "loss": 0.5954,
+      "step": 6176
+    },
+    {
+      "epoch": 1.1018642404781018,
+      "grad_norm": 0.4867410957813263,
+      "learning_rate": 0.00021009579170424758,
+      "loss": 0.7753,
+      "step": 6177
+    },
+    {
+      "epoch": 1.102042636696102,
+      "grad_norm": 0.42600226402282715,
+      "learning_rate": 0.00021002662925712827,
+      "loss": 0.6506,
+      "step": 6178
+    },
+    {
+      "epoch": 1.1022210329141022,
+      "grad_norm": 0.4646792709827423,
+      "learning_rate": 0.00020995746994950036,
+      "loss": 0.722,
+      "step": 6179
+    },
+    {
+      "epoch": 1.1023994291321024,
+      "grad_norm": 0.4408302903175354,
+      "learning_rate": 0.00020988831378679536,
+      "loss": 0.5475,
+      "step": 6180
+    },
+    {
+      "epoch": 1.1025778253501026,
+      "grad_norm": 0.5261194705963135,
+      "learning_rate": 0.000209819160774445,
+      "loss": 0.8598,
+      "step": 6181
+    },
+    {
+      "epoch": 1.1027562215681028,
+      "grad_norm": 0.4445323944091797,
+      "learning_rate": 0.00020975001091788048,
+      "loss": 0.5923,
+      "step": 6182
+    },
+    {
+      "epoch": 1.102934617786103,
+      "grad_norm": 0.8575343489646912,
+      "learning_rate": 0.0002096808642225328,
+      "loss": 0.6453,
+      "step": 6183
+    },
+    {
+      "epoch": 1.1031130140041032,
+      "grad_norm": 0.47943076491355896,
+      "learning_rate": 0.00020961172069383275,
+      "loss": 0.6593,
+      "step": 6184
+    },
+    {
+      "epoch": 1.1032914102221032,
+      "grad_norm": 0.4768347442150116,
+      "learning_rate": 0.00020954258033721072,
+      "loss": 0.81,
+      "step": 6185
+    },
+    {
+      "epoch": 1.1034698064401034,
+      "grad_norm": 0.5110130906105042,
+      "learning_rate": 0.00020947344315809703,
+      "loss": 0.7159,
+      "step": 6186
+    },
+    {
+      "epoch": 1.1036482026581036,
+      "grad_norm": 0.5536441206932068,
+      "learning_rate": 0.00020940430916192165,
+      "loss": 0.7023,
+      "step": 6187
+    },
+    {
+      "epoch": 1.1038265988761038,
+      "grad_norm": 0.5131134986877441,
+      "learning_rate": 0.00020933517835411436,
+      "loss": 0.7933,
+      "step": 6188
+    },
+    {
+      "epoch": 1.104004995094104,
+      "grad_norm": 0.4293597936630249,
+      "learning_rate": 0.0002092660507401047,
+      "loss": 0.5803,
+      "step": 6189
+    },
+    {
+      "epoch": 1.1041833913121042,
+      "grad_norm": 0.5110434293746948,
+      "learning_rate": 0.00020919692632532182,
+      "loss": 0.7934,
+      "step": 6190
+    },
+    {
+      "epoch": 1.1043617875301044,
+      "grad_norm": 0.5271182656288147,
+      "learning_rate": 0.00020912780511519484,
+      "loss": 0.8077,
+      "step": 6191
+    },
+    {
+      "epoch": 1.1045401837481046,
+      "grad_norm": 0.4269711673259735,
+      "learning_rate": 0.00020905868711515248,
+      "loss": 0.6145,
+      "step": 6192
+    },
+    {
+      "epoch": 1.1047185799661048,
+      "grad_norm": 0.4809115529060364,
+      "learning_rate": 0.0002089895723306232,
+      "loss": 0.6951,
+      "step": 6193
+    },
+    {
+      "epoch": 1.104896976184105,
+      "grad_norm": 0.5060098171234131,
+      "learning_rate": 0.00020892046076703523,
+      "loss": 0.9445,
+      "step": 6194
+    },
+    {
+      "epoch": 1.105075372402105,
+      "grad_norm": 0.5290777683258057,
+      "learning_rate": 0.00020885135242981647,
+      "loss": 0.7443,
+      "step": 6195
+    },
+    {
+      "epoch": 1.1052537686201052,
+      "grad_norm": 0.47494322061538696,
+      "learning_rate": 0.00020878224732439493,
+      "loss": 0.7927,
+      "step": 6196
+    },
+    {
+      "epoch": 1.1054321648381054,
+      "grad_norm": 0.44407710433006287,
+      "learning_rate": 0.000208713145456198,
+      "loss": 0.5888,
+      "step": 6197
+    },
+    {
+      "epoch": 1.1056105610561056,
+      "grad_norm": 0.45321714878082275,
+      "learning_rate": 0.00020864404683065276,
+      "loss": 0.6145,
+      "step": 6198
+    },
+    {
+      "epoch": 1.1057889572741058,
+      "grad_norm": 0.4794192910194397,
+      "learning_rate": 0.00020857495145318634,
+      "loss": 0.7451,
+      "step": 6199
+    },
+    {
+      "epoch": 1.105967353492106,
+      "grad_norm": 0.4514195919036865,
+      "learning_rate": 0.0002085058593292254,
+      "loss": 0.6112,
+      "step": 6200
+    },
+    {
+      "epoch": 1.1061457497101062,
+      "grad_norm": 0.48285120725631714,
+      "learning_rate": 0.00020843677046419637,
+      "loss": 0.7315,
+      "step": 6201
+    },
+    {
+      "epoch": 1.1063241459281064,
+      "grad_norm": 0.4666329324245453,
+      "learning_rate": 0.00020836768486352553,
+      "loss": 0.5705,
+      "step": 6202
+    },
+    {
+      "epoch": 1.1065025421461065,
+      "grad_norm": 0.5958988070487976,
+      "learning_rate": 0.0002082986025326388,
+      "loss": 0.6226,
+      "step": 6203
+    },
+    {
+      "epoch": 1.1066809383641067,
+      "grad_norm": 0.5066222548484802,
+      "learning_rate": 0.00020822952347696188,
+      "loss": 0.5811,
+      "step": 6204
+    },
+    {
+      "epoch": 1.106859334582107,
+      "grad_norm": 0.4955041706562042,
+      "learning_rate": 0.00020816044770192028,
+      "loss": 0.8117,
+      "step": 6205
+    },
+    {
+      "epoch": 1.1070377308001071,
+      "grad_norm": 0.4577980637550354,
+      "learning_rate": 0.00020809137521293902,
+      "loss": 0.5482,
+      "step": 6206
+    },
+    {
+      "epoch": 1.1072161270181071,
+      "grad_norm": 0.581256091594696,
+      "learning_rate": 0.00020802230601544314,
+      "loss": 0.9406,
+      "step": 6207
+    },
+    {
+      "epoch": 1.1073945232361073,
+      "grad_norm": 0.43365752696990967,
+      "learning_rate": 0.00020795324011485728,
+      "loss": 0.5084,
+      "step": 6208
+    },
+    {
+      "epoch": 1.1075729194541075,
+      "grad_norm": 0.4679083526134491,
+      "learning_rate": 0.00020788417751660588,
+      "loss": 0.6819,
+      "step": 6209
+    },
+    {
+      "epoch": 1.1077513156721077,
+      "grad_norm": 0.41970065236091614,
+      "learning_rate": 0.00020781511822611296,
+      "loss": 0.4688,
+      "step": 6210
+    },
+    {
+      "epoch": 1.107929711890108,
+      "grad_norm": 0.4997738301753998,
+      "learning_rate": 0.00020774606224880255,
+      "loss": 0.5245,
+      "step": 6211
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 0.4666350781917572,
+      "learning_rate": 0.00020767700959009824,
+      "loss": 0.7218,
+      "step": 6212
+    },
+    {
+      "epoch": 1.1082865043261083,
+      "grad_norm": 0.4743230938911438,
+      "learning_rate": 0.0002076079602554234,
+      "loss": 0.6804,
+      "step": 6213
+    },
+    {
+      "epoch": 1.1084649005441085,
+      "grad_norm": 0.5221150517463684,
+      "learning_rate": 0.0002075389142502011,
+      "loss": 0.8767,
+      "step": 6214
+    },
+    {
+      "epoch": 1.1086432967621087,
+      "grad_norm": 0.4845503866672516,
+      "learning_rate": 0.0002074698715798542,
+      "loss": 0.6975,
+      "step": 6215
+    },
+    {
+      "epoch": 1.108821692980109,
+      "grad_norm": 0.424908846616745,
+      "learning_rate": 0.0002074008322498053,
+      "loss": 0.4744,
+      "step": 6216
+    },
+    {
+      "epoch": 1.1090000891981089,
+      "grad_norm": 0.44817301630973816,
+      "learning_rate": 0.00020733179626547667,
+      "loss": 0.5131,
+      "step": 6217
+    },
+    {
+      "epoch": 1.109178485416109,
+      "grad_norm": 0.49306273460388184,
+      "learning_rate": 0.0002072627636322905,
+      "loss": 0.7012,
+      "step": 6218
+    },
+    {
+      "epoch": 1.1093568816341093,
+      "grad_norm": 0.4819309413433075,
+      "learning_rate": 0.00020719373435566842,
+      "loss": 0.6232,
+      "step": 6219
+    },
+    {
+      "epoch": 1.1095352778521095,
+      "grad_norm": 0.5482651591300964,
+      "learning_rate": 0.00020712470844103198,
+      "loss": 0.8316,
+      "step": 6220
+    },
+    {
+      "epoch": 1.1097136740701097,
+      "grad_norm": 0.47659358382225037,
+      "learning_rate": 0.00020705568589380252,
+      "loss": 0.7383,
+      "step": 6221
+    },
+    {
+      "epoch": 1.1098920702881099,
+      "grad_norm": 0.4586566686630249,
+      "learning_rate": 0.00020698666671940103,
+      "loss": 0.6438,
+      "step": 6222
+    },
+    {
+      "epoch": 1.11007046650611,
+      "grad_norm": 0.5156270861625671,
+      "learning_rate": 0.0002069176509232482,
+      "loss": 0.7463,
+      "step": 6223
+    },
+    {
+      "epoch": 1.1102488627241103,
+      "grad_norm": 0.4540422260761261,
+      "learning_rate": 0.0002068486385107645,
+      "loss": 0.7645,
+      "step": 6224
+    },
+    {
+      "epoch": 1.1104272589421105,
+      "grad_norm": 0.39585331082344055,
+      "learning_rate": 0.00020677962948737022,
+      "loss": 0.5105,
+      "step": 6225
+    },
+    {
+      "epoch": 1.1106056551601107,
+      "grad_norm": 0.4956061840057373,
+      "learning_rate": 0.00020671062385848517,
+      "loss": 0.7741,
+      "step": 6226
+    },
+    {
+      "epoch": 1.1107840513781109,
+      "grad_norm": 0.45627540349960327,
+      "learning_rate": 0.00020664162162952913,
+      "loss": 0.6525,
+      "step": 6227
+    },
+    {
+      "epoch": 1.110962447596111,
+      "grad_norm": 0.45656707882881165,
+      "learning_rate": 0.00020657262280592147,
+      "loss": 0.6371,
+      "step": 6228
+    },
+    {
+      "epoch": 1.111140843814111,
+      "grad_norm": 0.5140594840049744,
+      "learning_rate": 0.0002065036273930813,
+      "loss": 0.8361,
+      "step": 6229
+    },
+    {
+      "epoch": 1.1113192400321112,
+      "grad_norm": 0.4385344386100769,
+      "learning_rate": 0.00020643463539642766,
+      "loss": 0.5181,
+      "step": 6230
+    },
+    {
+      "epoch": 1.1114976362501114,
+      "grad_norm": 0.5141750574111938,
+      "learning_rate": 0.00020636564682137887,
+      "loss": 0.5044,
+      "step": 6231
+    },
+    {
+      "epoch": 1.1116760324681116,
+      "grad_norm": 0.5330828428268433,
+      "learning_rate": 0.00020629666167335342,
+      "loss": 0.7067,
+      "step": 6232
+    },
+    {
+      "epoch": 1.1118544286861118,
+      "grad_norm": 0.5701097249984741,
+      "learning_rate": 0.00020622767995776936,
+      "loss": 0.8013,
+      "step": 6233
+    },
+    {
+      "epoch": 1.112032824904112,
+      "grad_norm": 0.5556646585464478,
+      "learning_rate": 0.00020615870168004449,
+      "loss": 0.8375,
+      "step": 6234
+    },
+    {
+      "epoch": 1.1122112211221122,
+      "grad_norm": 0.5408580303192139,
+      "learning_rate": 0.0002060897268455963,
+      "loss": 0.5886,
+      "step": 6235
+    },
+    {
+      "epoch": 1.1123896173401124,
+      "grad_norm": 0.4653705954551697,
+      "learning_rate": 0.0002060207554598421,
+      "loss": 0.6007,
+      "step": 6236
+    },
+    {
+      "epoch": 1.1125680135581126,
+      "grad_norm": 0.485100656747818,
+      "learning_rate": 0.00020595178752819883,
+      "loss": 0.6615,
+      "step": 6237
+    },
+    {
+      "epoch": 1.1127464097761128,
+      "grad_norm": 0.4773385226726532,
+      "learning_rate": 0.00020588282305608325,
+      "loss": 0.7062,
+      "step": 6238
+    },
+    {
+      "epoch": 1.1129248059941128,
+      "grad_norm": 0.538426399230957,
+      "learning_rate": 0.00020581386204891172,
+      "loss": 0.8382,
+      "step": 6239
+    },
+    {
+      "epoch": 1.113103202212113,
+      "grad_norm": 0.45640110969543457,
+      "learning_rate": 0.00020574490451210045,
+      "loss": 0.6707,
+      "step": 6240
+    },
+    {
+      "epoch": 1.1132815984301132,
+      "grad_norm": 0.5065681338310242,
+      "learning_rate": 0.00020567595045106523,
+      "loss": 0.7169,
+      "step": 6241
+    },
+    {
+      "epoch": 1.1134599946481134,
+      "grad_norm": 0.49156734347343445,
+      "learning_rate": 0.0002056069998712219,
+      "loss": 0.6726,
+      "step": 6242
+    },
+    {
+      "epoch": 1.1136383908661136,
+      "grad_norm": 0.4353838860988617,
+      "learning_rate": 0.00020553805277798574,
+      "loss": 0.5635,
+      "step": 6243
+    },
+    {
+      "epoch": 1.1138167870841138,
+      "grad_norm": 0.5049417614936829,
+      "learning_rate": 0.00020546910917677172,
+      "loss": 0.8726,
+      "step": 6244
+    },
+    {
+      "epoch": 1.113995183302114,
+      "grad_norm": 0.46186742186546326,
+      "learning_rate": 0.00020540016907299473,
+      "loss": 0.5505,
+      "step": 6245
+    },
+    {
+      "epoch": 1.1141735795201142,
+      "grad_norm": 0.38499900698661804,
+      "learning_rate": 0.0002053312324720692,
+      "loss": 0.4884,
+      "step": 6246
+    },
+    {
+      "epoch": 1.1143519757381144,
+      "grad_norm": 0.4557759165763855,
+      "learning_rate": 0.00020526229937940956,
+      "loss": 0.5865,
+      "step": 6247
+    },
+    {
+      "epoch": 1.1145303719561146,
+      "grad_norm": 0.4711809754371643,
+      "learning_rate": 0.00020519336980042956,
+      "loss": 0.6212,
+      "step": 6248
+    },
+    {
+      "epoch": 1.1147087681741148,
+      "grad_norm": 0.5073553323745728,
+      "learning_rate": 0.00020512444374054309,
+      "loss": 0.7891,
+      "step": 6249
+    },
+    {
+      "epoch": 1.114887164392115,
+      "grad_norm": 0.5361490845680237,
+      "learning_rate": 0.00020505552120516347,
+      "loss": 0.8727,
+      "step": 6250
+    },
+    {
+      "epoch": 1.115065560610115,
+      "grad_norm": 0.49489954113960266,
+      "learning_rate": 0.00020498660219970394,
+      "loss": 0.6391,
+      "step": 6251
+    },
+    {
+      "epoch": 1.1152439568281152,
+      "grad_norm": 0.44518017768859863,
+      "learning_rate": 0.00020491768672957722,
+      "loss": 0.6854,
+      "step": 6252
+    },
+    {
+      "epoch": 1.1154223530461154,
+      "grad_norm": 0.501320481300354,
+      "learning_rate": 0.00020484877480019602,
+      "loss": 0.8994,
+      "step": 6253
+    },
+    {
+      "epoch": 1.1156007492641156,
+      "grad_norm": 0.5842064023017883,
+      "learning_rate": 0.0002047798664169726,
+      "loss": 0.988,
+      "step": 6254
+    },
+    {
+      "epoch": 1.1157791454821158,
+      "grad_norm": 0.4702877998352051,
+      "learning_rate": 0.000204710961585319,
+      "loss": 0.8212,
+      "step": 6255
+    },
+    {
+      "epoch": 1.115957541700116,
+      "grad_norm": 0.4764789640903473,
+      "learning_rate": 0.00020464206031064694,
+      "loss": 0.6702,
+      "step": 6256
+    },
+    {
+      "epoch": 1.1161359379181162,
+      "grad_norm": 0.4664061367511749,
+      "learning_rate": 0.000204573162598368,
+      "loss": 0.6985,
+      "step": 6257
+    },
+    {
+      "epoch": 1.1163143341361164,
+      "grad_norm": 1.0998966693878174,
+      "learning_rate": 0.00020450426845389333,
+      "loss": 0.6394,
+      "step": 6258
+    },
+    {
+      "epoch": 1.1164927303541166,
+      "grad_norm": 0.4073130786418915,
+      "learning_rate": 0.00020443537788263384,
+      "loss": 0.5515,
+      "step": 6259
+    },
+    {
+      "epoch": 1.1166711265721168,
+      "grad_norm": 0.48286718130111694,
+      "learning_rate": 0.00020436649089000013,
+      "loss": 0.7877,
+      "step": 6260
+    },
+    {
+      "epoch": 1.1168495227901167,
+      "grad_norm": 0.505884051322937,
+      "learning_rate": 0.00020429760748140262,
+      "loss": 0.7128,
+      "step": 6261
+    },
+    {
+      "epoch": 1.117027919008117,
+      "grad_norm": 0.6370548605918884,
+      "learning_rate": 0.00020422872766225137,
+      "loss": 0.9009,
+      "step": 6262
+    },
+    {
+      "epoch": 1.1172063152261171,
+      "grad_norm": 0.4987682104110718,
+      "learning_rate": 0.00020415985143795612,
+      "loss": 0.6506,
+      "step": 6263
+    },
+    {
+      "epoch": 1.1173847114441173,
+      "grad_norm": 0.7356500029563904,
+      "learning_rate": 0.00020409097881392646,
+      "loss": 0.7871,
+      "step": 6264
+    },
+    {
+      "epoch": 1.1175631076621175,
+      "grad_norm": 0.578613817691803,
+      "learning_rate": 0.00020402210979557153,
+      "loss": 0.7906,
+      "step": 6265
+    },
+    {
+      "epoch": 1.1177415038801177,
+      "grad_norm": 0.4457342028617859,
+      "learning_rate": 0.00020395324438830033,
+      "loss": 0.6093,
+      "step": 6266
+    },
+    {
+      "epoch": 1.117919900098118,
+      "grad_norm": 0.4944089949131012,
+      "learning_rate": 0.00020388438259752147,
+      "loss": 0.7959,
+      "step": 6267
+    },
+    {
+      "epoch": 1.1180982963161181,
+      "grad_norm": 0.46992331743240356,
+      "learning_rate": 0.00020381552442864337,
+      "loss": 0.6263,
+      "step": 6268
+    },
+    {
+      "epoch": 1.1182766925341183,
+      "grad_norm": 0.42646077275276184,
+      "learning_rate": 0.00020374666988707407,
+      "loss": 0.5327,
+      "step": 6269
+    },
+    {
+      "epoch": 1.1184550887521185,
+      "grad_norm": 0.48363184928894043,
+      "learning_rate": 0.00020367781897822146,
+      "loss": 0.7761,
+      "step": 6270
+    },
+    {
+      "epoch": 1.1186334849701187,
+      "grad_norm": 0.47337979078292847,
+      "learning_rate": 0.00020360897170749299,
+      "loss": 0.6731,
+      "step": 6271
+    },
+    {
+      "epoch": 1.118811881188119,
+      "grad_norm": 0.4817976653575897,
+      "learning_rate": 0.00020354012808029587,
+      "loss": 0.6265,
+      "step": 6272
+    },
+    {
+      "epoch": 1.118990277406119,
+      "grad_norm": 0.45678937435150146,
+      "learning_rate": 0.00020347128810203717,
+      "loss": 0.6553,
+      "step": 6273
+    },
+    {
+      "epoch": 1.119168673624119,
+      "grad_norm": 0.483571857213974,
+      "learning_rate": 0.00020340245177812344,
+      "loss": 0.746,
+      "step": 6274
+    },
+    {
+      "epoch": 1.1193470698421193,
+      "grad_norm": 0.4845406115055084,
+      "learning_rate": 0.00020333361911396112,
+      "loss": 0.6712,
+      "step": 6275
+    },
+    {
+      "epoch": 1.1195254660601195,
+      "grad_norm": 0.464175820350647,
+      "learning_rate": 0.00020326479011495627,
+      "loss": 0.5484,
+      "step": 6276
+    },
+    {
+      "epoch": 1.1197038622781197,
+      "grad_norm": 0.48887899518013,
+      "learning_rate": 0.00020319596478651477,
+      "loss": 0.6694,
+      "step": 6277
+    },
+    {
+      "epoch": 1.1198822584961199,
+      "grad_norm": 0.4447461664676666,
+      "learning_rate": 0.00020312714313404197,
+      "loss": 0.4637,
+      "step": 6278
+    },
+    {
+      "epoch": 1.12006065471412,
+      "grad_norm": 0.5541086792945862,
+      "learning_rate": 0.00020305832516294314,
+      "loss": 0.7051,
+      "step": 6279
+    },
+    {
+      "epoch": 1.1202390509321203,
+      "grad_norm": 0.43977195024490356,
+      "learning_rate": 0.00020298951087862333,
+      "loss": 0.5394,
+      "step": 6280
+    },
+    {
+      "epoch": 1.1204174471501205,
+      "grad_norm": 0.5157051682472229,
+      "learning_rate": 0.00020292070028648707,
+      "loss": 0.7626,
+      "step": 6281
+    },
+    {
+      "epoch": 1.1205958433681207,
+      "grad_norm": 0.4950587749481201,
+      "learning_rate": 0.00020285189339193873,
+      "loss": 0.704,
+      "step": 6282
+    },
+    {
+      "epoch": 1.1207742395861207,
+      "grad_norm": 0.4870520532131195,
+      "learning_rate": 0.0002027830902003824,
+      "loss": 0.6548,
+      "step": 6283
+    },
+    {
+      "epoch": 1.1209526358041209,
+      "grad_norm": 0.5210698246955872,
+      "learning_rate": 0.00020271429071722186,
+      "loss": 0.6496,
+      "step": 6284
+    },
+    {
+      "epoch": 1.121131032022121,
+      "grad_norm": 0.4650278091430664,
+      "learning_rate": 0.00020264549494786066,
+      "loss": 0.5775,
+      "step": 6285
+    },
+    {
+      "epoch": 1.1213094282401213,
+      "grad_norm": 0.5056297779083252,
+      "learning_rate": 0.00020257670289770181,
+      "loss": 0.7124,
+      "step": 6286
+    },
+    {
+      "epoch": 1.1214878244581215,
+      "grad_norm": 0.4929437041282654,
+      "learning_rate": 0.00020250791457214823,
+      "loss": 0.9018,
+      "step": 6287
+    },
+    {
+      "epoch": 1.1216662206761217,
+      "grad_norm": 0.5105372071266174,
+      "learning_rate": 0.0002024391299766027,
+      "loss": 0.7514,
+      "step": 6288
+    },
+    {
+      "epoch": 1.1218446168941218,
+      "grad_norm": 0.4985578656196594,
+      "learning_rate": 0.00020237034911646745,
+      "loss": 0.6705,
+      "step": 6289
+    },
+    {
+      "epoch": 1.122023013112122,
+      "grad_norm": 0.5785099864006042,
+      "learning_rate": 0.0002023015719971445,
+      "loss": 1.1394,
+      "step": 6290
+    },
+    {
+      "epoch": 1.1222014093301222,
+      "grad_norm": 0.5343894362449646,
+      "learning_rate": 0.0002022327986240355,
+      "loss": 0.8509,
+      "step": 6291
+    },
+    {
+      "epoch": 1.1223798055481224,
+      "grad_norm": 0.4681146442890167,
+      "learning_rate": 0.00020216402900254197,
+      "loss": 0.5766,
+      "step": 6292
+    },
+    {
+      "epoch": 1.1225582017661226,
+      "grad_norm": 0.4778856039047241,
+      "learning_rate": 0.000202095263138065,
+      "loss": 0.8314,
+      "step": 6293
+    },
+    {
+      "epoch": 1.1227365979841228,
+      "grad_norm": 0.44656750559806824,
+      "learning_rate": 0.00020202650103600544,
+      "loss": 0.6106,
+      "step": 6294
+    },
+    {
+      "epoch": 1.1229149942021228,
+      "grad_norm": 0.5104446411132812,
+      "learning_rate": 0.00020195774270176386,
+      "loss": 0.6599,
+      "step": 6295
+    },
+    {
+      "epoch": 1.123093390420123,
+      "grad_norm": 0.44691547751426697,
+      "learning_rate": 0.0002018889881407405,
+      "loss": 0.5945,
+      "step": 6296
+    },
+    {
+      "epoch": 1.1232717866381232,
+      "grad_norm": 0.5005384683609009,
+      "learning_rate": 0.00020182023735833531,
+      "loss": 0.5805,
+      "step": 6297
+    },
+    {
+      "epoch": 1.1234501828561234,
+      "grad_norm": 0.5261300802230835,
+      "learning_rate": 0.000201751490359948,
+      "loss": 0.825,
+      "step": 6298
+    },
+    {
+      "epoch": 1.1236285790741236,
+      "grad_norm": 0.47424525022506714,
+      "learning_rate": 0.00020168274715097782,
+      "loss": 0.6754,
+      "step": 6299
+    },
+    {
+      "epoch": 1.1238069752921238,
+      "grad_norm": 0.48731860518455505,
+      "learning_rate": 0.00020161400773682387,
+      "loss": 0.6607,
+      "step": 6300
+    },
+    {
+      "epoch": 1.123985371510124,
+      "grad_norm": 0.4794895350933075,
+      "learning_rate": 0.00020154527212288493,
+      "loss": 0.6262,
+      "step": 6301
+    },
+    {
+      "epoch": 1.1241637677281242,
+      "grad_norm": 0.4823654592037201,
+      "learning_rate": 0.00020147654031455942,
+      "loss": 0.6762,
+      "step": 6302
+    },
+    {
+      "epoch": 1.1243421639461244,
+      "grad_norm": 0.5071163177490234,
+      "learning_rate": 0.0002014078123172456,
+      "loss": 0.7975,
+      "step": 6303
+    },
+    {
+      "epoch": 1.1245205601641246,
+      "grad_norm": 0.4861220419406891,
+      "learning_rate": 0.0002013390881363413,
+      "loss": 0.6742,
+      "step": 6304
+    },
+    {
+      "epoch": 1.1246989563821246,
+      "grad_norm": 0.5352680087089539,
+      "learning_rate": 0.00020127036777724407,
+      "loss": 0.7588,
+      "step": 6305
+    },
+    {
+      "epoch": 1.1248773526001248,
+      "grad_norm": 0.4422907829284668,
+      "learning_rate": 0.00020120165124535119,
+      "loss": 0.6386,
+      "step": 6306
+    },
+    {
+      "epoch": 1.125055748818125,
+      "grad_norm": 0.4726184606552124,
+      "learning_rate": 0.00020113293854605963,
+      "loss": 0.6324,
+      "step": 6307
+    },
+    {
+      "epoch": 1.1252341450361252,
+      "grad_norm": 0.5098016858100891,
+      "learning_rate": 0.00020106422968476604,
+      "loss": 0.7632,
+      "step": 6308
+    },
+    {
+      "epoch": 1.1254125412541254,
+      "grad_norm": 0.4972569942474365,
+      "learning_rate": 0.00020099552466686677,
+      "loss": 0.6985,
+      "step": 6309
+    },
+    {
+      "epoch": 1.1255909374721256,
+      "grad_norm": 0.4584605097770691,
+      "learning_rate": 0.00020092682349775797,
+      "loss": 0.6985,
+      "step": 6310
+    },
+    {
+      "epoch": 1.1257693336901258,
+      "grad_norm": 0.49379658699035645,
+      "learning_rate": 0.0002008581261828353,
+      "loss": 0.5424,
+      "step": 6311
+    },
+    {
+      "epoch": 1.125947729908126,
+      "grad_norm": 0.5629817247390747,
+      "learning_rate": 0.00020078943272749426,
+      "loss": 0.6988,
+      "step": 6312
+    },
+    {
+      "epoch": 1.1261261261261262,
+      "grad_norm": 0.49711212515830994,
+      "learning_rate": 0.00020072074313712995,
+      "loss": 0.6466,
+      "step": 6313
+    },
+    {
+      "epoch": 1.1263045223441264,
+      "grad_norm": 0.5518889427185059,
+      "learning_rate": 0.00020065205741713732,
+      "loss": 0.7526,
+      "step": 6314
+    },
+    {
+      "epoch": 1.1264829185621266,
+      "grad_norm": 0.44839757680892944,
+      "learning_rate": 0.00020058337557291085,
+      "loss": 0.6743,
+      "step": 6315
+    },
+    {
+      "epoch": 1.1266613147801268,
+      "grad_norm": 0.4963269531726837,
+      "learning_rate": 0.00020051469760984475,
+      "loss": 0.7584,
+      "step": 6316
+    },
+    {
+      "epoch": 1.126839710998127,
+      "grad_norm": 0.4421939551830292,
+      "learning_rate": 0.00020044602353333304,
+      "loss": 0.6155,
+      "step": 6317
+    },
+    {
+      "epoch": 1.127018107216127,
+      "grad_norm": 0.47304272651672363,
+      "learning_rate": 0.00020037735334876928,
+      "loss": 0.843,
+      "step": 6318
+    },
+    {
+      "epoch": 1.1271965034341271,
+      "grad_norm": 0.4691406786441803,
+      "learning_rate": 0.00020030868706154688,
+      "loss": 0.8837,
+      "step": 6319
+    },
+    {
+      "epoch": 1.1273748996521273,
+      "grad_norm": 0.470851868391037,
+      "learning_rate": 0.00020024002467705878,
+      "loss": 0.7696,
+      "step": 6320
+    },
+    {
+      "epoch": 1.1275532958701275,
+      "grad_norm": 0.5345527529716492,
+      "learning_rate": 0.00020017136620069777,
+      "loss": 0.9637,
+      "step": 6321
+    },
+    {
+      "epoch": 1.1277316920881277,
+      "grad_norm": 0.47198501229286194,
+      "learning_rate": 0.00020010271163785622,
+      "loss": 0.705,
+      "step": 6322
+    },
+    {
+      "epoch": 1.127910088306128,
+      "grad_norm": 0.5204360485076904,
+      "learning_rate": 0.00020003406099392625,
+      "loss": 0.6943,
+      "step": 6323
+    },
+    {
+      "epoch": 1.1280884845241281,
+      "grad_norm": 0.49591392278671265,
+      "learning_rate": 0.00019996541427429957,
+      "loss": 0.6205,
+      "step": 6324
+    },
+    {
+      "epoch": 1.1282668807421283,
+      "grad_norm": 0.45050108432769775,
+      "learning_rate": 0.0001998967714843677,
+      "loss": 0.5172,
+      "step": 6325
+    },
+    {
+      "epoch": 1.1284452769601285,
+      "grad_norm": 0.45870310068130493,
+      "learning_rate": 0.0001998281326295219,
+      "loss": 0.6089,
+      "step": 6326
+    },
+    {
+      "epoch": 1.1286236731781285,
+      "grad_norm": 0.4750981032848358,
+      "learning_rate": 0.00019975949771515296,
+      "loss": 0.6546,
+      "step": 6327
+    },
+    {
+      "epoch": 1.1288020693961287,
+      "grad_norm": 1.4412025213241577,
+      "learning_rate": 0.0001996908667466514,
+      "loss": 0.7441,
+      "step": 6328
+    },
+    {
+      "epoch": 1.128980465614129,
+      "grad_norm": 0.49182528257369995,
+      "learning_rate": 0.00019962223972940757,
+      "loss": 0.6518,
+      "step": 6329
+    },
+    {
+      "epoch": 1.129158861832129,
+      "grad_norm": 0.5143123269081116,
+      "learning_rate": 0.00019955361666881133,
+      "loss": 0.6357,
+      "step": 6330
+    },
+    {
+      "epoch": 1.1293372580501293,
+      "grad_norm": 1.9303398132324219,
+      "learning_rate": 0.00019948499757025239,
+      "loss": 0.5802,
+      "step": 6331
+    },
+    {
+      "epoch": 1.1295156542681295,
+      "grad_norm": 0.5342963337898254,
+      "learning_rate": 0.00019941638243911993,
+      "loss": 0.6309,
+      "step": 6332
+    },
+    {
+      "epoch": 1.1296940504861297,
+      "grad_norm": 0.4394116997718811,
+      "learning_rate": 0.00019934777128080292,
+      "loss": 0.6286,
+      "step": 6333
+    },
+    {
+      "epoch": 1.12987244670413,
+      "grad_norm": 0.5031057596206665,
+      "learning_rate": 0.00019927916410069027,
+      "loss": 0.6877,
+      "step": 6334
+    },
+    {
+      "epoch": 1.13005084292213,
+      "grad_norm": 0.4916764795780182,
+      "learning_rate": 0.00019921056090417026,
+      "loss": 0.7054,
+      "step": 6335
+    },
+    {
+      "epoch": 1.1302292391401303,
+      "grad_norm": 0.535476803779602,
+      "learning_rate": 0.00019914196169663095,
+      "loss": 0.7493,
+      "step": 6336
+    },
+    {
+      "epoch": 1.1304076353581305,
+      "grad_norm": 0.5123772025108337,
+      "learning_rate": 0.00019907336648346008,
+      "loss": 0.6594,
+      "step": 6337
+    },
+    {
+      "epoch": 1.1305860315761307,
+      "grad_norm": 0.5296303629875183,
+      "learning_rate": 0.00019900477527004507,
+      "loss": 0.7239,
+      "step": 6338
+    },
+    {
+      "epoch": 1.1307644277941309,
+      "grad_norm": 0.5358021259307861,
+      "learning_rate": 0.00019893618806177306,
+      "loss": 0.9083,
+      "step": 6339
+    },
+    {
+      "epoch": 1.1309428240121309,
+      "grad_norm": 0.47676989436149597,
+      "learning_rate": 0.00019886760486403088,
+      "loss": 0.5314,
+      "step": 6340
+    },
+    {
+      "epoch": 1.131121220230131,
+      "grad_norm": 0.5259501934051514,
+      "learning_rate": 0.00019879902568220497,
+      "loss": 0.7819,
+      "step": 6341
+    },
+    {
+      "epoch": 1.1312996164481313,
+      "grad_norm": 0.4975660741329193,
+      "learning_rate": 0.00019873045052168158,
+      "loss": 0.6725,
+      "step": 6342
+    },
+    {
+      "epoch": 1.1314780126661315,
+      "grad_norm": 0.5157343149185181,
+      "learning_rate": 0.00019866187938784657,
+      "loss": 0.6001,
+      "step": 6343
+    },
+    {
+      "epoch": 1.1316564088841317,
+      "grad_norm": 0.43325984477996826,
+      "learning_rate": 0.00019859331228608547,
+      "loss": 0.6975,
+      "step": 6344
+    },
+    {
+      "epoch": 1.1318348051021319,
+      "grad_norm": 0.49547433853149414,
+      "learning_rate": 0.0001985247492217835,
+      "loss": 0.6571,
+      "step": 6345
+    },
+    {
+      "epoch": 1.132013201320132,
+      "grad_norm": 0.49157583713531494,
+      "learning_rate": 0.00019845619020032553,
+      "loss": 0.7212,
+      "step": 6346
+    },
+    {
+      "epoch": 1.1321915975381323,
+      "grad_norm": 0.48966532945632935,
+      "learning_rate": 0.0001983876352270962,
+      "loss": 0.5914,
+      "step": 6347
+    },
+    {
+      "epoch": 1.1323699937561325,
+      "grad_norm": 0.5003472566604614,
+      "learning_rate": 0.0001983190843074797,
+      "loss": 0.5599,
+      "step": 6348
+    },
+    {
+      "epoch": 1.1325483899741324,
+      "grad_norm": 0.48509982228279114,
+      "learning_rate": 0.00019825053744686023,
+      "loss": 0.6925,
+      "step": 6349
+    },
+    {
+      "epoch": 1.1327267861921326,
+      "grad_norm": 0.4695090651512146,
+      "learning_rate": 0.00019818199465062122,
+      "loss": 0.7031,
+      "step": 6350
+    },
+    {
+      "epoch": 1.1329051824101328,
+      "grad_norm": 0.494907408952713,
+      "learning_rate": 0.00019811345592414607,
+      "loss": 0.6428,
+      "step": 6351
+    },
+    {
+      "epoch": 1.133083578628133,
+      "grad_norm": 0.5415506958961487,
+      "learning_rate": 0.00019804492127281772,
+      "loss": 0.9687,
+      "step": 6352
+    },
+    {
+      "epoch": 1.1332619748461332,
+      "grad_norm": 0.45336905121803284,
+      "learning_rate": 0.00019797639070201896,
+      "loss": 0.6972,
+      "step": 6353
+    },
+    {
+      "epoch": 1.1334403710641334,
+      "grad_norm": 0.4650866687297821,
+      "learning_rate": 0.00019790786421713204,
+      "loss": 0.6297,
+      "step": 6354
+    },
+    {
+      "epoch": 1.1336187672821336,
+      "grad_norm": 0.4537782669067383,
+      "learning_rate": 0.00019783934182353904,
+      "loss": 0.5633,
+      "step": 6355
+    },
+    {
+      "epoch": 1.1337971635001338,
+      "grad_norm": 0.45536094903945923,
+      "learning_rate": 0.00019777082352662173,
+      "loss": 0.6543,
+      "step": 6356
+    },
+    {
+      "epoch": 1.133975559718134,
+      "grad_norm": 0.4556442201137543,
+      "learning_rate": 0.00019770230933176147,
+      "loss": 0.6048,
+      "step": 6357
+    },
+    {
+      "epoch": 1.1341539559361342,
+      "grad_norm": 0.44016602635383606,
+      "learning_rate": 0.00019763379924433934,
+      "loss": 0.6132,
+      "step": 6358
+    },
+    {
+      "epoch": 1.1343323521541344,
+      "grad_norm": 0.5556267499923706,
+      "learning_rate": 0.00019756529326973602,
+      "loss": 0.8597,
+      "step": 6359
+    },
+    {
+      "epoch": 1.1345107483721346,
+      "grad_norm": 0.5245857834815979,
+      "learning_rate": 0.00019749679141333205,
+      "loss": 0.6483,
+      "step": 6360
+    },
+    {
+      "epoch": 1.1346891445901348,
+      "grad_norm": 0.4881051182746887,
+      "learning_rate": 0.00019742829368050744,
+      "loss": 0.6523,
+      "step": 6361
+    },
+    {
+      "epoch": 1.1348675408081348,
+      "grad_norm": 0.4933343231678009,
+      "learning_rate": 0.00019735980007664207,
+      "loss": 0.6629,
+      "step": 6362
+    },
+    {
+      "epoch": 1.135045937026135,
+      "grad_norm": 0.4620760977268219,
+      "learning_rate": 0.0001972913106071153,
+      "loss": 0.6428,
+      "step": 6363
+    },
+    {
+      "epoch": 1.1352243332441352,
+      "grad_norm": 0.5003635883331299,
+      "learning_rate": 0.00019722282527730628,
+      "loss": 0.6564,
+      "step": 6364
+    },
+    {
+      "epoch": 1.1354027294621354,
+      "grad_norm": 0.41504591703414917,
+      "learning_rate": 0.0001971543440925939,
+      "loss": 0.441,
+      "step": 6365
+    },
+    {
+      "epoch": 1.1355811256801356,
+      "grad_norm": 0.49655354022979736,
+      "learning_rate": 0.0001970858670583566,
+      "loss": 0.6645,
+      "step": 6366
+    },
+    {
+      "epoch": 1.1357595218981358,
+      "grad_norm": 0.4752812087535858,
+      "learning_rate": 0.00019701739417997256,
+      "loss": 0.6179,
+      "step": 6367
+    },
+    {
+      "epoch": 1.135937918116136,
+      "grad_norm": 0.46678200364112854,
+      "learning_rate": 0.00019694892546281954,
+      "loss": 0.5577,
+      "step": 6368
+    },
+    {
+      "epoch": 1.1361163143341362,
+      "grad_norm": 0.5003397464752197,
+      "learning_rate": 0.0001968804609122751,
+      "loss": 0.6509,
+      "step": 6369
+    },
+    {
+      "epoch": 1.1362947105521364,
+      "grad_norm": 0.5790354013442993,
+      "learning_rate": 0.00019681200053371645,
+      "loss": 0.6953,
+      "step": 6370
+    },
+    {
+      "epoch": 1.1364731067701364,
+      "grad_norm": 0.5248673558235168,
+      "learning_rate": 0.00019674354433252034,
+      "loss": 0.6928,
+      "step": 6371
+    },
+    {
+      "epoch": 1.1366515029881366,
+      "grad_norm": 0.5383015871047974,
+      "learning_rate": 0.00019667509231406332,
+      "loss": 0.6522,
+      "step": 6372
+    },
+    {
+      "epoch": 1.1368298992061368,
+      "grad_norm": 0.5394713282585144,
+      "learning_rate": 0.00019660664448372162,
+      "loss": 0.6593,
+      "step": 6373
+    },
+    {
+      "epoch": 1.137008295424137,
+      "grad_norm": 0.5048239827156067,
+      "learning_rate": 0.00019653820084687107,
+      "loss": 0.8368,
+      "step": 6374
+    },
+    {
+      "epoch": 1.1371866916421371,
+      "grad_norm": 0.4014268219470978,
+      "learning_rate": 0.00019646976140888725,
+      "loss": 0.5085,
+      "step": 6375
+    },
+    {
+      "epoch": 1.1373650878601373,
+      "grad_norm": 0.4501616954803467,
+      "learning_rate": 0.00019640132617514534,
+      "loss": 0.5781,
+      "step": 6376
+    },
+    {
+      "epoch": 1.1375434840781375,
+      "grad_norm": 0.5049312114715576,
+      "learning_rate": 0.00019633289515102017,
+      "loss": 0.8279,
+      "step": 6377
+    },
+    {
+      "epoch": 1.1377218802961377,
+      "grad_norm": 0.4441901743412018,
+      "learning_rate": 0.00019626446834188638,
+      "loss": 0.58,
+      "step": 6378
+    },
+    {
+      "epoch": 1.137900276514138,
+      "grad_norm": 0.4739856719970703,
+      "learning_rate": 0.00019619604575311797,
+      "loss": 0.755,
+      "step": 6379
+    },
+    {
+      "epoch": 1.1380786727321381,
+      "grad_norm": 0.4583461880683899,
+      "learning_rate": 0.0001961276273900891,
+      "loss": 0.5062,
+      "step": 6380
+    },
+    {
+      "epoch": 1.1382570689501383,
+      "grad_norm": 0.4415271282196045,
+      "learning_rate": 0.00019605921325817317,
+      "loss": 0.5786,
+      "step": 6381
+    },
+    {
+      "epoch": 1.1384354651681385,
+      "grad_norm": 0.5233315229415894,
+      "learning_rate": 0.00019599080336274343,
+      "loss": 0.8495,
+      "step": 6382
+    },
+    {
+      "epoch": 1.1386138613861387,
+      "grad_norm": 0.5158662796020508,
+      "learning_rate": 0.00019592239770917276,
+      "loss": 0.7314,
+      "step": 6383
+    },
+    {
+      "epoch": 1.1387922576041387,
+      "grad_norm": 0.6792814135551453,
+      "learning_rate": 0.00019585399630283367,
+      "loss": 0.7851,
+      "step": 6384
+    },
+    {
+      "epoch": 1.138970653822139,
+      "grad_norm": 0.44133511185646057,
+      "learning_rate": 0.00019578559914909844,
+      "loss": 0.6067,
+      "step": 6385
+    },
+    {
+      "epoch": 1.139149050040139,
+      "grad_norm": 0.4633074700832367,
+      "learning_rate": 0.00019571720625333888,
+      "loss": 0.6206,
+      "step": 6386
+    },
+    {
+      "epoch": 1.1393274462581393,
+      "grad_norm": 0.7552552223205566,
+      "learning_rate": 0.00019564881762092662,
+      "loss": 0.659,
+      "step": 6387
+    },
+    {
+      "epoch": 1.1395058424761395,
+      "grad_norm": 0.5166911482810974,
+      "learning_rate": 0.00019558043325723282,
+      "loss": 0.6211,
+      "step": 6388
+    },
+    {
+      "epoch": 1.1396842386941397,
+      "grad_norm": 0.5080016851425171,
+      "learning_rate": 0.00019551205316762838,
+      "loss": 0.8162,
+      "step": 6389
+    },
+    {
+      "epoch": 1.13986263491214,
+      "grad_norm": 0.5829038619995117,
+      "learning_rate": 0.00019544367735748388,
+      "loss": 0.9879,
+      "step": 6390
+    },
+    {
+      "epoch": 1.14004103113014,
+      "grad_norm": 0.49968674778938293,
+      "learning_rate": 0.00019537530583216945,
+      "loss": 0.6904,
+      "step": 6391
+    },
+    {
+      "epoch": 1.1402194273481403,
+      "grad_norm": 0.469959557056427,
+      "learning_rate": 0.00019530693859705497,
+      "loss": 0.748,
+      "step": 6392
+    },
+    {
+      "epoch": 1.1403978235661403,
+      "grad_norm": 0.5133938193321228,
+      "learning_rate": 0.00019523857565751003,
+      "loss": 0.7729,
+      "step": 6393
+    },
+    {
+      "epoch": 1.1405762197841405,
+      "grad_norm": 0.543165385723114,
+      "learning_rate": 0.00019517021701890365,
+      "loss": 0.7534,
+      "step": 6394
+    },
+    {
+      "epoch": 1.1407546160021407,
+      "grad_norm": 0.4762142300605774,
+      "learning_rate": 0.00019510186268660497,
+      "loss": 0.4908,
+      "step": 6395
+    },
+    {
+      "epoch": 1.1409330122201409,
+      "grad_norm": 0.5711979866027832,
+      "learning_rate": 0.00019503351266598234,
+      "loss": 0.7407,
+      "step": 6396
+    },
+    {
+      "epoch": 1.141111408438141,
+      "grad_norm": 0.4412236511707306,
+      "learning_rate": 0.00019496516696240399,
+      "loss": 0.6312,
+      "step": 6397
+    },
+    {
+      "epoch": 1.1412898046561413,
+      "grad_norm": 0.5444515347480774,
+      "learning_rate": 0.00019489682558123772,
+      "loss": 0.6745,
+      "step": 6398
+    },
+    {
+      "epoch": 1.1414682008741415,
+      "grad_norm": 0.4231199026107788,
+      "learning_rate": 0.00019482848852785107,
+      "loss": 0.4643,
+      "step": 6399
+    },
+    {
+      "epoch": 1.1416465970921417,
+      "grad_norm": 0.49194931983947754,
+      "learning_rate": 0.00019476015580761118,
+      "loss": 0.6983,
+      "step": 6400
+    },
+    {
+      "epoch": 1.1418249933101419,
+      "grad_norm": 0.5072386860847473,
+      "learning_rate": 0.0001946918274258849,
+      "loss": 0.7764,
+      "step": 6401
+    },
+    {
+      "epoch": 1.142003389528142,
+      "grad_norm": 0.4746543765068054,
+      "learning_rate": 0.0001946235033880387,
+      "loss": 0.7924,
+      "step": 6402
+    },
+    {
+      "epoch": 1.1421817857461423,
+      "grad_norm": 0.4350649416446686,
+      "learning_rate": 0.00019455518369943873,
+      "loss": 0.4785,
+      "step": 6403
+    },
+    {
+      "epoch": 1.1423601819641425,
+      "grad_norm": 0.5181750655174255,
+      "learning_rate": 0.00019448686836545073,
+      "loss": 0.8109,
+      "step": 6404
+    },
+    {
+      "epoch": 1.1425385781821427,
+      "grad_norm": 0.6082465052604675,
+      "learning_rate": 0.0001944185573914402,
+      "loss": 0.6327,
+      "step": 6405
+    },
+    {
+      "epoch": 1.1427169744001426,
+      "grad_norm": 0.5113035440444946,
+      "learning_rate": 0.00019435025078277227,
+      "loss": 0.7438,
+      "step": 6406
+    },
+    {
+      "epoch": 1.1428953706181428,
+      "grad_norm": 0.4369472861289978,
+      "learning_rate": 0.00019428194854481169,
+      "loss": 0.5909,
+      "step": 6407
+    },
+    {
+      "epoch": 1.143073766836143,
+      "grad_norm": 0.4993836283683777,
+      "learning_rate": 0.00019421365068292287,
+      "loss": 0.6107,
+      "step": 6408
+    },
+    {
+      "epoch": 1.1432521630541432,
+      "grad_norm": 0.41479700803756714,
+      "learning_rate": 0.0001941453572024699,
+      "loss": 0.496,
+      "step": 6409
+    },
+    {
+      "epoch": 1.1434305592721434,
+      "grad_norm": 0.5059956312179565,
+      "learning_rate": 0.00019407706810881657,
+      "loss": 0.7022,
+      "step": 6410
+    },
+    {
+      "epoch": 1.1436089554901436,
+      "grad_norm": 0.46626099944114685,
+      "learning_rate": 0.00019400878340732625,
+      "loss": 0.5752,
+      "step": 6411
+    },
+    {
+      "epoch": 1.1437873517081438,
+      "grad_norm": 0.44759035110473633,
+      "learning_rate": 0.00019394050310336198,
+      "loss": 0.5391,
+      "step": 6412
+    },
+    {
+      "epoch": 1.143965747926144,
+      "grad_norm": 0.4467226564884186,
+      "learning_rate": 0.0001938722272022865,
+      "loss": 0.4655,
+      "step": 6413
+    },
+    {
+      "epoch": 1.1441441441441442,
+      "grad_norm": 0.49334871768951416,
+      "learning_rate": 0.0001938039557094621,
+      "loss": 0.6975,
+      "step": 6414
+    },
+    {
+      "epoch": 1.1443225403621442,
+      "grad_norm": 0.5575069785118103,
+      "learning_rate": 0.00019373568863025086,
+      "loss": 0.7331,
+      "step": 6415
+    },
+    {
+      "epoch": 1.1445009365801444,
+      "grad_norm": 0.5732327699661255,
+      "learning_rate": 0.00019366742597001446,
+      "loss": 0.9117,
+      "step": 6416
+    },
+    {
+      "epoch": 1.1446793327981446,
+      "grad_norm": 0.5360227227210999,
+      "learning_rate": 0.00019359916773411414,
+      "loss": 0.8448,
+      "step": 6417
+    },
+    {
+      "epoch": 1.1448577290161448,
+      "grad_norm": 0.4648105204105377,
+      "learning_rate": 0.00019353091392791094,
+      "loss": 0.5582,
+      "step": 6418
+    },
+    {
+      "epoch": 1.145036125234145,
+      "grad_norm": 0.5667071342468262,
+      "learning_rate": 0.00019346266455676542,
+      "loss": 0.7995,
+      "step": 6419
+    },
+    {
+      "epoch": 1.1452145214521452,
+      "grad_norm": 0.4563814699649811,
+      "learning_rate": 0.00019339441962603794,
+      "loss": 0.7236,
+      "step": 6420
+    },
+    {
+      "epoch": 1.1453929176701454,
+      "grad_norm": 0.4414633512496948,
+      "learning_rate": 0.00019332617914108834,
+      "loss": 0.5218,
+      "step": 6421
+    },
+    {
+      "epoch": 1.1455713138881456,
+      "grad_norm": 0.46729329228401184,
+      "learning_rate": 0.00019325794310727626,
+      "loss": 0.7908,
+      "step": 6422
+    },
+    {
+      "epoch": 1.1457497101061458,
+      "grad_norm": 0.4846903383731842,
+      "learning_rate": 0.0001931897115299609,
+      "loss": 0.6743,
+      "step": 6423
+    },
+    {
+      "epoch": 1.145928106324146,
+      "grad_norm": 0.4348449409008026,
+      "learning_rate": 0.00019312148441450122,
+      "loss": 0.5317,
+      "step": 6424
+    },
+    {
+      "epoch": 1.1461065025421462,
+      "grad_norm": 0.4379520118236542,
+      "learning_rate": 0.0001930532617662555,
+      "loss": 0.6015,
+      "step": 6425
+    },
+    {
+      "epoch": 1.1462848987601464,
+      "grad_norm": 0.42976364493370056,
+      "learning_rate": 0.00019298504359058222,
+      "loss": 0.4379,
+      "step": 6426
+    },
+    {
+      "epoch": 1.1464632949781466,
+      "grad_norm": 0.48329517245292664,
+      "learning_rate": 0.00019291682989283908,
+      "loss": 0.6796,
+      "step": 6427
+    },
+    {
+      "epoch": 1.1466416911961466,
+      "grad_norm": 0.4555102288722992,
+      "learning_rate": 0.00019284862067838352,
+      "loss": 0.5488,
+      "step": 6428
+    },
+    {
+      "epoch": 1.1468200874141468,
+      "grad_norm": 0.49463093280792236,
+      "learning_rate": 0.0001927804159525728,
+      "loss": 0.7097,
+      "step": 6429
+    },
+    {
+      "epoch": 1.146998483632147,
+      "grad_norm": 0.5678471922874451,
+      "learning_rate": 0.0001927122157207635,
+      "loss": 0.9311,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1471768798501472,
+      "grad_norm": 0.5055651068687439,
+      "learning_rate": 0.00019264401998831212,
+      "loss": 0.737,
+      "step": 6431
+    },
+    {
+      "epoch": 1.1473552760681474,
+      "grad_norm": 0.5009707808494568,
+      "learning_rate": 0.00019257582876057474,
+      "loss": 0.6349,
+      "step": 6432
+    },
+    {
+      "epoch": 1.1475336722861476,
+      "grad_norm": 0.4183083176612854,
+      "learning_rate": 0.00019250764204290709,
+      "loss": 0.5802,
+      "step": 6433
+    },
+    {
+      "epoch": 1.1477120685041478,
+      "grad_norm": 0.44495829939842224,
+      "learning_rate": 0.00019243945984066444,
+      "loss": 0.5597,
+      "step": 6434
+    },
+    {
+      "epoch": 1.147890464722148,
+      "grad_norm": 0.5188366770744324,
+      "learning_rate": 0.00019237128215920187,
+      "loss": 0.7409,
+      "step": 6435
+    },
+    {
+      "epoch": 1.1480688609401482,
+      "grad_norm": 0.49368321895599365,
+      "learning_rate": 0.000192303109003874,
+      "loss": 0.6452,
+      "step": 6436
+    },
+    {
+      "epoch": 1.1482472571581481,
+      "grad_norm": 0.41561809182167053,
+      "learning_rate": 0.00019223494038003516,
+      "loss": 0.6164,
+      "step": 6437
+    },
+    {
+      "epoch": 1.1484256533761483,
+      "grad_norm": 0.49593281745910645,
+      "learning_rate": 0.00019216677629303923,
+      "loss": 0.7516,
+      "step": 6438
+    },
+    {
+      "epoch": 1.1486040495941485,
+      "grad_norm": 0.5000485181808472,
+      "learning_rate": 0.00019209861674823975,
+      "loss": 0.6881,
+      "step": 6439
+    },
+    {
+      "epoch": 1.1487824458121487,
+      "grad_norm": 0.41080528497695923,
+      "learning_rate": 0.00019203046175098992,
+      "loss": 0.5175,
+      "step": 6440
+    },
+    {
+      "epoch": 1.148960842030149,
+      "grad_norm": 0.46818774938583374,
+      "learning_rate": 0.00019196231130664282,
+      "loss": 0.7355,
+      "step": 6441
+    },
+    {
+      "epoch": 1.1491392382481491,
+      "grad_norm": 0.4806887209415436,
+      "learning_rate": 0.00019189416542055078,
+      "loss": 0.6491,
+      "step": 6442
+    },
+    {
+      "epoch": 1.1493176344661493,
+      "grad_norm": 0.46855100989341736,
+      "learning_rate": 0.00019182602409806597,
+      "loss": 0.7616,
+      "step": 6443
+    },
+    {
+      "epoch": 1.1494960306841495,
+      "grad_norm": 0.49849364161491394,
+      "learning_rate": 0.00019175788734454019,
+      "loss": 0.722,
+      "step": 6444
+    },
+    {
+      "epoch": 1.1496744269021497,
+      "grad_norm": 0.48306459188461304,
+      "learning_rate": 0.0001916897551653249,
+      "loss": 0.6938,
+      "step": 6445
+    },
+    {
+      "epoch": 1.14985282312015,
+      "grad_norm": 0.4553467035293579,
+      "learning_rate": 0.0001916216275657711,
+      "loss": 0.5726,
+      "step": 6446
+    },
+    {
+      "epoch": 1.1500312193381501,
+      "grad_norm": 0.4541803300380707,
+      "learning_rate": 0.0001915535045512296,
+      "loss": 0.5465,
+      "step": 6447
+    },
+    {
+      "epoch": 1.1502096155561503,
+      "grad_norm": 0.5156320333480835,
+      "learning_rate": 0.00019148538612705066,
+      "loss": 0.6868,
+      "step": 6448
+    },
+    {
+      "epoch": 1.1503880117741505,
+      "grad_norm": 0.4388979375362396,
+      "learning_rate": 0.00019141727229858433,
+      "loss": 0.5943,
+      "step": 6449
+    },
+    {
+      "epoch": 1.1505664079921505,
+      "grad_norm": 0.4948444366455078,
+      "learning_rate": 0.00019134916307118028,
+      "loss": 0.6421,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1507448042101507,
+      "grad_norm": 0.46321436762809753,
+      "learning_rate": 0.00019128105845018766,
+      "loss": 0.5047,
+      "step": 6451
+    },
+    {
+      "epoch": 1.1509232004281509,
+      "grad_norm": 0.46855947375297546,
+      "learning_rate": 0.00019121295844095544,
+      "loss": 0.6025,
+      "step": 6452
+    },
+    {
+      "epoch": 1.151101596646151,
+      "grad_norm": 0.45880499482154846,
+      "learning_rate": 0.00019114486304883216,
+      "loss": 0.7395,
+      "step": 6453
+    },
+    {
+      "epoch": 1.1512799928641513,
+      "grad_norm": 0.4672548472881317,
+      "learning_rate": 0.00019107677227916603,
+      "loss": 0.7361,
+      "step": 6454
+    },
+    {
+      "epoch": 1.1514583890821515,
+      "grad_norm": 0.4436180591583252,
+      "learning_rate": 0.0001910086861373048,
+      "loss": 0.5226,
+      "step": 6455
+    },
+    {
+      "epoch": 1.1516367853001517,
+      "grad_norm": 0.47873300313949585,
+      "learning_rate": 0.000190940604628596,
+      "loss": 0.7932,
+      "step": 6456
+    },
+    {
+      "epoch": 1.1518151815181519,
+      "grad_norm": 0.5310173034667969,
+      "learning_rate": 0.00019087252775838671,
+      "loss": 0.8148,
+      "step": 6457
+    },
+    {
+      "epoch": 1.151993577736152,
+      "grad_norm": 0.4951856732368469,
+      "learning_rate": 0.00019080445553202363,
+      "loss": 0.7325,
+      "step": 6458
+    },
+    {
+      "epoch": 1.152171973954152,
+      "grad_norm": 0.5043248534202576,
+      "learning_rate": 0.0001907363879548532,
+      "loss": 0.6899,
+      "step": 6459
+    },
+    {
+      "epoch": 1.1523503701721522,
+      "grad_norm": 0.5174239873886108,
+      "learning_rate": 0.00019066832503222128,
+      "loss": 0.6925,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1525287663901524,
+      "grad_norm": 0.47568047046661377,
+      "learning_rate": 0.00019060026676947362,
+      "loss": 0.7057,
+      "step": 6461
+    },
+    {
+      "epoch": 1.1527071626081526,
+      "grad_norm": 0.4853077828884125,
+      "learning_rate": 0.0001905322131719555,
+      "loss": 0.7131,
+      "step": 6462
+    },
+    {
+      "epoch": 1.1528855588261528,
+      "grad_norm": 0.4645387828350067,
+      "learning_rate": 0.0001904641642450117,
+      "loss": 0.6627,
+      "step": 6463
+    },
+    {
+      "epoch": 1.153063955044153,
+      "grad_norm": 0.4763006269931793,
+      "learning_rate": 0.00019039611999398682,
+      "loss": 0.6816,
+      "step": 6464
+    },
+    {
+      "epoch": 1.1532423512621532,
+      "grad_norm": 0.5304715633392334,
+      "learning_rate": 0.00019032808042422503,
+      "loss": 0.7631,
+      "step": 6465
+    },
+    {
+      "epoch": 1.1534207474801534,
+      "grad_norm": 0.4912330210208893,
+      "learning_rate": 0.0001902600455410701,
+      "loss": 0.6471,
+      "step": 6466
+    },
+    {
+      "epoch": 1.1535991436981536,
+      "grad_norm": 0.4972264766693115,
+      "learning_rate": 0.00019019201534986553,
+      "loss": 0.7728,
+      "step": 6467
+    },
+    {
+      "epoch": 1.1537775399161538,
+      "grad_norm": 2.746565341949463,
+      "learning_rate": 0.0001901239898559543,
+      "loss": 0.6513,
+      "step": 6468
+    },
+    {
+      "epoch": 1.153955936134154,
+      "grad_norm": 0.5050514340400696,
+      "learning_rate": 0.00019005596906467918,
+      "loss": 0.8309,
+      "step": 6469
+    },
+    {
+      "epoch": 1.1541343323521542,
+      "grad_norm": 0.5243608951568604,
+      "learning_rate": 0.0001899879529813825,
+      "loss": 0.8084,
+      "step": 6470
+    },
+    {
+      "epoch": 1.1543127285701544,
+      "grad_norm": 0.4700818955898285,
+      "learning_rate": 0.00018991994161140596,
+      "loss": 0.5866,
+      "step": 6471
+    },
+    {
+      "epoch": 1.1544911247881544,
+      "grad_norm": 0.46044886112213135,
+      "learning_rate": 0.00018985193496009152,
+      "loss": 0.6026,
+      "step": 6472
+    },
+    {
+      "epoch": 1.1546695210061546,
+      "grad_norm": 0.47794002294540405,
+      "learning_rate": 0.0001897839330327802,
+      "loss": 0.593,
+      "step": 6473
+    },
+    {
+      "epoch": 1.1548479172241548,
+      "grad_norm": 0.49822142720222473,
+      "learning_rate": 0.00018971593583481282,
+      "loss": 0.8087,
+      "step": 6474
+    },
+    {
+      "epoch": 1.155026313442155,
+      "grad_norm": 0.49274197220802307,
+      "learning_rate": 0.00018964794337153002,
+      "loss": 0.6831,
+      "step": 6475
+    },
+    {
+      "epoch": 1.1552047096601552,
+      "grad_norm": 0.49197664856910706,
+      "learning_rate": 0.0001895799556482717,
+      "loss": 0.5862,
+      "step": 6476
+    },
+    {
+      "epoch": 1.1553831058781554,
+      "grad_norm": 0.49241068959236145,
+      "learning_rate": 0.00018951197267037765,
+      "loss": 0.6444,
+      "step": 6477
+    },
+    {
+      "epoch": 1.1555615020961556,
+      "grad_norm": 0.5059458017349243,
+      "learning_rate": 0.00018944399444318724,
+      "loss": 0.7429,
+      "step": 6478
+    },
+    {
+      "epoch": 1.1557398983141558,
+      "grad_norm": 0.4923364222049713,
+      "learning_rate": 0.00018937602097203942,
+      "loss": 0.6462,
+      "step": 6479
+    },
+    {
+      "epoch": 1.155918294532156,
+      "grad_norm": 0.5192193388938904,
+      "learning_rate": 0.0001893080522622729,
+      "loss": 0.7404,
+      "step": 6480
+    },
+    {
+      "epoch": 1.156096690750156,
+      "grad_norm": 0.5662673711776733,
+      "learning_rate": 0.0001892400883192258,
+      "loss": 0.7242,
+      "step": 6481
+    },
+    {
+      "epoch": 1.1562750869681562,
+      "grad_norm": 0.4407677948474884,
+      "learning_rate": 0.00018917212914823598,
+      "loss": 0.5281,
+      "step": 6482
+    },
+    {
+      "epoch": 1.1564534831861564,
+      "grad_norm": 0.4470885694026947,
+      "learning_rate": 0.00018910417475464104,
+      "loss": 0.6045,
+      "step": 6483
+    },
+    {
+      "epoch": 1.1566318794041566,
+      "grad_norm": 0.477516233921051,
+      "learning_rate": 0.00018903622514377798,
+      "loss": 0.6069,
+      "step": 6484
+    },
+    {
+      "epoch": 1.1568102756221568,
+      "grad_norm": 0.5461446046829224,
+      "learning_rate": 0.00018896828032098352,
+      "loss": 0.7073,
+      "step": 6485
+    },
+    {
+      "epoch": 1.156988671840157,
+      "grad_norm": 0.5411002039909363,
+      "learning_rate": 0.00018890034029159399,
+      "loss": 0.8649,
+      "step": 6486
+    },
+    {
+      "epoch": 1.1571670680581572,
+      "grad_norm": 0.5572760701179504,
+      "learning_rate": 0.0001888324050609455,
+      "loss": 0.9441,
+      "step": 6487
+    },
+    {
+      "epoch": 1.1573454642761574,
+      "grad_norm": 0.44463208317756653,
+      "learning_rate": 0.00018876447463437367,
+      "loss": 0.5227,
+      "step": 6488
+    },
+    {
+      "epoch": 1.1575238604941576,
+      "grad_norm": 0.47355902194976807,
+      "learning_rate": 0.0001886965490172136,
+      "loss": 0.4937,
+      "step": 6489
+    },
+    {
+      "epoch": 1.1577022567121578,
+      "grad_norm": 0.44894346594810486,
+      "learning_rate": 0.00018862862821480023,
+      "loss": 0.5107,
+      "step": 6490
+    },
+    {
+      "epoch": 1.157880652930158,
+      "grad_norm": 0.5426954627037048,
+      "learning_rate": 0.00018856071223246796,
+      "loss": 0.6947,
+      "step": 6491
+    },
+    {
+      "epoch": 1.1580590491481582,
+      "grad_norm": 0.44600164890289307,
+      "learning_rate": 0.0001884928010755509,
+      "loss": 0.6412,
+      "step": 6492
+    },
+    {
+      "epoch": 1.1582374453661584,
+      "grad_norm": 0.4314921498298645,
+      "learning_rate": 0.0001884248947493828,
+      "loss": 0.5661,
+      "step": 6493
+    },
+    {
+      "epoch": 1.1584158415841583,
+      "grad_norm": 0.44217392802238464,
+      "learning_rate": 0.00018835699325929692,
+      "loss": 0.5272,
+      "step": 6494
+    },
+    {
+      "epoch": 1.1585942378021585,
+      "grad_norm": 0.42524126172065735,
+      "learning_rate": 0.0001882890966106263,
+      "loss": 0.5549,
+      "step": 6495
+    },
+    {
+      "epoch": 1.1587726340201587,
+      "grad_norm": 0.4192121922969818,
+      "learning_rate": 0.00018822120480870352,
+      "loss": 0.4929,
+      "step": 6496
+    },
+    {
+      "epoch": 1.158951030238159,
+      "grad_norm": 0.4756525456905365,
+      "learning_rate": 0.00018815331785886066,
+      "loss": 0.6762,
+      "step": 6497
+    },
+    {
+      "epoch": 1.1591294264561591,
+      "grad_norm": 0.5104422569274902,
+      "learning_rate": 0.00018808543576642964,
+      "loss": 0.7025,
+      "step": 6498
+    },
+    {
+      "epoch": 1.1593078226741593,
+      "grad_norm": 0.46176689863204956,
+      "learning_rate": 0.00018801755853674183,
+      "loss": 0.5673,
+      "step": 6499
+    },
+    {
+      "epoch": 1.1594862188921595,
+      "grad_norm": 0.49885469675064087,
+      "learning_rate": 0.00018794968617512827,
+      "loss": 0.6967,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1596646151101597,
+      "grad_norm": 0.4945884943008423,
+      "learning_rate": 0.00018788181868691965,
+      "loss": 0.7192,
+      "step": 6501
+    },
+    {
+      "epoch": 1.15984301132816,
+      "grad_norm": 0.4873616099357605,
+      "learning_rate": 0.00018781395607744627,
+      "loss": 0.6867,
+      "step": 6502
+    },
+    {
+      "epoch": 1.16002140754616,
+      "grad_norm": 0.393449068069458,
+      "learning_rate": 0.00018774609835203808,
+      "loss": 0.4874,
+      "step": 6503
+    },
+    {
+      "epoch": 1.16019980376416,
+      "grad_norm": 0.42069491744041443,
+      "learning_rate": 0.00018767824551602446,
+      "loss": 0.493,
+      "step": 6504
+    },
+    {
+      "epoch": 1.1603781999821603,
+      "grad_norm": 0.44530075788497925,
+      "learning_rate": 0.00018761039757473466,
+      "loss": 0.4976,
+      "step": 6505
+    },
+    {
+      "epoch": 1.1605565962001605,
+      "grad_norm": 0.4841354191303253,
+      "learning_rate": 0.0001875425545334974,
+      "loss": 0.7774,
+      "step": 6506
+    },
+    {
+      "epoch": 1.1607349924181607,
+      "grad_norm": 0.44460541009902954,
+      "learning_rate": 0.00018747471639764103,
+      "loss": 0.6535,
+      "step": 6507
+    },
+    {
+      "epoch": 1.160913388636161,
+      "grad_norm": 0.5261197686195374,
+      "learning_rate": 0.00018740688317249349,
+      "loss": 0.7753,
+      "step": 6508
+    },
+    {
+      "epoch": 1.161091784854161,
+      "grad_norm": 0.4835450351238251,
+      "learning_rate": 0.0001873390548633825,
+      "loss": 0.9085,
+      "step": 6509
+    },
+    {
+      "epoch": 1.1612701810721613,
+      "grad_norm": 0.4938504993915558,
+      "learning_rate": 0.00018727123147563508,
+      "loss": 0.7482,
+      "step": 6510
+    },
+    {
+      "epoch": 1.1614485772901615,
+      "grad_norm": 0.4450148642063141,
+      "learning_rate": 0.00018720341301457815,
+      "loss": 0.5129,
+      "step": 6511
+    },
+    {
+      "epoch": 1.1616269735081617,
+      "grad_norm": 0.49409425258636475,
+      "learning_rate": 0.00018713559948553815,
+      "loss": 0.6571,
+      "step": 6512
+    },
+    {
+      "epoch": 1.1618053697261619,
+      "grad_norm": 0.42532381415367126,
+      "learning_rate": 0.00018706779089384113,
+      "loss": 0.508,
+      "step": 6513
+    },
+    {
+      "epoch": 1.161983765944162,
+      "grad_norm": 0.5003641843795776,
+      "learning_rate": 0.0001869999872448127,
+      "loss": 0.6938,
+      "step": 6514
+    },
+    {
+      "epoch": 1.1621621621621623,
+      "grad_norm": 0.4792661666870117,
+      "learning_rate": 0.0001869321885437782,
+      "loss": 0.6612,
+      "step": 6515
+    },
+    {
+      "epoch": 1.1623405583801623,
+      "grad_norm": 0.44840648770332336,
+      "learning_rate": 0.00018686439479606245,
+      "loss": 0.609,
+      "step": 6516
+    },
+    {
+      "epoch": 1.1625189545981625,
+      "grad_norm": 0.48601728677749634,
+      "learning_rate": 0.00018679660600698996,
+      "loss": 0.747,
+      "step": 6517
+    },
+    {
+      "epoch": 1.1626973508161627,
+      "grad_norm": 0.5069820284843445,
+      "learning_rate": 0.0001867288221818848,
+      "loss": 0.5674,
+      "step": 6518
+    },
+    {
+      "epoch": 1.1628757470341629,
+      "grad_norm": 0.4526546001434326,
+      "learning_rate": 0.00018666104332607075,
+      "loss": 0.5845,
+      "step": 6519
+    },
+    {
+      "epoch": 1.163054143252163,
+      "grad_norm": 0.48952043056488037,
+      "learning_rate": 0.00018659326944487115,
+      "loss": 0.7082,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1632325394701633,
+      "grad_norm": 0.5443047881126404,
+      "learning_rate": 0.0001865255005436089,
+      "loss": 0.718,
+      "step": 6521
+    },
+    {
+      "epoch": 1.1634109356881635,
+      "grad_norm": 0.5090129375457764,
+      "learning_rate": 0.00018645773662760647,
+      "loss": 0.9893,
+      "step": 6522
+    },
+    {
+      "epoch": 1.1635893319061636,
+      "grad_norm": 0.5263794660568237,
+      "learning_rate": 0.00018638997770218602,
+      "loss": 0.6613,
+      "step": 6523
+    },
+    {
+      "epoch": 1.1637677281241638,
+      "grad_norm": 0.6042819619178772,
+      "learning_rate": 0.0001863222237726694,
+      "loss": 0.7601,
+      "step": 6524
+    },
+    {
+      "epoch": 1.1639461243421638,
+      "grad_norm": 0.5052105784416199,
+      "learning_rate": 0.0001862544748443779,
+      "loss": 0.7175,
+      "step": 6525
+    },
+    {
+      "epoch": 1.164124520560164,
+      "grad_norm": 0.4983629882335663,
+      "learning_rate": 0.00018618673092263253,
+      "loss": 0.682,
+      "step": 6526
+    },
+    {
+      "epoch": 1.1643029167781642,
+      "grad_norm": 0.4915952682495117,
+      "learning_rate": 0.00018611899201275385,
+      "loss": 0.6723,
+      "step": 6527
+    },
+    {
+      "epoch": 1.1644813129961644,
+      "grad_norm": 0.5149100422859192,
+      "learning_rate": 0.000186051258120062,
+      "loss": 0.6966,
+      "step": 6528
+    },
+    {
+      "epoch": 1.1646597092141646,
+      "grad_norm": 0.5236539244651794,
+      "learning_rate": 0.00018598352924987689,
+      "loss": 0.8879,
+      "step": 6529
+    },
+    {
+      "epoch": 1.1648381054321648,
+      "grad_norm": 0.4917599856853485,
+      "learning_rate": 0.00018591580540751784,
+      "loss": 0.7433,
+      "step": 6530
+    },
+    {
+      "epoch": 1.165016501650165,
+      "grad_norm": 0.47887739539146423,
+      "learning_rate": 0.00018584808659830385,
+      "loss": 0.7286,
+      "step": 6531
+    },
+    {
+      "epoch": 1.1651948978681652,
+      "grad_norm": 0.4636463224887848,
+      "learning_rate": 0.0001857803728275534,
+      "loss": 0.6089,
+      "step": 6532
+    },
+    {
+      "epoch": 1.1653732940861654,
+      "grad_norm": 0.5316303372383118,
+      "learning_rate": 0.00018571266410058492,
+      "loss": 0.6789,
+      "step": 6533
+    },
+    {
+      "epoch": 1.1655516903041656,
+      "grad_norm": 0.49224573373794556,
+      "learning_rate": 0.00018564496042271624,
+      "loss": 0.788,
+      "step": 6534
+    },
+    {
+      "epoch": 1.1657300865221658,
+      "grad_norm": 0.5619474053382874,
+      "learning_rate": 0.0001855772617992646,
+      "loss": 0.7761,
+      "step": 6535
+    },
+    {
+      "epoch": 1.165908482740166,
+      "grad_norm": 0.49133652448654175,
+      "learning_rate": 0.00018550956823554706,
+      "loss": 0.6949,
+      "step": 6536
+    },
+    {
+      "epoch": 1.1660868789581662,
+      "grad_norm": 0.4631252884864807,
+      "learning_rate": 0.00018544187973688032,
+      "loss": 0.5078,
+      "step": 6537
+    },
+    {
+      "epoch": 1.1662652751761662,
+      "grad_norm": 0.4481963515281677,
+      "learning_rate": 0.00018537419630858053,
+      "loss": 0.6133,
+      "step": 6538
+    },
+    {
+      "epoch": 1.1664436713941664,
+      "grad_norm": 0.49360591173171997,
+      "learning_rate": 0.0001853065179559636,
+      "loss": 0.6333,
+      "step": 6539
+    },
+    {
+      "epoch": 1.1666220676121666,
+      "grad_norm": 0.5313968062400818,
+      "learning_rate": 0.00018523884468434488,
+      "loss": 0.8331,
+      "step": 6540
+    },
+    {
+      "epoch": 1.1668004638301668,
+      "grad_norm": 0.5059915781021118,
+      "learning_rate": 0.00018517117649903943,
+      "loss": 0.7155,
+      "step": 6541
+    },
+    {
+      "epoch": 1.166978860048167,
+      "grad_norm": 0.4583684504032135,
+      "learning_rate": 0.00018510351340536192,
+      "loss": 0.6011,
+      "step": 6542
+    },
+    {
+      "epoch": 1.1671572562661672,
+      "grad_norm": 0.4290786385536194,
+      "learning_rate": 0.0001850358554086265,
+      "loss": 0.5471,
+      "step": 6543
+    },
+    {
+      "epoch": 1.1673356524841674,
+      "grad_norm": 0.4612555205821991,
+      "learning_rate": 0.00018496820251414703,
+      "loss": 0.6882,
+      "step": 6544
+    },
+    {
+      "epoch": 1.1675140487021676,
+      "grad_norm": 0.3945925831794739,
+      "learning_rate": 0.00018490055472723696,
+      "loss": 0.51,
+      "step": 6545
+    },
+    {
+      "epoch": 1.1676924449201678,
+      "grad_norm": 0.43021753430366516,
+      "learning_rate": 0.0001848329120532093,
+      "loss": 0.5537,
+      "step": 6546
+    },
+    {
+      "epoch": 1.1678708411381677,
+      "grad_norm": 0.46711066365242004,
+      "learning_rate": 0.00018476527449737666,
+      "loss": 0.7032,
+      "step": 6547
+    },
+    {
+      "epoch": 1.168049237356168,
+      "grad_norm": 0.4292619228363037,
+      "learning_rate": 0.0001846976420650513,
+      "loss": 0.569,
+      "step": 6548
+    },
+    {
+      "epoch": 1.1682276335741681,
+      "grad_norm": 0.45965656638145447,
+      "learning_rate": 0.00018463001476154507,
+      "loss": 0.7034,
+      "step": 6549
+    },
+    {
+      "epoch": 1.1684060297921683,
+      "grad_norm": 0.49549368023872375,
+      "learning_rate": 0.00018456239259216934,
+      "loss": 0.6292,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1685844260101685,
+      "grad_norm": 0.5122089385986328,
+      "learning_rate": 0.0001844947755622351,
+      "loss": 0.7911,
+      "step": 6551
+    },
+    {
+      "epoch": 1.1687628222281687,
+      "grad_norm": 1.3904839754104614,
+      "learning_rate": 0.00018442716367705303,
+      "loss": 0.6504,
+      "step": 6552
+    },
+    {
+      "epoch": 1.168941218446169,
+      "grad_norm": 0.5320877432823181,
+      "learning_rate": 0.0001843595569419333,
+      "loss": 0.6984,
+      "step": 6553
+    },
+    {
+      "epoch": 1.1691196146641691,
+      "grad_norm": 0.44560760259628296,
+      "learning_rate": 0.0001842919553621857,
+      "loss": 0.5054,
+      "step": 6554
+    },
+    {
+      "epoch": 1.1692980108821693,
+      "grad_norm": 0.4876161217689514,
+      "learning_rate": 0.00018422435894311972,
+      "loss": 0.6423,
+      "step": 6555
+    },
+    {
+      "epoch": 1.1694764071001695,
+      "grad_norm": 0.4854297339916229,
+      "learning_rate": 0.00018415676769004426,
+      "loss": 0.6986,
+      "step": 6556
+    },
+    {
+      "epoch": 1.1696548033181697,
+      "grad_norm": 0.4554065465927124,
+      "learning_rate": 0.00018408918160826789,
+      "loss": 0.5424,
+      "step": 6557
+    },
+    {
+      "epoch": 1.16983319953617,
+      "grad_norm": 0.5047731399536133,
+      "learning_rate": 0.00018402160070309884,
+      "loss": 0.7302,
+      "step": 6558
+    },
+    {
+      "epoch": 1.1700115957541701,
+      "grad_norm": 0.47815465927124023,
+      "learning_rate": 0.0001839540249798449,
+      "loss": 0.7022,
+      "step": 6559
+    },
+    {
+      "epoch": 1.17018999197217,
+      "grad_norm": 0.6329243183135986,
+      "learning_rate": 0.0001838864544438134,
+      "loss": 0.5491,
+      "step": 6560
+    },
+    {
+      "epoch": 1.1703683881901703,
+      "grad_norm": 0.4645484685897827,
+      "learning_rate": 0.0001838188891003113,
+      "loss": 0.6384,
+      "step": 6561
+    },
+    {
+      "epoch": 1.1705467844081705,
+      "grad_norm": 0.49263012409210205,
+      "learning_rate": 0.0001837513289546452,
+      "loss": 0.7574,
+      "step": 6562
+    },
+    {
+      "epoch": 1.1707251806261707,
+      "grad_norm": 0.44261929392814636,
+      "learning_rate": 0.00018368377401212116,
+      "loss": 0.5943,
+      "step": 6563
+    },
+    {
+      "epoch": 1.170903576844171,
+      "grad_norm": 0.4851824641227722,
+      "learning_rate": 0.000183616224278045,
+      "loss": 0.5772,
+      "step": 6564
+    },
+    {
+      "epoch": 1.171081973062171,
+      "grad_norm": 0.4466243088245392,
+      "learning_rate": 0.00018354867975772205,
+      "loss": 0.5803,
+      "step": 6565
+    },
+    {
+      "epoch": 1.1712603692801713,
+      "grad_norm": 0.5128055810928345,
+      "learning_rate": 0.00018348114045645713,
+      "loss": 0.7322,
+      "step": 6566
+    },
+    {
+      "epoch": 1.1714387654981715,
+      "grad_norm": 0.5170561075210571,
+      "learning_rate": 0.00018341360637955489,
+      "loss": 0.7447,
+      "step": 6567
+    },
+    {
+      "epoch": 1.1716171617161717,
+      "grad_norm": 0.4576283097267151,
+      "learning_rate": 0.00018334607753231935,
+      "loss": 0.6132,
+      "step": 6568
+    },
+    {
+      "epoch": 1.1717955579341717,
+      "grad_norm": 0.44425490498542786,
+      "learning_rate": 0.00018327855392005418,
+      "loss": 0.5558,
+      "step": 6569
+    },
+    {
+      "epoch": 1.1719739541521719,
+      "grad_norm": 0.5051802396774292,
+      "learning_rate": 0.00018321103554806267,
+      "loss": 0.6139,
+      "step": 6570
+    },
+    {
+      "epoch": 1.172152350370172,
+      "grad_norm": 0.4622550904750824,
+      "learning_rate": 0.00018314352242164767,
+      "loss": 0.6846,
+      "step": 6571
+    },
+    {
+      "epoch": 1.1723307465881723,
+      "grad_norm": 0.47944626212120056,
+      "learning_rate": 0.00018307601454611166,
+      "loss": 0.6305,
+      "step": 6572
+    },
+    {
+      "epoch": 1.1725091428061725,
+      "grad_norm": 0.4514850080013275,
+      "learning_rate": 0.00018300851192675665,
+      "loss": 0.6158,
+      "step": 6573
+    },
+    {
+      "epoch": 1.1726875390241727,
+      "grad_norm": 0.4573253095149994,
+      "learning_rate": 0.00018294101456888432,
+      "loss": 0.6954,
+      "step": 6574
+    },
+    {
+      "epoch": 1.1728659352421729,
+      "grad_norm": 0.428072065114975,
+      "learning_rate": 0.00018287352247779582,
+      "loss": 0.4599,
+      "step": 6575
+    },
+    {
+      "epoch": 1.173044331460173,
+      "grad_norm": 0.4548456072807312,
+      "learning_rate": 0.00018280603565879207,
+      "loss": 0.6041,
+      "step": 6576
+    },
+    {
+      "epoch": 1.1732227276781733,
+      "grad_norm": 0.4897709786891937,
+      "learning_rate": 0.00018273855411717323,
+      "loss": 0.6279,
+      "step": 6577
+    },
+    {
+      "epoch": 1.1734011238961735,
+      "grad_norm": 0.5311153531074524,
+      "learning_rate": 0.00018267107785823936,
+      "loss": 0.645,
+      "step": 6578
+    },
+    {
+      "epoch": 1.1735795201141737,
+      "grad_norm": 0.48785173892974854,
+      "learning_rate": 0.00018260360688729016,
+      "loss": 0.5661,
+      "step": 6579
+    },
+    {
+      "epoch": 1.1737579163321739,
+      "grad_norm": 0.5114821791648865,
+      "learning_rate": 0.00018253614120962463,
+      "loss": 0.7209,
+      "step": 6580
+    },
+    {
+      "epoch": 1.173936312550174,
+      "grad_norm": 0.5291317105293274,
+      "learning_rate": 0.00018246868083054166,
+      "loss": 0.8096,
+      "step": 6581
+    },
+    {
+      "epoch": 1.174114708768174,
+      "grad_norm": 0.5284007787704468,
+      "learning_rate": 0.00018240122575533934,
+      "loss": 0.7968,
+      "step": 6582
+    },
+    {
+      "epoch": 1.1742931049861742,
+      "grad_norm": 0.49136999249458313,
+      "learning_rate": 0.00018233377598931566,
+      "loss": 0.6637,
+      "step": 6583
+    },
+    {
+      "epoch": 1.1744715012041744,
+      "grad_norm": 0.4023672342300415,
+      "learning_rate": 0.00018226633153776812,
+      "loss": 0.4708,
+      "step": 6584
+    },
+    {
+      "epoch": 1.1746498974221746,
+      "grad_norm": 0.39422962069511414,
+      "learning_rate": 0.00018219889240599375,
+      "loss": 0.5222,
+      "step": 6585
+    },
+    {
+      "epoch": 1.1748282936401748,
+      "grad_norm": 0.5104691386222839,
+      "learning_rate": 0.00018213145859928914,
+      "loss": 0.6543,
+      "step": 6586
+    },
+    {
+      "epoch": 1.175006689858175,
+      "grad_norm": 0.4648468792438507,
+      "learning_rate": 0.00018206403012295064,
+      "loss": 0.6523,
+      "step": 6587
+    },
+    {
+      "epoch": 1.1751850860761752,
+      "grad_norm": 0.5008534789085388,
+      "learning_rate": 0.00018199660698227393,
+      "loss": 0.7357,
+      "step": 6588
+    },
+    {
+      "epoch": 1.1753634822941754,
+      "grad_norm": 0.5412165522575378,
+      "learning_rate": 0.00018192918918255452,
+      "loss": 0.7517,
+      "step": 6589
+    },
+    {
+      "epoch": 1.1755418785121756,
+      "grad_norm": 0.4949028193950653,
+      "learning_rate": 0.0001818617767290872,
+      "loss": 0.636,
+      "step": 6590
+    },
+    {
+      "epoch": 1.1757202747301756,
+      "grad_norm": 0.49255791306495667,
+      "learning_rate": 0.00018179436962716667,
+      "loss": 0.6114,
+      "step": 6591
+    },
+    {
+      "epoch": 1.1758986709481758,
+      "grad_norm": 0.5153682827949524,
+      "learning_rate": 0.00018172696788208698,
+      "loss": 0.8209,
+      "step": 6592
+    },
+    {
+      "epoch": 1.176077067166176,
+      "grad_norm": 0.4873290956020355,
+      "learning_rate": 0.0001816595714991418,
+      "loss": 0.6777,
+      "step": 6593
+    },
+    {
+      "epoch": 1.1762554633841762,
+      "grad_norm": 0.4800300598144531,
+      "learning_rate": 0.00018159218048362452,
+      "loss": 0.659,
+      "step": 6594
+    },
+    {
+      "epoch": 1.1764338596021764,
+      "grad_norm": 0.511806070804596,
+      "learning_rate": 0.00018152479484082797,
+      "loss": 0.7344,
+      "step": 6595
+    },
+    {
+      "epoch": 1.1766122558201766,
+      "grad_norm": 0.5367228388786316,
+      "learning_rate": 0.0001814574145760445,
+      "loss": 0.9972,
+      "step": 6596
+    },
+    {
+      "epoch": 1.1767906520381768,
+      "grad_norm": 0.5340408682823181,
+      "learning_rate": 0.00018139003969456623,
+      "loss": 0.8547,
+      "step": 6597
+    },
+    {
+      "epoch": 1.176969048256177,
+      "grad_norm": 0.44577550888061523,
+      "learning_rate": 0.00018132267020168471,
+      "loss": 0.5981,
+      "step": 6598
+    },
+    {
+      "epoch": 1.1771474444741772,
+      "grad_norm": 0.5312680006027222,
+      "learning_rate": 0.00018125530610269114,
+      "loss": 0.7921,
+      "step": 6599
+    },
+    {
+      "epoch": 1.1773258406921774,
+      "grad_norm": 0.4474240839481354,
+      "learning_rate": 0.00018118794740287625,
+      "loss": 0.532,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1775042369101776,
+      "grad_norm": 0.455321341753006,
+      "learning_rate": 0.00018112059410753034,
+      "loss": 0.5826,
+      "step": 6601
+    },
+    {
+      "epoch": 1.1776826331281778,
+      "grad_norm": 0.5083394646644592,
+      "learning_rate": 0.00018105324622194336,
+      "loss": 0.7073,
+      "step": 6602
+    },
+    {
+      "epoch": 1.177861029346178,
+      "grad_norm": 0.5182134509086609,
+      "learning_rate": 0.00018098590375140473,
+      "loss": 0.7853,
+      "step": 6603
+    },
+    {
+      "epoch": 1.178039425564178,
+      "grad_norm": 0.5327116250991821,
+      "learning_rate": 0.00018091856670120348,
+      "loss": 0.8714,
+      "step": 6604
+    },
+    {
+      "epoch": 1.1782178217821782,
+      "grad_norm": 0.4751187562942505,
+      "learning_rate": 0.00018085123507662832,
+      "loss": 0.6433,
+      "step": 6605
+    },
+    {
+      "epoch": 1.1783962180001784,
+      "grad_norm": 0.4442801773548126,
+      "learning_rate": 0.0001807839088829674,
+      "loss": 0.539,
+      "step": 6606
+    },
+    {
+      "epoch": 1.1785746142181786,
+      "grad_norm": 0.5099819302558899,
+      "learning_rate": 0.00018071658812550845,
+      "loss": 0.741,
+      "step": 6607
+    },
+    {
+      "epoch": 1.1787530104361787,
+      "grad_norm": 0.4700085520744324,
+      "learning_rate": 0.00018064927280953891,
+      "loss": 0.7668,
+      "step": 6608
+    },
+    {
+      "epoch": 1.178931406654179,
+      "grad_norm": 0.5318174362182617,
+      "learning_rate": 0.00018058196294034554,
+      "loss": 0.8966,
+      "step": 6609
+    },
+    {
+      "epoch": 1.1791098028721791,
+      "grad_norm": 0.46459850668907166,
+      "learning_rate": 0.000180514658523215,
+      "loss": 0.6678,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1792881990901793,
+      "grad_norm": 0.47724831104278564,
+      "learning_rate": 0.00018044735956343328,
+      "loss": 0.6598,
+      "step": 6611
+    },
+    {
+      "epoch": 1.1794665953081795,
+      "grad_norm": 0.46925148367881775,
+      "learning_rate": 0.000180380066066286,
+      "loss": 0.5628,
+      "step": 6612
+    },
+    {
+      "epoch": 1.1796449915261795,
+      "grad_norm": 0.4924178421497345,
+      "learning_rate": 0.00018031277803705835,
+      "loss": 0.703,
+      "step": 6613
+    },
+    {
+      "epoch": 1.1798233877441797,
+      "grad_norm": 0.460475891828537,
+      "learning_rate": 0.00018024549548103518,
+      "loss": 0.5927,
+      "step": 6614
+    },
+    {
+      "epoch": 1.18000178396218,
+      "grad_norm": 0.5572229027748108,
+      "learning_rate": 0.0001801782184035008,
+      "loss": 0.8473,
+      "step": 6615
+    },
+    {
+      "epoch": 1.1801801801801801,
+      "grad_norm": 0.5910245180130005,
+      "learning_rate": 0.00018011094680973902,
+      "loss": 0.7021,
+      "step": 6616
+    },
+    {
+      "epoch": 1.1803585763981803,
+      "grad_norm": 0.5176860094070435,
+      "learning_rate": 0.00018004368070503342,
+      "loss": 0.7578,
+      "step": 6617
+    },
+    {
+      "epoch": 1.1805369726161805,
+      "grad_norm": 0.5350726842880249,
+      "learning_rate": 0.00017997642009466702,
+      "loss": 0.8098,
+      "step": 6618
+    },
+    {
+      "epoch": 1.1807153688341807,
+      "grad_norm": 0.552034318447113,
+      "learning_rate": 0.0001799091649839225,
+      "loss": 0.8272,
+      "step": 6619
+    },
+    {
+      "epoch": 1.180893765052181,
+      "grad_norm": 0.5275470018386841,
+      "learning_rate": 0.00017984191537808198,
+      "loss": 0.6984,
+      "step": 6620
+    },
+    {
+      "epoch": 1.181072161270181,
+      "grad_norm": 0.5647268891334534,
+      "learning_rate": 0.0001797746712824272,
+      "loss": 0.935,
+      "step": 6621
+    },
+    {
+      "epoch": 1.1812505574881813,
+      "grad_norm": 0.4905039370059967,
+      "learning_rate": 0.0001797074327022396,
+      "loss": 0.645,
+      "step": 6622
+    },
+    {
+      "epoch": 1.1814289537061815,
+      "grad_norm": 0.4301641881465912,
+      "learning_rate": 0.00017964019964279994,
+      "loss": 0.4728,
+      "step": 6623
+    },
+    {
+      "epoch": 1.1816073499241817,
+      "grad_norm": 0.46979376673698425,
+      "learning_rate": 0.0001795729721093886,
+      "loss": 0.6428,
+      "step": 6624
+    },
+    {
+      "epoch": 1.181785746142182,
+      "grad_norm": 0.48805493116378784,
+      "learning_rate": 0.00017950575010728582,
+      "loss": 0.6446,
+      "step": 6625
+    },
+    {
+      "epoch": 1.1819641423601819,
+      "grad_norm": 0.49873703718185425,
+      "learning_rate": 0.00017943853364177112,
+      "loss": 0.7287,
+      "step": 6626
+    },
+    {
+      "epoch": 1.182142538578182,
+      "grad_norm": 0.422006756067276,
+      "learning_rate": 0.00017937132271812368,
+      "loss": 0.4329,
+      "step": 6627
+    },
+    {
+      "epoch": 1.1823209347961823,
+      "grad_norm": 0.46575191617012024,
+      "learning_rate": 0.00017930411734162204,
+      "loss": 0.582,
+      "step": 6628
+    },
+    {
+      "epoch": 1.1824993310141825,
+      "grad_norm": 0.5291426181793213,
+      "learning_rate": 0.00017923691751754468,
+      "loss": 0.7962,
+      "step": 6629
+    },
+    {
+      "epoch": 1.1826777272321827,
+      "grad_norm": 0.48327967524528503,
+      "learning_rate": 0.00017916972325116931,
+      "loss": 0.6514,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1828561234501829,
+      "grad_norm": 0.47633296251296997,
+      "learning_rate": 0.00017910253454777344,
+      "loss": 0.6475,
+      "step": 6631
+    },
+    {
+      "epoch": 1.183034519668183,
+      "grad_norm": 0.5271769165992737,
+      "learning_rate": 0.000179035351412634,
+      "loss": 0.7574,
+      "step": 6632
+    },
+    {
+      "epoch": 1.1832129158861833,
+      "grad_norm": 0.45109376311302185,
+      "learning_rate": 0.00017896817385102748,
+      "loss": 0.538,
+      "step": 6633
+    },
+    {
+      "epoch": 1.1833913121041835,
+      "grad_norm": 0.489512175321579,
+      "learning_rate": 0.00017890100186823004,
+      "loss": 0.7353,
+      "step": 6634
+    },
+    {
+      "epoch": 1.1835697083221837,
+      "grad_norm": 0.5405703186988831,
+      "learning_rate": 0.00017883383546951737,
+      "loss": 0.7591,
+      "step": 6635
+    },
+    {
+      "epoch": 1.1837481045401836,
+      "grad_norm": 0.4727650284767151,
+      "learning_rate": 0.00017876667466016458,
+      "loss": 0.7046,
+      "step": 6636
+    },
+    {
+      "epoch": 1.1839265007581838,
+      "grad_norm": 0.49231699109077454,
+      "learning_rate": 0.0001786995194454465,
+      "loss": 0.7205,
+      "step": 6637
+    },
+    {
+      "epoch": 1.184104896976184,
+      "grad_norm": 0.49565812945365906,
+      "learning_rate": 0.0001786323698306375,
+      "loss": 0.7502,
+      "step": 6638
+    },
+    {
+      "epoch": 1.1842832931941842,
+      "grad_norm": 0.4355218708515167,
+      "learning_rate": 0.00017856522582101133,
+      "loss": 0.5357,
+      "step": 6639
+    },
+    {
+      "epoch": 1.1844616894121844,
+      "grad_norm": 0.509213387966156,
+      "learning_rate": 0.00017849808742184176,
+      "loss": 0.8362,
+      "step": 6640
+    },
+    {
+      "epoch": 1.1846400856301846,
+      "grad_norm": 0.49817323684692383,
+      "learning_rate": 0.0001784309546384016,
+      "loss": 0.6041,
+      "step": 6641
+    },
+    {
+      "epoch": 1.1848184818481848,
+      "grad_norm": 0.44568195939064026,
+      "learning_rate": 0.00017836382747596341,
+      "loss": 0.6403,
+      "step": 6642
+    },
+    {
+      "epoch": 1.184996878066185,
+      "grad_norm": 0.42027372121810913,
+      "learning_rate": 0.00017829670593979944,
+      "loss": 0.5297,
+      "step": 6643
+    },
+    {
+      "epoch": 1.1851752742841852,
+      "grad_norm": 0.4120577871799469,
+      "learning_rate": 0.0001782295900351813,
+      "loss": 0.5559,
+      "step": 6644
+    },
+    {
+      "epoch": 1.1853536705021854,
+      "grad_norm": 0.4313491880893707,
+      "learning_rate": 0.00017816247976738025,
+      "loss": 0.629,
+      "step": 6645
+    },
+    {
+      "epoch": 1.1855320667201856,
+      "grad_norm": 0.45468538999557495,
+      "learning_rate": 0.00017809537514166718,
+      "loss": 0.6541,
+      "step": 6646
+    },
+    {
+      "epoch": 1.1857104629381858,
+      "grad_norm": 42.75632858276367,
+      "learning_rate": 0.00017802827616331235,
+      "loss": 1.0525,
+      "step": 6647
+    },
+    {
+      "epoch": 1.1858888591561858,
+      "grad_norm": 0.5608348846435547,
+      "learning_rate": 0.00017796118283758584,
+      "loss": 0.7031,
+      "step": 6648
+    },
+    {
+      "epoch": 1.186067255374186,
+      "grad_norm": 0.4569344222545624,
+      "learning_rate": 0.00017789409516975698,
+      "loss": 0.524,
+      "step": 6649
+    },
+    {
+      "epoch": 1.1862456515921862,
+      "grad_norm": 0.5035879015922546,
+      "learning_rate": 0.0001778270131650948,
+      "loss": 0.681,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1864240478101864,
+      "grad_norm": 0.4764963984489441,
+      "learning_rate": 0.000177759936828868,
+      "loss": 0.6345,
+      "step": 6651
+    },
+    {
+      "epoch": 1.1866024440281866,
+      "grad_norm": 0.46746543049812317,
+      "learning_rate": 0.00017769286616634461,
+      "loss": 0.6118,
+      "step": 6652
+    },
+    {
+      "epoch": 1.1867808402461868,
+      "grad_norm": 0.4906679689884186,
+      "learning_rate": 0.00017762580118279244,
+      "loss": 0.6521,
+      "step": 6653
+    },
+    {
+      "epoch": 1.186959236464187,
+      "grad_norm": 0.5302233099937439,
+      "learning_rate": 0.0001775587418834787,
+      "loss": 0.6907,
+      "step": 6654
+    },
+    {
+      "epoch": 1.1871376326821872,
+      "grad_norm": 0.46473437547683716,
+      "learning_rate": 0.00017749168827367015,
+      "loss": 0.6949,
+      "step": 6655
+    },
+    {
+      "epoch": 1.1873160289001874,
+      "grad_norm": 0.43767476081848145,
+      "learning_rate": 0.00017742464035863325,
+      "loss": 0.5771,
+      "step": 6656
+    },
+    {
+      "epoch": 1.1874944251181876,
+      "grad_norm": 0.5204544067382812,
+      "learning_rate": 0.00017735759814363383,
+      "loss": 0.6368,
+      "step": 6657
+    },
+    {
+      "epoch": 1.1876728213361876,
+      "grad_norm": 0.49561235308647156,
+      "learning_rate": 0.00017729056163393744,
+      "loss": 0.5123,
+      "step": 6658
+    },
+    {
+      "epoch": 1.1878512175541878,
+      "grad_norm": 0.4810577929019928,
+      "learning_rate": 0.00017722353083480903,
+      "loss": 0.6761,
+      "step": 6659
+    },
+    {
+      "epoch": 1.188029613772188,
+      "grad_norm": 0.5160006284713745,
+      "learning_rate": 0.00017715650575151322,
+      "loss": 0.6518,
+      "step": 6660
+    },
+    {
+      "epoch": 1.1882080099901882,
+      "grad_norm": 0.42943114042282104,
+      "learning_rate": 0.0001770894863893141,
+      "loss": 0.529,
+      "step": 6661
+    },
+    {
+      "epoch": 1.1883864062081884,
+      "grad_norm": 0.4512742757797241,
+      "learning_rate": 0.00017702247275347532,
+      "loss": 0.5871,
+      "step": 6662
+    },
+    {
+      "epoch": 1.1885648024261886,
+      "grad_norm": 0.5028432011604309,
+      "learning_rate": 0.00017695546484926012,
+      "loss": 0.5853,
+      "step": 6663
+    },
+    {
+      "epoch": 1.1887431986441888,
+      "grad_norm": 0.5111497044563293,
+      "learning_rate": 0.00017688846268193125,
+      "loss": 0.6258,
+      "step": 6664
+    },
+    {
+      "epoch": 1.188921594862189,
+      "grad_norm": 0.4514205753803253,
+      "learning_rate": 0.0001768214662567511,
+      "loss": 0.5353,
+      "step": 6665
+    },
+    {
+      "epoch": 1.1890999910801892,
+      "grad_norm": 0.4557707607746124,
+      "learning_rate": 0.0001767544755789815,
+      "loss": 0.6054,
+      "step": 6666
+    },
+    {
+      "epoch": 1.1892783872981894,
+      "grad_norm": 0.5693210959434509,
+      "learning_rate": 0.00017668749065388384,
+      "loss": 0.7786,
+      "step": 6667
+    },
+    {
+      "epoch": 1.1894567835161896,
+      "grad_norm": 0.47501084208488464,
+      "learning_rate": 0.00017662051148671914,
+      "loss": 0.6898,
+      "step": 6668
+    },
+    {
+      "epoch": 1.1896351797341898,
+      "grad_norm": 0.4343969225883484,
+      "learning_rate": 0.00017655353808274793,
+      "loss": 0.5845,
+      "step": 6669
+    },
+    {
+      "epoch": 1.1898135759521897,
+      "grad_norm": 0.4890352487564087,
+      "learning_rate": 0.00017648657044723007,
+      "loss": 0.579,
+      "step": 6670
+    },
+    {
+      "epoch": 1.18999197217019,
+      "grad_norm": 0.4660862386226654,
+      "learning_rate": 0.00017641960858542544,
+      "loss": 0.5289,
+      "step": 6671
+    },
+    {
+      "epoch": 1.1901703683881901,
+      "grad_norm": 0.49059972167015076,
+      "learning_rate": 0.0001763526525025931,
+      "loss": 0.6006,
+      "step": 6672
+    },
+    {
+      "epoch": 1.1903487646061903,
+      "grad_norm": 0.4573802351951599,
+      "learning_rate": 0.00017628570220399177,
+      "loss": 0.5264,
+      "step": 6673
+    },
+    {
+      "epoch": 1.1905271608241905,
+      "grad_norm": 0.5544562935829163,
+      "learning_rate": 0.00017621875769487964,
+      "loss": 0.7761,
+      "step": 6674
+    },
+    {
+      "epoch": 1.1907055570421907,
+      "grad_norm": 0.4699319303035736,
+      "learning_rate": 0.00017615181898051452,
+      "loss": 0.604,
+      "step": 6675
+    },
+    {
+      "epoch": 1.190883953260191,
+      "grad_norm": 0.4405967891216278,
+      "learning_rate": 0.00017608488606615376,
+      "loss": 0.5839,
+      "step": 6676
+    },
+    {
+      "epoch": 1.1910623494781911,
+      "grad_norm": 0.4909313917160034,
+      "learning_rate": 0.00017601795895705422,
+      "loss": 0.6793,
+      "step": 6677
+    },
+    {
+      "epoch": 1.1912407456961913,
+      "grad_norm": 0.5177908539772034,
+      "learning_rate": 0.00017595103765847238,
+      "loss": 0.7589,
+      "step": 6678
+    },
+    {
+      "epoch": 1.1914191419141915,
+      "grad_norm": 0.4939672648906708,
+      "learning_rate": 0.00017588412217566413,
+      "loss": 0.6188,
+      "step": 6679
+    },
+    {
+      "epoch": 1.1915975381321915,
+      "grad_norm": 0.4704304039478302,
+      "learning_rate": 0.00017581721251388506,
+      "loss": 0.5345,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1917759343501917,
+      "grad_norm": 0.43384838104248047,
+      "learning_rate": 0.0001757503086783902,
+      "loss": 0.5269,
+      "step": 6681
+    },
+    {
+      "epoch": 1.1919543305681919,
+      "grad_norm": 0.4727664887905121,
+      "learning_rate": 0.0001756834106744342,
+      "loss": 0.7044,
+      "step": 6682
+    },
+    {
+      "epoch": 1.192132726786192,
+      "grad_norm": 0.5356529355049133,
+      "learning_rate": 0.00017561651850727105,
+      "loss": 0.8679,
+      "step": 6683
+    },
+    {
+      "epoch": 1.1923111230041923,
+      "grad_norm": 0.45256125926971436,
+      "learning_rate": 0.00017554963218215458,
+      "loss": 0.6471,
+      "step": 6684
+    },
+    {
+      "epoch": 1.1924895192221925,
+      "grad_norm": 0.6792163252830505,
+      "learning_rate": 0.00017548275170433783,
+      "loss": 0.713,
+      "step": 6685
+    },
+    {
+      "epoch": 1.1926679154401927,
+      "grad_norm": 0.42677950859069824,
+      "learning_rate": 0.00017541587707907387,
+      "loss": 0.4628,
+      "step": 6686
+    },
+    {
+      "epoch": 1.1928463116581929,
+      "grad_norm": 0.5247853994369507,
+      "learning_rate": 0.00017534900831161476,
+      "loss": 0.8362,
+      "step": 6687
+    },
+    {
+      "epoch": 1.193024707876193,
+      "grad_norm": 0.48497462272644043,
+      "learning_rate": 0.00017528214540721241,
+      "loss": 0.6612,
+      "step": 6688
+    },
+    {
+      "epoch": 1.1932031040941933,
+      "grad_norm": 0.43292883038520813,
+      "learning_rate": 0.0001752152883711182,
+      "loss": 0.5158,
+      "step": 6689
+    },
+    {
+      "epoch": 1.1933815003121935,
+      "grad_norm": 0.5147132277488708,
+      "learning_rate": 0.00017514843720858308,
+      "loss": 0.795,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1935598965301937,
+      "grad_norm": 0.5482494831085205,
+      "learning_rate": 0.00017508159192485746,
+      "loss": 0.6979,
+      "step": 6691
+    },
+    {
+      "epoch": 1.1937382927481937,
+      "grad_norm": 0.5331830978393555,
+      "learning_rate": 0.0001750147525251914,
+      "loss": 0.5722,
+      "step": 6692
+    },
+    {
+      "epoch": 1.1939166889661939,
+      "grad_norm": 0.5137856006622314,
+      "learning_rate": 0.0001749479190148344,
+      "loss": 0.7025,
+      "step": 6693
+    },
+    {
+      "epoch": 1.194095085184194,
+      "grad_norm": 0.4909903109073639,
+      "learning_rate": 0.00017488109139903558,
+      "loss": 0.5722,
+      "step": 6694
+    },
+    {
+      "epoch": 1.1942734814021942,
+      "grad_norm": 0.5011717081069946,
+      "learning_rate": 0.00017481426968304347,
+      "loss": 0.6142,
+      "step": 6695
+    },
+    {
+      "epoch": 1.1944518776201944,
+      "grad_norm": 0.635654628276825,
+      "learning_rate": 0.00017474745387210627,
+      "loss": 0.6943,
+      "step": 6696
+    },
+    {
+      "epoch": 1.1946302738381946,
+      "grad_norm": 0.5469489097595215,
+      "learning_rate": 0.00017468064397147166,
+      "loss": 0.6825,
+      "step": 6697
+    },
+    {
+      "epoch": 1.1948086700561948,
+      "grad_norm": 0.5090768337249756,
+      "learning_rate": 0.00017461383998638685,
+      "loss": 0.6433,
+      "step": 6698
+    },
+    {
+      "epoch": 1.194987066274195,
+      "grad_norm": 0.5082751512527466,
+      "learning_rate": 0.00017454704192209863,
+      "loss": 0.6789,
+      "step": 6699
+    },
+    {
+      "epoch": 1.1951654624921952,
+      "grad_norm": 0.4537877142429352,
+      "learning_rate": 0.0001744802497838532,
+      "loss": 0.6015,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1953438587101954,
+      "grad_norm": 0.4903229773044586,
+      "learning_rate": 0.00017441346357689651,
+      "loss": 0.6014,
+      "step": 6701
+    },
+    {
+      "epoch": 1.1955222549281954,
+      "grad_norm": 0.4441526532173157,
+      "learning_rate": 0.00017434668330647385,
+      "loss": 0.5323,
+      "step": 6702
+    },
+    {
+      "epoch": 1.1957006511461956,
+      "grad_norm": 0.532263994216919,
+      "learning_rate": 0.00017427990897783013,
+      "loss": 0.729,
+      "step": 6703
+    },
+    {
+      "epoch": 1.1958790473641958,
+      "grad_norm": 0.49677395820617676,
+      "learning_rate": 0.0001742131405962098,
+      "loss": 0.646,
+      "step": 6704
+    },
+    {
+      "epoch": 1.196057443582196,
+      "grad_norm": 0.5097184181213379,
+      "learning_rate": 0.00017414637816685677,
+      "loss": 0.688,
+      "step": 6705
+    },
+    {
+      "epoch": 1.1962358398001962,
+      "grad_norm": 0.4844436049461365,
+      "learning_rate": 0.00017407962169501456,
+      "loss": 0.6241,
+      "step": 6706
+    },
+    {
+      "epoch": 1.1964142360181964,
+      "grad_norm": 0.49923595786094666,
+      "learning_rate": 0.00017401287118592624,
+      "loss": 0.7081,
+      "step": 6707
+    },
+    {
+      "epoch": 1.1965926322361966,
+      "grad_norm": 0.49586620926856995,
+      "learning_rate": 0.00017394612664483429,
+      "loss": 0.6703,
+      "step": 6708
+    },
+    {
+      "epoch": 1.1967710284541968,
+      "grad_norm": 0.5127303004264832,
+      "learning_rate": 0.00017387938807698078,
+      "loss": 0.5966,
+      "step": 6709
+    },
+    {
+      "epoch": 1.196949424672197,
+      "grad_norm": 0.5159239172935486,
+      "learning_rate": 0.0001738126554876074,
+      "loss": 0.6792,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1971278208901972,
+      "grad_norm": 0.45261603593826294,
+      "learning_rate": 0.0001737459288819553,
+      "loss": 0.5966,
+      "step": 6711
+    },
+    {
+      "epoch": 1.1973062171081974,
+      "grad_norm": 0.4482129216194153,
+      "learning_rate": 0.00017367920826526508,
+      "loss": 0.5745,
+      "step": 6712
+    },
+    {
+      "epoch": 1.1974846133261976,
+      "grad_norm": 0.49750131368637085,
+      "learning_rate": 0.000173612493642777,
+      "loss": 0.6134,
+      "step": 6713
+    },
+    {
+      "epoch": 1.1976630095441976,
+      "grad_norm": 0.5099869966506958,
+      "learning_rate": 0.00017354578501973083,
+      "loss": 0.8435,
+      "step": 6714
+    },
+    {
+      "epoch": 1.1978414057621978,
+      "grad_norm": 0.4573536813259125,
+      "learning_rate": 0.00017347908240136585,
+      "loss": 0.5377,
+      "step": 6715
+    },
+    {
+      "epoch": 1.198019801980198,
+      "grad_norm": 0.5403563380241394,
+      "learning_rate": 0.00017341238579292063,
+      "loss": 0.792,
+      "step": 6716
+    },
+    {
+      "epoch": 1.1981981981981982,
+      "grad_norm": 0.44874098896980286,
+      "learning_rate": 0.00017334569519963377,
+      "loss": 0.6036,
+      "step": 6717
+    },
+    {
+      "epoch": 1.1983765944161984,
+      "grad_norm": 0.5061765313148499,
+      "learning_rate": 0.00017327901062674306,
+      "loss": 0.6612,
+      "step": 6718
+    },
+    {
+      "epoch": 1.1985549906341986,
+      "grad_norm": 0.4852476418018341,
+      "learning_rate": 0.00017321233207948583,
+      "loss": 0.5551,
+      "step": 6719
+    },
+    {
+      "epoch": 1.1987333868521988,
+      "grad_norm": 0.7161738276481628,
+      "learning_rate": 0.00017314565956309903,
+      "loss": 0.5223,
+      "step": 6720
+    },
+    {
+      "epoch": 1.198911783070199,
+      "grad_norm": 0.5031499266624451,
+      "learning_rate": 0.000173078993082819,
+      "loss": 0.7557,
+      "step": 6721
+    },
+    {
+      "epoch": 1.1990901792881992,
+      "grad_norm": 1.0971347093582153,
+      "learning_rate": 0.00017301233264388176,
+      "loss": 0.6716,
+      "step": 6722
+    },
+    {
+      "epoch": 1.1992685755061994,
+      "grad_norm": 0.6954957842826843,
+      "learning_rate": 0.0001729456782515228,
+      "loss": 0.669,
+      "step": 6723
+    },
+    {
+      "epoch": 1.1994469717241993,
+      "grad_norm": 0.5447986125946045,
+      "learning_rate": 0.0001728790299109771,
+      "loss": 0.7785,
+      "step": 6724
+    },
+    {
+      "epoch": 1.1996253679421995,
+      "grad_norm": 0.5347351431846619,
+      "learning_rate": 0.00017281238762747919,
+      "loss": 0.7014,
+      "step": 6725
+    },
+    {
+      "epoch": 1.1998037641601997,
+      "grad_norm": 0.5047118663787842,
+      "learning_rate": 0.00017274575140626317,
+      "loss": 0.574,
+      "step": 6726
+    },
+    {
+      "epoch": 1.1999821603782,
+      "grad_norm": 0.5407329797744751,
+      "learning_rate": 0.0001726791212525626,
+      "loss": 0.7382,
+      "step": 6727
+    },
+    {
+      "epoch": 1.2001605565962001,
+      "grad_norm": 0.51371830701828,
+      "learning_rate": 0.00017261249717161054,
+      "loss": 0.779,
+      "step": 6728
+    },
+    {
+      "epoch": 1.2003389528142003,
+      "grad_norm": 0.4466804563999176,
+      "learning_rate": 0.00017254587916863969,
+      "loss": 0.5177,
+      "step": 6729
+    },
+    {
+      "epoch": 1.2005173490322005,
+      "grad_norm": 0.45748934149742126,
+      "learning_rate": 0.0001724792672488821,
+      "loss": 0.6624,
+      "step": 6730
+    },
+    {
+      "epoch": 1.2006957452502007,
+      "grad_norm": 0.5126693844795227,
+      "learning_rate": 0.0001724126614175694,
+      "loss": 0.7792,
+      "step": 6731
+    },
+    {
+      "epoch": 1.200874141468201,
+      "grad_norm": 0.46355128288269043,
+      "learning_rate": 0.00017234606167993303,
+      "loss": 0.6419,
+      "step": 6732
+    },
+    {
+      "epoch": 1.2010525376862011,
+      "grad_norm": 0.4698454439640045,
+      "learning_rate": 0.0001722794680412036,
+      "loss": 0.6422,
+      "step": 6733
+    },
+    {
+      "epoch": 1.2012309339042013,
+      "grad_norm": 0.4981541335582733,
+      "learning_rate": 0.0001722128805066112,
+      "loss": 0.6687,
+      "step": 6734
+    },
+    {
+      "epoch": 1.2014093301222015,
+      "grad_norm": 0.5997437834739685,
+      "learning_rate": 0.00017214629908138574,
+      "loss": 0.9846,
+      "step": 6735
+    },
+    {
+      "epoch": 1.2015877263402015,
+      "grad_norm": 0.4991457164287567,
+      "learning_rate": 0.0001720797237707564,
+      "loss": 0.6268,
+      "step": 6736
+    },
+    {
+      "epoch": 1.2017661225582017,
+      "grad_norm": 0.4917805790901184,
+      "learning_rate": 0.0001720131545799521,
+      "loss": 0.62,
+      "step": 6737
+    },
+    {
+      "epoch": 1.201944518776202,
+      "grad_norm": 0.45788219571113586,
+      "learning_rate": 0.00017194659151420106,
+      "loss": 0.7157,
+      "step": 6738
+    },
+    {
+      "epoch": 1.202122914994202,
+      "grad_norm": 0.49863317608833313,
+      "learning_rate": 0.0001718800345787311,
+      "loss": 0.6979,
+      "step": 6739
+    },
+    {
+      "epoch": 1.2023013112122023,
+      "grad_norm": 0.591437578201294,
+      "learning_rate": 0.00017181348377876958,
+      "loss": 0.6895,
+      "step": 6740
+    },
+    {
+      "epoch": 1.2024797074302025,
+      "grad_norm": 0.4187926650047302,
+      "learning_rate": 0.00017174693911954354,
+      "loss": 0.5386,
+      "step": 6741
+    },
+    {
+      "epoch": 1.2026581036482027,
+      "grad_norm": 0.4731749892234802,
+      "learning_rate": 0.0001716804006062791,
+      "loss": 0.6097,
+      "step": 6742
+    },
+    {
+      "epoch": 1.202836499866203,
+      "grad_norm": 0.4960040748119354,
+      "learning_rate": 0.00017161386824420232,
+      "loss": 0.7277,
+      "step": 6743
+    },
+    {
+      "epoch": 1.203014896084203,
+      "grad_norm": 0.47977831959724426,
+      "learning_rate": 0.00017154734203853858,
+      "loss": 0.6146,
+      "step": 6744
+    },
+    {
+      "epoch": 1.2031932923022033,
+      "grad_norm": 0.6268126368522644,
+      "learning_rate": 0.00017148082199451288,
+      "loss": 0.6726,
+      "step": 6745
+    },
+    {
+      "epoch": 1.2033716885202033,
+      "grad_norm": 0.5123054385185242,
+      "learning_rate": 0.0001714143081173495,
+      "loss": 0.8116,
+      "step": 6746
+    },
+    {
+      "epoch": 1.2035500847382035,
+      "grad_norm": 0.45163649320602417,
+      "learning_rate": 0.00017134780041227265,
+      "loss": 0.6415,
+      "step": 6747
+    },
+    {
+      "epoch": 1.2037284809562037,
+      "grad_norm": 0.511045515537262,
+      "learning_rate": 0.00017128129888450573,
+      "loss": 0.5628,
+      "step": 6748
+    },
+    {
+      "epoch": 1.2039068771742039,
+      "grad_norm": 0.44558185338974,
+      "learning_rate": 0.00017121480353927165,
+      "loss": 0.5404,
+      "step": 6749
+    },
+    {
+      "epoch": 1.204085273392204,
+      "grad_norm": 0.5241187810897827,
+      "learning_rate": 0.00017114831438179304,
+      "loss": 0.7767,
+      "step": 6750
+    },
+    {
+      "epoch": 1.2042636696102043,
+      "grad_norm": 0.722863495349884,
+      "learning_rate": 0.00017108183141729188,
+      "loss": 0.685,
+      "step": 6751
+    },
+    {
+      "epoch": 1.2044420658282045,
+      "grad_norm": 0.6952433586120605,
+      "learning_rate": 0.00017101535465098973,
+      "loss": 0.6479,
+      "step": 6752
+    },
+    {
+      "epoch": 1.2046204620462047,
+      "grad_norm": 0.48947012424468994,
+      "learning_rate": 0.00017094888408810763,
+      "loss": 0.749,
+      "step": 6753
+    },
+    {
+      "epoch": 1.2047988582642049,
+      "grad_norm": 0.4884073734283447,
+      "learning_rate": 0.0001708824197338662,
+      "loss": 0.727,
+      "step": 6754
+    },
+    {
+      "epoch": 1.204977254482205,
+      "grad_norm": 0.5633352398872375,
+      "learning_rate": 0.00017081596159348544,
+      "loss": 0.8563,
+      "step": 6755
+    },
+    {
+      "epoch": 1.2051556507002052,
+      "grad_norm": 0.48596227169036865,
+      "learning_rate": 0.00017074950967218495,
+      "loss": 0.6633,
+      "step": 6756
+    },
+    {
+      "epoch": 1.2053340469182054,
+      "grad_norm": 0.45270687341690063,
+      "learning_rate": 0.0001706830639751839,
+      "loss": 0.5795,
+      "step": 6757
+    },
+    {
+      "epoch": 1.2055124431362054,
+      "grad_norm": 0.48366621136665344,
+      "learning_rate": 0.00017061662450770085,
+      "loss": 0.7047,
+      "step": 6758
+    },
+    {
+      "epoch": 1.2056908393542056,
+      "grad_norm": 0.48411786556243896,
+      "learning_rate": 0.00017055019127495396,
+      "loss": 0.7534,
+      "step": 6759
+    },
+    {
+      "epoch": 1.2058692355722058,
+      "grad_norm": 0.4793466627597809,
+      "learning_rate": 0.00017048376428216083,
+      "loss": 0.6564,
+      "step": 6760
+    },
+    {
+      "epoch": 1.206047631790206,
+      "grad_norm": 0.4573822021484375,
+      "learning_rate": 0.0001704173435345387,
+      "loss": 0.4675,
+      "step": 6761
+    },
+    {
+      "epoch": 1.2062260280082062,
+      "grad_norm": 0.540582537651062,
+      "learning_rate": 0.00017035092903730403,
+      "loss": 0.7691,
+      "step": 6762
+    },
+    {
+      "epoch": 1.2064044242262064,
+      "grad_norm": 0.5048598647117615,
+      "learning_rate": 0.00017028452079567325,
+      "loss": 0.7058,
+      "step": 6763
+    },
+    {
+      "epoch": 1.2065828204442066,
+      "grad_norm": 0.46692147850990295,
+      "learning_rate": 0.00017021811881486184,
+      "loss": 0.5641,
+      "step": 6764
+    },
+    {
+      "epoch": 1.2067612166622068,
+      "grad_norm": 0.4819284975528717,
+      "learning_rate": 0.00017015172310008508,
+      "loss": 0.733,
+      "step": 6765
+    },
+    {
+      "epoch": 1.206939612880207,
+      "grad_norm": 0.4647090435028076,
+      "learning_rate": 0.00017008533365655765,
+      "loss": 0.5602,
+      "step": 6766
+    },
+    {
+      "epoch": 1.2071180090982072,
+      "grad_norm": 0.42821335792541504,
+      "learning_rate": 0.0001700189504894937,
+      "loss": 0.5918,
+      "step": 6767
+    },
+    {
+      "epoch": 1.2072964053162072,
+      "grad_norm": 0.42195186018943787,
+      "learning_rate": 0.00016995257360410694,
+      "loss": 0.5887,
+      "step": 6768
+    },
+    {
+      "epoch": 1.2074748015342074,
+      "grad_norm": 0.5590615272521973,
+      "learning_rate": 0.00016988620300561065,
+      "loss": 0.5942,
+      "step": 6769
+    },
+    {
+      "epoch": 1.2076531977522076,
+      "grad_norm": 0.459682822227478,
+      "learning_rate": 0.0001698198386992174,
+      "loss": 0.6768,
+      "step": 6770
+    },
+    {
+      "epoch": 1.2078315939702078,
+      "grad_norm": 0.5003504157066345,
+      "learning_rate": 0.0001697534806901396,
+      "loss": 0.7413,
+      "step": 6771
+    },
+    {
+      "epoch": 1.208009990188208,
+      "grad_norm": 0.4521839916706085,
+      "learning_rate": 0.00016968712898358888,
+      "loss": 0.5958,
+      "step": 6772
+    },
+    {
+      "epoch": 1.2081883864062082,
+      "grad_norm": 0.47012245655059814,
+      "learning_rate": 0.00016962078358477648,
+      "loss": 0.6722,
+      "step": 6773
+    },
+    {
+      "epoch": 1.2083667826242084,
+      "grad_norm": 0.4755409359931946,
+      "learning_rate": 0.0001695544444989132,
+      "loss": 0.7082,
+      "step": 6774
+    },
+    {
+      "epoch": 1.2085451788422086,
+      "grad_norm": 0.44018933176994324,
+      "learning_rate": 0.00016948811173120914,
+      "loss": 0.6281,
+      "step": 6775
+    },
+    {
+      "epoch": 1.2087235750602088,
+      "grad_norm": 0.5177139639854431,
+      "learning_rate": 0.00016942178528687419,
+      "loss": 0.829,
+      "step": 6776
+    },
+    {
+      "epoch": 1.208901971278209,
+      "grad_norm": 0.42148521542549133,
+      "learning_rate": 0.00016935546517111744,
+      "loss": 0.5036,
+      "step": 6777
+    },
+    {
+      "epoch": 1.2090803674962092,
+      "grad_norm": 0.4840809404850006,
+      "learning_rate": 0.00016928915138914787,
+      "loss": 0.6851,
+      "step": 6778
+    },
+    {
+      "epoch": 1.2092587637142094,
+      "grad_norm": 0.4874308407306671,
+      "learning_rate": 0.0001692228439461736,
+      "loss": 0.7215,
+      "step": 6779
+    },
+    {
+      "epoch": 1.2094371599322093,
+      "grad_norm": 0.5354540944099426,
+      "learning_rate": 0.0001691565428474024,
+      "loss": 0.6564,
+      "step": 6780
+    },
+    {
+      "epoch": 1.2096155561502095,
+      "grad_norm": 0.47135040163993835,
+      "learning_rate": 0.0001690902480980415,
+      "loss": 0.6163,
+      "step": 6781
+    },
+    {
+      "epoch": 1.2097939523682097,
+      "grad_norm": 0.5024769902229309,
+      "learning_rate": 0.00016902395970329776,
+      "loss": 0.7379,
+      "step": 6782
+    },
+    {
+      "epoch": 1.20997234858621,
+      "grad_norm": 0.4286527633666992,
+      "learning_rate": 0.00016895767766837732,
+      "loss": 0.4737,
+      "step": 6783
+    },
+    {
+      "epoch": 1.2101507448042101,
+      "grad_norm": 0.4754869043827057,
+      "learning_rate": 0.00016889140199848605,
+      "loss": 0.5874,
+      "step": 6784
+    },
+    {
+      "epoch": 1.2103291410222103,
+      "grad_norm": 0.5316412448883057,
+      "learning_rate": 0.00016882513269882916,
+      "loss": 0.8397,
+      "step": 6785
+    },
+    {
+      "epoch": 1.2105075372402105,
+      "grad_norm": 0.46648043394088745,
+      "learning_rate": 0.00016875886977461136,
+      "loss": 0.6463,
+      "step": 6786
+    },
+    {
+      "epoch": 1.2106859334582107,
+      "grad_norm": 0.5333806276321411,
+      "learning_rate": 0.00016869261323103707,
+      "loss": 0.845,
+      "step": 6787
+    },
+    {
+      "epoch": 1.210864329676211,
+      "grad_norm": 0.48115333914756775,
+      "learning_rate": 0.00016862636307330987,
+      "loss": 0.6568,
+      "step": 6788
+    },
+    {
+      "epoch": 1.2110427258942111,
+      "grad_norm": 0.524509072303772,
+      "learning_rate": 0.00016856011930663312,
+      "loss": 0.8149,
+      "step": 6789
+    },
+    {
+      "epoch": 1.2112211221122111,
+      "grad_norm": 0.4812416732311249,
+      "learning_rate": 0.0001684938819362095,
+      "loss": 0.5716,
+      "step": 6790
+    },
+    {
+      "epoch": 1.2113995183302113,
+      "grad_norm": 0.4996643662452698,
+      "learning_rate": 0.0001684276509672414,
+      "loss": 0.8065,
+      "step": 6791
+    },
+    {
+      "epoch": 1.2115779145482115,
+      "grad_norm": 0.5265586376190186,
+      "learning_rate": 0.0001683614264049304,
+      "loss": 0.7074,
+      "step": 6792
+    },
+    {
+      "epoch": 1.2117563107662117,
+      "grad_norm": 0.4892607033252716,
+      "learning_rate": 0.00016829520825447787,
+      "loss": 0.7555,
+      "step": 6793
+    },
+    {
+      "epoch": 1.211934706984212,
+      "grad_norm": 0.493960440158844,
+      "learning_rate": 0.00016822899652108454,
+      "loss": 0.8655,
+      "step": 6794
+    },
+    {
+      "epoch": 1.212113103202212,
+      "grad_norm": 0.5139688849449158,
+      "learning_rate": 0.00016816279120995063,
+      "loss": 0.8795,
+      "step": 6795
+    },
+    {
+      "epoch": 1.2122914994202123,
+      "grad_norm": 0.41714024543762207,
+      "learning_rate": 0.00016809659232627588,
+      "loss": 0.4714,
+      "step": 6796
+    },
+    {
+      "epoch": 1.2124698956382125,
+      "grad_norm": 0.5181586742401123,
+      "learning_rate": 0.00016803039987525953,
+      "loss": 0.7304,
+      "step": 6797
+    },
+    {
+      "epoch": 1.2126482918562127,
+      "grad_norm": 0.47298526763916016,
+      "learning_rate": 0.00016796421386210028,
+      "loss": 0.5293,
+      "step": 6798
+    },
+    {
+      "epoch": 1.212826688074213,
+      "grad_norm": 0.47888705134391785,
+      "learning_rate": 0.0001678980342919964,
+      "loss": 0.5461,
+      "step": 6799
+    },
+    {
+      "epoch": 1.213005084292213,
+      "grad_norm": 0.5607367753982544,
+      "learning_rate": 0.0001678318611701456,
+      "loss": 0.8278,
+      "step": 6800
+    },
+    {
+      "epoch": 1.2131834805102133,
+      "grad_norm": 0.5705345869064331,
+      "learning_rate": 0.00016776569450174504,
+      "loss": 0.786,
+      "step": 6801
+    },
+    {
+      "epoch": 1.2133618767282133,
+      "grad_norm": 0.5020168423652649,
+      "learning_rate": 0.00016769953429199142,
+      "loss": 0.7604,
+      "step": 6802
+    },
+    {
+      "epoch": 1.2135402729462135,
+      "grad_norm": 0.4835543632507324,
+      "learning_rate": 0.00016763338054608096,
+      "loss": 0.6996,
+      "step": 6803
+    },
+    {
+      "epoch": 1.2137186691642137,
+      "grad_norm": 0.44707992672920227,
+      "learning_rate": 0.00016756723326920937,
+      "loss": 0.6119,
+      "step": 6804
+    },
+    {
+      "epoch": 1.2138970653822139,
+      "grad_norm": 0.40079420804977417,
+      "learning_rate": 0.0001675010924665718,
+      "loss": 0.5236,
+      "step": 6805
+    },
+    {
+      "epoch": 1.214075461600214,
+      "grad_norm": 0.4688962697982788,
+      "learning_rate": 0.0001674349581433629,
+      "loss": 0.5646,
+      "step": 6806
+    },
+    {
+      "epoch": 1.2142538578182143,
+      "grad_norm": 0.46147796511650085,
+      "learning_rate": 0.0001673688303047769,
+      "loss": 0.5085,
+      "step": 6807
+    },
+    {
+      "epoch": 1.2144322540362145,
+      "grad_norm": 0.4987983703613281,
+      "learning_rate": 0.00016730270895600732,
+      "loss": 0.7102,
+      "step": 6808
+    },
+    {
+      "epoch": 1.2146106502542147,
+      "grad_norm": 0.4857620894908905,
+      "learning_rate": 0.00016723659410224746,
+      "loss": 0.6561,
+      "step": 6809
+    },
+    {
+      "epoch": 1.2147890464722149,
+      "grad_norm": 0.5185416340827942,
+      "learning_rate": 0.00016717048574868987,
+      "loss": 0.688,
+      "step": 6810
+    },
+    {
+      "epoch": 1.214967442690215,
+      "grad_norm": 0.43132469058036804,
+      "learning_rate": 0.0001671043839005267,
+      "loss": 0.6281,
+      "step": 6811
+    },
+    {
+      "epoch": 1.215145838908215,
+      "grad_norm": 0.5003576278686523,
+      "learning_rate": 0.00016703828856294955,
+      "loss": 0.7138,
+      "step": 6812
+    },
+    {
+      "epoch": 1.2153242351262152,
+      "grad_norm": 0.5227103233337402,
+      "learning_rate": 0.00016697219974114955,
+      "loss": 0.7263,
+      "step": 6813
+    },
+    {
+      "epoch": 1.2155026313442154,
+      "grad_norm": 0.46056926250457764,
+      "learning_rate": 0.0001669061174403172,
+      "loss": 0.6924,
+      "step": 6814
+    },
+    {
+      "epoch": 1.2156810275622156,
+      "grad_norm": 0.5291325449943542,
+      "learning_rate": 0.00016684004166564264,
+      "loss": 0.7044,
+      "step": 6815
+    },
+    {
+      "epoch": 1.2158594237802158,
+      "grad_norm": 0.3964986503124237,
+      "learning_rate": 0.0001667739724223154,
+      "loss": 0.4319,
+      "step": 6816
+    },
+    {
+      "epoch": 1.216037819998216,
+      "grad_norm": 0.49517232179641724,
+      "learning_rate": 0.00016670790971552458,
+      "loss": 0.5721,
+      "step": 6817
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 0.47603654861450195,
+      "learning_rate": 0.0001666418535504587,
+      "loss": 0.6336,
+      "step": 6818
+    },
+    {
+      "epoch": 1.2163946124342164,
+      "grad_norm": 0.4822855293750763,
+      "learning_rate": 0.00016657580393230573,
+      "loss": 0.7157,
+      "step": 6819
+    },
+    {
+      "epoch": 1.2165730086522166,
+      "grad_norm": 0.45155301690101624,
+      "learning_rate": 0.00016650976086625324,
+      "loss": 0.5622,
+      "step": 6820
+    },
+    {
+      "epoch": 1.2167514048702168,
+      "grad_norm": 0.4612867832183838,
+      "learning_rate": 0.00016644372435748822,
+      "loss": 0.5382,
+      "step": 6821
+    },
+    {
+      "epoch": 1.216929801088217,
+      "grad_norm": 0.44989264011383057,
+      "learning_rate": 0.00016637769441119713,
+      "loss": 0.5503,
+      "step": 6822
+    },
+    {
+      "epoch": 1.2171081973062172,
+      "grad_norm": 0.5135206580162048,
+      "learning_rate": 0.00016631167103256582,
+      "loss": 0.7834,
+      "step": 6823
+    },
+    {
+      "epoch": 1.2172865935242172,
+      "grad_norm": 0.5145620107650757,
+      "learning_rate": 0.00016624565422677996,
+      "loss": 0.8026,
+      "step": 6824
+    },
+    {
+      "epoch": 1.2174649897422174,
+      "grad_norm": 0.4703541398048401,
+      "learning_rate": 0.0001661796439990244,
+      "loss": 0.677,
+      "step": 6825
+    },
+    {
+      "epoch": 1.2176433859602176,
+      "grad_norm": 0.4866558611392975,
+      "learning_rate": 0.00016611364035448348,
+      "loss": 0.6063,
+      "step": 6826
+    },
+    {
+      "epoch": 1.2178217821782178,
+      "grad_norm": 0.5036291480064392,
+      "learning_rate": 0.00016604764329834117,
+      "loss": 0.8213,
+      "step": 6827
+    },
+    {
+      "epoch": 1.218000178396218,
+      "grad_norm": 0.4724942743778229,
+      "learning_rate": 0.00016598165283578082,
+      "loss": 0.6396,
+      "step": 6828
+    },
+    {
+      "epoch": 1.2181785746142182,
+      "grad_norm": 0.44807612895965576,
+      "learning_rate": 0.0001659156689719853,
+      "loss": 0.5442,
+      "step": 6829
+    },
+    {
+      "epoch": 1.2183569708322184,
+      "grad_norm": 0.42465436458587646,
+      "learning_rate": 0.00016584969171213693,
+      "loss": 0.4956,
+      "step": 6830
+    },
+    {
+      "epoch": 1.2185353670502186,
+      "grad_norm": 0.9363518357276917,
+      "learning_rate": 0.0001657837210614176,
+      "loss": 0.8942,
+      "step": 6831
+    },
+    {
+      "epoch": 1.2187137632682188,
+      "grad_norm": 0.44213372468948364,
+      "learning_rate": 0.00016571775702500856,
+      "loss": 0.4695,
+      "step": 6832
+    },
+    {
+      "epoch": 1.218892159486219,
+      "grad_norm": 0.4889032542705536,
+      "learning_rate": 0.0001656517996080906,
+      "loss": 0.4976,
+      "step": 6833
+    },
+    {
+      "epoch": 1.219070555704219,
+      "grad_norm": 0.5484008193016052,
+      "learning_rate": 0.00016558584881584408,
+      "loss": 0.7479,
+      "step": 6834
+    },
+    {
+      "epoch": 1.2192489519222192,
+      "grad_norm": 0.4762975573539734,
+      "learning_rate": 0.00016551990465344857,
+      "loss": 0.6117,
+      "step": 6835
+    },
+    {
+      "epoch": 1.2194273481402194,
+      "grad_norm": 0.5832328200340271,
+      "learning_rate": 0.00016545396712608346,
+      "loss": 0.7153,
+      "step": 6836
+    },
+    {
+      "epoch": 1.2196057443582196,
+      "grad_norm": 0.46144014596939087,
+      "learning_rate": 0.00016538803623892734,
+      "loss": 0.568,
+      "step": 6837
+    },
+    {
+      "epoch": 1.2197841405762198,
+      "grad_norm": 0.4600122570991516,
+      "learning_rate": 0.0001653221119971584,
+      "loss": 0.5168,
+      "step": 6838
+    },
+    {
+      "epoch": 1.21996253679422,
+      "grad_norm": 0.51224684715271,
+      "learning_rate": 0.0001652561944059544,
+      "loss": 0.77,
+      "step": 6839
+    },
+    {
+      "epoch": 1.2201409330122202,
+      "grad_norm": 0.529111921787262,
+      "learning_rate": 0.00016519028347049242,
+      "loss": 0.7114,
+      "step": 6840
+    },
+    {
+      "epoch": 1.2203193292302204,
+      "grad_norm": 0.49239465594291687,
+      "learning_rate": 0.00016512437919594908,
+      "loss": 0.7044,
+      "step": 6841
+    },
+    {
+      "epoch": 1.2204977254482205,
+      "grad_norm": 0.4696052670478821,
+      "learning_rate": 0.00016505848158750047,
+      "loss": 0.647,
+      "step": 6842
+    },
+    {
+      "epoch": 1.2206761216662207,
+      "grad_norm": 0.4317275285720825,
+      "learning_rate": 0.00016499259065032217,
+      "loss": 0.5817,
+      "step": 6843
+    },
+    {
+      "epoch": 1.220854517884221,
+      "grad_norm": 0.5241662263870239,
+      "learning_rate": 0.00016492670638958924,
+      "loss": 0.7357,
+      "step": 6844
+    },
+    {
+      "epoch": 1.2210329141022211,
+      "grad_norm": 0.5198963284492493,
+      "learning_rate": 0.00016486082881047616,
+      "loss": 0.5708,
+      "step": 6845
+    },
+    {
+      "epoch": 1.2212113103202211,
+      "grad_norm": 0.49722328782081604,
+      "learning_rate": 0.00016479495791815702,
+      "loss": 0.6887,
+      "step": 6846
+    },
+    {
+      "epoch": 1.2213897065382213,
+      "grad_norm": 0.5908113121986389,
+      "learning_rate": 0.00016472909371780512,
+      "loss": 0.9454,
+      "step": 6847
+    },
+    {
+      "epoch": 1.2215681027562215,
+      "grad_norm": 0.5063602328300476,
+      "learning_rate": 0.00016466323621459352,
+      "loss": 0.6783,
+      "step": 6848
+    },
+    {
+      "epoch": 1.2217464989742217,
+      "grad_norm": 0.47944772243499756,
+      "learning_rate": 0.00016459738541369466,
+      "loss": 0.6294,
+      "step": 6849
+    },
+    {
+      "epoch": 1.221924895192222,
+      "grad_norm": 0.48459392786026,
+      "learning_rate": 0.00016453154132028036,
+      "loss": 0.6708,
+      "step": 6850
+    },
+    {
+      "epoch": 1.2221032914102221,
+      "grad_norm": 0.5024449229240417,
+      "learning_rate": 0.00016446570393952205,
+      "loss": 0.6274,
+      "step": 6851
+    },
+    {
+      "epoch": 1.2222816876282223,
+      "grad_norm": 0.4757004976272583,
+      "learning_rate": 0.0001643998732765905,
+      "loss": 0.6813,
+      "step": 6852
+    },
+    {
+      "epoch": 1.2224600838462225,
+      "grad_norm": 0.6229151487350464,
+      "learning_rate": 0.0001643340493366561,
+      "loss": 0.5887,
+      "step": 6853
+    },
+    {
+      "epoch": 1.2226384800642227,
+      "grad_norm": 0.5339584946632385,
+      "learning_rate": 0.00016426823212488856,
+      "loss": 0.8846,
+      "step": 6854
+    },
+    {
+      "epoch": 1.222816876282223,
+      "grad_norm": 0.4785784184932709,
+      "learning_rate": 0.0001642024216464572,
+      "loss": 0.5552,
+      "step": 6855
+    },
+    {
+      "epoch": 1.2229952725002229,
+      "grad_norm": 0.45622581243515015,
+      "learning_rate": 0.00016413661790653074,
+      "loss": 0.5302,
+      "step": 6856
+    },
+    {
+      "epoch": 1.223173668718223,
+      "grad_norm": 0.4625503420829773,
+      "learning_rate": 0.00016407082091027736,
+      "loss": 0.5837,
+      "step": 6857
+    },
+    {
+      "epoch": 1.2233520649362233,
+      "grad_norm": 0.49430111050605774,
+      "learning_rate": 0.00016400503066286472,
+      "loss": 0.5746,
+      "step": 6858
+    },
+    {
+      "epoch": 1.2235304611542235,
+      "grad_norm": 0.6408543586730957,
+      "learning_rate": 0.00016393924716946002,
+      "loss": 0.6712,
+      "step": 6859
+    },
+    {
+      "epoch": 1.2237088573722237,
+      "grad_norm": 0.5817059874534607,
+      "learning_rate": 0.00016387347043522976,
+      "loss": 0.8973,
+      "step": 6860
+    },
+    {
+      "epoch": 1.2238872535902239,
+      "grad_norm": 0.7297067642211914,
+      "learning_rate": 0.00016380770046534005,
+      "loss": 0.9795,
+      "step": 6861
+    },
+    {
+      "epoch": 1.224065649808224,
+      "grad_norm": 0.5279180407524109,
+      "learning_rate": 0.00016374193726495647,
+      "loss": 0.6417,
+      "step": 6862
+    },
+    {
+      "epoch": 1.2242440460262243,
+      "grad_norm": 0.45359596610069275,
+      "learning_rate": 0.00016367618083924402,
+      "loss": 0.6208,
+      "step": 6863
+    },
+    {
+      "epoch": 1.2244224422442245,
+      "grad_norm": 0.47759270668029785,
+      "learning_rate": 0.00016361043119336719,
+      "loss": 0.4829,
+      "step": 6864
+    },
+    {
+      "epoch": 1.2246008384622247,
+      "grad_norm": 0.4140789210796356,
+      "learning_rate": 0.00016354468833248992,
+      "loss": 0.3787,
+      "step": 6865
+    },
+    {
+      "epoch": 1.2247792346802249,
+      "grad_norm": 0.4323157072067261,
+      "learning_rate": 0.00016347895226177561,
+      "loss": 0.5246,
+      "step": 6866
+    },
+    {
+      "epoch": 1.224957630898225,
+      "grad_norm": 0.42192304134368896,
+      "learning_rate": 0.0001634132229863872,
+      "loss": 0.5802,
+      "step": 6867
+    },
+    {
+      "epoch": 1.225136027116225,
+      "grad_norm": 0.39509162306785583,
+      "learning_rate": 0.00016334750051148696,
+      "loss": 0.3766,
+      "step": 6868
+    },
+    {
+      "epoch": 1.2253144233342252,
+      "grad_norm": 0.5555423498153687,
+      "learning_rate": 0.0001632817848422366,
+      "loss": 0.8368,
+      "step": 6869
+    },
+    {
+      "epoch": 1.2254928195522254,
+      "grad_norm": 0.489187628030777,
+      "learning_rate": 0.00016321607598379767,
+      "loss": 0.6287,
+      "step": 6870
+    },
+    {
+      "epoch": 1.2256712157702256,
+      "grad_norm": 0.4797876477241516,
+      "learning_rate": 0.00016315037394133082,
+      "loss": 0.5631,
+      "step": 6871
+    },
+    {
+      "epoch": 1.2258496119882258,
+      "grad_norm": 0.5317756533622742,
+      "learning_rate": 0.00016308467871999622,
+      "loss": 0.6423,
+      "step": 6872
+    },
+    {
+      "epoch": 1.226028008206226,
+      "grad_norm": 0.5093753933906555,
+      "learning_rate": 0.00016301899032495354,
+      "loss": 0.7641,
+      "step": 6873
+    },
+    {
+      "epoch": 1.2262064044242262,
+      "grad_norm": 0.4402000606060028,
+      "learning_rate": 0.0001629533087613619,
+      "loss": 0.5262,
+      "step": 6874
+    },
+    {
+      "epoch": 1.2263848006422264,
+      "grad_norm": 0.605444073677063,
+      "learning_rate": 0.00016288763403437994,
+      "loss": 0.8924,
+      "step": 6875
+    },
+    {
+      "epoch": 1.2265631968602266,
+      "grad_norm": 0.5263209342956543,
+      "learning_rate": 0.00016282196614916572,
+      "loss": 0.5979,
+      "step": 6876
+    },
+    {
+      "epoch": 1.2267415930782268,
+      "grad_norm": 0.49599960446357727,
+      "learning_rate": 0.0001627563051108768,
+      "loss": 0.554,
+      "step": 6877
+    },
+    {
+      "epoch": 1.2269199892962268,
+      "grad_norm": 0.5215542912483215,
+      "learning_rate": 0.0001626906509246701,
+      "loss": 0.673,
+      "step": 6878
+    },
+    {
+      "epoch": 1.227098385514227,
+      "grad_norm": 0.48019513487815857,
+      "learning_rate": 0.0001626250035957021,
+      "loss": 0.6647,
+      "step": 6879
+    },
+    {
+      "epoch": 1.2272767817322272,
+      "grad_norm": 0.5026001334190369,
+      "learning_rate": 0.00016255936312912876,
+      "loss": 0.5802,
+      "step": 6880
+    },
+    {
+      "epoch": 1.2274551779502274,
+      "grad_norm": 0.5892630815505981,
+      "learning_rate": 0.00016249372953010537,
+      "loss": 0.7349,
+      "step": 6881
+    },
+    {
+      "epoch": 1.2276335741682276,
+      "grad_norm": 0.49261564016342163,
+      "learning_rate": 0.00016242810280378678,
+      "loss": 0.6885,
+      "step": 6882
+    },
+    {
+      "epoch": 1.2278119703862278,
+      "grad_norm": 0.5151751637458801,
+      "learning_rate": 0.00016236248295532736,
+      "loss": 0.7812,
+      "step": 6883
+    },
+    {
+      "epoch": 1.227990366604228,
+      "grad_norm": 1.1146833896636963,
+      "learning_rate": 0.00016229686998988068,
+      "loss": 0.5243,
+      "step": 6884
+    },
+    {
+      "epoch": 1.2281687628222282,
+      "grad_norm": 0.5391646027565002,
+      "learning_rate": 0.00016223126391260023,
+      "loss": 0.8716,
+      "step": 6885
+    },
+    {
+      "epoch": 1.2283471590402284,
+      "grad_norm": 0.4731014370918274,
+      "learning_rate": 0.00016216566472863854,
+      "loss": 0.5053,
+      "step": 6886
+    },
+    {
+      "epoch": 1.2285255552582286,
+      "grad_norm": 0.5936664938926697,
+      "learning_rate": 0.00016210007244314774,
+      "loss": 0.8034,
+      "step": 6887
+    },
+    {
+      "epoch": 1.2287039514762288,
+      "grad_norm": 0.5338798761367798,
+      "learning_rate": 0.00016203448706127938,
+      "loss": 0.7982,
+      "step": 6888
+    },
+    {
+      "epoch": 1.228882347694229,
+      "grad_norm": 0.5306500792503357,
+      "learning_rate": 0.00016196890858818458,
+      "loss": 0.6466,
+      "step": 6889
+    },
+    {
+      "epoch": 1.229060743912229,
+      "grad_norm": 0.44917765259742737,
+      "learning_rate": 0.0001619033370290138,
+      "loss": 0.5473,
+      "step": 6890
+    },
+    {
+      "epoch": 1.2292391401302292,
+      "grad_norm": 0.520117998123169,
+      "learning_rate": 0.00016183777238891703,
+      "loss": 0.6827,
+      "step": 6891
+    },
+    {
+      "epoch": 1.2294175363482294,
+      "grad_norm": 0.5320703387260437,
+      "learning_rate": 0.0001617722146730437,
+      "loss": 0.8353,
+      "step": 6892
+    },
+    {
+      "epoch": 1.2295959325662296,
+      "grad_norm": 0.47005659341812134,
+      "learning_rate": 0.00016170666388654265,
+      "loss": 0.6324,
+      "step": 6893
+    },
+    {
+      "epoch": 1.2297743287842298,
+      "grad_norm": 0.45758256316185,
+      "learning_rate": 0.00016164112003456223,
+      "loss": 0.6812,
+      "step": 6894
+    },
+    {
+      "epoch": 1.22995272500223,
+      "grad_norm": 0.4741186201572418,
+      "learning_rate": 0.00016157558312225018,
+      "loss": 0.588,
+      "step": 6895
+    },
+    {
+      "epoch": 1.2301311212202302,
+      "grad_norm": 0.4784983694553375,
+      "learning_rate": 0.0001615100531547538,
+      "loss": 0.601,
+      "step": 6896
+    },
+    {
+      "epoch": 1.2303095174382304,
+      "grad_norm": 0.45370981097221375,
+      "learning_rate": 0.00016144453013721978,
+      "loss": 0.4963,
+      "step": 6897
+    },
+    {
+      "epoch": 1.2304879136562306,
+      "grad_norm": 0.5208538770675659,
+      "learning_rate": 0.00016137901407479421,
+      "loss": 0.5965,
+      "step": 6898
+    },
+    {
+      "epoch": 1.2306663098742308,
+      "grad_norm": 0.45619308948516846,
+      "learning_rate": 0.00016131350497262278,
+      "loss": 0.5699,
+      "step": 6899
+    },
+    {
+      "epoch": 1.2308447060922307,
+      "grad_norm": 0.5480870604515076,
+      "learning_rate": 0.00016124800283585044,
+      "loss": 0.6709,
+      "step": 6900
+    },
+    {
+      "epoch": 1.231023102310231,
+      "grad_norm": 0.5735124945640564,
+      "learning_rate": 0.00016118250766962184,
+      "loss": 0.7542,
+      "step": 6901
+    },
+    {
+      "epoch": 1.2312014985282311,
+      "grad_norm": 0.4865647852420807,
+      "learning_rate": 0.00016111701947908085,
+      "loss": 0.6056,
+      "step": 6902
+    },
+    {
+      "epoch": 1.2313798947462313,
+      "grad_norm": 0.7100339531898499,
+      "learning_rate": 0.00016105153826937086,
+      "loss": 0.9174,
+      "step": 6903
+    },
+    {
+      "epoch": 1.2315582909642315,
+      "grad_norm": 0.5096773505210876,
+      "learning_rate": 0.00016098606404563482,
+      "loss": 0.6564,
+      "step": 6904
+    },
+    {
+      "epoch": 1.2317366871822317,
+      "grad_norm": 0.5119658708572388,
+      "learning_rate": 0.000160920596813015,
+      "loss": 0.7196,
+      "step": 6905
+    },
+    {
+      "epoch": 1.231915083400232,
+      "grad_norm": 0.4193865656852722,
+      "learning_rate": 0.0001608551365766532,
+      "loss": 0.5337,
+      "step": 6906
+    },
+    {
+      "epoch": 1.2320934796182321,
+      "grad_norm": 0.47937509417533875,
+      "learning_rate": 0.00016078968334169057,
+      "loss": 0.6803,
+      "step": 6907
+    },
+    {
+      "epoch": 1.2322718758362323,
+      "grad_norm": 0.5272152423858643,
+      "learning_rate": 0.00016072423711326782,
+      "loss": 0.6065,
+      "step": 6908
+    },
+    {
+      "epoch": 1.2324502720542325,
+      "grad_norm": 0.40397560596466064,
+      "learning_rate": 0.0001606587978965251,
+      "loss": 0.546,
+      "step": 6909
+    },
+    {
+      "epoch": 1.2326286682722327,
+      "grad_norm": 0.428237646818161,
+      "learning_rate": 0.00016059336569660193,
+      "loss": 0.4717,
+      "step": 6910
+    },
+    {
+      "epoch": 1.232807064490233,
+      "grad_norm": 0.4891768991947174,
+      "learning_rate": 0.00016052794051863733,
+      "loss": 0.6167,
+      "step": 6911
+    },
+    {
+      "epoch": 1.232985460708233,
+      "grad_norm": 0.5653678178787231,
+      "learning_rate": 0.00016046252236776978,
+      "loss": 0.8338,
+      "step": 6912
+    },
+    {
+      "epoch": 1.233163856926233,
+      "grad_norm": 0.4607936441898346,
+      "learning_rate": 0.00016039711124913718,
+      "loss": 0.6023,
+      "step": 6913
+    },
+    {
+      "epoch": 1.2333422531442333,
+      "grad_norm": 0.5353094339370728,
+      "learning_rate": 0.00016033170716787698,
+      "loss": 0.785,
+      "step": 6914
+    },
+    {
+      "epoch": 1.2335206493622335,
+      "grad_norm": 0.47947603464126587,
+      "learning_rate": 0.00016026631012912578,
+      "loss": 0.6174,
+      "step": 6915
+    },
+    {
+      "epoch": 1.2336990455802337,
+      "grad_norm": 0.5786678791046143,
+      "learning_rate": 0.00016020092013802002,
+      "loss": 0.9001,
+      "step": 6916
+    },
+    {
+      "epoch": 1.2338774417982339,
+      "grad_norm": 0.4535296559333801,
+      "learning_rate": 0.00016013553719969537,
+      "loss": 0.5654,
+      "step": 6917
+    },
+    {
+      "epoch": 1.234055838016234,
+      "grad_norm": 0.5149961113929749,
+      "learning_rate": 0.00016007016131928703,
+      "loss": 0.6585,
+      "step": 6918
+    },
+    {
+      "epoch": 1.2342342342342343,
+      "grad_norm": 0.4400131106376648,
+      "learning_rate": 0.00016000479250192942,
+      "loss": 0.5134,
+      "step": 6919
+    },
+    {
+      "epoch": 1.2344126304522345,
+      "grad_norm": 0.4986150562763214,
+      "learning_rate": 0.0001599394307527567,
+      "loss": 0.6103,
+      "step": 6920
+    },
+    {
+      "epoch": 1.2345910266702347,
+      "grad_norm": 0.487212210893631,
+      "learning_rate": 0.00015987407607690235,
+      "loss": 0.7115,
+      "step": 6921
+    },
+    {
+      "epoch": 1.2347694228882347,
+      "grad_norm": 0.46623316407203674,
+      "learning_rate": 0.0001598087284794993,
+      "loss": 0.5599,
+      "step": 6922
+    },
+    {
+      "epoch": 1.2349478191062349,
+      "grad_norm": 0.4859059453010559,
+      "learning_rate": 0.00015974338796567982,
+      "loss": 0.6134,
+      "step": 6923
+    },
+    {
+      "epoch": 1.235126215324235,
+      "grad_norm": 0.5123870968818665,
+      "learning_rate": 0.00015967805454057587,
+      "loss": 0.7607,
+      "step": 6924
+    },
+    {
+      "epoch": 1.2353046115422353,
+      "grad_norm": 0.49646908044815063,
+      "learning_rate": 0.00015961272820931867,
+      "loss": 0.6267,
+      "step": 6925
+    },
+    {
+      "epoch": 1.2354830077602355,
+      "grad_norm": 0.4790204167366028,
+      "learning_rate": 0.0001595474089770389,
+      "loss": 0.6403,
+      "step": 6926
+    },
+    {
+      "epoch": 1.2356614039782357,
+      "grad_norm": 0.49380695819854736,
+      "learning_rate": 0.00015948209684886667,
+      "loss": 0.6544,
+      "step": 6927
+    },
+    {
+      "epoch": 1.2358398001962358,
+      "grad_norm": 0.5434766411781311,
+      "learning_rate": 0.00015941679182993157,
+      "loss": 0.8058,
+      "step": 6928
+    },
+    {
+      "epoch": 1.236018196414236,
+      "grad_norm": 0.6229785680770874,
+      "learning_rate": 0.00015935149392536273,
+      "loss": 0.7764,
+      "step": 6929
+    },
+    {
+      "epoch": 1.2361965926322362,
+      "grad_norm": 0.5308107733726501,
+      "learning_rate": 0.00015928620314028838,
+      "loss": 0.6995,
+      "step": 6930
+    },
+    {
+      "epoch": 1.2363749888502364,
+      "grad_norm": 0.6573651432991028,
+      "learning_rate": 0.0001592209194798368,
+      "loss": 0.7208,
+      "step": 6931
+    },
+    {
+      "epoch": 1.2365533850682366,
+      "grad_norm": 0.5968353748321533,
+      "learning_rate": 0.00015915564294913503,
+      "loss": 0.8999,
+      "step": 6932
+    },
+    {
+      "epoch": 1.2367317812862368,
+      "grad_norm": 0.45062559843063354,
+      "learning_rate": 0.00015909037355331003,
+      "loss": 0.5419,
+      "step": 6933
+    },
+    {
+      "epoch": 1.2369101775042368,
+      "grad_norm": 0.5150581002235413,
+      "learning_rate": 0.0001590251112974879,
+      "loss": 0.6129,
+      "step": 6934
+    },
+    {
+      "epoch": 1.237088573722237,
+      "grad_norm": 0.4370402693748474,
+      "learning_rate": 0.00015895985618679444,
+      "loss": 0.6262,
+      "step": 6935
+    },
+    {
+      "epoch": 1.2372669699402372,
+      "grad_norm": 0.5415884852409363,
+      "learning_rate": 0.0001588946082263547,
+      "loss": 0.8408,
+      "step": 6936
+    },
+    {
+      "epoch": 1.2374453661582374,
+      "grad_norm": 0.49960747361183167,
+      "learning_rate": 0.0001588293674212932,
+      "loss": 0.7211,
+      "step": 6937
+    },
+    {
+      "epoch": 1.2376237623762376,
+      "grad_norm": 0.47522178292274475,
+      "learning_rate": 0.00015876413377673395,
+      "loss": 0.4949,
+      "step": 6938
+    },
+    {
+      "epoch": 1.2378021585942378,
+      "grad_norm": 0.5218939781188965,
+      "learning_rate": 0.00015869890729780045,
+      "loss": 0.5721,
+      "step": 6939
+    },
+    {
+      "epoch": 1.237980554812238,
+      "grad_norm": 0.5287362933158875,
+      "learning_rate": 0.00015863368798961538,
+      "loss": 0.9253,
+      "step": 6940
+    },
+    {
+      "epoch": 1.2381589510302382,
+      "grad_norm": 0.5668135285377502,
+      "learning_rate": 0.00015856847585730117,
+      "loss": 0.8171,
+      "step": 6941
+    },
+    {
+      "epoch": 1.2383373472482384,
+      "grad_norm": 0.5259323716163635,
+      "learning_rate": 0.00015850327090597953,
+      "loss": 0.6072,
+      "step": 6942
+    },
+    {
+      "epoch": 1.2385157434662386,
+      "grad_norm": 0.4646325409412384,
+      "learning_rate": 0.00015843807314077157,
+      "loss": 0.5816,
+      "step": 6943
+    },
+    {
+      "epoch": 1.2386941396842386,
+      "grad_norm": 0.4163079857826233,
+      "learning_rate": 0.00015837288256679799,
+      "loss": 0.4746,
+      "step": 6944
+    },
+    {
+      "epoch": 1.2388725359022388,
+      "grad_norm": 0.5029598474502563,
+      "learning_rate": 0.00015830769918917872,
+      "loss": 0.7697,
+      "step": 6945
+    },
+    {
+      "epoch": 1.239050932120239,
+      "grad_norm": 0.5194973349571228,
+      "learning_rate": 0.00015824252301303336,
+      "loss": 0.7874,
+      "step": 6946
+    },
+    {
+      "epoch": 1.2392293283382392,
+      "grad_norm": 0.4863186478614807,
+      "learning_rate": 0.00015817735404348072,
+      "loss": 0.715,
+      "step": 6947
+    },
+    {
+      "epoch": 1.2394077245562394,
+      "grad_norm": 0.4885355830192566,
+      "learning_rate": 0.00015811219228563923,
+      "loss": 0.6721,
+      "step": 6948
+    },
+    {
+      "epoch": 1.2395861207742396,
+      "grad_norm": 0.46745818853378296,
+      "learning_rate": 0.00015804703774462657,
+      "loss": 0.6217,
+      "step": 6949
+    },
+    {
+      "epoch": 1.2397645169922398,
+      "grad_norm": 0.5246323347091675,
+      "learning_rate": 0.00015798189042556,
+      "loss": 0.7571,
+      "step": 6950
+    },
+    {
+      "epoch": 1.23994291321024,
+      "grad_norm": 0.5258684158325195,
+      "learning_rate": 0.0001579167503335562,
+      "loss": 0.8582,
+      "step": 6951
+    },
+    {
+      "epoch": 1.2401213094282402,
+      "grad_norm": 0.49560466408729553,
+      "learning_rate": 0.0001578516174737312,
+      "loss": 0.6666,
+      "step": 6952
+    },
+    {
+      "epoch": 1.2402997056462404,
+      "grad_norm": 0.4392854571342468,
+      "learning_rate": 0.00015778649185120048,
+      "loss": 0.5967,
+      "step": 6953
+    },
+    {
+      "epoch": 1.2404781018642406,
+      "grad_norm": 0.49791455268859863,
+      "learning_rate": 0.000157721373471079,
+      "loss": 0.6931,
+      "step": 6954
+    },
+    {
+      "epoch": 1.2406564980822408,
+      "grad_norm": 0.48088231682777405,
+      "learning_rate": 0.00015765626233848115,
+      "loss": 0.733,
+      "step": 6955
+    },
+    {
+      "epoch": 1.2408348943002407,
+      "grad_norm": 0.5019537806510925,
+      "learning_rate": 0.00015759115845852072,
+      "loss": 0.8829,
+      "step": 6956
+    },
+    {
+      "epoch": 1.241013290518241,
+      "grad_norm": 0.5134669542312622,
+      "learning_rate": 0.0001575260618363109,
+      "loss": 0.7432,
+      "step": 6957
+    },
+    {
+      "epoch": 1.2411916867362411,
+      "grad_norm": 0.5204741358757019,
+      "learning_rate": 0.00015746097247696443,
+      "loss": 0.7888,
+      "step": 6958
+    },
+    {
+      "epoch": 1.2413700829542413,
+      "grad_norm": 0.5397398471832275,
+      "learning_rate": 0.0001573958903855934,
+      "loss": 0.7761,
+      "step": 6959
+    },
+    {
+      "epoch": 1.2415484791722415,
+      "grad_norm": 0.4783801734447479,
+      "learning_rate": 0.00015733081556730926,
+      "loss": 0.7478,
+      "step": 6960
+    },
+    {
+      "epoch": 1.2417268753902417,
+      "grad_norm": 0.6109618544578552,
+      "learning_rate": 0.00015726574802722284,
+      "loss": 0.7975,
+      "step": 6961
+    },
+    {
+      "epoch": 1.241905271608242,
+      "grad_norm": 0.48145678639411926,
+      "learning_rate": 0.00015720068777044476,
+      "loss": 0.5358,
+      "step": 6962
+    },
+    {
+      "epoch": 1.2420836678262421,
+      "grad_norm": 0.5572999715805054,
+      "learning_rate": 0.0001571356348020848,
+      "loss": 0.6544,
+      "step": 6963
+    },
+    {
+      "epoch": 1.2422620640442423,
+      "grad_norm": 0.503771185874939,
+      "learning_rate": 0.00015707058912725207,
+      "loss": 0.5603,
+      "step": 6964
+    },
+    {
+      "epoch": 1.2424404602622425,
+      "grad_norm": 0.5243847370147705,
+      "learning_rate": 0.00015700555075105532,
+      "loss": 0.6095,
+      "step": 6965
+    },
+    {
+      "epoch": 1.2426188564802425,
+      "grad_norm": 0.5380844473838806,
+      "learning_rate": 0.00015694051967860256,
+      "loss": 0.6285,
+      "step": 6966
+    },
+    {
+      "epoch": 1.2427972526982427,
+      "grad_norm": 0.5606120824813843,
+      "learning_rate": 0.00015687549591500134,
+      "loss": 0.8092,
+      "step": 6967
+    },
+    {
+      "epoch": 1.242975648916243,
+      "grad_norm": 0.5026856660842896,
+      "learning_rate": 0.0001568104794653586,
+      "loss": 0.6013,
+      "step": 6968
+    },
+    {
+      "epoch": 1.243154045134243,
+      "grad_norm": 0.47534608840942383,
+      "learning_rate": 0.00015674547033478072,
+      "loss": 0.5524,
+      "step": 6969
+    },
+    {
+      "epoch": 1.2433324413522433,
+      "grad_norm": 0.5083494186401367,
+      "learning_rate": 0.00015668046852837343,
+      "loss": 0.7611,
+      "step": 6970
+    },
+    {
+      "epoch": 1.2435108375702435,
+      "grad_norm": 0.4332534372806549,
+      "learning_rate": 0.000156615474051242,
+      "loss": 0.4897,
+      "step": 6971
+    },
+    {
+      "epoch": 1.2436892337882437,
+      "grad_norm": 0.4423793852329254,
+      "learning_rate": 0.00015655048690849102,
+      "loss": 0.5234,
+      "step": 6972
+    },
+    {
+      "epoch": 1.243867630006244,
+      "grad_norm": 0.47594019770622253,
+      "learning_rate": 0.00015648550710522466,
+      "loss": 0.6987,
+      "step": 6973
+    },
+    {
+      "epoch": 1.244046026224244,
+      "grad_norm": 0.5025418400764465,
+      "learning_rate": 0.00015642053464654627,
+      "loss": 0.5931,
+      "step": 6974
+    },
+    {
+      "epoch": 1.2442244224422443,
+      "grad_norm": 0.48870721459388733,
+      "learning_rate": 0.00015635556953755883,
+      "loss": 0.7561,
+      "step": 6975
+    },
+    {
+      "epoch": 1.2444028186602445,
+      "grad_norm": 0.5276898741722107,
+      "learning_rate": 0.00015629061178336456,
+      "loss": 0.5714,
+      "step": 6976
+    },
+    {
+      "epoch": 1.2445812148782447,
+      "grad_norm": 0.5232852697372437,
+      "learning_rate": 0.00015622566138906547,
+      "loss": 0.6564,
+      "step": 6977
+    },
+    {
+      "epoch": 1.2447596110962447,
+      "grad_norm": 0.41942137479782104,
+      "learning_rate": 0.00015616071835976254,
+      "loss": 0.4946,
+      "step": 6978
+    },
+    {
+      "epoch": 1.2449380073142449,
+      "grad_norm": 0.6095328330993652,
+      "learning_rate": 0.00015609578270055636,
+      "loss": 0.6298,
+      "step": 6979
+    },
+    {
+      "epoch": 1.245116403532245,
+      "grad_norm": 0.4936719238758087,
+      "learning_rate": 0.00015603085441654702,
+      "loss": 0.744,
+      "step": 6980
+    },
+    {
+      "epoch": 1.2452947997502453,
+      "grad_norm": 0.4040975868701935,
+      "learning_rate": 0.00015596593351283394,
+      "loss": 0.5204,
+      "step": 6981
+    },
+    {
+      "epoch": 1.2454731959682455,
+      "grad_norm": 0.5431026816368103,
+      "learning_rate": 0.000155901019994516,
+      "loss": 0.7077,
+      "step": 6982
+    },
+    {
+      "epoch": 1.2456515921862457,
+      "grad_norm": 0.5668693780899048,
+      "learning_rate": 0.00015583611386669143,
+      "loss": 0.7685,
+      "step": 6983
+    },
+    {
+      "epoch": 1.2458299884042459,
+      "grad_norm": 0.4577692747116089,
+      "learning_rate": 0.00015577121513445796,
+      "loss": 0.4953,
+      "step": 6984
+    },
+    {
+      "epoch": 1.246008384622246,
+      "grad_norm": 0.5005066394805908,
+      "learning_rate": 0.0001557063238029127,
+      "loss": 0.6231,
+      "step": 6985
+    },
+    {
+      "epoch": 1.2461867808402463,
+      "grad_norm": 0.4887985289096832,
+      "learning_rate": 0.00015564143987715224,
+      "loss": 0.5986,
+      "step": 6986
+    },
+    {
+      "epoch": 1.2463651770582465,
+      "grad_norm": 0.5014304518699646,
+      "learning_rate": 0.00015557656336227243,
+      "loss": 0.6816,
+      "step": 6987
+    },
+    {
+      "epoch": 1.2465435732762464,
+      "grad_norm": 0.4619125425815582,
+      "learning_rate": 0.00015551169426336874,
+      "loss": 0.608,
+      "step": 6988
+    },
+    {
+      "epoch": 1.2467219694942466,
+      "grad_norm": 0.4961630702018738,
+      "learning_rate": 0.0001554468325855359,
+      "loss": 0.6304,
+      "step": 6989
+    },
+    {
+      "epoch": 1.2469003657122468,
+      "grad_norm": 0.5148522853851318,
+      "learning_rate": 0.00015538197833386814,
+      "loss": 0.81,
+      "step": 6990
+    },
+    {
+      "epoch": 1.247078761930247,
+      "grad_norm": 0.414350688457489,
+      "learning_rate": 0.000155317131513459,
+      "loss": 0.4424,
+      "step": 6991
+    },
+    {
+      "epoch": 1.2472571581482472,
+      "grad_norm": 0.4806058406829834,
+      "learning_rate": 0.00015525229212940168,
+      "loss": 0.6765,
+      "step": 6992
+    },
+    {
+      "epoch": 1.2474355543662474,
+      "grad_norm": 0.5416685342788696,
+      "learning_rate": 0.0001551874601867886,
+      "loss": 0.7167,
+      "step": 6993
+    },
+    {
+      "epoch": 1.2476139505842476,
+      "grad_norm": 0.4818612039089203,
+      "learning_rate": 0.00015512263569071152,
+      "loss": 0.6258,
+      "step": 6994
+    },
+    {
+      "epoch": 1.2477923468022478,
+      "grad_norm": 0.5225372910499573,
+      "learning_rate": 0.00015505781864626184,
+      "loss": 0.8175,
+      "step": 6995
+    },
+    {
+      "epoch": 1.247970743020248,
+      "grad_norm": 0.45656535029411316,
+      "learning_rate": 0.00015499300905853026,
+      "loss": 0.6183,
+      "step": 6996
+    },
+    {
+      "epoch": 1.2481491392382482,
+      "grad_norm": 0.5095564126968384,
+      "learning_rate": 0.00015492820693260682,
+      "loss": 0.3982,
+      "step": 6997
+    },
+    {
+      "epoch": 1.2483275354562484,
+      "grad_norm": 0.4879413843154907,
+      "learning_rate": 0.00015486341227358114,
+      "loss": 0.5668,
+      "step": 6998
+    },
+    {
+      "epoch": 1.2485059316742486,
+      "grad_norm": 0.463022381067276,
+      "learning_rate": 0.00015479862508654212,
+      "loss": 0.6153,
+      "step": 6999
+    },
+    {
+      "epoch": 1.2486843278922486,
+      "grad_norm": 52.42532730102539,
+      "learning_rate": 0.0001547338453765781,
+      "loss": 1.0213,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2488627241102488,
+      "grad_norm": 0.45679882168769836,
+      "learning_rate": 0.00015466907314877682,
+      "loss": 0.5221,
+      "step": 7001
+    },
+    {
+      "epoch": 1.249041120328249,
+      "grad_norm": 0.5205798745155334,
+      "learning_rate": 0.00015460430840822552,
+      "loss": 0.6813,
+      "step": 7002
+    },
+    {
+      "epoch": 1.2492195165462492,
+      "grad_norm": 0.510740339756012,
+      "learning_rate": 0.00015453955116001084,
+      "loss": 0.6022,
+      "step": 7003
+    },
+    {
+      "epoch": 1.2493979127642494,
+      "grad_norm": 0.5229342579841614,
+      "learning_rate": 0.0001544748014092187,
+      "loss": 0.6381,
+      "step": 7004
+    },
+    {
+      "epoch": 1.2495763089822496,
+      "grad_norm": 0.48941877484321594,
+      "learning_rate": 0.00015441005916093454,
+      "loss": 0.6733,
+      "step": 7005
+    },
+    {
+      "epoch": 1.2497547052002498,
+      "grad_norm": 0.5428256392478943,
+      "learning_rate": 0.0001543453244202433,
+      "loss": 0.7319,
+      "step": 7006
+    },
+    {
+      "epoch": 1.24993310141825,
+      "grad_norm": 0.4766102433204651,
+      "learning_rate": 0.0001542805971922289,
+      "loss": 0.6504,
+      "step": 7007
+    },
+    {
+      "epoch": 1.2501114976362502,
+      "grad_norm": 0.573760986328125,
+      "learning_rate": 0.00015421587748197532,
+      "loss": 0.7209,
+      "step": 7008
+    },
+    {
+      "epoch": 1.2502898938542502,
+      "grad_norm": 0.5121127963066101,
+      "learning_rate": 0.00015415116529456552,
+      "loss": 0.7118,
+      "step": 7009
+    },
+    {
+      "epoch": 1.2504682900722504,
+      "grad_norm": 0.5492568612098694,
+      "learning_rate": 0.00015408646063508197,
+      "loss": 0.7907,
+      "step": 7010
+    },
+    {
+      "epoch": 1.2506466862902506,
+      "grad_norm": 0.4394141733646393,
+      "learning_rate": 0.00015402176350860653,
+      "loss": 0.5036,
+      "step": 7011
+    },
+    {
+      "epoch": 1.2508250825082508,
+      "grad_norm": 0.4782421886920929,
+      "learning_rate": 0.00015395707392022045,
+      "loss": 0.6055,
+      "step": 7012
+    },
+    {
+      "epoch": 1.251003478726251,
+      "grad_norm": 0.5013086795806885,
+      "learning_rate": 0.0001538923918750045,
+      "loss": 0.6325,
+      "step": 7013
+    },
+    {
+      "epoch": 1.2511818749442511,
+      "grad_norm": 0.5063600540161133,
+      "learning_rate": 0.00015382771737803866,
+      "loss": 0.7373,
+      "step": 7014
+    },
+    {
+      "epoch": 1.2513602711622513,
+      "grad_norm": 0.48954328894615173,
+      "learning_rate": 0.00015376305043440254,
+      "loss": 0.5793,
+      "step": 7015
+    },
+    {
+      "epoch": 1.2515386673802515,
+      "grad_norm": 0.4994099736213684,
+      "learning_rate": 0.00015369839104917505,
+      "loss": 0.5759,
+      "step": 7016
+    },
+    {
+      "epoch": 1.2517170635982517,
+      "grad_norm": 0.5017921924591064,
+      "learning_rate": 0.00015363373922743444,
+      "loss": 0.6372,
+      "step": 7017
+    },
+    {
+      "epoch": 1.251895459816252,
+      "grad_norm": 0.4896945059299469,
+      "learning_rate": 0.0001535690949742585,
+      "loss": 0.5314,
+      "step": 7018
+    },
+    {
+      "epoch": 1.2520738560342521,
+      "grad_norm": 0.524643063545227,
+      "learning_rate": 0.00015350445829472444,
+      "loss": 0.7481,
+      "step": 7019
+    },
+    {
+      "epoch": 1.2522522522522523,
+      "grad_norm": 0.42429521679878235,
+      "learning_rate": 0.00015343982919390858,
+      "loss": 0.5296,
+      "step": 7020
+    },
+    {
+      "epoch": 1.2524306484702525,
+      "grad_norm": 0.44809257984161377,
+      "learning_rate": 0.000153375207676887,
+      "loss": 0.5499,
+      "step": 7021
+    },
+    {
+      "epoch": 1.2526090446882527,
+      "grad_norm": 0.45676136016845703,
+      "learning_rate": 0.00015331059374873495,
+      "loss": 0.537,
+      "step": 7022
+    },
+    {
+      "epoch": 1.2527874409062527,
+      "grad_norm": 0.5065332651138306,
+      "learning_rate": 0.00015324598741452733,
+      "loss": 0.7571,
+      "step": 7023
+    },
+    {
+      "epoch": 1.252965837124253,
+      "grad_norm": 0.4751022160053253,
+      "learning_rate": 0.0001531813886793383,
+      "loss": 0.5616,
+      "step": 7024
+    },
+    {
+      "epoch": 1.253144233342253,
+      "grad_norm": 0.5255012512207031,
+      "learning_rate": 0.00015311679754824125,
+      "loss": 0.832,
+      "step": 7025
+    },
+    {
+      "epoch": 1.2533226295602533,
+      "grad_norm": 0.4546463191509247,
+      "learning_rate": 0.00015305221402630925,
+      "loss": 0.5618,
+      "step": 7026
+    },
+    {
+      "epoch": 1.2535010257782535,
+      "grad_norm": 0.46323102712631226,
+      "learning_rate": 0.00015298763811861466,
+      "loss": 0.501,
+      "step": 7027
+    },
+    {
+      "epoch": 1.2536794219962537,
+      "grad_norm": 0.5189341902732849,
+      "learning_rate": 0.0001529230698302292,
+      "loss": 0.6688,
+      "step": 7028
+    },
+    {
+      "epoch": 1.253857818214254,
+      "grad_norm": 0.49833574891090393,
+      "learning_rate": 0.0001528585091662241,
+      "loss": 0.6542,
+      "step": 7029
+    },
+    {
+      "epoch": 1.254036214432254,
+      "grad_norm": 0.5910803079605103,
+      "learning_rate": 0.00015279395613166985,
+      "loss": 0.8463,
+      "step": 7030
+    },
+    {
+      "epoch": 1.254214610650254,
+      "grad_norm": 0.4795231223106384,
+      "learning_rate": 0.00015272941073163647,
+      "loss": 0.709,
+      "step": 7031
+    },
+    {
+      "epoch": 1.2543930068682543,
+      "grad_norm": 0.520872950553894,
+      "learning_rate": 0.00015266487297119335,
+      "loss": 0.7132,
+      "step": 7032
+    },
+    {
+      "epoch": 1.2545714030862545,
+      "grad_norm": 0.44845151901245117,
+      "learning_rate": 0.00015260034285540915,
+      "loss": 0.5002,
+      "step": 7033
+    },
+    {
+      "epoch": 1.2547497993042547,
+      "grad_norm": 0.4645794928073883,
+      "learning_rate": 0.00015253582038935216,
+      "loss": 0.5065,
+      "step": 7034
+    },
+    {
+      "epoch": 1.2549281955222549,
+      "grad_norm": 0.4405880272388458,
+      "learning_rate": 0.00015247130557808985,
+      "loss": 0.5645,
+      "step": 7035
+    },
+    {
+      "epoch": 1.255106591740255,
+      "grad_norm": 0.4625834822654724,
+      "learning_rate": 0.00015240679842668924,
+      "loss": 0.4968,
+      "step": 7036
+    },
+    {
+      "epoch": 1.2552849879582553,
+      "grad_norm": 0.4433553218841553,
+      "learning_rate": 0.00015234229894021666,
+      "loss": 0.5323,
+      "step": 7037
+    },
+    {
+      "epoch": 1.2554633841762555,
+      "grad_norm": 0.510197639465332,
+      "learning_rate": 0.0001522778071237379,
+      "loss": 0.7702,
+      "step": 7038
+    },
+    {
+      "epoch": 1.2556417803942557,
+      "grad_norm": 0.5122009515762329,
+      "learning_rate": 0.00015221332298231816,
+      "loss": 0.6138,
+      "step": 7039
+    },
+    {
+      "epoch": 1.2558201766122559,
+      "grad_norm": 0.47657519578933716,
+      "learning_rate": 0.00015214884652102193,
+      "loss": 0.6609,
+      "step": 7040
+    },
+    {
+      "epoch": 1.255998572830256,
+      "grad_norm": 0.4612807631492615,
+      "learning_rate": 0.00015208437774491318,
+      "loss": 0.466,
+      "step": 7041
+    },
+    {
+      "epoch": 1.2561769690482563,
+      "grad_norm": 0.5131711959838867,
+      "learning_rate": 0.00015201991665905528,
+      "loss": 0.6583,
+      "step": 7042
+    },
+    {
+      "epoch": 1.2563553652662565,
+      "grad_norm": 0.4903299808502197,
+      "learning_rate": 0.00015195546326851096,
+      "loss": 0.5683,
+      "step": 7043
+    },
+    {
+      "epoch": 1.2565337614842567,
+      "grad_norm": 0.772181510925293,
+      "learning_rate": 0.00015189101757834235,
+      "loss": 0.58,
+      "step": 7044
+    },
+    {
+      "epoch": 1.2567121577022566,
+      "grad_norm": 0.43410158157348633,
+      "learning_rate": 0.00015182657959361107,
+      "loss": 0.5278,
+      "step": 7045
+    },
+    {
+      "epoch": 1.2568905539202568,
+      "grad_norm": 0.5446970462799072,
+      "learning_rate": 0.00015176214931937794,
+      "loss": 0.7534,
+      "step": 7046
+    },
+    {
+      "epoch": 1.257068950138257,
+      "grad_norm": 0.4579203426837921,
+      "learning_rate": 0.00015169772676070328,
+      "loss": 0.6194,
+      "step": 7047
+    },
+    {
+      "epoch": 1.2572473463562572,
+      "grad_norm": 0.4995640516281128,
+      "learning_rate": 0.0001516333119226469,
+      "loss": 0.6539,
+      "step": 7048
+    },
+    {
+      "epoch": 1.2574257425742574,
+      "grad_norm": 0.4806896150112152,
+      "learning_rate": 0.00015156890481026787,
+      "loss": 0.4902,
+      "step": 7049
+    },
+    {
+      "epoch": 1.2576041387922576,
+      "grad_norm": 0.5618352293968201,
+      "learning_rate": 0.00015150450542862466,
+      "loss": 0.7631,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2577825350102578,
+      "grad_norm": 0.679996907711029,
+      "learning_rate": 0.00015144011378277522,
+      "loss": 0.8891,
+      "step": 7051
+    },
+    {
+      "epoch": 1.257960931228258,
+      "grad_norm": 0.47723329067230225,
+      "learning_rate": 0.00015137572987777688,
+      "loss": 0.526,
+      "step": 7052
+    },
+    {
+      "epoch": 1.258139327446258,
+      "grad_norm": 0.53426593542099,
+      "learning_rate": 0.00015131135371868615,
+      "loss": 0.5509,
+      "step": 7053
+    },
+    {
+      "epoch": 1.2583177236642582,
+      "grad_norm": 0.5936741232872009,
+      "learning_rate": 0.0001512469853105593,
+      "loss": 0.6007,
+      "step": 7054
+    },
+    {
+      "epoch": 1.2584961198822584,
+      "grad_norm": 0.5116682052612305,
+      "learning_rate": 0.00015118262465845179,
+      "loss": 0.7028,
+      "step": 7055
+    },
+    {
+      "epoch": 1.2586745161002586,
+      "grad_norm": 0.5201318264007568,
+      "learning_rate": 0.00015111827176741833,
+      "loss": 0.7666,
+      "step": 7056
+    },
+    {
+      "epoch": 1.2588529123182588,
+      "grad_norm": 0.4887928068637848,
+      "learning_rate": 0.0001510539266425133,
+      "loss": 0.6696,
+      "step": 7057
+    },
+    {
+      "epoch": 1.259031308536259,
+      "grad_norm": 0.499969482421875,
+      "learning_rate": 0.00015098958928879035,
+      "loss": 0.5775,
+      "step": 7058
+    },
+    {
+      "epoch": 1.2592097047542592,
+      "grad_norm": 0.4900054335594177,
+      "learning_rate": 0.0001509252597113024,
+      "loss": 0.7217,
+      "step": 7059
+    },
+    {
+      "epoch": 1.2593881009722594,
+      "grad_norm": 0.45761123299598694,
+      "learning_rate": 0.00015086093791510187,
+      "loss": 0.4968,
+      "step": 7060
+    },
+    {
+      "epoch": 1.2595664971902596,
+      "grad_norm": 0.4074775278568268,
+      "learning_rate": 0.00015079662390524062,
+      "loss": 0.457,
+      "step": 7061
+    },
+    {
+      "epoch": 1.2597448934082598,
+      "grad_norm": 0.5297835469245911,
+      "learning_rate": 0.00015073231768676987,
+      "loss": 0.6604,
+      "step": 7062
+    },
+    {
+      "epoch": 1.25992328962626,
+      "grad_norm": 0.5857179164886475,
+      "learning_rate": 0.00015066801926474015,
+      "loss": 0.8372,
+      "step": 7063
+    },
+    {
+      "epoch": 1.2601016858442602,
+      "grad_norm": 0.5072324872016907,
+      "learning_rate": 0.00015060372864420147,
+      "loss": 0.6661,
+      "step": 7064
+    },
+    {
+      "epoch": 1.2602800820622604,
+      "grad_norm": 0.5058489441871643,
+      "learning_rate": 0.00015053944583020318,
+      "loss": 0.6894,
+      "step": 7065
+    },
+    {
+      "epoch": 1.2604584782802606,
+      "grad_norm": 0.4949549436569214,
+      "learning_rate": 0.00015047517082779406,
+      "loss": 0.6834,
+      "step": 7066
+    },
+    {
+      "epoch": 1.2606368744982606,
+      "grad_norm": 0.47919708490371704,
+      "learning_rate": 0.0001504109036420221,
+      "loss": 0.6148,
+      "step": 7067
+    },
+    {
+      "epoch": 1.2608152707162608,
+      "grad_norm": 0.5355616211891174,
+      "learning_rate": 0.00015034664427793484,
+      "loss": 0.6993,
+      "step": 7068
+    },
+    {
+      "epoch": 1.260993666934261,
+      "grad_norm": 0.4291095435619354,
+      "learning_rate": 0.00015028239274057938,
+      "loss": 0.5921,
+      "step": 7069
+    },
+    {
+      "epoch": 1.2611720631522612,
+      "grad_norm": 0.5145418047904968,
+      "learning_rate": 0.0001502181490350019,
+      "loss": 0.715,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2613504593702614,
+      "grad_norm": 0.48409318923950195,
+      "learning_rate": 0.00015015391316624798,
+      "loss": 0.6167,
+      "step": 7071
+    },
+    {
+      "epoch": 1.2615288555882616,
+      "grad_norm": 0.5300613641738892,
+      "learning_rate": 0.0001500896851393628,
+      "loss": 0.7082,
+      "step": 7072
+    },
+    {
+      "epoch": 1.2617072518062618,
+      "grad_norm": 0.5523662567138672,
+      "learning_rate": 0.00015002546495939073,
+      "loss": 0.6498,
+      "step": 7073
+    },
+    {
+      "epoch": 1.261885648024262,
+      "grad_norm": 0.48872825503349304,
+      "learning_rate": 0.00014996125263137564,
+      "loss": 0.6851,
+      "step": 7074
+    },
+    {
+      "epoch": 1.262064044242262,
+      "grad_norm": 0.4612981677055359,
+      "learning_rate": 0.0001498970481603607,
+      "loss": 0.7026,
+      "step": 7075
+    },
+    {
+      "epoch": 1.2622424404602621,
+      "grad_norm": 0.47795626521110535,
+      "learning_rate": 0.00014983285155138854,
+      "loss": 0.7194,
+      "step": 7076
+    },
+    {
+      "epoch": 1.2624208366782623,
+      "grad_norm": 0.4560318887233734,
+      "learning_rate": 0.00014976866280950107,
+      "loss": 0.587,
+      "step": 7077
+    },
+    {
+      "epoch": 1.2625992328962625,
+      "grad_norm": 0.4796944856643677,
+      "learning_rate": 0.00014970448193973979,
+      "loss": 0.6386,
+      "step": 7078
+    },
+    {
+      "epoch": 1.2627776291142627,
+      "grad_norm": 0.528049647808075,
+      "learning_rate": 0.00014964030894714525,
+      "loss": 0.5845,
+      "step": 7079
+    },
+    {
+      "epoch": 1.262956025332263,
+      "grad_norm": 0.46722978353500366,
+      "learning_rate": 0.0001495761438367577,
+      "loss": 0.6446,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2631344215502631,
+      "grad_norm": 0.5087683200836182,
+      "learning_rate": 0.00014951198661361656,
+      "loss": 0.589,
+      "step": 7081
+    },
+    {
+      "epoch": 1.2633128177682633,
+      "grad_norm": 0.5003060698509216,
+      "learning_rate": 0.00014944783728276076,
+      "loss": 0.7562,
+      "step": 7082
+    },
+    {
+      "epoch": 1.2634912139862635,
+      "grad_norm": 0.5159687399864197,
+      "learning_rate": 0.0001493836958492285,
+      "loss": 0.6961,
+      "step": 7083
+    },
+    {
+      "epoch": 1.2636696102042637,
+      "grad_norm": 0.6173499822616577,
+      "learning_rate": 0.0001493195623180575,
+      "loss": 0.7186,
+      "step": 7084
+    },
+    {
+      "epoch": 1.263848006422264,
+      "grad_norm": 0.5042861104011536,
+      "learning_rate": 0.00014925543669428478,
+      "loss": 0.6666,
+      "step": 7085
+    },
+    {
+      "epoch": 1.2640264026402641,
+      "grad_norm": 0.5728045105934143,
+      "learning_rate": 0.00014919131898294668,
+      "loss": 0.7712,
+      "step": 7086
+    },
+    {
+      "epoch": 1.2642047988582643,
+      "grad_norm": 0.46245428919792175,
+      "learning_rate": 0.00014912720918907905,
+      "loss": 0.5197,
+      "step": 7087
+    },
+    {
+      "epoch": 1.2643831950762645,
+      "grad_norm": 0.4901646077632904,
+      "learning_rate": 0.00014906310731771697,
+      "loss": 0.6486,
+      "step": 7088
+    },
+    {
+      "epoch": 1.2645615912942645,
+      "grad_norm": 0.5244214534759521,
+      "learning_rate": 0.000148999013373895,
+      "loss": 0.6701,
+      "step": 7089
+    },
+    {
+      "epoch": 1.2647399875122647,
+      "grad_norm": 0.6227227449417114,
+      "learning_rate": 0.00014893492736264708,
+      "loss": 0.5513,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2649183837302649,
+      "grad_norm": 0.4480903148651123,
+      "learning_rate": 0.00014887084928900653,
+      "loss": 0.5382,
+      "step": 7091
+    },
+    {
+      "epoch": 1.265096779948265,
+      "grad_norm": 0.41941893100738525,
+      "learning_rate": 0.00014880677915800585,
+      "loss": 0.451,
+      "step": 7092
+    },
+    {
+      "epoch": 1.2652751761662653,
+      "grad_norm": 0.44205811619758606,
+      "learning_rate": 0.00014874271697467724,
+      "loss": 0.5072,
+      "step": 7093
+    },
+    {
+      "epoch": 1.2654535723842655,
+      "grad_norm": 0.4529136121273041,
+      "learning_rate": 0.00014867866274405204,
+      "loss": 0.5873,
+      "step": 7094
+    },
+    {
+      "epoch": 1.2656319686022657,
+      "grad_norm": 0.4797188639640808,
+      "learning_rate": 0.00014861461647116105,
+      "loss": 0.579,
+      "step": 7095
+    },
+    {
+      "epoch": 1.2658103648202659,
+      "grad_norm": 0.47483915090560913,
+      "learning_rate": 0.00014855057816103452,
+      "loss": 0.6112,
+      "step": 7096
+    },
+    {
+      "epoch": 1.2659887610382659,
+      "grad_norm": 0.6069769263267517,
+      "learning_rate": 0.00014848654781870186,
+      "loss": 0.6348,
+      "step": 7097
+    },
+    {
+      "epoch": 1.266167157256266,
+      "grad_norm": 0.6138676404953003,
+      "learning_rate": 0.00014842252544919205,
+      "loss": 0.5382,
+      "step": 7098
+    },
+    {
+      "epoch": 1.2663455534742663,
+      "grad_norm": 0.5163396596908569,
+      "learning_rate": 0.00014835851105753333,
+      "loss": 0.6694,
+      "step": 7099
+    },
+    {
+      "epoch": 1.2665239496922664,
+      "grad_norm": 0.5151557922363281,
+      "learning_rate": 0.0001482945046487535,
+      "loss": 0.7113,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2667023459102666,
+      "grad_norm": 0.5661108493804932,
+      "learning_rate": 0.00014823050622787948,
+      "loss": 0.8239,
+      "step": 7101
+    },
+    {
+      "epoch": 1.2668807421282668,
+      "grad_norm": 0.7180517315864563,
+      "learning_rate": 0.00014816651579993773,
+      "loss": 0.8975,
+      "step": 7102
+    },
+    {
+      "epoch": 1.267059138346267,
+      "grad_norm": 0.5672160983085632,
+      "learning_rate": 0.000148102533369954,
+      "loss": 0.6035,
+      "step": 7103
+    },
+    {
+      "epoch": 1.2672375345642672,
+      "grad_norm": 0.47604504227638245,
+      "learning_rate": 0.0001480385589429535,
+      "loss": 0.5376,
+      "step": 7104
+    },
+    {
+      "epoch": 1.2674159307822674,
+      "grad_norm": 0.6531439423561096,
+      "learning_rate": 0.0001479745925239606,
+      "loss": 0.5673,
+      "step": 7105
+    },
+    {
+      "epoch": 1.2675943270002676,
+      "grad_norm": 0.5715680122375488,
+      "learning_rate": 0.0001479106341179994,
+      "loss": 0.7327,
+      "step": 7106
+    },
+    {
+      "epoch": 1.2677727232182678,
+      "grad_norm": 0.5460495948791504,
+      "learning_rate": 0.00014784668373009298,
+      "loss": 0.5201,
+      "step": 7107
+    },
+    {
+      "epoch": 1.267951119436268,
+      "grad_norm": 0.5381412506103516,
+      "learning_rate": 0.00014778274136526408,
+      "loss": 0.7042,
+      "step": 7108
+    },
+    {
+      "epoch": 1.2681295156542682,
+      "grad_norm": 0.49676334857940674,
+      "learning_rate": 0.00014771880702853468,
+      "loss": 0.5428,
+      "step": 7109
+    },
+    {
+      "epoch": 1.2683079118722684,
+      "grad_norm": 0.726411759853363,
+      "learning_rate": 0.00014765488072492617,
+      "loss": 0.6615,
+      "step": 7110
+    },
+    {
+      "epoch": 1.2684863080902684,
+      "grad_norm": 0.5252700448036194,
+      "learning_rate": 0.00014759096245945929,
+      "loss": 0.6958,
+      "step": 7111
+    },
+    {
+      "epoch": 1.2686647043082686,
+      "grad_norm": 1.2057543992996216,
+      "learning_rate": 0.0001475270522371542,
+      "loss": 0.4825,
+      "step": 7112
+    },
+    {
+      "epoch": 1.2688431005262688,
+      "grad_norm": 0.4835338592529297,
+      "learning_rate": 0.00014746315006303027,
+      "loss": 0.6651,
+      "step": 7113
+    },
+    {
+      "epoch": 1.269021496744269,
+      "grad_norm": 3.452477216720581,
+      "learning_rate": 0.0001473992559421063,
+      "loss": 0.7079,
+      "step": 7114
+    },
+    {
+      "epoch": 1.2691998929622692,
+      "grad_norm": 0.5416814088821411,
+      "learning_rate": 0.00014733536987940075,
+      "loss": 0.5033,
+      "step": 7115
+    },
+    {
+      "epoch": 1.2693782891802694,
+      "grad_norm": 0.46173161268234253,
+      "learning_rate": 0.00014727149187993105,
+      "loss": 0.6123,
+      "step": 7116
+    },
+    {
+      "epoch": 1.2695566853982696,
+      "grad_norm": 0.4671329855918884,
+      "learning_rate": 0.00014720762194871424,
+      "loss": 0.6949,
+      "step": 7117
+    },
+    {
+      "epoch": 1.2697350816162698,
+      "grad_norm": 0.5171025395393372,
+      "learning_rate": 0.00014714376009076647,
+      "loss": 0.6301,
+      "step": 7118
+    },
+    {
+      "epoch": 1.2699134778342698,
+      "grad_norm": 0.47935229539871216,
+      "learning_rate": 0.00014707990631110355,
+      "loss": 0.6334,
+      "step": 7119
+    },
+    {
+      "epoch": 1.27009187405227,
+      "grad_norm": 0.4533785283565521,
+      "learning_rate": 0.0001470160606147405,
+      "loss": 0.5618,
+      "step": 7120
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 0.5093548893928528,
+      "learning_rate": 0.0001469522230066917,
+      "loss": 0.7452,
+      "step": 7121
+    },
+    {
+      "epoch": 1.2704486664882704,
+      "grad_norm": 0.43215861916542053,
+      "learning_rate": 0.000146888393491971,
+      "loss": 0.599,
+      "step": 7122
+    },
+    {
+      "epoch": 1.2706270627062706,
+      "grad_norm": 0.46486908197402954,
+      "learning_rate": 0.0001468245720755915,
+      "loss": 0.6116,
+      "step": 7123
+    },
+    {
+      "epoch": 1.2708054589242708,
+      "grad_norm": 0.5231932401657104,
+      "learning_rate": 0.00014676075876256567,
+      "loss": 0.5941,
+      "step": 7124
+    },
+    {
+      "epoch": 1.270983855142271,
+      "grad_norm": 0.5373736619949341,
+      "learning_rate": 0.00014669695355790552,
+      "loss": 0.5963,
+      "step": 7125
+    },
+    {
+      "epoch": 1.2711622513602712,
+      "grad_norm": 0.5684579014778137,
+      "learning_rate": 0.00014663315646662212,
+      "loss": 0.6317,
+      "step": 7126
+    },
+    {
+      "epoch": 1.2713406475782714,
+      "grad_norm": 0.5242096185684204,
+      "learning_rate": 0.00014656936749372614,
+      "loss": 0.8106,
+      "step": 7127
+    },
+    {
+      "epoch": 1.2715190437962716,
+      "grad_norm": 0.5143678188323975,
+      "learning_rate": 0.00014650558664422748,
+      "loss": 0.665,
+      "step": 7128
+    },
+    {
+      "epoch": 1.2716974400142718,
+      "grad_norm": 0.5314942598342896,
+      "learning_rate": 0.0001464418139231355,
+      "loss": 0.633,
+      "step": 7129
+    },
+    {
+      "epoch": 1.271875836232272,
+      "grad_norm": 0.5419167876243591,
+      "learning_rate": 0.0001463780493354589,
+      "loss": 0.6588,
+      "step": 7130
+    },
+    {
+      "epoch": 1.2720542324502722,
+      "grad_norm": 1.082300066947937,
+      "learning_rate": 0.00014631429288620575,
+      "loss": 0.7633,
+      "step": 7131
+    },
+    {
+      "epoch": 1.2722326286682724,
+      "grad_norm": 0.5155816674232483,
+      "learning_rate": 0.0001462505445803834,
+      "loss": 0.6571,
+      "step": 7132
+    },
+    {
+      "epoch": 1.2724110248862723,
+      "grad_norm": 0.7917349338531494,
+      "learning_rate": 0.00014618680442299864,
+      "loss": 0.5941,
+      "step": 7133
+    },
+    {
+      "epoch": 1.2725894211042725,
+      "grad_norm": 0.5167495012283325,
+      "learning_rate": 0.00014612307241905758,
+      "loss": 0.7426,
+      "step": 7134
+    },
+    {
+      "epoch": 1.2727678173222727,
+      "grad_norm": 0.6022337675094604,
+      "learning_rate": 0.00014605934857356571,
+      "loss": 0.7413,
+      "step": 7135
+    },
+    {
+      "epoch": 1.272946213540273,
+      "grad_norm": 0.4629674553871155,
+      "learning_rate": 0.0001459956328915279,
+      "loss": 0.5869,
+      "step": 7136
+    },
+    {
+      "epoch": 1.2731246097582731,
+      "grad_norm": 0.4646163582801819,
+      "learning_rate": 0.00014593192537794834,
+      "loss": 0.52,
+      "step": 7137
+    },
+    {
+      "epoch": 1.2733030059762733,
+      "grad_norm": 0.4827001988887787,
+      "learning_rate": 0.00014586822603783047,
+      "loss": 0.6536,
+      "step": 7138
+    },
+    {
+      "epoch": 1.2734814021942735,
+      "grad_norm": 0.442564457654953,
+      "learning_rate": 0.00014580453487617745,
+      "loss": 0.5261,
+      "step": 7139
+    },
+    {
+      "epoch": 1.2736597984122737,
+      "grad_norm": 0.5968412756919861,
+      "learning_rate": 0.0001457408518979913,
+      "loss": 0.8212,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2738381946302737,
+      "grad_norm": 0.5755636692047119,
+      "learning_rate": 0.00014567717710827388,
+      "loss": 0.8274,
+      "step": 7141
+    },
+    {
+      "epoch": 1.274016590848274,
+      "grad_norm": 0.5358353853225708,
+      "learning_rate": 0.0001456135105120261,
+      "loss": 0.7647,
+      "step": 7142
+    },
+    {
+      "epoch": 1.274194987066274,
+      "grad_norm": 0.45210355520248413,
+      "learning_rate": 0.00014554985211424814,
+      "loss": 0.582,
+      "step": 7143
+    },
+    {
+      "epoch": 1.2743733832842743,
+      "grad_norm": 0.5111218094825745,
+      "learning_rate": 0.00014548620191994,
+      "loss": 0.7459,
+      "step": 7144
+    },
+    {
+      "epoch": 1.2745517795022745,
+      "grad_norm": 0.4721735715866089,
+      "learning_rate": 0.00014542255993410034,
+      "loss": 0.6472,
+      "step": 7145
+    },
+    {
+      "epoch": 1.2747301757202747,
+      "grad_norm": 0.4758826196193695,
+      "learning_rate": 0.000145358926161728,
+      "loss": 0.5357,
+      "step": 7146
+    },
+    {
+      "epoch": 1.274908571938275,
+      "grad_norm": 0.5536307096481323,
+      "learning_rate": 0.00014529530060782066,
+      "loss": 0.7217,
+      "step": 7147
+    },
+    {
+      "epoch": 1.275086968156275,
+      "grad_norm": 0.465385377407074,
+      "learning_rate": 0.00014523168327737517,
+      "loss": 0.5545,
+      "step": 7148
+    },
+    {
+      "epoch": 1.2752653643742753,
+      "grad_norm": 0.47780296206474304,
+      "learning_rate": 0.0001451680741753883,
+      "loss": 0.6572,
+      "step": 7149
+    },
+    {
+      "epoch": 1.2754437605922755,
+      "grad_norm": 0.4876740276813507,
+      "learning_rate": 0.00014510447330685572,
+      "loss": 0.6728,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2756221568102757,
+      "grad_norm": 0.5187113881111145,
+      "learning_rate": 0.00014504088067677273,
+      "loss": 0.573,
+      "step": 7151
+    },
+    {
+      "epoch": 1.2758005530282759,
+      "grad_norm": 0.46036651730537415,
+      "learning_rate": 0.00014497729629013367,
+      "loss": 0.6419,
+      "step": 7152
+    },
+    {
+      "epoch": 1.275978949246276,
+      "grad_norm": 0.49554046988487244,
+      "learning_rate": 0.0001449137201519327,
+      "loss": 0.7062,
+      "step": 7153
+    },
+    {
+      "epoch": 1.2761573454642763,
+      "grad_norm": 0.44751298427581787,
+      "learning_rate": 0.00014485015226716296,
+      "loss": 0.5182,
+      "step": 7154
+    },
+    {
+      "epoch": 1.2763357416822763,
+      "grad_norm": 0.4852393567562103,
+      "learning_rate": 0.0001447865926408169,
+      "loss": 0.6061,
+      "step": 7155
+    },
+    {
+      "epoch": 1.2765141379002765,
+      "grad_norm": 0.49661117792129517,
+      "learning_rate": 0.00014472304127788663,
+      "loss": 0.7156,
+      "step": 7156
+    },
+    {
+      "epoch": 1.2766925341182767,
+      "grad_norm": 0.4922674298286438,
+      "learning_rate": 0.00014465949818336332,
+      "loss": 0.6471,
+      "step": 7157
+    },
+    {
+      "epoch": 1.2768709303362769,
+      "grad_norm": 0.4639200270175934,
+      "learning_rate": 0.0001445959633622378,
+      "loss": 0.5754,
+      "step": 7158
+    },
+    {
+      "epoch": 1.277049326554277,
+      "grad_norm": 0.4920192360877991,
+      "learning_rate": 0.00014453243681949985,
+      "loss": 0.5749,
+      "step": 7159
+    },
+    {
+      "epoch": 1.2772277227722773,
+      "grad_norm": 0.5195376873016357,
+      "learning_rate": 0.00014446891856013895,
+      "loss": 0.6442,
+      "step": 7160
+    },
+    {
+      "epoch": 1.2774061189902775,
+      "grad_norm": 0.5274832844734192,
+      "learning_rate": 0.00014440540858914384,
+      "loss": 0.6963,
+      "step": 7161
+    },
+    {
+      "epoch": 1.2775845152082776,
+      "grad_norm": 0.538936972618103,
+      "learning_rate": 0.0001443419069115024,
+      "loss": 0.7555,
+      "step": 7162
+    },
+    {
+      "epoch": 1.2777629114262776,
+      "grad_norm": 0.4810023009777069,
+      "learning_rate": 0.00014427841353220223,
+      "loss": 0.5429,
+      "step": 7163
+    },
+    {
+      "epoch": 1.2779413076442778,
+      "grad_norm": 0.49055951833724976,
+      "learning_rate": 0.00014421492845622985,
+      "loss": 0.6905,
+      "step": 7164
+    },
+    {
+      "epoch": 1.278119703862278,
+      "grad_norm": 0.4986650049686432,
+      "learning_rate": 0.0001441514516885716,
+      "loss": 0.8251,
+      "step": 7165
+    },
+    {
+      "epoch": 1.2782981000802782,
+      "grad_norm": 0.47552618384361267,
+      "learning_rate": 0.00014408798323421268,
+      "loss": 0.5881,
+      "step": 7166
+    },
+    {
+      "epoch": 1.2784764962982784,
+      "grad_norm": 0.4542260468006134,
+      "learning_rate": 0.00014402452309813808,
+      "loss": 0.549,
+      "step": 7167
+    },
+    {
+      "epoch": 1.2786548925162786,
+      "grad_norm": 0.44126781821250916,
+      "learning_rate": 0.00014396107128533182,
+      "loss": 0.499,
+      "step": 7168
+    },
+    {
+      "epoch": 1.2788332887342788,
+      "grad_norm": 0.45107805728912354,
+      "learning_rate": 0.00014389762780077725,
+      "loss": 0.5451,
+      "step": 7169
+    },
+    {
+      "epoch": 1.279011684952279,
+      "grad_norm": 0.5026717185974121,
+      "learning_rate": 0.00014383419264945747,
+      "loss": 0.634,
+      "step": 7170
+    },
+    {
+      "epoch": 1.2791900811702792,
+      "grad_norm": 0.4510766267776489,
+      "learning_rate": 0.00014377076583635442,
+      "loss": 0.5339,
+      "step": 7171
+    },
+    {
+      "epoch": 1.2793684773882794,
+      "grad_norm": 0.40873438119888306,
+      "learning_rate": 0.0001437073473664498,
+      "loss": 0.4696,
+      "step": 7172
+    },
+    {
+      "epoch": 1.2795468736062796,
+      "grad_norm": 0.48713698983192444,
+      "learning_rate": 0.0001436439372447243,
+      "loss": 0.5792,
+      "step": 7173
+    },
+    {
+      "epoch": 1.2797252698242798,
+      "grad_norm": 0.45994314551353455,
+      "learning_rate": 0.00014358053547615824,
+      "loss": 0.5733,
+      "step": 7174
+    },
+    {
+      "epoch": 1.27990366604228,
+      "grad_norm": 0.5316993594169617,
+      "learning_rate": 0.00014351714206573107,
+      "loss": 0.726,
+      "step": 7175
+    },
+    {
+      "epoch": 1.2800820622602802,
+      "grad_norm": 0.6356591582298279,
+      "learning_rate": 0.00014345375701842173,
+      "loss": 0.5762,
+      "step": 7176
+    },
+    {
+      "epoch": 1.2802604584782802,
+      "grad_norm": 0.5784276723861694,
+      "learning_rate": 0.00014339038033920858,
+      "loss": 0.6256,
+      "step": 7177
+    },
+    {
+      "epoch": 1.2804388546962804,
+      "grad_norm": 0.4795292615890503,
+      "learning_rate": 0.00014332701203306896,
+      "loss": 0.6083,
+      "step": 7178
+    },
+    {
+      "epoch": 1.2806172509142806,
+      "grad_norm": 0.46686503291130066,
+      "learning_rate": 0.00014326365210498001,
+      "loss": 0.5921,
+      "step": 7179
+    },
+    {
+      "epoch": 1.2807956471322808,
+      "grad_norm": 0.4302233159542084,
+      "learning_rate": 0.0001432003005599179,
+      "loss": 0.5002,
+      "step": 7180
+    },
+    {
+      "epoch": 1.280974043350281,
+      "grad_norm": 0.5263996124267578,
+      "learning_rate": 0.00014313695740285814,
+      "loss": 0.7031,
+      "step": 7181
+    },
+    {
+      "epoch": 1.2811524395682812,
+      "grad_norm": 0.48142942786216736,
+      "learning_rate": 0.00014307362263877581,
+      "loss": 0.5069,
+      "step": 7182
+    },
+    {
+      "epoch": 1.2813308357862814,
+      "grad_norm": 0.503164529800415,
+      "learning_rate": 0.00014301029627264512,
+      "loss": 0.6391,
+      "step": 7183
+    },
+    {
+      "epoch": 1.2815092320042816,
+      "grad_norm": 0.4538707137107849,
+      "learning_rate": 0.00014294697830943975,
+      "loss": 0.4964,
+      "step": 7184
+    },
+    {
+      "epoch": 1.2816876282222815,
+      "grad_norm": 0.5014795660972595,
+      "learning_rate": 0.00014288366875413256,
+      "loss": 0.7765,
+      "step": 7185
+    },
+    {
+      "epoch": 1.2818660244402817,
+      "grad_norm": 0.5531452298164368,
+      "learning_rate": 0.00014282036761169604,
+      "loss": 0.6846,
+      "step": 7186
+    },
+    {
+      "epoch": 1.282044420658282,
+      "grad_norm": 0.6034454703330994,
+      "learning_rate": 0.0001427570748871016,
+      "loss": 0.7885,
+      "step": 7187
+    },
+    {
+      "epoch": 1.2822228168762821,
+      "grad_norm": 0.4677254557609558,
+      "learning_rate": 0.0001426937905853205,
+      "loss": 0.6689,
+      "step": 7188
+    },
+    {
+      "epoch": 1.2824012130942823,
+      "grad_norm": 0.4888259172439575,
+      "learning_rate": 0.00014263051471132286,
+      "loss": 0.5989,
+      "step": 7189
+    },
+    {
+      "epoch": 1.2825796093122825,
+      "grad_norm": 0.5150851607322693,
+      "learning_rate": 0.0001425672472700783,
+      "loss": 0.7595,
+      "step": 7190
+    },
+    {
+      "epoch": 1.2827580055302827,
+      "grad_norm": 1.0994735956192017,
+      "learning_rate": 0.00014250398826655593,
+      "loss": 0.7208,
+      "step": 7191
+    },
+    {
+      "epoch": 1.282936401748283,
+      "grad_norm": 0.5045804977416992,
+      "learning_rate": 0.00014244073770572403,
+      "loss": 0.7697,
+      "step": 7192
+    },
+    {
+      "epoch": 1.2831147979662831,
+      "grad_norm": 0.7239548563957214,
+      "learning_rate": 0.00014237749559255043,
+      "loss": 0.7041,
+      "step": 7193
+    },
+    {
+      "epoch": 1.2832931941842833,
+      "grad_norm": 0.4901469647884369,
+      "learning_rate": 0.000142314261932002,
+      "loss": 0.7087,
+      "step": 7194
+    },
+    {
+      "epoch": 1.2834715904022835,
+      "grad_norm": 0.5142937302589417,
+      "learning_rate": 0.000142251036729045,
+      "loss": 0.6189,
+      "step": 7195
+    },
+    {
+      "epoch": 1.2836499866202837,
+      "grad_norm": 0.6226754784584045,
+      "learning_rate": 0.00014218781998864526,
+      "loss": 0.8731,
+      "step": 7196
+    },
+    {
+      "epoch": 1.283828382838284,
+      "grad_norm": 0.48372164368629456,
+      "learning_rate": 0.00014212461171576768,
+      "loss": 0.7479,
+      "step": 7197
+    },
+    {
+      "epoch": 1.2840067790562841,
+      "grad_norm": 0.5454527139663696,
+      "learning_rate": 0.0001420614119153768,
+      "loss": 0.6334,
+      "step": 7198
+    },
+    {
+      "epoch": 1.284185175274284,
+      "grad_norm": 0.5037645101547241,
+      "learning_rate": 0.00014199822059243606,
+      "loss": 0.5815,
+      "step": 7199
+    },
+    {
+      "epoch": 1.2843635714922843,
+      "grad_norm": 0.5184837579727173,
+      "learning_rate": 0.00014193503775190868,
+      "loss": 0.7455,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2845419677102845,
+      "grad_norm": 0.52362459897995,
+      "learning_rate": 0.00014187186339875696,
+      "loss": 0.6227,
+      "step": 7201
+    },
+    {
+      "epoch": 1.2847203639282847,
+      "grad_norm": 0.3797712028026581,
+      "learning_rate": 0.00014180869753794247,
+      "loss": 0.4564,
+      "step": 7202
+    },
+    {
+      "epoch": 1.284898760146285,
+      "grad_norm": 0.49934831261634827,
+      "learning_rate": 0.00014174554017442638,
+      "loss": 0.6207,
+      "step": 7203
+    },
+    {
+      "epoch": 1.285077156364285,
+      "grad_norm": 0.5866370797157288,
+      "learning_rate": 0.0001416823913131689,
+      "loss": 0.6251,
+      "step": 7204
+    },
+    {
+      "epoch": 1.2852555525822853,
+      "grad_norm": 0.5229966640472412,
+      "learning_rate": 0.00014161925095912986,
+      "loss": 0.6002,
+      "step": 7205
+    },
+    {
+      "epoch": 1.2854339488002855,
+      "grad_norm": 0.5066050887107849,
+      "learning_rate": 0.00014155611911726814,
+      "loss": 0.6648,
+      "step": 7206
+    },
+    {
+      "epoch": 1.2856123450182855,
+      "grad_norm": 0.547694981098175,
+      "learning_rate": 0.00014149299579254215,
+      "loss": 0.7228,
+      "step": 7207
+    },
+    {
+      "epoch": 1.2857907412362857,
+      "grad_norm": 0.475382000207901,
+      "learning_rate": 0.00014142988098990968,
+      "loss": 0.6569,
+      "step": 7208
+    },
+    {
+      "epoch": 1.2859691374542859,
+      "grad_norm": 0.4719434082508087,
+      "learning_rate": 0.00014136677471432755,
+      "loss": 0.6066,
+      "step": 7209
+    },
+    {
+      "epoch": 1.286147533672286,
+      "grad_norm": 0.4972936511039734,
+      "learning_rate": 0.00014130367697075225,
+      "loss": 0.6338,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2863259298902863,
+      "grad_norm": 0.4694492220878601,
+      "learning_rate": 0.0001412405877641393,
+      "loss": 0.6117,
+      "step": 7211
+    },
+    {
+      "epoch": 1.2865043261082865,
+      "grad_norm": 0.4956999719142914,
+      "learning_rate": 0.00014117750709944388,
+      "loss": 0.706,
+      "step": 7212
+    },
+    {
+      "epoch": 1.2866827223262867,
+      "grad_norm": 0.49469372630119324,
+      "learning_rate": 0.00014111443498162013,
+      "loss": 0.7405,
+      "step": 7213
+    },
+    {
+      "epoch": 1.2868611185442869,
+      "grad_norm": 0.48914238810539246,
+      "learning_rate": 0.00014105137141562192,
+      "loss": 0.6415,
+      "step": 7214
+    },
+    {
+      "epoch": 1.287039514762287,
+      "grad_norm": 0.43339380621910095,
+      "learning_rate": 0.0001409883164064021,
+      "loss": 0.4761,
+      "step": 7215
+    },
+    {
+      "epoch": 1.2872179109802873,
+      "grad_norm": 0.5167938470840454,
+      "learning_rate": 0.0001409252699589129,
+      "loss": 0.7187,
+      "step": 7216
+    },
+    {
+      "epoch": 1.2873963071982875,
+      "grad_norm": 0.45789870619773865,
+      "learning_rate": 0.00014086223207810614,
+      "loss": 0.5403,
+      "step": 7217
+    },
+    {
+      "epoch": 1.2875747034162877,
+      "grad_norm": 0.49316656589508057,
+      "learning_rate": 0.00014079920276893263,
+      "loss": 0.6763,
+      "step": 7218
+    },
+    {
+      "epoch": 1.2877530996342879,
+      "grad_norm": 0.4639813303947449,
+      "learning_rate": 0.00014073618203634282,
+      "loss": 0.5242,
+      "step": 7219
+    },
+    {
+      "epoch": 1.287931495852288,
+      "grad_norm": 0.47487959265708923,
+      "learning_rate": 0.00014067316988528616,
+      "loss": 0.5721,
+      "step": 7220
+    },
+    {
+      "epoch": 1.288109892070288,
+      "grad_norm": 0.4365386664867401,
+      "learning_rate": 0.00014061016632071173,
+      "loss": 0.4473,
+      "step": 7221
+    },
+    {
+      "epoch": 1.2882882882882882,
+      "grad_norm": 0.5029173493385315,
+      "learning_rate": 0.0001405471713475678,
+      "loss": 0.7293,
+      "step": 7222
+    },
+    {
+      "epoch": 1.2884666845062884,
+      "grad_norm": 0.6281735301017761,
+      "learning_rate": 0.00014048418497080185,
+      "loss": 0.7677,
+      "step": 7223
+    },
+    {
+      "epoch": 1.2886450807242886,
+      "grad_norm": 0.5290473103523254,
+      "learning_rate": 0.000140421207195361,
+      "loss": 0.6217,
+      "step": 7224
+    },
+    {
+      "epoch": 1.2888234769422888,
+      "grad_norm": 0.4867291748523712,
+      "learning_rate": 0.00014035823802619127,
+      "loss": 0.6605,
+      "step": 7225
+    },
+    {
+      "epoch": 1.289001873160289,
+      "grad_norm": 0.5893452167510986,
+      "learning_rate": 0.00014029527746823846,
+      "loss": 0.6862,
+      "step": 7226
+    },
+    {
+      "epoch": 1.2891802693782892,
+      "grad_norm": 0.5156494975090027,
+      "learning_rate": 0.00014023232552644733,
+      "loss": 0.658,
+      "step": 7227
+    },
+    {
+      "epoch": 1.2893586655962894,
+      "grad_norm": 0.48713207244873047,
+      "learning_rate": 0.00014016938220576204,
+      "loss": 0.6116,
+      "step": 7228
+    },
+    {
+      "epoch": 1.2895370618142896,
+      "grad_norm": 0.4963529109954834,
+      "learning_rate": 0.00014010644751112628,
+      "loss": 0.5374,
+      "step": 7229
+    },
+    {
+      "epoch": 1.2897154580322896,
+      "grad_norm": 0.5015867948532104,
+      "learning_rate": 0.00014004352144748273,
+      "loss": 0.6357,
+      "step": 7230
+    },
+    {
+      "epoch": 1.2898938542502898,
+      "grad_norm": 0.5196330547332764,
+      "learning_rate": 0.0001399806040197738,
+      "loss": 0.6529,
+      "step": 7231
+    },
+    {
+      "epoch": 1.29007225046829,
+      "grad_norm": 0.6642472147941589,
+      "learning_rate": 0.00013991769523294078,
+      "loss": 0.6634,
+      "step": 7232
+    },
+    {
+      "epoch": 1.2902506466862902,
+      "grad_norm": 0.5459299087524414,
+      "learning_rate": 0.00013985479509192472,
+      "loss": 0.7802,
+      "step": 7233
+    },
+    {
+      "epoch": 1.2904290429042904,
+      "grad_norm": 0.5235913991928101,
+      "learning_rate": 0.00013979190360166566,
+      "loss": 0.6494,
+      "step": 7234
+    },
+    {
+      "epoch": 1.2906074391222906,
+      "grad_norm": 0.4864553213119507,
+      "learning_rate": 0.00013972902076710297,
+      "loss": 0.5321,
+      "step": 7235
+    },
+    {
+      "epoch": 1.2907858353402908,
+      "grad_norm": 0.4859999418258667,
+      "learning_rate": 0.0001396661465931755,
+      "loss": 0.5942,
+      "step": 7236
+    },
+    {
+      "epoch": 1.290964231558291,
+      "grad_norm": 0.4459374248981476,
+      "learning_rate": 0.00013960328108482146,
+      "loss": 0.5264,
+      "step": 7237
+    },
+    {
+      "epoch": 1.2911426277762912,
+      "grad_norm": 0.5186498165130615,
+      "learning_rate": 0.00013954042424697827,
+      "loss": 0.7167,
+      "step": 7238
+    },
+    {
+      "epoch": 1.2913210239942914,
+      "grad_norm": 0.49951282143592834,
+      "learning_rate": 0.00013947757608458262,
+      "loss": 0.5215,
+      "step": 7239
+    },
+    {
+      "epoch": 1.2914994202122916,
+      "grad_norm": 0.4620944857597351,
+      "learning_rate": 0.00013941473660257047,
+      "loss": 0.7108,
+      "step": 7240
+    },
+    {
+      "epoch": 1.2916778164302918,
+      "grad_norm": 0.4689721167087555,
+      "learning_rate": 0.00013935190580587745,
+      "loss": 0.6203,
+      "step": 7241
+    },
+    {
+      "epoch": 1.291856212648292,
+      "grad_norm": 0.43373170495033264,
+      "learning_rate": 0.00013928908369943802,
+      "loss": 0.5734,
+      "step": 7242
+    },
+    {
+      "epoch": 1.292034608866292,
+      "grad_norm": 0.48551568388938904,
+      "learning_rate": 0.00013922627028818642,
+      "loss": 0.6898,
+      "step": 7243
+    },
+    {
+      "epoch": 1.2922130050842922,
+      "grad_norm": 0.4706454873085022,
+      "learning_rate": 0.00013916346557705579,
+      "loss": 0.6658,
+      "step": 7244
+    },
+    {
+      "epoch": 1.2923914013022924,
+      "grad_norm": 0.47324392199516296,
+      "learning_rate": 0.00013910066957097895,
+      "loss": 0.6189,
+      "step": 7245
+    },
+    {
+      "epoch": 1.2925697975202926,
+      "grad_norm": 0.4348990321159363,
+      "learning_rate": 0.00013903788227488773,
+      "loss": 0.5601,
+      "step": 7246
+    },
+    {
+      "epoch": 1.2927481937382928,
+      "grad_norm": 0.4262705445289612,
+      "learning_rate": 0.00013897510369371359,
+      "loss": 0.4819,
+      "step": 7247
+    },
+    {
+      "epoch": 1.292926589956293,
+      "grad_norm": 0.5076077580451965,
+      "learning_rate": 0.000138912333832387,
+      "loss": 0.8249,
+      "step": 7248
+    },
+    {
+      "epoch": 1.2931049861742931,
+      "grad_norm": 0.44083496928215027,
+      "learning_rate": 0.00013884957269583777,
+      "loss": 0.5899,
+      "step": 7249
+    },
+    {
+      "epoch": 1.2932833823922933,
+      "grad_norm": 0.5185807347297668,
+      "learning_rate": 0.00013878682028899543,
+      "loss": 0.6274,
+      "step": 7250
+    },
+    {
+      "epoch": 1.2934617786102935,
+      "grad_norm": 0.46207600831985474,
+      "learning_rate": 0.00013872407661678825,
+      "loss": 0.6729,
+      "step": 7251
+    },
+    {
+      "epoch": 1.2936401748282935,
+      "grad_norm": 0.49033087491989136,
+      "learning_rate": 0.00013866134168414421,
+      "loss": 0.5321,
+      "step": 7252
+    },
+    {
+      "epoch": 1.2938185710462937,
+      "grad_norm": 0.728157103061676,
+      "learning_rate": 0.00013859861549599058,
+      "loss": 0.4031,
+      "step": 7253
+    },
+    {
+      "epoch": 1.293996967264294,
+      "grad_norm": 0.5120450258255005,
+      "learning_rate": 0.00013853589805725363,
+      "loss": 0.6634,
+      "step": 7254
+    },
+    {
+      "epoch": 1.2941753634822941,
+      "grad_norm": 0.4849170446395874,
+      "learning_rate": 0.00013847318937285942,
+      "loss": 0.6161,
+      "step": 7255
+    },
+    {
+      "epoch": 1.2943537597002943,
+      "grad_norm": 0.5624775886535645,
+      "learning_rate": 0.00013841048944773278,
+      "loss": 0.6795,
+      "step": 7256
+    },
+    {
+      "epoch": 1.2945321559182945,
+      "grad_norm": 0.5056706070899963,
+      "learning_rate": 0.00013834779828679838,
+      "loss": 0.7651,
+      "step": 7257
+    },
+    {
+      "epoch": 1.2947105521362947,
+      "grad_norm": 0.5136926174163818,
+      "learning_rate": 0.00013828511589497977,
+      "loss": 0.8467,
+      "step": 7258
+    },
+    {
+      "epoch": 1.294888948354295,
+      "grad_norm": 0.4533271789550781,
+      "learning_rate": 0.0001382224422772002,
+      "loss": 0.536,
+      "step": 7259
+    },
+    {
+      "epoch": 1.295067344572295,
+      "grad_norm": 0.4658295214176178,
+      "learning_rate": 0.00013815977743838188,
+      "loss": 0.5347,
+      "step": 7260
+    },
+    {
+      "epoch": 1.2952457407902953,
+      "grad_norm": 0.4228724539279938,
+      "learning_rate": 0.00013809712138344643,
+      "loss": 0.5316,
+      "step": 7261
+    },
+    {
+      "epoch": 1.2954241370082955,
+      "grad_norm": 0.4298241436481476,
+      "learning_rate": 0.000138034474117315,
+      "loss": 0.4455,
+      "step": 7262
+    },
+    {
+      "epoch": 1.2956025332262957,
+      "grad_norm": 0.4992910623550415,
+      "learning_rate": 0.00013797183564490773,
+      "loss": 0.6047,
+      "step": 7263
+    },
+    {
+      "epoch": 1.295780929444296,
+      "grad_norm": 0.49004730582237244,
+      "learning_rate": 0.00013790920597114433,
+      "loss": 0.6561,
+      "step": 7264
+    },
+    {
+      "epoch": 1.2959593256622959,
+      "grad_norm": 1.2394171953201294,
+      "learning_rate": 0.00013784658510094356,
+      "loss": 0.5966,
+      "step": 7265
+    },
+    {
+      "epoch": 1.296137721880296,
+      "grad_norm": 0.40918824076652527,
+      "learning_rate": 0.00013778397303922387,
+      "loss": 0.4365,
+      "step": 7266
+    },
+    {
+      "epoch": 1.2963161180982963,
+      "grad_norm": 0.521958589553833,
+      "learning_rate": 0.0001377213697909025,
+      "loss": 0.6175,
+      "step": 7267
+    },
+    {
+      "epoch": 1.2964945143162965,
+      "grad_norm": 0.594232976436615,
+      "learning_rate": 0.00013765877536089648,
+      "loss": 0.7357,
+      "step": 7268
+    },
+    {
+      "epoch": 1.2966729105342967,
+      "grad_norm": 0.6050662398338318,
+      "learning_rate": 0.00013759618975412198,
+      "loss": 0.8643,
+      "step": 7269
+    },
+    {
+      "epoch": 1.2968513067522969,
+      "grad_norm": 0.4976935088634491,
+      "learning_rate": 0.00013753361297549421,
+      "loss": 0.5672,
+      "step": 7270
+    },
+    {
+      "epoch": 1.297029702970297,
+      "grad_norm": 0.4690896272659302,
+      "learning_rate": 0.00013747104502992823,
+      "loss": 0.5142,
+      "step": 7271
+    },
+    {
+      "epoch": 1.2972080991882973,
+      "grad_norm": 0.5143970251083374,
+      "learning_rate": 0.00013740848592233785,
+      "loss": 0.6846,
+      "step": 7272
+    },
+    {
+      "epoch": 1.2973864954062975,
+      "grad_norm": 0.5174264907836914,
+      "learning_rate": 0.00013734593565763664,
+      "loss": 0.6434,
+      "step": 7273
+    },
+    {
+      "epoch": 1.2975648916242974,
+      "grad_norm": 0.4763343334197998,
+      "learning_rate": 0.00013728339424073715,
+      "loss": 0.6206,
+      "step": 7274
+    },
+    {
+      "epoch": 1.2977432878422976,
+      "grad_norm": 0.44902318716049194,
+      "learning_rate": 0.00013722086167655128,
+      "loss": 0.4841,
+      "step": 7275
+    },
+    {
+      "epoch": 1.2979216840602978,
+      "grad_norm": 0.6365222930908203,
+      "learning_rate": 0.0001371583379699905,
+      "loss": 0.7602,
+      "step": 7276
+    },
+    {
+      "epoch": 1.298100080278298,
+      "grad_norm": 0.49694737792015076,
+      "learning_rate": 0.0001370958231259652,
+      "loss": 0.5375,
+      "step": 7277
+    },
+    {
+      "epoch": 1.2982784764962982,
+      "grad_norm": 0.507412850856781,
+      "learning_rate": 0.00013703331714938545,
+      "loss": 0.6576,
+      "step": 7278
+    },
+    {
+      "epoch": 1.2984568727142984,
+      "grad_norm": 0.45106565952301025,
+      "learning_rate": 0.00013697082004516026,
+      "loss": 0.5242,
+      "step": 7279
+    },
+    {
+      "epoch": 1.2986352689322986,
+      "grad_norm": 0.4997382164001465,
+      "learning_rate": 0.00013690833181819834,
+      "loss": 0.6324,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2988136651502988,
+      "grad_norm": 0.4807584583759308,
+      "learning_rate": 0.00013684585247340734,
+      "loss": 0.6275,
+      "step": 7281
+    },
+    {
+      "epoch": 1.298992061368299,
+      "grad_norm": 0.5849037170410156,
+      "learning_rate": 0.00013678338201569422,
+      "loss": 0.7473,
+      "step": 7282
+    },
+    {
+      "epoch": 1.2991704575862992,
+      "grad_norm": 0.5092651844024658,
+      "learning_rate": 0.00013672092044996576,
+      "loss": 0.6442,
+      "step": 7283
+    },
+    {
+      "epoch": 1.2993488538042994,
+      "grad_norm": 0.45579296350479126,
+      "learning_rate": 0.00013665846778112734,
+      "loss": 0.4597,
+      "step": 7284
+    },
+    {
+      "epoch": 1.2995272500222996,
+      "grad_norm": 0.5193167924880981,
+      "learning_rate": 0.00013659602401408416,
+      "loss": 0.614,
+      "step": 7285
+    },
+    {
+      "epoch": 1.2997056462402998,
+      "grad_norm": 0.511915922164917,
+      "learning_rate": 0.0001365335891537405,
+      "loss": 0.69,
+      "step": 7286
+    },
+    {
+      "epoch": 1.2998840424582998,
+      "grad_norm": 0.47553443908691406,
+      "learning_rate": 0.0001364711632049998,
+      "loss": 0.5546,
+      "step": 7287
+    },
+    {
+      "epoch": 1.3000624386763,
+      "grad_norm": 0.4340362250804901,
+      "learning_rate": 0.00013640874617276523,
+      "loss": 0.4619,
+      "step": 7288
+    },
+    {
+      "epoch": 1.3002408348943002,
+      "grad_norm": 0.558949887752533,
+      "learning_rate": 0.00013634633806193868,
+      "loss": 1.1185,
+      "step": 7289
+    },
+    {
+      "epoch": 1.3004192311123004,
+      "grad_norm": 0.45011255145072937,
+      "learning_rate": 0.00013628393887742197,
+      "loss": 0.5726,
+      "step": 7290
+    },
+    {
+      "epoch": 1.3005976273303006,
+      "grad_norm": 0.44671180844306946,
+      "learning_rate": 0.00013622154862411568,
+      "loss": 0.549,
+      "step": 7291
+    },
+    {
+      "epoch": 1.3007760235483008,
+      "grad_norm": 0.5005772709846497,
+      "learning_rate": 0.00013615916730692006,
+      "loss": 0.6959,
+      "step": 7292
+    },
+    {
+      "epoch": 1.300954419766301,
+      "grad_norm": 0.5107164978981018,
+      "learning_rate": 0.00013609679493073435,
+      "loss": 0.5958,
+      "step": 7293
+    },
+    {
+      "epoch": 1.3011328159843012,
+      "grad_norm": 0.46262556314468384,
+      "learning_rate": 0.00013603443150045745,
+      "loss": 0.531,
+      "step": 7294
+    },
+    {
+      "epoch": 1.3013112122023014,
+      "grad_norm": 0.49760428071022034,
+      "learning_rate": 0.0001359720770209873,
+      "loss": 0.5535,
+      "step": 7295
+    },
+    {
+      "epoch": 1.3014896084203014,
+      "grad_norm": 0.5045909881591797,
+      "learning_rate": 0.00013590973149722103,
+      "loss": 0.737,
+      "step": 7296
+    },
+    {
+      "epoch": 1.3016680046383016,
+      "grad_norm": 0.4710780680179596,
+      "learning_rate": 0.00013584739493405546,
+      "loss": 0.6419,
+      "step": 7297
+    },
+    {
+      "epoch": 1.3018464008563018,
+      "grad_norm": 0.5229588747024536,
+      "learning_rate": 0.00013578506733638622,
+      "loss": 0.6623,
+      "step": 7298
+    },
+    {
+      "epoch": 1.302024797074302,
+      "grad_norm": 0.48090076446533203,
+      "learning_rate": 0.0001357227487091087,
+      "loss": 0.645,
+      "step": 7299
+    },
+    {
+      "epoch": 1.3022031932923022,
+      "grad_norm": 0.4656026363372803,
+      "learning_rate": 0.0001356604390571174,
+      "loss": 0.482,
+      "step": 7300
+    },
+    {
+      "epoch": 1.3023815895103024,
+      "grad_norm": 0.4530593454837799,
+      "learning_rate": 0.00013559813838530588,
+      "loss": 0.4968,
+      "step": 7301
+    },
+    {
+      "epoch": 1.3025599857283026,
+      "grad_norm": 0.5168888568878174,
+      "learning_rate": 0.0001355358466985675,
+      "loss": 0.6378,
+      "step": 7302
+    },
+    {
+      "epoch": 1.3027383819463028,
+      "grad_norm": 0.45193690061569214,
+      "learning_rate": 0.00013547356400179432,
+      "loss": 0.4601,
+      "step": 7303
+    },
+    {
+      "epoch": 1.302916778164303,
+      "grad_norm": 0.5073840618133545,
+      "learning_rate": 0.00013541129029987826,
+      "loss": 0.5461,
+      "step": 7304
+    },
+    {
+      "epoch": 1.3030951743823032,
+      "grad_norm": 0.5524479150772095,
+      "learning_rate": 0.00013534902559771,
+      "loss": 0.6818,
+      "step": 7305
+    },
+    {
+      "epoch": 1.3032735706003034,
+      "grad_norm": 0.5372066497802734,
+      "learning_rate": 0.00013528676990018007,
+      "loss": 0.7605,
+      "step": 7306
+    },
+    {
+      "epoch": 1.3034519668183036,
+      "grad_norm": 0.5457733869552612,
+      "learning_rate": 0.00013522452321217788,
+      "loss": 0.7541,
+      "step": 7307
+    },
+    {
+      "epoch": 1.3036303630363038,
+      "grad_norm": 0.5292448997497559,
+      "learning_rate": 0.00013516228553859212,
+      "loss": 0.6048,
+      "step": 7308
+    },
+    {
+      "epoch": 1.3038087592543037,
+      "grad_norm": 0.45932427048683167,
+      "learning_rate": 0.00013510005688431115,
+      "loss": 0.5494,
+      "step": 7309
+    },
+    {
+      "epoch": 1.303987155472304,
+      "grad_norm": 0.4594475030899048,
+      "learning_rate": 0.00013503783725422216,
+      "loss": 0.4971,
+      "step": 7310
+    },
+    {
+      "epoch": 1.3041655516903041,
+      "grad_norm": 0.5574771165847778,
+      "learning_rate": 0.00013497562665321206,
+      "loss": 0.5786,
+      "step": 7311
+    },
+    {
+      "epoch": 1.3043439479083043,
+      "grad_norm": 0.46712732315063477,
+      "learning_rate": 0.00013491342508616667,
+      "loss": 0.5883,
+      "step": 7312
+    },
+    {
+      "epoch": 1.3045223441263045,
+      "grad_norm": 0.47527366876602173,
+      "learning_rate": 0.00013485123255797132,
+      "loss": 0.6923,
+      "step": 7313
+    },
+    {
+      "epoch": 1.3047007403443047,
+      "grad_norm": 0.45830777287483215,
+      "learning_rate": 0.0001347890490735107,
+      "loss": 0.5369,
+      "step": 7314
+    },
+    {
+      "epoch": 1.304879136562305,
+      "grad_norm": 0.5052610039710999,
+      "learning_rate": 0.0001347268746376685,
+      "loss": 0.5744,
+      "step": 7315
+    },
+    {
+      "epoch": 1.3050575327803051,
+      "grad_norm": 0.4734193682670593,
+      "learning_rate": 0.00013466470925532808,
+      "loss": 0.5339,
+      "step": 7316
+    },
+    {
+      "epoch": 1.3052359289983053,
+      "grad_norm": 0.4872669279575348,
+      "learning_rate": 0.00013460255293137164,
+      "loss": 0.5804,
+      "step": 7317
+    },
+    {
+      "epoch": 1.3054143252163053,
+      "grad_norm": 0.5103761553764343,
+      "learning_rate": 0.00013454040567068113,
+      "loss": 0.6856,
+      "step": 7318
+    },
+    {
+      "epoch": 1.3055927214343055,
+      "grad_norm": 0.5222267508506775,
+      "learning_rate": 0.00013447826747813748,
+      "loss": 0.7514,
+      "step": 7319
+    },
+    {
+      "epoch": 1.3057711176523057,
+      "grad_norm": 0.4690699875354767,
+      "learning_rate": 0.0001344161383586209,
+      "loss": 0.6305,
+      "step": 7320
+    },
+    {
+      "epoch": 1.305949513870306,
+      "grad_norm": 0.5072498917579651,
+      "learning_rate": 0.00013435401831701115,
+      "loss": 0.6352,
+      "step": 7321
+    },
+    {
+      "epoch": 1.306127910088306,
+      "grad_norm": 0.4983615577220917,
+      "learning_rate": 0.00013429190735818696,
+      "loss": 0.5993,
+      "step": 7322
+    },
+    {
+      "epoch": 1.3063063063063063,
+      "grad_norm": 0.5258157253265381,
+      "learning_rate": 0.0001342298054870267,
+      "loss": 0.7972,
+      "step": 7323
+    },
+    {
+      "epoch": 1.3064847025243065,
+      "grad_norm": 0.4537825882434845,
+      "learning_rate": 0.00013416771270840751,
+      "loss": 0.5255,
+      "step": 7324
+    },
+    {
+      "epoch": 1.3066630987423067,
+      "grad_norm": 0.47432446479797363,
+      "learning_rate": 0.00013410562902720647,
+      "loss": 0.4699,
+      "step": 7325
+    },
+    {
+      "epoch": 1.3068414949603069,
+      "grad_norm": 0.5209951996803284,
+      "learning_rate": 0.00013404355444829934,
+      "loss": 0.6747,
+      "step": 7326
+    },
+    {
+      "epoch": 1.307019891178307,
+      "grad_norm": 0.4715597927570343,
+      "learning_rate": 0.00013398148897656164,
+      "loss": 0.5054,
+      "step": 7327
+    },
+    {
+      "epoch": 1.3071982873963073,
+      "grad_norm": 0.8090279698371887,
+      "learning_rate": 0.00013391943261686782,
+      "loss": 0.4708,
+      "step": 7328
+    },
+    {
+      "epoch": 1.3073766836143075,
+      "grad_norm": 0.500850260257721,
+      "learning_rate": 0.00013385738537409174,
+      "loss": 0.6052,
+      "step": 7329
+    },
+    {
+      "epoch": 1.3075550798323077,
+      "grad_norm": 0.5126155018806458,
+      "learning_rate": 0.00013379534725310678,
+      "loss": 0.5697,
+      "step": 7330
+    },
+    {
+      "epoch": 1.3077334760503077,
+      "grad_norm": 0.5333918929100037,
+      "learning_rate": 0.00013373331825878516,
+      "loss": 0.695,
+      "step": 7331
+    },
+    {
+      "epoch": 1.3079118722683079,
+      "grad_norm": 0.5330007076263428,
+      "learning_rate": 0.00013367129839599872,
+      "loss": 0.5267,
+      "step": 7332
+    },
+    {
+      "epoch": 1.308090268486308,
+      "grad_norm": 0.5808795690536499,
+      "learning_rate": 0.0001336092876696185,
+      "loss": 0.8733,
+      "step": 7333
+    },
+    {
+      "epoch": 1.3082686647043082,
+      "grad_norm": 0.5341413021087646,
+      "learning_rate": 0.0001335472860845146,
+      "loss": 0.778,
+      "step": 7334
+    },
+    {
+      "epoch": 1.3084470609223084,
+      "grad_norm": 0.4854552447795868,
+      "learning_rate": 0.00013348529364555685,
+      "loss": 0.5773,
+      "step": 7335
+    },
+    {
+      "epoch": 1.3086254571403086,
+      "grad_norm": 0.4257410168647766,
+      "learning_rate": 0.0001334233103576139,
+      "loss": 0.5018,
+      "step": 7336
+    },
+    {
+      "epoch": 1.3088038533583088,
+      "grad_norm": 0.5281109809875488,
+      "learning_rate": 0.0001333613362255541,
+      "loss": 0.633,
+      "step": 7337
+    },
+    {
+      "epoch": 1.308982249576309,
+      "grad_norm": 0.45057862997055054,
+      "learning_rate": 0.00013329937125424466,
+      "loss": 0.5499,
+      "step": 7338
+    },
+    {
+      "epoch": 1.3091606457943092,
+      "grad_norm": 0.568513035774231,
+      "learning_rate": 0.00013323741544855246,
+      "loss": 0.7561,
+      "step": 7339
+    },
+    {
+      "epoch": 1.3093390420123092,
+      "grad_norm": 0.4433366358280182,
+      "learning_rate": 0.00013317546881334342,
+      "loss": 0.4844,
+      "step": 7340
+    },
+    {
+      "epoch": 1.3095174382303094,
+      "grad_norm": 0.4924166798591614,
+      "learning_rate": 0.00013311353135348267,
+      "loss": 0.5766,
+      "step": 7341
+    },
+    {
+      "epoch": 1.3096958344483096,
+      "grad_norm": 0.531856894493103,
+      "learning_rate": 0.00013305160307383495,
+      "loss": 0.6314,
+      "step": 7342
+    },
+    {
+      "epoch": 1.3098742306663098,
+      "grad_norm": 0.5930870175361633,
+      "learning_rate": 0.00013298968397926398,
+      "loss": 0.7782,
+      "step": 7343
+    },
+    {
+      "epoch": 1.31005262688431,
+      "grad_norm": 0.5085564255714417,
+      "learning_rate": 0.0001329277740746328,
+      "loss": 0.6933,
+      "step": 7344
+    },
+    {
+      "epoch": 1.3102310231023102,
+      "grad_norm": 0.5098043084144592,
+      "learning_rate": 0.000132865873364804,
+      "loss": 0.6053,
+      "step": 7345
+    },
+    {
+      "epoch": 1.3104094193203104,
+      "grad_norm": 0.49799683690071106,
+      "learning_rate": 0.00013280398185463898,
+      "loss": 0.6819,
+      "step": 7346
+    },
+    {
+      "epoch": 1.3105878155383106,
+      "grad_norm": 0.4547264277935028,
+      "learning_rate": 0.00013274209954899888,
+      "loss": 0.4916,
+      "step": 7347
+    },
+    {
+      "epoch": 1.3107662117563108,
+      "grad_norm": 0.48524683713912964,
+      "learning_rate": 0.00013268022645274375,
+      "loss": 0.7242,
+      "step": 7348
+    },
+    {
+      "epoch": 1.310944607974311,
+      "grad_norm": 0.4403549134731293,
+      "learning_rate": 0.00013261836257073324,
+      "loss": 0.5146,
+      "step": 7349
+    },
+    {
+      "epoch": 1.3111230041923112,
+      "grad_norm": 0.4839078485965729,
+      "learning_rate": 0.00013255650790782591,
+      "loss": 0.5517,
+      "step": 7350
+    },
+    {
+      "epoch": 1.3113014004103114,
+      "grad_norm": 0.4646989107131958,
+      "learning_rate": 0.00013249466246888,
+      "loss": 0.5391,
+      "step": 7351
+    },
+    {
+      "epoch": 1.3114797966283116,
+      "grad_norm": 0.46415239572525024,
+      "learning_rate": 0.00013243282625875267,
+      "loss": 0.6404,
+      "step": 7352
+    },
+    {
+      "epoch": 1.3116581928463116,
+      "grad_norm": 0.5185125470161438,
+      "learning_rate": 0.00013237099928230066,
+      "loss": 0.7389,
+      "step": 7353
+    },
+    {
+      "epoch": 1.3118365890643118,
+      "grad_norm": 0.4275851547718048,
+      "learning_rate": 0.0001323091815443797,
+      "loss": 0.5303,
+      "step": 7354
+    },
+    {
+      "epoch": 1.312014985282312,
+      "grad_norm": 0.41446274518966675,
+      "learning_rate": 0.00013224737304984494,
+      "loss": 0.4631,
+      "step": 7355
+    },
+    {
+      "epoch": 1.3121933815003122,
+      "grad_norm": 0.4250105917453766,
+      "learning_rate": 0.0001321855738035509,
+      "loss": 0.483,
+      "step": 7356
+    },
+    {
+      "epoch": 1.3123717777183124,
+      "grad_norm": 0.501487135887146,
+      "learning_rate": 0.0001321237838103511,
+      "loss": 0.6878,
+      "step": 7357
+    },
+    {
+      "epoch": 1.3125501739363126,
+      "grad_norm": 0.5788127183914185,
+      "learning_rate": 0.0001320620030750987,
+      "loss": 0.5583,
+      "step": 7358
+    },
+    {
+      "epoch": 1.3127285701543128,
+      "grad_norm": 0.5718269348144531,
+      "learning_rate": 0.00013200023160264568,
+      "loss": 0.7069,
+      "step": 7359
+    },
+    {
+      "epoch": 1.312906966372313,
+      "grad_norm": 0.5326660871505737,
+      "learning_rate": 0.00013193846939784374,
+      "loss": 0.7465,
+      "step": 7360
+    },
+    {
+      "epoch": 1.3130853625903132,
+      "grad_norm": 0.5201784372329712,
+      "learning_rate": 0.00013187671646554367,
+      "loss": 0.6788,
+      "step": 7361
+    },
+    {
+      "epoch": 1.3132637588083131,
+      "grad_norm": 0.5949232578277588,
+      "learning_rate": 0.0001318149728105954,
+      "loss": 0.6459,
+      "step": 7362
+    },
+    {
+      "epoch": 1.3134421550263133,
+      "grad_norm": 0.4743814766407013,
+      "learning_rate": 0.00013175323843784837,
+      "loss": 0.705,
+      "step": 7363
+    },
+    {
+      "epoch": 1.3136205512443135,
+      "grad_norm": 0.4792519211769104,
+      "learning_rate": 0.00013169151335215101,
+      "loss": 0.6492,
+      "step": 7364
+    },
+    {
+      "epoch": 1.3137989474623137,
+      "grad_norm": 0.4962617754936218,
+      "learning_rate": 0.00013162979755835142,
+      "loss": 0.5276,
+      "step": 7365
+    },
+    {
+      "epoch": 1.313977343680314,
+      "grad_norm": 0.438672810792923,
+      "learning_rate": 0.00013156809106129656,
+      "loss": 0.4126,
+      "step": 7366
+    },
+    {
+      "epoch": 1.3141557398983141,
+      "grad_norm": 0.43237635493278503,
+      "learning_rate": 0.00013150639386583278,
+      "loss": 0.5041,
+      "step": 7367
+    },
+    {
+      "epoch": 1.3143341361163143,
+      "grad_norm": 0.4956110119819641,
+      "learning_rate": 0.00013144470597680592,
+      "loss": 0.6361,
+      "step": 7368
+    },
+    {
+      "epoch": 1.3145125323343145,
+      "grad_norm": 0.4873587191104889,
+      "learning_rate": 0.00013138302739906072,
+      "loss": 0.7563,
+      "step": 7369
+    },
+    {
+      "epoch": 1.3146909285523147,
+      "grad_norm": 0.4835222363471985,
+      "learning_rate": 0.0001313213581374416,
+      "loss": 0.5999,
+      "step": 7370
+    },
+    {
+      "epoch": 1.314869324770315,
+      "grad_norm": 0.4809320867061615,
+      "learning_rate": 0.00013125969819679188,
+      "loss": 0.6086,
+      "step": 7371
+    },
+    {
+      "epoch": 1.3150477209883151,
+      "grad_norm": 0.4588382840156555,
+      "learning_rate": 0.00013119804758195442,
+      "loss": 0.4699,
+      "step": 7372
+    },
+    {
+      "epoch": 1.3152261172063153,
+      "grad_norm": 0.4489077031612396,
+      "learning_rate": 0.00013113640629777113,
+      "loss": 0.3956,
+      "step": 7373
+    },
+    {
+      "epoch": 1.3154045134243155,
+      "grad_norm": 0.5158323049545288,
+      "learning_rate": 0.0001310747743490833,
+      "loss": 0.6929,
+      "step": 7374
+    },
+    {
+      "epoch": 1.3155829096423155,
+      "grad_norm": 0.5298815965652466,
+      "learning_rate": 0.0001310131517407316,
+      "loss": 0.7762,
+      "step": 7375
+    },
+    {
+      "epoch": 1.3157613058603157,
+      "grad_norm": 0.5620563626289368,
+      "learning_rate": 0.0001309515384775557,
+      "loss": 0.5016,
+      "step": 7376
+    },
+    {
+      "epoch": 1.315939702078316,
+      "grad_norm": 0.5272053480148315,
+      "learning_rate": 0.0001308899345643948,
+      "loss": 0.8066,
+      "step": 7377
+    },
+    {
+      "epoch": 1.316118098296316,
+      "grad_norm": 0.4727185368537903,
+      "learning_rate": 0.00013082834000608724,
+      "loss": 0.5287,
+      "step": 7378
+    },
+    {
+      "epoch": 1.3162964945143163,
+      "grad_norm": 0.4828498959541321,
+      "learning_rate": 0.00013076675480747042,
+      "loss": 0.6108,
+      "step": 7379
+    },
+    {
+      "epoch": 1.3164748907323165,
+      "grad_norm": 0.45956501364707947,
+      "learning_rate": 0.00013070517897338147,
+      "loss": 0.5884,
+      "step": 7380
+    },
+    {
+      "epoch": 1.3166532869503167,
+      "grad_norm": 0.49907490611076355,
+      "learning_rate": 0.00013064361250865637,
+      "loss": 0.5318,
+      "step": 7381
+    },
+    {
+      "epoch": 1.316831683168317,
+      "grad_norm": 0.48141568899154663,
+      "learning_rate": 0.0001305820554181306,
+      "loss": 0.7263,
+      "step": 7382
+    },
+    {
+      "epoch": 1.317010079386317,
+      "grad_norm": 0.4059731960296631,
+      "learning_rate": 0.0001305205077066388,
+      "loss": 0.4103,
+      "step": 7383
+    },
+    {
+      "epoch": 1.317188475604317,
+      "grad_norm": 0.48073357343673706,
+      "learning_rate": 0.00013045896937901496,
+      "loss": 0.5717,
+      "step": 7384
+    },
+    {
+      "epoch": 1.3173668718223173,
+      "grad_norm": 0.4782210886478424,
+      "learning_rate": 0.00013039744044009212,
+      "loss": 0.6653,
+      "step": 7385
+    },
+    {
+      "epoch": 1.3175452680403175,
+      "grad_norm": 0.5318115949630737,
+      "learning_rate": 0.00013033592089470295,
+      "loss": 0.5354,
+      "step": 7386
+    },
+    {
+      "epoch": 1.3177236642583177,
+      "grad_norm": 0.5100518465042114,
+      "learning_rate": 0.00013027441074767903,
+      "loss": 0.6491,
+      "step": 7387
+    },
+    {
+      "epoch": 1.3179020604763179,
+      "grad_norm": 0.4386556148529053,
+      "learning_rate": 0.00013021291000385132,
+      "loss": 0.4745,
+      "step": 7388
+    },
+    {
+      "epoch": 1.318080456694318,
+      "grad_norm": 0.5141617655754089,
+      "learning_rate": 0.0001301514186680502,
+      "loss": 0.5839,
+      "step": 7389
+    },
+    {
+      "epoch": 1.3182588529123183,
+      "grad_norm": 7.867745399475098,
+      "learning_rate": 0.00013008993674510483,
+      "loss": 0.5498,
+      "step": 7390
+    },
+    {
+      "epoch": 1.3184372491303185,
+      "grad_norm": 0.5267961621284485,
+      "learning_rate": 0.00013002846423984448,
+      "loss": 0.6516,
+      "step": 7391
+    },
+    {
+      "epoch": 1.3186156453483187,
+      "grad_norm": 0.47720450162887573,
+      "learning_rate": 0.00012996700115709692,
+      "loss": 0.5947,
+      "step": 7392
+    },
+    {
+      "epoch": 1.3187940415663189,
+      "grad_norm": 0.514872133731842,
+      "learning_rate": 0.00012990554750168931,
+      "loss": 0.6416,
+      "step": 7393
+    },
+    {
+      "epoch": 1.318972437784319,
+      "grad_norm": 0.443766713142395,
+      "learning_rate": 0.00012984410327844843,
+      "loss": 0.4777,
+      "step": 7394
+    },
+    {
+      "epoch": 1.3191508340023193,
+      "grad_norm": 0.42529335618019104,
+      "learning_rate": 0.00012978266849219985,
+      "loss": 0.5089,
+      "step": 7395
+    },
+    {
+      "epoch": 1.3193292302203194,
+      "grad_norm": 0.540155291557312,
+      "learning_rate": 0.00012972124314776886,
+      "loss": 0.7208,
+      "step": 7396
+    },
+    {
+      "epoch": 1.3195076264383194,
+      "grad_norm": 0.4460678994655609,
+      "learning_rate": 0.0001296598272499796,
+      "loss": 0.5367,
+      "step": 7397
+    },
+    {
+      "epoch": 1.3196860226563196,
+      "grad_norm": 0.5822709798812866,
+      "learning_rate": 0.0001295984208036558,
+      "loss": 0.5829,
+      "step": 7398
+    },
+    {
+      "epoch": 1.3198644188743198,
+      "grad_norm": 0.4621860086917877,
+      "learning_rate": 0.00012953702381362023,
+      "loss": 0.5372,
+      "step": 7399
+    },
+    {
+      "epoch": 1.32004281509232,
+      "grad_norm": 0.5000967979431152,
+      "learning_rate": 0.00012947563628469487,
+      "loss": 0.5573,
+      "step": 7400
+    },
+    {
+      "epoch": 1.3202212113103202,
+      "grad_norm": 0.4751860797405243,
+      "learning_rate": 0.00012941425822170124,
+      "loss": 0.6386,
+      "step": 7401
+    },
+    {
+      "epoch": 1.3203996075283204,
+      "grad_norm": 0.5089471340179443,
+      "learning_rate": 0.0001293528896294598,
+      "loss": 0.664,
+      "step": 7402
+    },
+    {
+      "epoch": 1.3205780037463206,
+      "grad_norm": 0.5417841076850891,
+      "learning_rate": 0.00012929153051279062,
+      "loss": 0.5501,
+      "step": 7403
+    },
+    {
+      "epoch": 1.3207563999643208,
+      "grad_norm": 0.5360578298568726,
+      "learning_rate": 0.00012923018087651256,
+      "loss": 0.6679,
+      "step": 7404
+    },
+    {
+      "epoch": 1.320934796182321,
+      "grad_norm": 0.48479247093200684,
+      "learning_rate": 0.0001291688407254441,
+      "loss": 0.4484,
+      "step": 7405
+    },
+    {
+      "epoch": 1.321113192400321,
+      "grad_norm": 0.5442363023757935,
+      "learning_rate": 0.000129107510064403,
+      "loss": 0.6294,
+      "step": 7406
+    },
+    {
+      "epoch": 1.3212915886183212,
+      "grad_norm": 0.41899165511131287,
+      "learning_rate": 0.00012904618889820595,
+      "loss": 0.5342,
+      "step": 7407
+    },
+    {
+      "epoch": 1.3214699848363214,
+      "grad_norm": 0.49729251861572266,
+      "learning_rate": 0.0001289848772316693,
+      "loss": 0.7296,
+      "step": 7408
+    },
+    {
+      "epoch": 1.3216483810543216,
+      "grad_norm": 0.49834224581718445,
+      "learning_rate": 0.00012892357506960817,
+      "loss": 0.654,
+      "step": 7409
+    },
+    {
+      "epoch": 1.3218267772723218,
+      "grad_norm": 0.47109490633010864,
+      "learning_rate": 0.0001288622824168375,
+      "loss": 0.6164,
+      "step": 7410
+    },
+    {
+      "epoch": 1.322005173490322,
+      "grad_norm": 0.43739229440689087,
+      "learning_rate": 0.0001288009992781709,
+      "loss": 0.5467,
+      "step": 7411
+    },
+    {
+      "epoch": 1.3221835697083222,
+      "grad_norm": 0.4927254617214203,
+      "learning_rate": 0.00012873972565842173,
+      "loss": 0.5437,
+      "step": 7412
+    },
+    {
+      "epoch": 1.3223619659263224,
+      "grad_norm": 0.4604239761829376,
+      "learning_rate": 0.00012867846156240238,
+      "loss": 0.5555,
+      "step": 7413
+    },
+    {
+      "epoch": 1.3225403621443226,
+      "grad_norm": 0.4850262999534607,
+      "learning_rate": 0.00012861720699492435,
+      "loss": 0.5776,
+      "step": 7414
+    },
+    {
+      "epoch": 1.3227187583623228,
+      "grad_norm": 0.5016975998878479,
+      "learning_rate": 0.00012855596196079873,
+      "loss": 0.7419,
+      "step": 7415
+    },
+    {
+      "epoch": 1.322897154580323,
+      "grad_norm": 0.49091964960098267,
+      "learning_rate": 0.0001284947264648355,
+      "loss": 0.7162,
+      "step": 7416
+    },
+    {
+      "epoch": 1.3230755507983232,
+      "grad_norm": 0.4609431326389313,
+      "learning_rate": 0.00012843350051184425,
+      "loss": 0.5136,
+      "step": 7417
+    },
+    {
+      "epoch": 1.3232539470163234,
+      "grad_norm": 0.4845166802406311,
+      "learning_rate": 0.00012837228410663348,
+      "loss": 0.5677,
+      "step": 7418
+    },
+    {
+      "epoch": 1.3234323432343233,
+      "grad_norm": 0.5238844156265259,
+      "learning_rate": 0.00012831107725401125,
+      "loss": 0.6283,
+      "step": 7419
+    },
+    {
+      "epoch": 1.3236107394523235,
+      "grad_norm": 0.5406588912010193,
+      "learning_rate": 0.00012824987995878456,
+      "loss": 0.5869,
+      "step": 7420
+    },
+    {
+      "epoch": 1.3237891356703237,
+      "grad_norm": 0.526140570640564,
+      "learning_rate": 0.0001281886922257599,
+      "loss": 0.7051,
+      "step": 7421
+    },
+    {
+      "epoch": 1.323967531888324,
+      "grad_norm": 0.4669652581214905,
+      "learning_rate": 0.00012812751405974306,
+      "loss": 0.5571,
+      "step": 7422
+    },
+    {
+      "epoch": 1.3241459281063241,
+      "grad_norm": 0.47372832894325256,
+      "learning_rate": 0.0001280663454655387,
+      "loss": 0.5887,
+      "step": 7423
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 0.4857545793056488,
+      "learning_rate": 0.00012800518644795117,
+      "loss": 0.6542,
+      "step": 7424
+    },
+    {
+      "epoch": 1.3245027205423245,
+      "grad_norm": 0.5294866561889648,
+      "learning_rate": 0.0001279440370117838,
+      "loss": 0.6887,
+      "step": 7425
+    },
+    {
+      "epoch": 1.3246811167603247,
+      "grad_norm": 0.5368157029151917,
+      "learning_rate": 0.00012788289716183918,
+      "loss": 0.6384,
+      "step": 7426
+    },
+    {
+      "epoch": 1.324859512978325,
+      "grad_norm": 0.4165150225162506,
+      "learning_rate": 0.00012782176690291936,
+      "loss": 0.4424,
+      "step": 7427
+    },
+    {
+      "epoch": 1.325037909196325,
+      "grad_norm": 0.5419753789901733,
+      "learning_rate": 0.00012776064623982525,
+      "loss": 0.65,
+      "step": 7428
+    },
+    {
+      "epoch": 1.3252163054143251,
+      "grad_norm": 0.5392053127288818,
+      "learning_rate": 0.0001276995351773575,
+      "loss": 0.6592,
+      "step": 7429
+    },
+    {
+      "epoch": 1.3253947016323253,
+      "grad_norm": 0.49515146017074585,
+      "learning_rate": 0.00012763843372031554,
+      "loss": 0.6608,
+      "step": 7430
+    },
+    {
+      "epoch": 1.3255730978503255,
+      "grad_norm": 0.4864502549171448,
+      "learning_rate": 0.00012757734187349843,
+      "loss": 0.5878,
+      "step": 7431
+    },
+    {
+      "epoch": 1.3257514940683257,
+      "grad_norm": 0.5112424492835999,
+      "learning_rate": 0.0001275162596417041,
+      "loss": 0.4266,
+      "step": 7432
+    },
+    {
+      "epoch": 1.325929890286326,
+      "grad_norm": 0.5315030813217163,
+      "learning_rate": 0.00012745518702973014,
+      "loss": 0.7017,
+      "step": 7433
+    },
+    {
+      "epoch": 1.326108286504326,
+      "grad_norm": 0.5505911111831665,
+      "learning_rate": 0.00012739412404237305,
+      "loss": 0.7733,
+      "step": 7434
+    },
+    {
+      "epoch": 1.3262866827223263,
+      "grad_norm": 0.5162804126739502,
+      "learning_rate": 0.00012733307068442862,
+      "loss": 0.6944,
+      "step": 7435
+    },
+    {
+      "epoch": 1.3264650789403265,
+      "grad_norm": 0.4766071140766144,
+      "learning_rate": 0.000127272026960692,
+      "loss": 0.573,
+      "step": 7436
+    },
+    {
+      "epoch": 1.3266434751583267,
+      "grad_norm": 0.4825485944747925,
+      "learning_rate": 0.00012721099287595766,
+      "loss": 0.6408,
+      "step": 7437
+    },
+    {
+      "epoch": 1.326821871376327,
+      "grad_norm": 0.40906190872192383,
+      "learning_rate": 0.00012714996843501904,
+      "loss": 0.507,
+      "step": 7438
+    },
+    {
+      "epoch": 1.327000267594327,
+      "grad_norm": 0.5242511630058289,
+      "learning_rate": 0.0001270889536426691,
+      "loss": 0.7517,
+      "step": 7439
+    },
+    {
+      "epoch": 1.3271786638123273,
+      "grad_norm": 0.5269128084182739,
+      "learning_rate": 0.00012702794850369975,
+      "loss": 0.7014,
+      "step": 7440
+    },
+    {
+      "epoch": 1.3273570600303275,
+      "grad_norm": 0.5377137064933777,
+      "learning_rate": 0.00012696695302290251,
+      "loss": 0.7327,
+      "step": 7441
+    },
+    {
+      "epoch": 1.3275354562483275,
+      "grad_norm": 0.4645613729953766,
+      "learning_rate": 0.00012690596720506776,
+      "loss": 0.5923,
+      "step": 7442
+    },
+    {
+      "epoch": 1.3277138524663277,
+      "grad_norm": 0.5495232343673706,
+      "learning_rate": 0.00012684499105498543,
+      "loss": 0.7983,
+      "step": 7443
+    },
+    {
+      "epoch": 1.3278922486843279,
+      "grad_norm": 0.5172943472862244,
+      "learning_rate": 0.00012678402457744442,
+      "loss": 0.5517,
+      "step": 7444
+    },
+    {
+      "epoch": 1.328070644902328,
+      "grad_norm": 0.44186466932296753,
+      "learning_rate": 0.0001267230677772332,
+      "loss": 0.4636,
+      "step": 7445
+    },
+    {
+      "epoch": 1.3282490411203283,
+      "grad_norm": 0.4601621925830841,
+      "learning_rate": 0.00012666212065913922,
+      "loss": 0.5381,
+      "step": 7446
+    },
+    {
+      "epoch": 1.3284274373383285,
+      "grad_norm": 0.540649950504303,
+      "learning_rate": 0.00012660118322794907,
+      "loss": 0.584,
+      "step": 7447
+    },
+    {
+      "epoch": 1.3286058335563287,
+      "grad_norm": 0.45804736018180847,
+      "learning_rate": 0.000126540255488449,
+      "loss": 0.5367,
+      "step": 7448
+    },
+    {
+      "epoch": 1.3287842297743289,
+      "grad_norm": 0.46976906061172485,
+      "learning_rate": 0.000126479337445424,
+      "loss": 0.5498,
+      "step": 7449
+    },
+    {
+      "epoch": 1.3289626259923288,
+      "grad_norm": 0.42822468280792236,
+      "learning_rate": 0.0001264184291036588,
+      "loss": 0.519,
+      "step": 7450
+    },
+    {
+      "epoch": 1.329141022210329,
+      "grad_norm": 0.5430753827095032,
+      "learning_rate": 0.00012635753046793692,
+      "loss": 0.7222,
+      "step": 7451
+    },
+    {
+      "epoch": 1.3293194184283292,
+      "grad_norm": 0.4692500829696655,
+      "learning_rate": 0.00012629664154304137,
+      "loss": 0.4665,
+      "step": 7452
+    },
+    {
+      "epoch": 1.3294978146463294,
+      "grad_norm": 0.45048218965530396,
+      "learning_rate": 0.00012623576233375449,
+      "loss": 0.4468,
+      "step": 7453
+    },
+    {
+      "epoch": 1.3296762108643296,
+      "grad_norm": 0.5404579043388367,
+      "learning_rate": 0.00012617489284485746,
+      "loss": 0.6637,
+      "step": 7454
+    },
+    {
+      "epoch": 1.3298546070823298,
+      "grad_norm": 0.5273029208183289,
+      "learning_rate": 0.00012611403308113113,
+      "loss": 0.7473,
+      "step": 7455
+    },
+    {
+      "epoch": 1.33003300330033,
+      "grad_norm": 0.4636683762073517,
+      "learning_rate": 0.00012605318304735524,
+      "loss": 0.4647,
+      "step": 7456
+    },
+    {
+      "epoch": 1.3302113995183302,
+      "grad_norm": 0.5994560718536377,
+      "learning_rate": 0.00012599234274830913,
+      "loss": 0.7458,
+      "step": 7457
+    },
+    {
+      "epoch": 1.3303897957363304,
+      "grad_norm": 0.4628642201423645,
+      "learning_rate": 0.00012593151218877105,
+      "loss": 0.523,
+      "step": 7458
+    },
+    {
+      "epoch": 1.3305681919543306,
+      "grad_norm": 0.5867023468017578,
+      "learning_rate": 0.00012587069137351853,
+      "loss": 0.8639,
+      "step": 7459
+    },
+    {
+      "epoch": 1.3307465881723308,
+      "grad_norm": 0.5695323944091797,
+      "learning_rate": 0.00012580988030732858,
+      "loss": 0.7051,
+      "step": 7460
+    },
+    {
+      "epoch": 1.330924984390331,
+      "grad_norm": 0.5253010392189026,
+      "learning_rate": 0.00012574907899497707,
+      "loss": 0.7779,
+      "step": 7461
+    },
+    {
+      "epoch": 1.3311033806083312,
+      "grad_norm": 0.4339081346988678,
+      "learning_rate": 0.00012568828744123956,
+      "loss": 0.4361,
+      "step": 7462
+    },
+    {
+      "epoch": 1.3312817768263314,
+      "grad_norm": 0.4968469738960266,
+      "learning_rate": 0.0001256275056508903,
+      "loss": 0.6992,
+      "step": 7463
+    },
+    {
+      "epoch": 1.3314601730443314,
+      "grad_norm": 0.4775252044200897,
+      "learning_rate": 0.00012556673362870338,
+      "loss": 0.582,
+      "step": 7464
+    },
+    {
+      "epoch": 1.3316385692623316,
+      "grad_norm": 0.482811838388443,
+      "learning_rate": 0.00012550597137945152,
+      "loss": 0.6071,
+      "step": 7465
+    },
+    {
+      "epoch": 1.3318169654803318,
+      "grad_norm": 0.4188464879989624,
+      "learning_rate": 0.00012544521890790712,
+      "loss": 0.4437,
+      "step": 7466
+    },
+    {
+      "epoch": 1.331995361698332,
+      "grad_norm": 0.49079805612564087,
+      "learning_rate": 0.0001253844762188417,
+      "loss": 0.5477,
+      "step": 7467
+    },
+    {
+      "epoch": 1.3321737579163322,
+      "grad_norm": 0.4884779751300812,
+      "learning_rate": 0.00012532374331702584,
+      "loss": 0.5643,
+      "step": 7468
+    },
+    {
+      "epoch": 1.3323521541343324,
+      "grad_norm": 0.9924476742744446,
+      "learning_rate": 0.00012526302020722958,
+      "loss": 0.5498,
+      "step": 7469
+    },
+    {
+      "epoch": 1.3325305503523326,
+      "grad_norm": 0.4724234640598297,
+      "learning_rate": 0.00012520230689422196,
+      "loss": 0.5189,
+      "step": 7470
+    },
+    {
+      "epoch": 1.3327089465703328,
+      "grad_norm": 0.4852631390094757,
+      "learning_rate": 0.00012514160338277154,
+      "loss": 0.515,
+      "step": 7471
+    },
+    {
+      "epoch": 1.3328873427883328,
+      "grad_norm": 0.4969404637813568,
+      "learning_rate": 0.00012508090967764586,
+      "loss": 0.6241,
+      "step": 7472
+    },
+    {
+      "epoch": 1.333065739006333,
+      "grad_norm": 0.47249653935432434,
+      "learning_rate": 0.00012502022578361166,
+      "loss": 0.6354,
+      "step": 7473
+    },
+    {
+      "epoch": 1.3332441352243332,
+      "grad_norm": 0.4671791195869446,
+      "learning_rate": 0.00012495955170543528,
+      "loss": 0.5399,
+      "step": 7474
+    },
+    {
+      "epoch": 1.3334225314423334,
+      "grad_norm": 0.45596200227737427,
+      "learning_rate": 0.00012489888744788178,
+      "loss": 0.4377,
+      "step": 7475
+    },
+    {
+      "epoch": 1.3336009276603336,
+      "grad_norm": 0.48205843567848206,
+      "learning_rate": 0.00012483823301571593,
+      "loss": 0.6145,
+      "step": 7476
+    },
+    {
+      "epoch": 1.3337793238783338,
+      "grad_norm": 0.6052742004394531,
+      "learning_rate": 0.0001247775884137013,
+      "loss": 0.7105,
+      "step": 7477
+    },
+    {
+      "epoch": 1.333957720096334,
+      "grad_norm": 0.4180543124675751,
+      "learning_rate": 0.00012471695364660106,
+      "loss": 0.6102,
+      "step": 7478
+    },
+    {
+      "epoch": 1.3341361163143342,
+      "grad_norm": 0.5577853322029114,
+      "learning_rate": 0.0001246563287191774,
+      "loss": 0.5936,
+      "step": 7479
+    },
+    {
+      "epoch": 1.3343145125323344,
+      "grad_norm": 0.5047107934951782,
+      "learning_rate": 0.00012459571363619167,
+      "loss": 0.5753,
+      "step": 7480
+    },
+    {
+      "epoch": 1.3344929087503345,
+      "grad_norm": 0.4621174931526184,
+      "learning_rate": 0.00012453510840240457,
+      "loss": 0.5743,
+      "step": 7481
+    },
+    {
+      "epoch": 1.3346713049683347,
+      "grad_norm": 0.456528902053833,
+      "learning_rate": 0.00012447451302257607,
+      "loss": 0.5022,
+      "step": 7482
+    },
+    {
+      "epoch": 1.334849701186335,
+      "grad_norm": 0.43456900119781494,
+      "learning_rate": 0.00012441392750146542,
+      "loss": 0.4349,
+      "step": 7483
+    },
+    {
+      "epoch": 1.3350280974043351,
+      "grad_norm": 0.46913912892341614,
+      "learning_rate": 0.00012435335184383085,
+      "loss": 0.709,
+      "step": 7484
+    },
+    {
+      "epoch": 1.3352064936223353,
+      "grad_norm": 0.42998531460762024,
+      "learning_rate": 0.00012429278605442988,
+      "loss": 0.5149,
+      "step": 7485
+    },
+    {
+      "epoch": 1.3353848898403353,
+      "grad_norm": 0.5419842600822449,
+      "learning_rate": 0.00012423223013801945,
+      "loss": 0.7627,
+      "step": 7486
+    },
+    {
+      "epoch": 1.3355632860583355,
+      "grad_norm": 0.4358643889427185,
+      "learning_rate": 0.00012417168409935547,
+      "loss": 0.4494,
+      "step": 7487
+    },
+    {
+      "epoch": 1.3357416822763357,
+      "grad_norm": 0.6511504650115967,
+      "learning_rate": 0.00012411114794319336,
+      "loss": 0.7345,
+      "step": 7488
+    },
+    {
+      "epoch": 1.335920078494336,
+      "grad_norm": 0.40742728114128113,
+      "learning_rate": 0.00012405062167428744,
+      "loss": 0.4618,
+      "step": 7489
+    },
+    {
+      "epoch": 1.3360984747123361,
+      "grad_norm": 0.4425670802593231,
+      "learning_rate": 0.00012399010529739158,
+      "loss": 0.5878,
+      "step": 7490
+    },
+    {
+      "epoch": 1.3362768709303363,
+      "grad_norm": 0.4261229634284973,
+      "learning_rate": 0.00012392959881725853,
+      "loss": 0.4344,
+      "step": 7491
+    },
+    {
+      "epoch": 1.3364552671483365,
+      "grad_norm": 0.5334789156913757,
+      "learning_rate": 0.00012386910223864062,
+      "loss": 0.6877,
+      "step": 7492
+    },
+    {
+      "epoch": 1.3366336633663367,
+      "grad_norm": 0.4170753061771393,
+      "learning_rate": 0.00012380861556628915,
+      "loss": 0.4582,
+      "step": 7493
+    },
+    {
+      "epoch": 1.3368120595843367,
+      "grad_norm": 0.4649193286895752,
+      "learning_rate": 0.0001237481388049546,
+      "loss": 0.5534,
+      "step": 7494
+    },
+    {
+      "epoch": 1.3369904558023369,
+      "grad_norm": 0.4809529483318329,
+      "learning_rate": 0.00012368767195938701,
+      "loss": 0.4642,
+      "step": 7495
+    },
+    {
+      "epoch": 1.337168852020337,
+      "grad_norm": 0.467174232006073,
+      "learning_rate": 0.00012362721503433521,
+      "loss": 0.5001,
+      "step": 7496
+    },
+    {
+      "epoch": 1.3373472482383373,
+      "grad_norm": 0.5081356167793274,
+      "learning_rate": 0.00012356676803454758,
+      "loss": 0.6698,
+      "step": 7497
+    },
+    {
+      "epoch": 1.3375256444563375,
+      "grad_norm": 0.5005205273628235,
+      "learning_rate": 0.00012350633096477165,
+      "loss": 0.6773,
+      "step": 7498
+    },
+    {
+      "epoch": 1.3377040406743377,
+      "grad_norm": 0.5624291896820068,
+      "learning_rate": 0.00012344590382975395,
+      "loss": 0.7196,
+      "step": 7499
+    },
+    {
+      "epoch": 1.3378824368923379,
+      "grad_norm": 0.5750653147697449,
+      "learning_rate": 0.00012338548663424063,
+      "loss": 0.7924,
+      "step": 7500
+    },
+    {
+      "epoch": 1.338060833110338,
+      "grad_norm": 0.47942960262298584,
+      "learning_rate": 0.00012332507938297657,
+      "loss": 0.5147,
+      "step": 7501
+    },
+    {
+      "epoch": 1.3382392293283383,
+      "grad_norm": 0.4365304112434387,
+      "learning_rate": 0.0001232646820807064,
+      "loss": 0.5417,
+      "step": 7502
+    },
+    {
+      "epoch": 1.3384176255463385,
+      "grad_norm": 0.4866253435611725,
+      "learning_rate": 0.0001232042947321734,
+      "loss": 0.5912,
+      "step": 7503
+    },
+    {
+      "epoch": 1.3385960217643387,
+      "grad_norm": 0.42192235589027405,
+      "learning_rate": 0.00012314391734212068,
+      "loss": 0.4446,
+      "step": 7504
+    },
+    {
+      "epoch": 1.3387744179823389,
+      "grad_norm": 0.45548638701438904,
+      "learning_rate": 0.00012308354991529008,
+      "loss": 0.5527,
+      "step": 7505
+    },
+    {
+      "epoch": 1.338952814200339,
+      "grad_norm": 0.4403938949108124,
+      "learning_rate": 0.00012302319245642278,
+      "loss": 0.4453,
+      "step": 7506
+    },
+    {
+      "epoch": 1.3391312104183393,
+      "grad_norm": 0.4472365975379944,
+      "learning_rate": 0.00012296284497025938,
+      "loss": 0.485,
+      "step": 7507
+    },
+    {
+      "epoch": 1.3393096066363392,
+      "grad_norm": 0.49241968989372253,
+      "learning_rate": 0.00012290250746153935,
+      "loss": 0.674,
+      "step": 7508
+    },
+    {
+      "epoch": 1.3394880028543394,
+      "grad_norm": 0.5484983921051025,
+      "learning_rate": 0.0001228421799350018,
+      "loss": 0.6603,
+      "step": 7509
+    },
+    {
+      "epoch": 1.3396663990723396,
+      "grad_norm": 0.666890025138855,
+      "learning_rate": 0.00012278186239538463,
+      "loss": 0.5482,
+      "step": 7510
+    },
+    {
+      "epoch": 1.3398447952903398,
+      "grad_norm": 0.45667779445648193,
+      "learning_rate": 0.00012272155484742534,
+      "loss": 0.5867,
+      "step": 7511
+    },
+    {
+      "epoch": 1.34002319150834,
+      "grad_norm": 0.4096013009548187,
+      "learning_rate": 0.00012266125729586025,
+      "loss": 0.4004,
+      "step": 7512
+    },
+    {
+      "epoch": 1.3402015877263402,
+      "grad_norm": 0.4495585262775421,
+      "learning_rate": 0.00012260096974542524,
+      "loss": 0.4578,
+      "step": 7513
+    },
+    {
+      "epoch": 1.3403799839443404,
+      "grad_norm": 0.49030259251594543,
+      "learning_rate": 0.0001225406922008553,
+      "loss": 0.7084,
+      "step": 7514
+    },
+    {
+      "epoch": 1.3405583801623406,
+      "grad_norm": 0.49130570888519287,
+      "learning_rate": 0.0001224804246668845,
+      "loss": 0.6026,
+      "step": 7515
+    },
+    {
+      "epoch": 1.3407367763803406,
+      "grad_norm": 0.4424048364162445,
+      "learning_rate": 0.00012242016714824632,
+      "loss": 0.4755,
+      "step": 7516
+    },
+    {
+      "epoch": 1.3409151725983408,
+      "grad_norm": 0.5130663514137268,
+      "learning_rate": 0.00012235991964967325,
+      "loss": 0.6028,
+      "step": 7517
+    },
+    {
+      "epoch": 1.341093568816341,
+      "grad_norm": 0.4313303232192993,
+      "learning_rate": 0.0001222996821758972,
+      "loss": 0.4998,
+      "step": 7518
+    },
+    {
+      "epoch": 1.3412719650343412,
+      "grad_norm": 0.5465921759605408,
+      "learning_rate": 0.0001222394547316492,
+      "loss": 0.8042,
+      "step": 7519
+    },
+    {
+      "epoch": 1.3414503612523414,
+      "grad_norm": 0.5101335048675537,
+      "learning_rate": 0.00012217923732165938,
+      "loss": 0.6261,
+      "step": 7520
+    },
+    {
+      "epoch": 1.3416287574703416,
+      "grad_norm": 0.5270312428474426,
+      "learning_rate": 0.00012211902995065728,
+      "loss": 0.6911,
+      "step": 7521
+    },
+    {
+      "epoch": 1.3418071536883418,
+      "grad_norm": 0.5207296013832092,
+      "learning_rate": 0.0001220588326233715,
+      "loss": 0.7429,
+      "step": 7522
+    },
+    {
+      "epoch": 1.341985549906342,
+      "grad_norm": 0.4834536015987396,
+      "learning_rate": 0.00012199864534453003,
+      "loss": 0.594,
+      "step": 7523
+    },
+    {
+      "epoch": 1.3421639461243422,
+      "grad_norm": 0.46801358461380005,
+      "learning_rate": 0.00012193846811885978,
+      "loss": 0.5481,
+      "step": 7524
+    },
+    {
+      "epoch": 1.3423423423423424,
+      "grad_norm": 0.4802463948726654,
+      "learning_rate": 0.00012187830095108721,
+      "loss": 0.7895,
+      "step": 7525
+    },
+    {
+      "epoch": 1.3425207385603426,
+      "grad_norm": 0.49674612283706665,
+      "learning_rate": 0.00012181814384593776,
+      "loss": 0.693,
+      "step": 7526
+    },
+    {
+      "epoch": 1.3426991347783428,
+      "grad_norm": 0.44266417622566223,
+      "learning_rate": 0.00012175799680813593,
+      "loss": 0.4989,
+      "step": 7527
+    },
+    {
+      "epoch": 1.342877530996343,
+      "grad_norm": 0.5586349964141846,
+      "learning_rate": 0.00012169785984240605,
+      "loss": 0.556,
+      "step": 7528
+    },
+    {
+      "epoch": 1.3430559272143432,
+      "grad_norm": 0.485568106174469,
+      "learning_rate": 0.00012163773295347095,
+      "loss": 0.6456,
+      "step": 7529
+    },
+    {
+      "epoch": 1.3432343234323432,
+      "grad_norm": 0.47074824571609497,
+      "learning_rate": 0.00012157761614605314,
+      "loss": 0.5116,
+      "step": 7530
+    },
+    {
+      "epoch": 1.3434127196503434,
+      "grad_norm": 0.48266589641571045,
+      "learning_rate": 0.0001215175094248741,
+      "loss": 0.5583,
+      "step": 7531
+    },
+    {
+      "epoch": 1.3435911158683436,
+      "grad_norm": 0.5991159677505493,
+      "learning_rate": 0.0001214574127946545,
+      "loss": 0.8355,
+      "step": 7532
+    },
+    {
+      "epoch": 1.3437695120863438,
+      "grad_norm": 0.47576338052749634,
+      "learning_rate": 0.00012139732626011446,
+      "loss": 0.6401,
+      "step": 7533
+    },
+    {
+      "epoch": 1.343947908304344,
+      "grad_norm": 0.5145847201347351,
+      "learning_rate": 0.000121337249825973,
+      "loss": 0.6782,
+      "step": 7534
+    },
+    {
+      "epoch": 1.3441263045223442,
+      "grad_norm": 0.6812740564346313,
+      "learning_rate": 0.00012127718349694863,
+      "loss": 0.5941,
+      "step": 7535
+    },
+    {
+      "epoch": 1.3443047007403444,
+      "grad_norm": 0.5664840936660767,
+      "learning_rate": 0.00012121712727775882,
+      "loss": 0.755,
+      "step": 7536
+    },
+    {
+      "epoch": 1.3444830969583446,
+      "grad_norm": 0.43093249201774597,
+      "learning_rate": 0.00012115708117312049,
+      "loss": 0.4926,
+      "step": 7537
+    },
+    {
+      "epoch": 1.3446614931763445,
+      "grad_norm": 0.4539382755756378,
+      "learning_rate": 0.00012109704518774956,
+      "loss": 0.5437,
+      "step": 7538
+    },
+    {
+      "epoch": 1.3448398893943447,
+      "grad_norm": 0.46809741854667664,
+      "learning_rate": 0.00012103701932636114,
+      "loss": 0.4688,
+      "step": 7539
+    },
+    {
+      "epoch": 1.345018285612345,
+      "grad_norm": 0.479373037815094,
+      "learning_rate": 0.00012097700359366981,
+      "loss": 0.6389,
+      "step": 7540
+    },
+    {
+      "epoch": 1.3451966818303451,
+      "grad_norm": 0.3928679823875427,
+      "learning_rate": 0.00012091699799438899,
+      "loss": 0.36,
+      "step": 7541
+    },
+    {
+      "epoch": 1.3453750780483453,
+      "grad_norm": 0.4709312915802002,
+      "learning_rate": 0.00012085700253323173,
+      "loss": 0.5524,
+      "step": 7542
+    },
+    {
+      "epoch": 1.3455534742663455,
+      "grad_norm": 0.4242085814476013,
+      "learning_rate": 0.0001207970172149098,
+      "loss": 0.4975,
+      "step": 7543
+    },
+    {
+      "epoch": 1.3457318704843457,
+      "grad_norm": 0.4750572741031647,
+      "learning_rate": 0.00012073704204413452,
+      "loss": 0.6045,
+      "step": 7544
+    },
+    {
+      "epoch": 1.345910266702346,
+      "grad_norm": 0.5419875383377075,
+      "learning_rate": 0.00012067707702561645,
+      "loss": 0.7744,
+      "step": 7545
+    },
+    {
+      "epoch": 1.3460886629203461,
+      "grad_norm": 0.5909712910652161,
+      "learning_rate": 0.00012061712216406501,
+      "loss": 0.8503,
+      "step": 7546
+    },
+    {
+      "epoch": 1.3462670591383463,
+      "grad_norm": 0.4636007249355316,
+      "learning_rate": 0.00012055717746418918,
+      "loss": 0.5914,
+      "step": 7547
+    },
+    {
+      "epoch": 1.3464454553563465,
+      "grad_norm": 0.5388915538787842,
+      "learning_rate": 0.00012049724293069686,
+      "loss": 0.7088,
+      "step": 7548
+    },
+    {
+      "epoch": 1.3466238515743467,
+      "grad_norm": 0.5035228729248047,
+      "learning_rate": 0.00012043731856829543,
+      "loss": 0.6403,
+      "step": 7549
+    },
+    {
+      "epoch": 1.346802247792347,
+      "grad_norm": 0.4420143961906433,
+      "learning_rate": 0.00012037740438169118,
+      "loss": 0.5323,
+      "step": 7550
+    },
+    {
+      "epoch": 1.3469806440103471,
+      "grad_norm": 0.5462261438369751,
+      "learning_rate": 0.00012031750037558986,
+      "loss": 0.8891,
+      "step": 7551
+    },
+    {
+      "epoch": 1.347159040228347,
+      "grad_norm": 0.45456668734550476,
+      "learning_rate": 0.00012025760655469628,
+      "loss": 0.4707,
+      "step": 7552
+    },
+    {
+      "epoch": 1.3473374364463473,
+      "grad_norm": 0.4722398519515991,
+      "learning_rate": 0.00012019772292371437,
+      "loss": 0.5228,
+      "step": 7553
+    },
+    {
+      "epoch": 1.3475158326643475,
+      "grad_norm": 0.53815096616745,
+      "learning_rate": 0.0001201378494873475,
+      "loss": 0.804,
+      "step": 7554
+    },
+    {
+      "epoch": 1.3476942288823477,
+      "grad_norm": 0.48279356956481934,
+      "learning_rate": 0.00012007798625029798,
+      "loss": 0.7179,
+      "step": 7555
+    },
+    {
+      "epoch": 1.3478726251003479,
+      "grad_norm": 1.062436819076538,
+      "learning_rate": 0.0001200181332172676,
+      "loss": 0.4516,
+      "step": 7556
+    },
+    {
+      "epoch": 1.348051021318348,
+      "grad_norm": 0.44580498337745667,
+      "learning_rate": 0.000119958290392957,
+      "loss": 0.5279,
+      "step": 7557
+    },
+    {
+      "epoch": 1.3482294175363483,
+      "grad_norm": 0.48104509711265564,
+      "learning_rate": 0.00011989845778206629,
+      "loss": 0.6148,
+      "step": 7558
+    },
+    {
+      "epoch": 1.3484078137543485,
+      "grad_norm": 0.5047935843467712,
+      "learning_rate": 0.00011983863538929485,
+      "loss": 0.6991,
+      "step": 7559
+    },
+    {
+      "epoch": 1.3485862099723485,
+      "grad_norm": 0.5424177646636963,
+      "learning_rate": 0.00011977882321934086,
+      "loss": 0.6587,
+      "step": 7560
+    },
+    {
+      "epoch": 1.3487646061903487,
+      "grad_norm": 0.5127599239349365,
+      "learning_rate": 0.00011971902127690215,
+      "loss": 0.7465,
+      "step": 7561
+    },
+    {
+      "epoch": 1.3489430024083489,
+      "grad_norm": 0.4546005427837372,
+      "learning_rate": 0.00011965922956667535,
+      "loss": 0.5514,
+      "step": 7562
+    },
+    {
+      "epoch": 1.349121398626349,
+      "grad_norm": 0.49438631534576416,
+      "learning_rate": 0.00011959944809335668,
+      "loss": 0.6156,
+      "step": 7563
+    },
+    {
+      "epoch": 1.3492997948443493,
+      "grad_norm": 0.5039214491844177,
+      "learning_rate": 0.00011953967686164125,
+      "loss": 0.532,
+      "step": 7564
+    },
+    {
+      "epoch": 1.3494781910623495,
+      "grad_norm": 0.5272899270057678,
+      "learning_rate": 0.00011947991587622334,
+      "loss": 0.7789,
+      "step": 7565
+    },
+    {
+      "epoch": 1.3496565872803497,
+      "grad_norm": 0.5394687652587891,
+      "learning_rate": 0.00011942016514179677,
+      "loss": 0.7146,
+      "step": 7566
+    },
+    {
+      "epoch": 1.3498349834983498,
+      "grad_norm": 0.4609784185886383,
+      "learning_rate": 0.00011936042466305413,
+      "loss": 0.555,
+      "step": 7567
+    },
+    {
+      "epoch": 1.35001337971635,
+      "grad_norm": 0.48477205634117126,
+      "learning_rate": 0.00011930069444468764,
+      "loss": 0.5976,
+      "step": 7568
+    },
+    {
+      "epoch": 1.3501917759343502,
+      "grad_norm": 0.47605961561203003,
+      "learning_rate": 0.00011924097449138824,
+      "loss": 0.5233,
+      "step": 7569
+    },
+    {
+      "epoch": 1.3503701721523504,
+      "grad_norm": 0.46146827936172485,
+      "learning_rate": 0.00011918126480784655,
+      "loss": 0.4737,
+      "step": 7570
+    },
+    {
+      "epoch": 1.3505485683703506,
+      "grad_norm": 0.48921510577201843,
+      "learning_rate": 0.0001191215653987519,
+      "loss": 0.6749,
+      "step": 7571
+    },
+    {
+      "epoch": 1.3507269645883508,
+      "grad_norm": 0.49213096499443054,
+      "learning_rate": 0.0001190618762687933,
+      "loss": 0.6015,
+      "step": 7572
+    },
+    {
+      "epoch": 1.350905360806351,
+      "grad_norm": 0.5104438662528992,
+      "learning_rate": 0.0001190021974226585,
+      "loss": 0.5599,
+      "step": 7573
+    },
+    {
+      "epoch": 1.351083757024351,
+      "grad_norm": 0.55939120054245,
+      "learning_rate": 0.00011894252886503476,
+      "loss": 0.6426,
+      "step": 7574
+    },
+    {
+      "epoch": 1.3512621532423512,
+      "grad_norm": 0.5199212431907654,
+      "learning_rate": 0.00011888287060060845,
+      "loss": 0.7041,
+      "step": 7575
+    },
+    {
+      "epoch": 1.3514405494603514,
+      "grad_norm": 0.48316359519958496,
+      "learning_rate": 0.000118823222634065,
+      "loss": 0.5074,
+      "step": 7576
+    },
+    {
+      "epoch": 1.3516189456783516,
+      "grad_norm": 0.630157470703125,
+      "learning_rate": 0.0001187635849700893,
+      "loss": 0.5867,
+      "step": 7577
+    },
+    {
+      "epoch": 1.3517973418963518,
+      "grad_norm": 0.48873627185821533,
+      "learning_rate": 0.00011870395761336514,
+      "loss": 0.5635,
+      "step": 7578
+    },
+    {
+      "epoch": 1.351975738114352,
+      "grad_norm": 0.4779564142227173,
+      "learning_rate": 0.00011864434056857554,
+      "loss": 0.4933,
+      "step": 7579
+    },
+    {
+      "epoch": 1.3521541343323522,
+      "grad_norm": 0.5071210861206055,
+      "learning_rate": 0.00011858473384040302,
+      "loss": 0.7165,
+      "step": 7580
+    },
+    {
+      "epoch": 1.3523325305503524,
+      "grad_norm": 0.5788468718528748,
+      "learning_rate": 0.00011852513743352885,
+      "loss": 0.8357,
+      "step": 7581
+    },
+    {
+      "epoch": 1.3525109267683524,
+      "grad_norm": 0.4292871356010437,
+      "learning_rate": 0.0001184655513526339,
+      "loss": 0.5256,
+      "step": 7582
+    },
+    {
+      "epoch": 1.3526893229863526,
+      "grad_norm": 0.4763195216655731,
+      "learning_rate": 0.00011840597560239785,
+      "loss": 0.5758,
+      "step": 7583
+    },
+    {
+      "epoch": 1.3528677192043528,
+      "grad_norm": 0.5240211486816406,
+      "learning_rate": 0.00011834641018749994,
+      "loss": 0.5393,
+      "step": 7584
+    },
+    {
+      "epoch": 1.353046115422353,
+      "grad_norm": 0.46273085474967957,
+      "learning_rate": 0.00011828685511261833,
+      "loss": 0.4274,
+      "step": 7585
+    },
+    {
+      "epoch": 1.3532245116403532,
+      "grad_norm": 0.5363104343414307,
+      "learning_rate": 0.00011822731038243035,
+      "loss": 0.6826,
+      "step": 7586
+    },
+    {
+      "epoch": 1.3534029078583534,
+      "grad_norm": 0.46597224473953247,
+      "learning_rate": 0.00011816777600161278,
+      "loss": 0.5414,
+      "step": 7587
+    },
+    {
+      "epoch": 1.3535813040763536,
+      "grad_norm": 0.48101913928985596,
+      "learning_rate": 0.00011810825197484126,
+      "loss": 0.4022,
+      "step": 7588
+    },
+    {
+      "epoch": 1.3537597002943538,
+      "grad_norm": 0.5787967443466187,
+      "learning_rate": 0.00011804873830679089,
+      "loss": 0.7585,
+      "step": 7589
+    },
+    {
+      "epoch": 1.353938096512354,
+      "grad_norm": 0.39563214778900146,
+      "learning_rate": 0.0001179892350021359,
+      "loss": 0.3329,
+      "step": 7590
+    },
+    {
+      "epoch": 1.3541164927303542,
+      "grad_norm": 0.5252716541290283,
+      "learning_rate": 0.00011792974206554949,
+      "loss": 0.6866,
+      "step": 7591
+    },
+    {
+      "epoch": 1.3542948889483544,
+      "grad_norm": 0.5136657953262329,
+      "learning_rate": 0.00011787025950170441,
+      "loss": 0.6072,
+      "step": 7592
+    },
+    {
+      "epoch": 1.3544732851663546,
+      "grad_norm": 0.503588080406189,
+      "learning_rate": 0.0001178107873152722,
+      "loss": 0.6746,
+      "step": 7593
+    },
+    {
+      "epoch": 1.3546516813843548,
+      "grad_norm": 0.5402643084526062,
+      "learning_rate": 0.00011775132551092397,
+      "loss": 0.73,
+      "step": 7594
+    },
+    {
+      "epoch": 1.354830077602355,
+      "grad_norm": 0.4741210639476776,
+      "learning_rate": 0.0001176918740933296,
+      "loss": 0.5221,
+      "step": 7595
+    },
+    {
+      "epoch": 1.355008473820355,
+      "grad_norm": 0.5205968618392944,
+      "learning_rate": 0.00011763243306715862,
+      "loss": 0.639,
+      "step": 7596
+    },
+    {
+      "epoch": 1.3551868700383551,
+      "grad_norm": 0.5402557253837585,
+      "learning_rate": 0.00011757300243707927,
+      "loss": 0.7846,
+      "step": 7597
+    },
+    {
+      "epoch": 1.3553652662563553,
+      "grad_norm": 0.46083173155784607,
+      "learning_rate": 0.00011751358220775943,
+      "loss": 0.5686,
+      "step": 7598
+    },
+    {
+      "epoch": 1.3555436624743555,
+      "grad_norm": 0.5107517242431641,
+      "learning_rate": 0.00011745417238386583,
+      "loss": 0.639,
+      "step": 7599
+    },
+    {
+      "epoch": 1.3557220586923557,
+      "grad_norm": 0.5665692687034607,
+      "learning_rate": 0.0001173947729700644,
+      "loss": 0.7715,
+      "step": 7600
+    },
+    {
+      "epoch": 1.355900454910356,
+      "grad_norm": 0.5142179727554321,
+      "learning_rate": 0.00011733538397102053,
+      "loss": 0.5198,
+      "step": 7601
+    },
+    {
+      "epoch": 1.3560788511283561,
+      "grad_norm": 0.49913138151168823,
+      "learning_rate": 0.00011727600539139841,
+      "loss": 0.6599,
+      "step": 7602
+    },
+    {
+      "epoch": 1.3562572473463563,
+      "grad_norm": 0.4611426293849945,
+      "learning_rate": 0.00011721663723586181,
+      "loss": 0.5418,
+      "step": 7603
+    },
+    {
+      "epoch": 1.3564356435643563,
+      "grad_norm": 0.49066436290740967,
+      "learning_rate": 0.00011715727950907329,
+      "loss": 0.6004,
+      "step": 7604
+    },
+    {
+      "epoch": 1.3566140397823565,
+      "grad_norm": 0.4376862645149231,
+      "learning_rate": 0.00011709793221569486,
+      "loss": 0.5262,
+      "step": 7605
+    },
+    {
+      "epoch": 1.3567924360003567,
+      "grad_norm": 0.46929794549942017,
+      "learning_rate": 0.00011703859536038774,
+      "loss": 0.4824,
+      "step": 7606
+    },
+    {
+      "epoch": 1.356970832218357,
+      "grad_norm": 0.5407469868659973,
+      "learning_rate": 0.00011697926894781205,
+      "loss": 0.692,
+      "step": 7607
+    },
+    {
+      "epoch": 1.357149228436357,
+      "grad_norm": 0.5103482007980347,
+      "learning_rate": 0.00011691995298262739,
+      "loss": 0.6721,
+      "step": 7608
+    },
+    {
+      "epoch": 1.3573276246543573,
+      "grad_norm": 0.532581090927124,
+      "learning_rate": 0.00011686064746949229,
+      "loss": 0.7574,
+      "step": 7609
+    },
+    {
+      "epoch": 1.3575060208723575,
+      "grad_norm": 0.4488193690776825,
+      "learning_rate": 0.00011680135241306472,
+      "loss": 0.5191,
+      "step": 7610
+    },
+    {
+      "epoch": 1.3576844170903577,
+      "grad_norm": 0.4854279160499573,
+      "learning_rate": 0.00011674206781800162,
+      "loss": 0.5963,
+      "step": 7611
+    },
+    {
+      "epoch": 1.357862813308358,
+      "grad_norm": 0.4275936186313629,
+      "learning_rate": 0.00011668279368895907,
+      "loss": 0.4465,
+      "step": 7612
+    },
+    {
+      "epoch": 1.358041209526358,
+      "grad_norm": 0.5791702270507812,
+      "learning_rate": 0.00011662353003059262,
+      "loss": 0.7993,
+      "step": 7613
+    },
+    {
+      "epoch": 1.3582196057443583,
+      "grad_norm": 0.5329555869102478,
+      "learning_rate": 0.00011656427684755666,
+      "loss": 0.6066,
+      "step": 7614
+    },
+    {
+      "epoch": 1.3583980019623585,
+      "grad_norm": 0.4500406086444855,
+      "learning_rate": 0.00011650503414450502,
+      "loss": 0.5459,
+      "step": 7615
+    },
+    {
+      "epoch": 1.3585763981803587,
+      "grad_norm": 0.5136424899101257,
+      "learning_rate": 0.0001164458019260905,
+      "loss": 0.6399,
+      "step": 7616
+    },
+    {
+      "epoch": 1.358754794398359,
+      "grad_norm": 0.49857786297798157,
+      "learning_rate": 0.0001163865801969653,
+      "loss": 0.632,
+      "step": 7617
+    },
+    {
+      "epoch": 1.3589331906163589,
+      "grad_norm": 0.529205858707428,
+      "learning_rate": 0.00011632736896178059,
+      "loss": 0.5917,
+      "step": 7618
+    },
+    {
+      "epoch": 1.359111586834359,
+      "grad_norm": 0.6193312406539917,
+      "learning_rate": 0.00011626816822518662,
+      "loss": 0.693,
+      "step": 7619
+    },
+    {
+      "epoch": 1.3592899830523593,
+      "grad_norm": 0.4899401068687439,
+      "learning_rate": 0.00011620897799183336,
+      "loss": 0.5771,
+      "step": 7620
+    },
+    {
+      "epoch": 1.3594683792703595,
+      "grad_norm": 0.5664353966712952,
+      "learning_rate": 0.0001161497982663693,
+      "loss": 0.7672,
+      "step": 7621
+    },
+    {
+      "epoch": 1.3596467754883597,
+      "grad_norm": 0.4969763159751892,
+      "learning_rate": 0.00011609062905344256,
+      "loss": 0.669,
+      "step": 7622
+    },
+    {
+      "epoch": 1.3598251717063599,
+      "grad_norm": 0.48748770356178284,
+      "learning_rate": 0.0001160314703577002,
+      "loss": 0.6251,
+      "step": 7623
+    },
+    {
+      "epoch": 1.36000356792436,
+      "grad_norm": 0.4871166944503784,
+      "learning_rate": 0.00011597232218378842,
+      "loss": 0.4868,
+      "step": 7624
+    },
+    {
+      "epoch": 1.3601819641423603,
+      "grad_norm": 0.4607223868370056,
+      "learning_rate": 0.00011591318453635286,
+      "loss": 0.617,
+      "step": 7625
+    },
+    {
+      "epoch": 1.3603603603603602,
+      "grad_norm": 0.5145749449729919,
+      "learning_rate": 0.000115854057420038,
+      "loss": 0.8115,
+      "step": 7626
+    },
+    {
+      "epoch": 1.3605387565783604,
+      "grad_norm": 0.4471273124217987,
+      "learning_rate": 0.00011579494083948783,
+      "loss": 0.4159,
+      "step": 7627
+    },
+    {
+      "epoch": 1.3607171527963606,
+      "grad_norm": 0.5210676193237305,
+      "learning_rate": 0.00011573583479934516,
+      "loss": 0.6254,
+      "step": 7628
+    },
+    {
+      "epoch": 1.3608955490143608,
+      "grad_norm": 0.44996219873428345,
+      "learning_rate": 0.00011567673930425232,
+      "loss": 0.4922,
+      "step": 7629
+    },
+    {
+      "epoch": 1.361073945232361,
+      "grad_norm": 0.5416077375411987,
+      "learning_rate": 0.0001156176543588505,
+      "loss": 0.6659,
+      "step": 7630
+    },
+    {
+      "epoch": 1.3612523414503612,
+      "grad_norm": 0.5629211068153381,
+      "learning_rate": 0.00011555857996778038,
+      "loss": 0.6857,
+      "step": 7631
+    },
+    {
+      "epoch": 1.3614307376683614,
+      "grad_norm": 0.552343487739563,
+      "learning_rate": 0.00011549951613568152,
+      "loss": 0.8255,
+      "step": 7632
+    },
+    {
+      "epoch": 1.3616091338863616,
+      "grad_norm": 0.4920686185359955,
+      "learning_rate": 0.0001154404628671927,
+      "loss": 0.5705,
+      "step": 7633
+    },
+    {
+      "epoch": 1.3617875301043618,
+      "grad_norm": 0.514735758304596,
+      "learning_rate": 0.0001153814201669521,
+      "loss": 0.5387,
+      "step": 7634
+    },
+    {
+      "epoch": 1.361965926322362,
+      "grad_norm": 0.6460976600646973,
+      "learning_rate": 0.00011532238803959666,
+      "loss": 0.5714,
+      "step": 7635
+    },
+    {
+      "epoch": 1.3621443225403622,
+      "grad_norm": 0.48432081937789917,
+      "learning_rate": 0.00011526336648976307,
+      "loss": 0.4946,
+      "step": 7636
+    },
+    {
+      "epoch": 1.3623227187583624,
+      "grad_norm": 0.5579152703285217,
+      "learning_rate": 0.00011520435552208672,
+      "loss": 0.6566,
+      "step": 7637
+    },
+    {
+      "epoch": 1.3625011149763626,
+      "grad_norm": 0.5772841572761536,
+      "learning_rate": 0.00011514535514120217,
+      "loss": 0.7587,
+      "step": 7638
+    },
+    {
+      "epoch": 1.3626795111943628,
+      "grad_norm": 0.43424656987190247,
+      "learning_rate": 0.00011508636535174349,
+      "loss": 0.4863,
+      "step": 7639
+    },
+    {
+      "epoch": 1.3628579074123628,
+      "grad_norm": 0.5111271739006042,
+      "learning_rate": 0.00011502738615834351,
+      "loss": 0.6528,
+      "step": 7640
+    },
+    {
+      "epoch": 1.363036303630363,
+      "grad_norm": 0.4701317548751831,
+      "learning_rate": 0.00011496841756563467,
+      "loss": 0.4773,
+      "step": 7641
+    },
+    {
+      "epoch": 1.3632146998483632,
+      "grad_norm": 0.5139216780662537,
+      "learning_rate": 0.00011490945957824808,
+      "loss": 0.647,
+      "step": 7642
+    },
+    {
+      "epoch": 1.3633930960663634,
+      "grad_norm": 0.5112646222114563,
+      "learning_rate": 0.00011485051220081449,
+      "loss": 0.582,
+      "step": 7643
+    },
+    {
+      "epoch": 1.3635714922843636,
+      "grad_norm": 0.5711879730224609,
+      "learning_rate": 0.00011479157543796353,
+      "loss": 0.7522,
+      "step": 7644
+    },
+    {
+      "epoch": 1.3637498885023638,
+      "grad_norm": 0.4305424988269806,
+      "learning_rate": 0.00011473264929432398,
+      "loss": 0.4716,
+      "step": 7645
+    },
+    {
+      "epoch": 1.363928284720364,
+      "grad_norm": 0.5155298113822937,
+      "learning_rate": 0.000114673733774524,
+      "loss": 0.6646,
+      "step": 7646
+    },
+    {
+      "epoch": 1.3641066809383642,
+      "grad_norm": 0.5119175910949707,
+      "learning_rate": 0.00011461482888319064,
+      "loss": 0.6522,
+      "step": 7647
+    },
+    {
+      "epoch": 1.3642850771563642,
+      "grad_norm": 0.5106381177902222,
+      "learning_rate": 0.00011455593462495047,
+      "loss": 0.7938,
+      "step": 7648
+    },
+    {
+      "epoch": 1.3644634733743644,
+      "grad_norm": 0.44368186593055725,
+      "learning_rate": 0.00011449705100442881,
+      "loss": 0.4743,
+      "step": 7649
+    },
+    {
+      "epoch": 1.3646418695923646,
+      "grad_norm": 0.4758698046207428,
+      "learning_rate": 0.00011443817802625044,
+      "loss": 0.5669,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3648202658103648,
+      "grad_norm": 0.41871505975723267,
+      "learning_rate": 0.00011437931569503935,
+      "loss": 0.4859,
+      "step": 7651
+    },
+    {
+      "epoch": 1.364998662028365,
+      "grad_norm": 0.44119468331336975,
+      "learning_rate": 0.00011432046401541835,
+      "loss": 0.57,
+      "step": 7652
+    },
+    {
+      "epoch": 1.3651770582463651,
+      "grad_norm": 0.5434176921844482,
+      "learning_rate": 0.0001142616229920098,
+      "loss": 0.7457,
+      "step": 7653
+    },
+    {
+      "epoch": 1.3653554544643653,
+      "grad_norm": 0.4746476411819458,
+      "learning_rate": 0.00011420279262943487,
+      "loss": 0.5716,
+      "step": 7654
+    },
+    {
+      "epoch": 1.3655338506823655,
+      "grad_norm": 0.4947410523891449,
+      "learning_rate": 0.00011414397293231424,
+      "loss": 0.6332,
+      "step": 7655
+    },
+    {
+      "epoch": 1.3657122469003657,
+      "grad_norm": 0.4677318036556244,
+      "learning_rate": 0.00011408516390526747,
+      "loss": 0.5055,
+      "step": 7656
+    },
+    {
+      "epoch": 1.365890643118366,
+      "grad_norm": 0.535660445690155,
+      "learning_rate": 0.00011402636555291348,
+      "loss": 0.8911,
+      "step": 7657
+    },
+    {
+      "epoch": 1.3660690393363661,
+      "grad_norm": 0.4632878601551056,
+      "learning_rate": 0.00011396757787987025,
+      "loss": 0.5509,
+      "step": 7658
+    },
+    {
+      "epoch": 1.3662474355543663,
+      "grad_norm": 0.5350466966629028,
+      "learning_rate": 0.00011390880089075483,
+      "loss": 0.7065,
+      "step": 7659
+    },
+    {
+      "epoch": 1.3664258317723665,
+      "grad_norm": 0.5175812840461731,
+      "learning_rate": 0.00011385003459018369,
+      "loss": 0.7321,
+      "step": 7660
+    },
+    {
+      "epoch": 1.3666042279903667,
+      "grad_norm": 0.4582364857196808,
+      "learning_rate": 0.00011379127898277217,
+      "loss": 0.5169,
+      "step": 7661
+    },
+    {
+      "epoch": 1.3667826242083667,
+      "grad_norm": 0.45724961161613464,
+      "learning_rate": 0.00011373253407313508,
+      "loss": 0.5547,
+      "step": 7662
+    },
+    {
+      "epoch": 1.366961020426367,
+      "grad_norm": 0.48377981781959534,
+      "learning_rate": 0.00011367379986588603,
+      "loss": 0.4886,
+      "step": 7663
+    },
+    {
+      "epoch": 1.3671394166443671,
+      "grad_norm": 0.4916684627532959,
+      "learning_rate": 0.00011361507636563817,
+      "loss": 0.4752,
+      "step": 7664
+    },
+    {
+      "epoch": 1.3673178128623673,
+      "grad_norm": 0.48477721214294434,
+      "learning_rate": 0.00011355636357700342,
+      "loss": 0.5644,
+      "step": 7665
+    },
+    {
+      "epoch": 1.3674962090803675,
+      "grad_norm": 0.5753141641616821,
+      "learning_rate": 0.00011349766150459314,
+      "loss": 0.7344,
+      "step": 7666
+    },
+    {
+      "epoch": 1.3676746052983677,
+      "grad_norm": 0.48007839918136597,
+      "learning_rate": 0.0001134389701530179,
+      "loss": 0.5605,
+      "step": 7667
+    },
+    {
+      "epoch": 1.367853001516368,
+      "grad_norm": 0.46499067544937134,
+      "learning_rate": 0.00011338028952688709,
+      "loss": 0.5273,
+      "step": 7668
+    },
+    {
+      "epoch": 1.368031397734368,
+      "grad_norm": 0.4519871473312378,
+      "learning_rate": 0.00011332161963080961,
+      "loss": 0.5694,
+      "step": 7669
+    },
+    {
+      "epoch": 1.368209793952368,
+      "grad_norm": 0.4737212061882019,
+      "learning_rate": 0.00011326296046939332,
+      "loss": 0.5597,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3683881901703683,
+      "grad_norm": 0.4954215884208679,
+      "learning_rate": 0.00011320431204724519,
+      "loss": 0.6356,
+      "step": 7671
+    },
+    {
+      "epoch": 1.3685665863883685,
+      "grad_norm": 0.4386094808578491,
+      "learning_rate": 0.00011314567436897161,
+      "loss": 0.6296,
+      "step": 7672
+    },
+    {
+      "epoch": 1.3687449826063687,
+      "grad_norm": 0.7949509620666504,
+      "learning_rate": 0.0001130870474391778,
+      "loss": 0.6376,
+      "step": 7673
+    },
+    {
+      "epoch": 1.3689233788243689,
+      "grad_norm": 0.47607821226119995,
+      "learning_rate": 0.00011302843126246842,
+      "loss": 0.5973,
+      "step": 7674
+    },
+    {
+      "epoch": 1.369101775042369,
+      "grad_norm": 0.4358568489551544,
+      "learning_rate": 0.00011296982584344704,
+      "loss": 0.4679,
+      "step": 7675
+    },
+    {
+      "epoch": 1.3692801712603693,
+      "grad_norm": 0.5030319690704346,
+      "learning_rate": 0.00011291123118671665,
+      "loss": 0.6087,
+      "step": 7676
+    },
+    {
+      "epoch": 1.3694585674783695,
+      "grad_norm": 0.5743786096572876,
+      "learning_rate": 0.00011285264729687908,
+      "loss": 0.8441,
+      "step": 7677
+    },
+    {
+      "epoch": 1.3696369636963697,
+      "grad_norm": 0.45528116822242737,
+      "learning_rate": 0.00011279407417853569,
+      "loss": 0.4794,
+      "step": 7678
+    },
+    {
+      "epoch": 1.3698153599143699,
+      "grad_norm": 0.538870632648468,
+      "learning_rate": 0.00011273551183628664,
+      "loss": 0.6597,
+      "step": 7679
+    },
+    {
+      "epoch": 1.36999375613237,
+      "grad_norm": 0.4873597025871277,
+      "learning_rate": 0.00011267696027473132,
+      "loss": 0.5062,
+      "step": 7680
+    },
+    {
+      "epoch": 1.3701721523503703,
+      "grad_norm": 0.573042094707489,
+      "learning_rate": 0.00011261841949846846,
+      "loss": 0.6462,
+      "step": 7681
+    },
+    {
+      "epoch": 1.3703505485683705,
+      "grad_norm": 0.4371779263019562,
+      "learning_rate": 0.00011255988951209589,
+      "loss": 0.4188,
+      "step": 7682
+    },
+    {
+      "epoch": 1.3705289447863707,
+      "grad_norm": 0.5239207744598389,
+      "learning_rate": 0.00011250137032021038,
+      "loss": 0.6176,
+      "step": 7683
+    },
+    {
+      "epoch": 1.3707073410043706,
+      "grad_norm": 0.4236181378364563,
+      "learning_rate": 0.00011244286192740815,
+      "loss": 0.4331,
+      "step": 7684
+    },
+    {
+      "epoch": 1.3708857372223708,
+      "grad_norm": 0.5079860091209412,
+      "learning_rate": 0.00011238436433828427,
+      "loss": 0.5725,
+      "step": 7685
+    },
+    {
+      "epoch": 1.371064133440371,
+      "grad_norm": 0.48492246866226196,
+      "learning_rate": 0.00011232587755743332,
+      "loss": 0.5257,
+      "step": 7686
+    },
+    {
+      "epoch": 1.3712425296583712,
+      "grad_norm": 0.43290892243385315,
+      "learning_rate": 0.00011226740158944856,
+      "loss": 0.4729,
+      "step": 7687
+    },
+    {
+      "epoch": 1.3714209258763714,
+      "grad_norm": 0.4931113123893738,
+      "learning_rate": 0.00011220893643892291,
+      "loss": 0.5433,
+      "step": 7688
+    },
+    {
+      "epoch": 1.3715993220943716,
+      "grad_norm": 0.492724746465683,
+      "learning_rate": 0.00011215048211044801,
+      "loss": 0.5631,
+      "step": 7689
+    },
+    {
+      "epoch": 1.3717777183123718,
+      "grad_norm": 0.48536619544029236,
+      "learning_rate": 0.000112092038608615,
+      "loss": 0.509,
+      "step": 7690
+    },
+    {
+      "epoch": 1.371956114530372,
+      "grad_norm": 0.43689486384391785,
+      "learning_rate": 0.00011203360593801396,
+      "loss": 0.5927,
+      "step": 7691
+    },
+    {
+      "epoch": 1.372134510748372,
+      "grad_norm": 0.40810397267341614,
+      "learning_rate": 0.00011197518410323401,
+      "loss": 0.4723,
+      "step": 7692
+    },
+    {
+      "epoch": 1.3723129069663722,
+      "grad_norm": 0.4644933044910431,
+      "learning_rate": 0.00011191677310886384,
+      "loss": 0.562,
+      "step": 7693
+    },
+    {
+      "epoch": 1.3724913031843724,
+      "grad_norm": 0.5269336104393005,
+      "learning_rate": 0.00011185837295949075,
+      "loss": 0.5993,
+      "step": 7694
+    },
+    {
+      "epoch": 1.3726696994023726,
+      "grad_norm": 0.5395172238349915,
+      "learning_rate": 0.00011179998365970174,
+      "loss": 0.7452,
+      "step": 7695
+    },
+    {
+      "epoch": 1.3728480956203728,
+      "grad_norm": 0.5476558804512024,
+      "learning_rate": 0.00011174160521408241,
+      "loss": 0.7672,
+      "step": 7696
+    },
+    {
+      "epoch": 1.373026491838373,
+      "grad_norm": 0.5317317247390747,
+      "learning_rate": 0.0001116832376272179,
+      "loss": 0.679,
+      "step": 7697
+    },
+    {
+      "epoch": 1.3732048880563732,
+      "grad_norm": 0.9869957566261292,
+      "learning_rate": 0.00011162488090369252,
+      "loss": 0.7226,
+      "step": 7698
+    },
+    {
+      "epoch": 1.3733832842743734,
+      "grad_norm": 0.5336429476737976,
+      "learning_rate": 0.00011156653504808934,
+      "loss": 0.7258,
+      "step": 7699
+    },
+    {
+      "epoch": 1.3735616804923736,
+      "grad_norm": 0.5142067670822144,
+      "learning_rate": 0.00011150820006499101,
+      "loss": 0.6718,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3737400767103738,
+      "grad_norm": 0.49160951375961304,
+      "learning_rate": 0.00011144987595897896,
+      "loss": 0.623,
+      "step": 7701
+    },
+    {
+      "epoch": 1.373918472928374,
+      "grad_norm": 0.4981287717819214,
+      "learning_rate": 0.00011139156273463411,
+      "loss": 0.6447,
+      "step": 7702
+    },
+    {
+      "epoch": 1.3740968691463742,
+      "grad_norm": 0.4330550730228424,
+      "learning_rate": 0.0001113332603965363,
+      "loss": 0.3944,
+      "step": 7703
+    },
+    {
+      "epoch": 1.3742752653643744,
+      "grad_norm": 0.5066832900047302,
+      "learning_rate": 0.00011127496894926442,
+      "loss": 0.5891,
+      "step": 7704
+    },
+    {
+      "epoch": 1.3744536615823746,
+      "grad_norm": 0.5171502232551575,
+      "learning_rate": 0.00011121668839739691,
+      "loss": 0.7084,
+      "step": 7705
+    },
+    {
+      "epoch": 1.3746320578003746,
+      "grad_norm": 0.48667481541633606,
+      "learning_rate": 0.00011115841874551084,
+      "loss": 0.5493,
+      "step": 7706
+    },
+    {
+      "epoch": 1.3748104540183748,
+      "grad_norm": 0.4992111623287201,
+      "learning_rate": 0.00011110015999818293,
+      "loss": 0.584,
+      "step": 7707
+    },
+    {
+      "epoch": 1.374988850236375,
+      "grad_norm": 0.4618065357208252,
+      "learning_rate": 0.00011104191215998857,
+      "loss": 0.6038,
+      "step": 7708
+    },
+    {
+      "epoch": 1.3751672464543752,
+      "grad_norm": 0.4726942777633667,
+      "learning_rate": 0.00011098367523550273,
+      "loss": 0.5524,
+      "step": 7709
+    },
+    {
+      "epoch": 1.3753456426723754,
+      "grad_norm": 0.4904707372188568,
+      "learning_rate": 0.00011092544922929914,
+      "loss": 0.6269,
+      "step": 7710
+    },
+    {
+      "epoch": 1.3755240388903756,
+      "grad_norm": 0.4519045054912567,
+      "learning_rate": 0.0001108672341459509,
+      "loss": 0.5587,
+      "step": 7711
+    },
+    {
+      "epoch": 1.3757024351083758,
+      "grad_norm": 0.45719850063323975,
+      "learning_rate": 0.00011080902999003032,
+      "loss": 0.5617,
+      "step": 7712
+    },
+    {
+      "epoch": 1.375880831326376,
+      "grad_norm": 0.4688970446586609,
+      "learning_rate": 0.00011075083676610853,
+      "loss": 0.4525,
+      "step": 7713
+    },
+    {
+      "epoch": 1.376059227544376,
+      "grad_norm": 0.4638102650642395,
+      "learning_rate": 0.00011069265447875617,
+      "loss": 0.5887,
+      "step": 7714
+    },
+    {
+      "epoch": 1.3762376237623761,
+      "grad_norm": 0.5201415419578552,
+      "learning_rate": 0.0001106344831325427,
+      "loss": 0.6049,
+      "step": 7715
+    },
+    {
+      "epoch": 1.3764160199803763,
+      "grad_norm": 0.5137766003608704,
+      "learning_rate": 0.00011057632273203708,
+      "loss": 0.6497,
+      "step": 7716
+    },
+    {
+      "epoch": 1.3765944161983765,
+      "grad_norm": 0.5084298849105835,
+      "learning_rate": 0.00011051817328180702,
+      "loss": 0.6516,
+      "step": 7717
+    },
+    {
+      "epoch": 1.3767728124163767,
+      "grad_norm": 0.4721417725086212,
+      "learning_rate": 0.00011046003478641955,
+      "loss": 0.5664,
+      "step": 7718
+    },
+    {
+      "epoch": 1.376951208634377,
+      "grad_norm": 0.4899318814277649,
+      "learning_rate": 0.00011040190725044097,
+      "loss": 0.6067,
+      "step": 7719
+    },
+    {
+      "epoch": 1.3771296048523771,
+      "grad_norm": 0.48559266328811646,
+      "learning_rate": 0.00011034379067843644,
+      "loss": 0.5541,
+      "step": 7720
+    },
+    {
+      "epoch": 1.3773080010703773,
+      "grad_norm": 0.46098917722702026,
+      "learning_rate": 0.00011028568507497058,
+      "loss": 0.5172,
+      "step": 7721
+    },
+    {
+      "epoch": 1.3774863972883775,
+      "grad_norm": 0.5564523935317993,
+      "learning_rate": 0.00011022759044460678,
+      "loss": 0.7203,
+      "step": 7722
+    },
+    {
+      "epoch": 1.3776647935063777,
+      "grad_norm": 0.5125877261161804,
+      "learning_rate": 0.00011016950679190798,
+      "loss": 0.5827,
+      "step": 7723
+    },
+    {
+      "epoch": 1.377843189724378,
+      "grad_norm": 0.4653429090976715,
+      "learning_rate": 0.00011011143412143596,
+      "loss": 0.593,
+      "step": 7724
+    },
+    {
+      "epoch": 1.3780215859423781,
+      "grad_norm": 0.44072821736335754,
+      "learning_rate": 0.0001100533724377516,
+      "loss": 0.5665,
+      "step": 7725
+    },
+    {
+      "epoch": 1.3781999821603783,
+      "grad_norm": 0.48405829071998596,
+      "learning_rate": 0.00010999532174541524,
+      "loss": 0.6273,
+      "step": 7726
+    },
+    {
+      "epoch": 1.3783783783783785,
+      "grad_norm": 0.5710448622703552,
+      "learning_rate": 0.0001099372820489859,
+      "loss": 0.6321,
+      "step": 7727
+    },
+    {
+      "epoch": 1.3785567745963785,
+      "grad_norm": 0.47928139567375183,
+      "learning_rate": 0.00010987925335302229,
+      "loss": 0.6109,
+      "step": 7728
+    },
+    {
+      "epoch": 1.3787351708143787,
+      "grad_norm": 0.47921472787857056,
+      "learning_rate": 0.00010982123566208185,
+      "loss": 0.5726,
+      "step": 7729
+    },
+    {
+      "epoch": 1.3789135670323789,
+      "grad_norm": 0.4955821633338928,
+      "learning_rate": 0.00010976322898072117,
+      "loss": 0.5733,
+      "step": 7730
+    },
+    {
+      "epoch": 1.379091963250379,
+      "grad_norm": 0.45048630237579346,
+      "learning_rate": 0.00010970523331349619,
+      "loss": 0.5002,
+      "step": 7731
+    },
+    {
+      "epoch": 1.3792703594683793,
+      "grad_norm": 0.4999293088912964,
+      "learning_rate": 0.00010964724866496173,
+      "loss": 0.6949,
+      "step": 7732
+    },
+    {
+      "epoch": 1.3794487556863795,
+      "grad_norm": 0.5359913110733032,
+      "learning_rate": 0.00010958927503967206,
+      "loss": 0.6405,
+      "step": 7733
+    },
+    {
+      "epoch": 1.3796271519043797,
+      "grad_norm": 0.4533331096172333,
+      "learning_rate": 0.0001095313124421802,
+      "loss": 0.5499,
+      "step": 7734
+    },
+    {
+      "epoch": 1.3798055481223799,
+      "grad_norm": 0.4622163772583008,
+      "learning_rate": 0.00010947336087703872,
+      "loss": 0.532,
+      "step": 7735
+    },
+    {
+      "epoch": 1.3799839443403799,
+      "grad_norm": 0.4350248873233795,
+      "learning_rate": 0.0001094154203487989,
+      "loss": 0.4439,
+      "step": 7736
+    },
+    {
+      "epoch": 1.38016234055838,
+      "grad_norm": 0.47929033637046814,
+      "learning_rate": 0.00010935749086201158,
+      "loss": 0.5798,
+      "step": 7737
+    },
+    {
+      "epoch": 1.3803407367763803,
+      "grad_norm": 0.5002579689025879,
+      "learning_rate": 0.00010929957242122637,
+      "loss": 0.6427,
+      "step": 7738
+    },
+    {
+      "epoch": 1.3805191329943804,
+      "grad_norm": 0.5057860612869263,
+      "learning_rate": 0.00010924166503099211,
+      "loss": 0.7614,
+      "step": 7739
+    },
+    {
+      "epoch": 1.3806975292123806,
+      "grad_norm": 0.4656047224998474,
+      "learning_rate": 0.00010918376869585702,
+      "loss": 0.5467,
+      "step": 7740
+    },
+    {
+      "epoch": 1.3808759254303808,
+      "grad_norm": 0.5039976835250854,
+      "learning_rate": 0.00010912588342036802,
+      "loss": 0.6859,
+      "step": 7741
+    },
+    {
+      "epoch": 1.381054321648381,
+      "grad_norm": 0.6870278716087341,
+      "learning_rate": 0.00010906800920907153,
+      "loss": 0.5815,
+      "step": 7742
+    },
+    {
+      "epoch": 1.3812327178663812,
+      "grad_norm": 0.43213722109794617,
+      "learning_rate": 0.00010901014606651305,
+      "loss": 0.3694,
+      "step": 7743
+    },
+    {
+      "epoch": 1.3814111140843814,
+      "grad_norm": 0.5106995701789856,
+      "learning_rate": 0.00010895229399723694,
+      "loss": 0.5759,
+      "step": 7744
+    },
+    {
+      "epoch": 1.3815895103023816,
+      "grad_norm": 0.45927223563194275,
+      "learning_rate": 0.00010889445300578701,
+      "loss": 0.5318,
+      "step": 7745
+    },
+    {
+      "epoch": 1.3817679065203818,
+      "grad_norm": 0.5364645719528198,
+      "learning_rate": 0.00010883662309670597,
+      "loss": 0.6094,
+      "step": 7746
+    },
+    {
+      "epoch": 1.381946302738382,
+      "grad_norm": 0.5728722214698792,
+      "learning_rate": 0.00010877880427453588,
+      "loss": 0.7887,
+      "step": 7747
+    },
+    {
+      "epoch": 1.3821246989563822,
+      "grad_norm": 0.49186766147613525,
+      "learning_rate": 0.00010872099654381762,
+      "loss": 0.5248,
+      "step": 7748
+    },
+    {
+      "epoch": 1.3823030951743824,
+      "grad_norm": 0.48165062069892883,
+      "learning_rate": 0.00010866319990909163,
+      "loss": 0.6226,
+      "step": 7749
+    },
+    {
+      "epoch": 1.3824814913923824,
+      "grad_norm": 0.49330055713653564,
+      "learning_rate": 0.00010860541437489705,
+      "loss": 0.6703,
+      "step": 7750
+    },
+    {
+      "epoch": 1.3826598876103826,
+      "grad_norm": 0.5186200141906738,
+      "learning_rate": 0.00010854763994577232,
+      "loss": 0.7396,
+      "step": 7751
+    },
+    {
+      "epoch": 1.3828382838283828,
+      "grad_norm": 0.5332891941070557,
+      "learning_rate": 0.00010848987662625515,
+      "loss": 0.7148,
+      "step": 7752
+    },
+    {
+      "epoch": 1.383016680046383,
+      "grad_norm": 0.5629260540008545,
+      "learning_rate": 0.00010843212442088207,
+      "loss": 0.7122,
+      "step": 7753
+    },
+    {
+      "epoch": 1.3831950762643832,
+      "grad_norm": 0.44756463170051575,
+      "learning_rate": 0.00010837438333418914,
+      "loss": 0.5578,
+      "step": 7754
+    },
+    {
+      "epoch": 1.3833734724823834,
+      "grad_norm": 0.48155879974365234,
+      "learning_rate": 0.00010831665337071109,
+      "loss": 0.5866,
+      "step": 7755
+    },
+    {
+      "epoch": 1.3835518687003836,
+      "grad_norm": 0.5249548554420471,
+      "learning_rate": 0.0001082589345349822,
+      "loss": 0.6235,
+      "step": 7756
+    },
+    {
+      "epoch": 1.3837302649183838,
+      "grad_norm": 0.4601089656352997,
+      "learning_rate": 0.00010820122683153552,
+      "loss": 0.4966,
+      "step": 7757
+    },
+    {
+      "epoch": 1.3839086611363838,
+      "grad_norm": 0.4255712330341339,
+      "learning_rate": 0.00010814353026490345,
+      "loss": 0.3554,
+      "step": 7758
+    },
+    {
+      "epoch": 1.384087057354384,
+      "grad_norm": 0.3792378902435303,
+      "learning_rate": 0.00010808584483961755,
+      "loss": 0.3656,
+      "step": 7759
+    },
+    {
+      "epoch": 1.3842654535723842,
+      "grad_norm": 0.5113689303398132,
+      "learning_rate": 0.00010802817056020825,
+      "loss": 0.6168,
+      "step": 7760
+    },
+    {
+      "epoch": 1.3844438497903844,
+      "grad_norm": 0.5441461205482483,
+      "learning_rate": 0.00010797050743120542,
+      "loss": 0.7846,
+      "step": 7761
+    },
+    {
+      "epoch": 1.3846222460083846,
+      "grad_norm": 0.47966670989990234,
+      "learning_rate": 0.00010791285545713783,
+      "loss": 0.5587,
+      "step": 7762
+    },
+    {
+      "epoch": 1.3848006422263848,
+      "grad_norm": 0.46844810247421265,
+      "learning_rate": 0.00010785521464253334,
+      "loss": 0.551,
+      "step": 7763
+    },
+    {
+      "epoch": 1.384979038444385,
+      "grad_norm": 1.28788161277771,
+      "learning_rate": 0.00010779758499191919,
+      "loss": 0.7065,
+      "step": 7764
+    },
+    {
+      "epoch": 1.3851574346623852,
+      "grad_norm": 0.5162654519081116,
+      "learning_rate": 0.00010773996650982146,
+      "loss": 0.6795,
+      "step": 7765
+    },
+    {
+      "epoch": 1.3853358308803854,
+      "grad_norm": 0.5239789485931396,
+      "learning_rate": 0.00010768235920076561,
+      "loss": 0.7062,
+      "step": 7766
+    },
+    {
+      "epoch": 1.3855142270983856,
+      "grad_norm": 0.4905536472797394,
+      "learning_rate": 0.00010762476306927594,
+      "loss": 0.5994,
+      "step": 7767
+    },
+    {
+      "epoch": 1.3856926233163858,
+      "grad_norm": 0.5204237699508667,
+      "learning_rate": 0.00010756717811987618,
+      "loss": 0.5825,
+      "step": 7768
+    },
+    {
+      "epoch": 1.385871019534386,
+      "grad_norm": 0.5072759389877319,
+      "learning_rate": 0.0001075096043570889,
+      "loss": 0.6562,
+      "step": 7769
+    },
+    {
+      "epoch": 1.3860494157523862,
+      "grad_norm": 0.45566022396087646,
+      "learning_rate": 0.00010745204178543605,
+      "loss": 0.5398,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3862278119703864,
+      "grad_norm": 0.4731004238128662,
+      "learning_rate": 0.00010739449040943849,
+      "loss": 0.6673,
+      "step": 7771
+    },
+    {
+      "epoch": 1.3864062081883863,
+      "grad_norm": 0.472790390253067,
+      "learning_rate": 0.0001073369502336161,
+      "loss": 0.499,
+      "step": 7772
+    },
+    {
+      "epoch": 1.3865846044063865,
+      "grad_norm": 0.46505823731422424,
+      "learning_rate": 0.00010727942126248843,
+      "loss": 0.5368,
+      "step": 7773
+    },
+    {
+      "epoch": 1.3867630006243867,
+      "grad_norm": 0.47594350576400757,
+      "learning_rate": 0.0001072219035005735,
+      "loss": 0.5684,
+      "step": 7774
+    },
+    {
+      "epoch": 1.386941396842387,
+      "grad_norm": 0.49051812291145325,
+      "learning_rate": 0.00010716439695238895,
+      "loss": 0.5684,
+      "step": 7775
+    },
+    {
+      "epoch": 1.3871197930603871,
+      "grad_norm": 0.5569090843200684,
+      "learning_rate": 0.00010710690162245118,
+      "loss": 0.666,
+      "step": 7776
+    },
+    {
+      "epoch": 1.3872981892783873,
+      "grad_norm": 0.4677387475967407,
+      "learning_rate": 0.00010704941751527578,
+      "loss": 0.5223,
+      "step": 7777
+    },
+    {
+      "epoch": 1.3874765854963875,
+      "grad_norm": 0.5612090229988098,
+      "learning_rate": 0.00010699194463537767,
+      "loss": 0.7503,
+      "step": 7778
+    },
+    {
+      "epoch": 1.3876549817143877,
+      "grad_norm": 0.5161556601524353,
+      "learning_rate": 0.00010693448298727062,
+      "loss": 0.6457,
+      "step": 7779
+    },
+    {
+      "epoch": 1.3878333779323877,
+      "grad_norm": 0.5419967174530029,
+      "learning_rate": 0.0001068770325754678,
+      "loss": 0.6822,
+      "step": 7780
+    },
+    {
+      "epoch": 1.388011774150388,
+      "grad_norm": 0.4641594886779785,
+      "learning_rate": 0.00010681959340448116,
+      "loss": 0.4709,
+      "step": 7781
+    },
+    {
+      "epoch": 1.388190170368388,
+      "grad_norm": 0.8507322669029236,
+      "learning_rate": 0.00010676216547882214,
+      "loss": 0.5894,
+      "step": 7782
+    },
+    {
+      "epoch": 1.3883685665863883,
+      "grad_norm": 0.5100333094596863,
+      "learning_rate": 0.00010670474880300098,
+      "loss": 0.6258,
+      "step": 7783
+    },
+    {
+      "epoch": 1.3885469628043885,
+      "grad_norm": 0.49771493673324585,
+      "learning_rate": 0.00010664734338152712,
+      "loss": 0.5853,
+      "step": 7784
+    },
+    {
+      "epoch": 1.3887253590223887,
+      "grad_norm": 0.5230436325073242,
+      "learning_rate": 0.0001065899492189093,
+      "loss": 0.6821,
+      "step": 7785
+    },
+    {
+      "epoch": 1.388903755240389,
+      "grad_norm": 0.5067762732505798,
+      "learning_rate": 0.00010653256631965505,
+      "loss": 0.7528,
+      "step": 7786
+    },
+    {
+      "epoch": 1.389082151458389,
+      "grad_norm": 0.48270925879478455,
+      "learning_rate": 0.0001064751946882714,
+      "loss": 0.5972,
+      "step": 7787
+    },
+    {
+      "epoch": 1.3892605476763893,
+      "grad_norm": 0.47450000047683716,
+      "learning_rate": 0.0001064178343292641,
+      "loss": 0.5595,
+      "step": 7788
+    },
+    {
+      "epoch": 1.3894389438943895,
+      "grad_norm": 0.48282551765441895,
+      "learning_rate": 0.00010636048524713832,
+      "loss": 0.515,
+      "step": 7789
+    },
+    {
+      "epoch": 1.3896173401123897,
+      "grad_norm": 0.5305141806602478,
+      "learning_rate": 0.0001063031474463983,
+      "loss": 0.5755,
+      "step": 7790
+    },
+    {
+      "epoch": 1.3897957363303899,
+      "grad_norm": 0.48532387614250183,
+      "learning_rate": 0.00010624582093154717,
+      "loss": 0.5692,
+      "step": 7791
+    },
+    {
+      "epoch": 1.38997413254839,
+      "grad_norm": 0.5484132170677185,
+      "learning_rate": 0.00010618850570708746,
+      "loss": 0.6092,
+      "step": 7792
+    },
+    {
+      "epoch": 1.3901525287663903,
+      "grad_norm": 0.5588973760604858,
+      "learning_rate": 0.00010613120177752056,
+      "loss": 0.7572,
+      "step": 7793
+    },
+    {
+      "epoch": 1.3903309249843903,
+      "grad_norm": 0.5241721272468567,
+      "learning_rate": 0.00010607390914734721,
+      "loss": 0.4951,
+      "step": 7794
+    },
+    {
+      "epoch": 1.3905093212023905,
+      "grad_norm": 0.49718037247657776,
+      "learning_rate": 0.00010601662782106705,
+      "loss": 0.7169,
+      "step": 7795
+    },
+    {
+      "epoch": 1.3906877174203907,
+      "grad_norm": 0.5253555774688721,
+      "learning_rate": 0.00010595935780317906,
+      "loss": 0.6851,
+      "step": 7796
+    },
+    {
+      "epoch": 1.3908661136383909,
+      "grad_norm": 0.5283133387565613,
+      "learning_rate": 0.00010590209909818113,
+      "loss": 0.6305,
+      "step": 7797
+    },
+    {
+      "epoch": 1.391044509856391,
+      "grad_norm": 0.5732675790786743,
+      "learning_rate": 0.0001058448517105702,
+      "loss": 0.781,
+      "step": 7798
+    },
+    {
+      "epoch": 1.3912229060743913,
+      "grad_norm": 0.4708344638347626,
+      "learning_rate": 0.0001057876156448427,
+      "loss": 0.5672,
+      "step": 7799
+    },
+    {
+      "epoch": 1.3914013022923915,
+      "grad_norm": 0.48784154653549194,
+      "learning_rate": 0.00010573039090549374,
+      "loss": 0.5962,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3915796985103916,
+      "grad_norm": 0.4613569974899292,
+      "learning_rate": 0.00010567317749701785,
+      "loss": 0.5355,
+      "step": 7801
+    },
+    {
+      "epoch": 1.3917580947283916,
+      "grad_norm": 0.51911461353302,
+      "learning_rate": 0.00010561597542390842,
+      "loss": 0.609,
+      "step": 7802
+    },
+    {
+      "epoch": 1.3919364909463918,
+      "grad_norm": 0.478140652179718,
+      "learning_rate": 0.00010555878469065814,
+      "loss": 0.5447,
+      "step": 7803
+    },
+    {
+      "epoch": 1.392114887164392,
+      "grad_norm": 0.4372231662273407,
+      "learning_rate": 0.00010550160530175883,
+      "loss": 0.5127,
+      "step": 7804
+    },
+    {
+      "epoch": 1.3922932833823922,
+      "grad_norm": 0.4996614456176758,
+      "learning_rate": 0.00010544443726170119,
+      "loss": 0.5593,
+      "step": 7805
+    },
+    {
+      "epoch": 1.3924716796003924,
+      "grad_norm": 0.47102734446525574,
+      "learning_rate": 0.00010538728057497532,
+      "loss": 0.6029,
+      "step": 7806
+    },
+    {
+      "epoch": 1.3926500758183926,
+      "grad_norm": 0.5277090072631836,
+      "learning_rate": 0.0001053301352460701,
+      "loss": 0.6538,
+      "step": 7807
+    },
+    {
+      "epoch": 1.3928284720363928,
+      "grad_norm": 0.4949512183666229,
+      "learning_rate": 0.0001052730012794739,
+      "loss": 0.5816,
+      "step": 7808
+    },
+    {
+      "epoch": 1.393006868254393,
+      "grad_norm": 0.4444587230682373,
+      "learning_rate": 0.00010521587867967389,
+      "loss": 0.4541,
+      "step": 7809
+    },
+    {
+      "epoch": 1.3931852644723932,
+      "grad_norm": 0.46152183413505554,
+      "learning_rate": 0.0001051587674511564,
+      "loss": 0.4362,
+      "step": 7810
+    },
+    {
+      "epoch": 1.3933636606903934,
+      "grad_norm": 0.4519818127155304,
+      "learning_rate": 0.00010510166759840705,
+      "loss": 0.4725,
+      "step": 7811
+    },
+    {
+      "epoch": 1.3935420569083936,
+      "grad_norm": 0.48864850401878357,
+      "learning_rate": 0.00010504457912591028,
+      "loss": 0.4834,
+      "step": 7812
+    },
+    {
+      "epoch": 1.3937204531263938,
+      "grad_norm": 0.5076178908348083,
+      "learning_rate": 0.00010498750203815,
+      "loss": 0.6161,
+      "step": 7813
+    },
+    {
+      "epoch": 1.393898849344394,
+      "grad_norm": 0.4903450012207031,
+      "learning_rate": 0.0001049304363396088,
+      "loss": 0.5639,
+      "step": 7814
+    },
+    {
+      "epoch": 1.3940772455623942,
+      "grad_norm": 0.44738972187042236,
+      "learning_rate": 0.0001048733820347688,
+      "loss": 0.4897,
+      "step": 7815
+    },
+    {
+      "epoch": 1.3942556417803942,
+      "grad_norm": 0.5322644710540771,
+      "learning_rate": 0.00010481633912811081,
+      "loss": 0.7432,
+      "step": 7816
+    },
+    {
+      "epoch": 1.3944340379983944,
+      "grad_norm": 0.5089389085769653,
+      "learning_rate": 0.00010475930762411518,
+      "loss": 0.5638,
+      "step": 7817
+    },
+    {
+      "epoch": 1.3946124342163946,
+      "grad_norm": 0.5720763206481934,
+      "learning_rate": 0.00010470228752726094,
+      "loss": 0.6849,
+      "step": 7818
+    },
+    {
+      "epoch": 1.3947908304343948,
+      "grad_norm": 0.5013535618782043,
+      "learning_rate": 0.00010464527884202648,
+      "loss": 0.6032,
+      "step": 7819
+    },
+    {
+      "epoch": 1.394969226652395,
+      "grad_norm": 0.523189127445221,
+      "learning_rate": 0.00010458828157288938,
+      "loss": 0.7473,
+      "step": 7820
+    },
+    {
+      "epoch": 1.3951476228703952,
+      "grad_norm": 0.42415139079093933,
+      "learning_rate": 0.00010453129572432599,
+      "loss": 0.4601,
+      "step": 7821
+    },
+    {
+      "epoch": 1.3953260190883954,
+      "grad_norm": 0.4629247784614563,
+      "learning_rate": 0.0001044743213008121,
+      "loss": 0.4898,
+      "step": 7822
+    },
+    {
+      "epoch": 1.3955044153063956,
+      "grad_norm": 0.7844071984291077,
+      "learning_rate": 0.00010441735830682242,
+      "loss": 0.8687,
+      "step": 7823
+    },
+    {
+      "epoch": 1.3956828115243956,
+      "grad_norm": 0.46940067410469055,
+      "learning_rate": 0.00010436040674683067,
+      "loss": 0.5528,
+      "step": 7824
+    },
+    {
+      "epoch": 1.3958612077423957,
+      "grad_norm": 0.49708932638168335,
+      "learning_rate": 0.00010430346662530999,
+      "loss": 0.6975,
+      "step": 7825
+    },
+    {
+      "epoch": 1.396039603960396,
+      "grad_norm": 0.47257721424102783,
+      "learning_rate": 0.00010424653794673222,
+      "loss": 0.5658,
+      "step": 7826
+    },
+    {
+      "epoch": 1.3962180001783961,
+      "grad_norm": 0.4764406383037567,
+      "learning_rate": 0.00010418962071556876,
+      "loss": 0.7406,
+      "step": 7827
+    },
+    {
+      "epoch": 1.3963963963963963,
+      "grad_norm": 0.4551503360271454,
+      "learning_rate": 0.00010413271493628965,
+      "loss": 0.5595,
+      "step": 7828
+    },
+    {
+      "epoch": 1.3965747926143965,
+      "grad_norm": 0.4807843267917633,
+      "learning_rate": 0.00010407582061336443,
+      "loss": 0.4765,
+      "step": 7829
+    },
+    {
+      "epoch": 1.3967531888323967,
+      "grad_norm": 0.46619912981987,
+      "learning_rate": 0.00010401893775126146,
+      "loss": 0.5123,
+      "step": 7830
+    },
+    {
+      "epoch": 1.396931585050397,
+      "grad_norm": 0.4833389222621918,
+      "learning_rate": 0.00010396206635444819,
+      "loss": 0.5746,
+      "step": 7831
+    },
+    {
+      "epoch": 1.3971099812683971,
+      "grad_norm": 1.4866968393325806,
+      "learning_rate": 0.00010390520642739149,
+      "loss": 0.4599,
+      "step": 7832
+    },
+    {
+      "epoch": 1.3972883774863973,
+      "grad_norm": 0.4797883927822113,
+      "learning_rate": 0.00010384835797455691,
+      "loss": 0.4975,
+      "step": 7833
+    },
+    {
+      "epoch": 1.3974667737043975,
+      "grad_norm": 0.543132483959198,
+      "learning_rate": 0.00010379152100040942,
+      "loss": 0.7443,
+      "step": 7834
+    },
+    {
+      "epoch": 1.3976451699223977,
+      "grad_norm": 0.49611344933509827,
+      "learning_rate": 0.00010373469550941304,
+      "loss": 0.5435,
+      "step": 7835
+    },
+    {
+      "epoch": 1.397823566140398,
+      "grad_norm": 0.549901008605957,
+      "learning_rate": 0.00010367788150603061,
+      "loss": 0.8226,
+      "step": 7836
+    },
+    {
+      "epoch": 1.3980019623583981,
+      "grad_norm": 0.48870500922203064,
+      "learning_rate": 0.0001036210789947245,
+      "loss": 0.6682,
+      "step": 7837
+    },
+    {
+      "epoch": 1.398180358576398,
+      "grad_norm": 0.4649103879928589,
+      "learning_rate": 0.00010356428797995579,
+      "loss": 0.5182,
+      "step": 7838
+    },
+    {
+      "epoch": 1.3983587547943983,
+      "grad_norm": 0.49515223503112793,
+      "learning_rate": 0.00010350750846618495,
+      "loss": 0.6104,
+      "step": 7839
+    },
+    {
+      "epoch": 1.3985371510123985,
+      "grad_norm": 0.4876445531845093,
+      "learning_rate": 0.00010345074045787128,
+      "loss": 0.6187,
+      "step": 7840
+    },
+    {
+      "epoch": 1.3987155472303987,
+      "grad_norm": 0.47437965869903564,
+      "learning_rate": 0.00010339398395947347,
+      "loss": 0.7047,
+      "step": 7841
+    },
+    {
+      "epoch": 1.398893943448399,
+      "grad_norm": 0.4776071608066559,
+      "learning_rate": 0.00010333723897544908,
+      "loss": 0.655,
+      "step": 7842
+    },
+    {
+      "epoch": 1.399072339666399,
+      "grad_norm": 0.48358532786369324,
+      "learning_rate": 0.00010328050551025472,
+      "loss": 0.6231,
+      "step": 7843
+    },
+    {
+      "epoch": 1.3992507358843993,
+      "grad_norm": 0.5266225934028625,
+      "learning_rate": 0.00010322378356834641,
+      "loss": 0.7969,
+      "step": 7844
+    },
+    {
+      "epoch": 1.3994291321023995,
+      "grad_norm": 0.42173731327056885,
+      "learning_rate": 0.00010316707315417892,
+      "loss": 0.4284,
+      "step": 7845
+    },
+    {
+      "epoch": 1.3996075283203995,
+      "grad_norm": 0.47026675939559937,
+      "learning_rate": 0.00010311037427220637,
+      "loss": 0.5441,
+      "step": 7846
+    },
+    {
+      "epoch": 1.3997859245383997,
+      "grad_norm": 0.4756429195404053,
+      "learning_rate": 0.00010305368692688174,
+      "loss": 0.4786,
+      "step": 7847
+    },
+    {
+      "epoch": 1.3999643207563999,
+      "grad_norm": 0.48206183314323425,
+      "learning_rate": 0.00010299701112265739,
+      "loss": 0.623,
+      "step": 7848
+    },
+    {
+      "epoch": 1.4001427169744,
+      "grad_norm": 0.47061610221862793,
+      "learning_rate": 0.00010294034686398443,
+      "loss": 0.5716,
+      "step": 7849
+    },
+    {
+      "epoch": 1.4003211131924003,
+      "grad_norm": 0.46232593059539795,
+      "learning_rate": 0.00010288369415531335,
+      "loss": 0.5713,
+      "step": 7850
+    },
+    {
+      "epoch": 1.4004995094104005,
+      "grad_norm": 0.415913850069046,
+      "learning_rate": 0.00010282705300109372,
+      "loss": 0.4381,
+      "step": 7851
+    },
+    {
+      "epoch": 1.4006779056284007,
+      "grad_norm": 0.5227693319320679,
+      "learning_rate": 0.00010277042340577388,
+      "loss": 0.6813,
+      "step": 7852
+    },
+    {
+      "epoch": 1.4008563018464009,
+      "grad_norm": 0.4625205099582672,
+      "learning_rate": 0.00010271380537380176,
+      "loss": 0.5038,
+      "step": 7853
+    },
+    {
+      "epoch": 1.401034698064401,
+      "grad_norm": 0.405355840921402,
+      "learning_rate": 0.00010265719890962388,
+      "loss": 0.4174,
+      "step": 7854
+    },
+    {
+      "epoch": 1.4012130942824013,
+      "grad_norm": 0.4783884584903717,
+      "learning_rate": 0.00010260060401768628,
+      "loss": 0.5663,
+      "step": 7855
+    },
+    {
+      "epoch": 1.4013914905004015,
+      "grad_norm": 0.49855825304985046,
+      "learning_rate": 0.00010254402070243383,
+      "loss": 0.6415,
+      "step": 7856
+    },
+    {
+      "epoch": 1.4015698867184017,
+      "grad_norm": 0.5020763278007507,
+      "learning_rate": 0.00010248744896831044,
+      "loss": 0.6289,
+      "step": 7857
+    },
+    {
+      "epoch": 1.4017482829364019,
+      "grad_norm": 0.5697162747383118,
+      "learning_rate": 0.0001024308888197594,
+      "loss": 0.7057,
+      "step": 7858
+    },
+    {
+      "epoch": 1.401926679154402,
+      "grad_norm": 0.5255553126335144,
+      "learning_rate": 0.00010237434026122278,
+      "loss": 0.699,
+      "step": 7859
+    },
+    {
+      "epoch": 1.402105075372402,
+      "grad_norm": 0.48212558031082153,
+      "learning_rate": 0.00010231780329714202,
+      "loss": 0.4816,
+      "step": 7860
+    },
+    {
+      "epoch": 1.4022834715904022,
+      "grad_norm": 0.55892014503479,
+      "learning_rate": 0.00010226127793195736,
+      "loss": 0.7858,
+      "step": 7861
+    },
+    {
+      "epoch": 1.4024618678084024,
+      "grad_norm": 0.473724365234375,
+      "learning_rate": 0.00010220476417010843,
+      "loss": 0.7018,
+      "step": 7862
+    },
+    {
+      "epoch": 1.4026402640264026,
+      "grad_norm": 0.5031070709228516,
+      "learning_rate": 0.00010214826201603372,
+      "loss": 0.5582,
+      "step": 7863
+    },
+    {
+      "epoch": 1.4028186602444028,
+      "grad_norm": 0.5168371200561523,
+      "learning_rate": 0.00010209177147417073,
+      "loss": 0.6044,
+      "step": 7864
+    },
+    {
+      "epoch": 1.402997056462403,
+      "grad_norm": 0.5146030783653259,
+      "learning_rate": 0.00010203529254895652,
+      "loss": 0.6976,
+      "step": 7865
+    },
+    {
+      "epoch": 1.4031754526804032,
+      "grad_norm": 0.5497780442237854,
+      "learning_rate": 0.00010197882524482669,
+      "loss": 0.6612,
+      "step": 7866
+    },
+    {
+      "epoch": 1.4033538488984034,
+      "grad_norm": 0.530221164226532,
+      "learning_rate": 0.00010192236956621628,
+      "loss": 0.6626,
+      "step": 7867
+    },
+    {
+      "epoch": 1.4035322451164034,
+      "grad_norm": 0.5098912715911865,
+      "learning_rate": 0.00010186592551755927,
+      "loss": 0.6494,
+      "step": 7868
+    },
+    {
+      "epoch": 1.4037106413344036,
+      "grad_norm": 0.44442427158355713,
+      "learning_rate": 0.0001018094931032886,
+      "loss": 0.4793,
+      "step": 7869
+    },
+    {
+      "epoch": 1.4038890375524038,
+      "grad_norm": 0.5126553773880005,
+      "learning_rate": 0.0001017530723278367,
+      "loss": 0.6441,
+      "step": 7870
+    },
+    {
+      "epoch": 1.404067433770404,
+      "grad_norm": 0.568909227848053,
+      "learning_rate": 0.00010169666319563458,
+      "loss": 0.8517,
+      "step": 7871
+    },
+    {
+      "epoch": 1.4042458299884042,
+      "grad_norm": 0.500556230545044,
+      "learning_rate": 0.00010164026571111284,
+      "loss": 0.6562,
+      "step": 7872
+    },
+    {
+      "epoch": 1.4044242262064044,
+      "grad_norm": 0.4273224472999573,
+      "learning_rate": 0.00010158387987870065,
+      "loss": 0.4493,
+      "step": 7873
+    },
+    {
+      "epoch": 1.4046026224244046,
+      "grad_norm": 0.49929437041282654,
+      "learning_rate": 0.00010152750570282679,
+      "loss": 0.7073,
+      "step": 7874
+    },
+    {
+      "epoch": 1.4047810186424048,
+      "grad_norm": 0.47794604301452637,
+      "learning_rate": 0.00010147114318791864,
+      "loss": 0.5762,
+      "step": 7875
+    },
+    {
+      "epoch": 1.404959414860405,
+      "grad_norm": 0.4568422734737396,
+      "learning_rate": 0.00010141479233840309,
+      "loss": 0.5284,
+      "step": 7876
+    },
+    {
+      "epoch": 1.4051378110784052,
+      "grad_norm": 0.5395530462265015,
+      "learning_rate": 0.00010135845315870579,
+      "loss": 0.7679,
+      "step": 7877
+    },
+    {
+      "epoch": 1.4053162072964054,
+      "grad_norm": 0.5494503974914551,
+      "learning_rate": 0.00010130212565325153,
+      "loss": 0.7338,
+      "step": 7878
+    },
+    {
+      "epoch": 1.4054946035144056,
+      "grad_norm": 0.4741472899913788,
+      "learning_rate": 0.00010124580982646442,
+      "loss": 0.5062,
+      "step": 7879
+    },
+    {
+      "epoch": 1.4056729997324058,
+      "grad_norm": 0.47103115916252136,
+      "learning_rate": 0.00010118950568276722,
+      "loss": 0.548,
+      "step": 7880
+    },
+    {
+      "epoch": 1.405851395950406,
+      "grad_norm": 0.48441192507743835,
+      "learning_rate": 0.0001011332132265824,
+      "loss": 0.6215,
+      "step": 7881
+    },
+    {
+      "epoch": 1.406029792168406,
+      "grad_norm": 0.5685884356498718,
+      "learning_rate": 0.0001010769324623309,
+      "loss": 0.5983,
+      "step": 7882
+    },
+    {
+      "epoch": 1.4062081883864062,
+      "grad_norm": 0.5327664613723755,
+      "learning_rate": 0.00010102066339443299,
+      "loss": 0.418,
+      "step": 7883
+    },
+    {
+      "epoch": 1.4063865846044064,
+      "grad_norm": 0.4698556363582611,
+      "learning_rate": 0.00010096440602730816,
+      "loss": 0.6731,
+      "step": 7884
+    },
+    {
+      "epoch": 1.4065649808224066,
+      "grad_norm": 0.46319758892059326,
+      "learning_rate": 0.00010090816036537462,
+      "loss": 0.4945,
+      "step": 7885
+    },
+    {
+      "epoch": 1.4067433770404068,
+      "grad_norm": 0.5858322381973267,
+      "learning_rate": 0.00010085192641305013,
+      "loss": 0.7778,
+      "step": 7886
+    },
+    {
+      "epoch": 1.406921773258407,
+      "grad_norm": 0.4609849750995636,
+      "learning_rate": 0.00010079570417475106,
+      "loss": 0.521,
+      "step": 7887
+    },
+    {
+      "epoch": 1.4071001694764071,
+      "grad_norm": 0.6530886888504028,
+      "learning_rate": 0.00010073949365489323,
+      "loss": 0.3922,
+      "step": 7888
+    },
+    {
+      "epoch": 1.4072785656944073,
+      "grad_norm": 0.7445523142814636,
+      "learning_rate": 0.00010068329485789138,
+      "loss": 0.6121,
+      "step": 7889
+    },
+    {
+      "epoch": 1.4074569619124073,
+      "grad_norm": 0.5070978999137878,
+      "learning_rate": 0.0001006271077881592,
+      "loss": 0.5824,
+      "step": 7890
+    },
+    {
+      "epoch": 1.4076353581304075,
+      "grad_norm": 0.46640315651893616,
+      "learning_rate": 0.00010057093245010975,
+      "loss": 0.5733,
+      "step": 7891
+    },
+    {
+      "epoch": 1.4078137543484077,
+      "grad_norm": 0.5237272381782532,
+      "learning_rate": 0.00010051476884815491,
+      "loss": 0.5761,
+      "step": 7892
+    },
+    {
+      "epoch": 1.407992150566408,
+      "grad_norm": 0.4710404872894287,
+      "learning_rate": 0.00010045861698670589,
+      "loss": 0.5382,
+      "step": 7893
+    },
+    {
+      "epoch": 1.4081705467844081,
+      "grad_norm": 0.46644294261932373,
+      "learning_rate": 0.00010040247687017263,
+      "loss": 0.3815,
+      "step": 7894
+    },
+    {
+      "epoch": 1.4083489430024083,
+      "grad_norm": 0.4705592393875122,
+      "learning_rate": 0.00010034634850296445,
+      "loss": 0.4517,
+      "step": 7895
+    },
+    {
+      "epoch": 1.4085273392204085,
+      "grad_norm": 0.5408839583396912,
+      "learning_rate": 0.00010029023188948976,
+      "loss": 0.8073,
+      "step": 7896
+    },
+    {
+      "epoch": 1.4087057354384087,
+      "grad_norm": 0.5179523825645447,
+      "learning_rate": 0.00010023412703415574,
+      "loss": 0.6676,
+      "step": 7897
+    },
+    {
+      "epoch": 1.408884131656409,
+      "grad_norm": 0.49431368708610535,
+      "learning_rate": 0.00010017803394136902,
+      "loss": 0.4934,
+      "step": 7898
+    },
+    {
+      "epoch": 1.409062527874409,
+      "grad_norm": 0.5114967226982117,
+      "learning_rate": 0.00010012195261553494,
+      "loss": 0.6566,
+      "step": 7899
+    },
+    {
+      "epoch": 1.4092409240924093,
+      "grad_norm": 0.47235408425331116,
+      "learning_rate": 0.00010006588306105832,
+      "loss": 0.5071,
+      "step": 7900
+    },
+    {
+      "epoch": 1.4094193203104095,
+      "grad_norm": 0.4692701995372772,
+      "learning_rate": 0.00010000982528234262,
+      "loss": 0.5268,
+      "step": 7901
+    },
+    {
+      "epoch": 1.4095977165284097,
+      "grad_norm": 0.5551561117172241,
+      "learning_rate": 9.995377928379079e-05,
+      "loss": 0.6615,
+      "step": 7902
+    },
+    {
+      "epoch": 1.40977611274641,
+      "grad_norm": 0.48021402955055237,
+      "learning_rate": 9.989774506980457e-05,
+      "loss": 0.601,
+      "step": 7903
+    },
+    {
+      "epoch": 1.4099545089644099,
+      "grad_norm": 0.5263514518737793,
+      "learning_rate": 9.984172264478475e-05,
+      "loss": 0.6353,
+      "step": 7904
+    },
+    {
+      "epoch": 1.41013290518241,
+      "grad_norm": 0.48175185918807983,
+      "learning_rate": 9.978571201313153e-05,
+      "loss": 0.5538,
+      "step": 7905
+    },
+    {
+      "epoch": 1.4103113014004103,
+      "grad_norm": 0.47728538513183594,
+      "learning_rate": 9.972971317924374e-05,
+      "loss": 0.5215,
+      "step": 7906
+    },
+    {
+      "epoch": 1.4104896976184105,
+      "grad_norm": 0.49182960391044617,
+      "learning_rate": 9.967372614751971e-05,
+      "loss": 0.4886,
+      "step": 7907
+    },
+    {
+      "epoch": 1.4106680938364107,
+      "grad_norm": 0.532971203327179,
+      "learning_rate": 9.961775092235642e-05,
+      "loss": 0.6214,
+      "step": 7908
+    },
+    {
+      "epoch": 1.4108464900544109,
+      "grad_norm": 0.4523656368255615,
+      "learning_rate": 9.956178750815037e-05,
+      "loss": 0.4945,
+      "step": 7909
+    },
+    {
+      "epoch": 1.411024886272411,
+      "grad_norm": 0.49244529008865356,
+      "learning_rate": 9.950583590929671e-05,
+      "loss": 0.565,
+      "step": 7910
+    },
+    {
+      "epoch": 1.4112032824904113,
+      "grad_norm": 0.4732704162597656,
+      "learning_rate": 9.944989613018993e-05,
+      "loss": 0.471,
+      "step": 7911
+    },
+    {
+      "epoch": 1.4113816787084112,
+      "grad_norm": 0.5447127223014832,
+      "learning_rate": 9.939396817522362e-05,
+      "loss": 0.6094,
+      "step": 7912
+    },
+    {
+      "epoch": 1.4115600749264114,
+      "grad_norm": 0.5061089992523193,
+      "learning_rate": 9.933805204879013e-05,
+      "loss": 0.553,
+      "step": 7913
+    },
+    {
+      "epoch": 1.4117384711444116,
+      "grad_norm": 0.5438100695610046,
+      "learning_rate": 9.928214775528127e-05,
+      "loss": 0.6007,
+      "step": 7914
+    },
+    {
+      "epoch": 1.4119168673624118,
+      "grad_norm": 0.49287474155426025,
+      "learning_rate": 9.922625529908769e-05,
+      "loss": 0.5443,
+      "step": 7915
+    },
+    {
+      "epoch": 1.412095263580412,
+      "grad_norm": 0.5883775353431702,
+      "learning_rate": 9.917037468459905e-05,
+      "loss": 0.7577,
+      "step": 7916
+    },
+    {
+      "epoch": 1.4122736597984122,
+      "grad_norm": 0.540412187576294,
+      "learning_rate": 9.911450591620436e-05,
+      "loss": 0.6534,
+      "step": 7917
+    },
+    {
+      "epoch": 1.4124520560164124,
+      "grad_norm": 0.5728485584259033,
+      "learning_rate": 9.905864899829135e-05,
+      "loss": 0.7171,
+      "step": 7918
+    },
+    {
+      "epoch": 1.4126304522344126,
+      "grad_norm": 0.4712778329849243,
+      "learning_rate": 9.900280393524719e-05,
+      "loss": 0.5482,
+      "step": 7919
+    },
+    {
+      "epoch": 1.4128088484524128,
+      "grad_norm": 0.5077334642410278,
+      "learning_rate": 9.894697073145773e-05,
+      "loss": 0.7018,
+      "step": 7920
+    },
+    {
+      "epoch": 1.412987244670413,
+      "grad_norm": 0.489620566368103,
+      "learning_rate": 9.889114939130828e-05,
+      "loss": 0.6211,
+      "step": 7921
+    },
+    {
+      "epoch": 1.4131656408884132,
+      "grad_norm": 0.4577091634273529,
+      "learning_rate": 9.883533991918291e-05,
+      "loss": 0.5056,
+      "step": 7922
+    },
+    {
+      "epoch": 1.4133440371064134,
+      "grad_norm": 0.5330988764762878,
+      "learning_rate": 9.877954231946485e-05,
+      "loss": 0.5441,
+      "step": 7923
+    },
+    {
+      "epoch": 1.4135224333244136,
+      "grad_norm": 0.46641597151756287,
+      "learning_rate": 9.872375659653652e-05,
+      "loss": 0.5333,
+      "step": 7924
+    },
+    {
+      "epoch": 1.4137008295424138,
+      "grad_norm": 0.47872334718704224,
+      "learning_rate": 9.866798275477915e-05,
+      "loss": 0.5553,
+      "step": 7925
+    },
+    {
+      "epoch": 1.4138792257604138,
+      "grad_norm": 0.4439626634120941,
+      "learning_rate": 9.861222079857332e-05,
+      "loss": 0.4818,
+      "step": 7926
+    },
+    {
+      "epoch": 1.414057621978414,
+      "grad_norm": 0.5242343544960022,
+      "learning_rate": 9.85564707322986e-05,
+      "loss": 0.6662,
+      "step": 7927
+    },
+    {
+      "epoch": 1.4142360181964142,
+      "grad_norm": 0.47222810983657837,
+      "learning_rate": 9.850073256033337e-05,
+      "loss": 0.5828,
+      "step": 7928
+    },
+    {
+      "epoch": 1.4144144144144144,
+      "grad_norm": 0.4517582952976227,
+      "learning_rate": 9.844500628705555e-05,
+      "loss": 0.464,
+      "step": 7929
+    },
+    {
+      "epoch": 1.4145928106324146,
+      "grad_norm": 0.5746625065803528,
+      "learning_rate": 9.83892919168416e-05,
+      "loss": 0.7129,
+      "step": 7930
+    },
+    {
+      "epoch": 1.4147712068504148,
+      "grad_norm": 0.41956639289855957,
+      "learning_rate": 9.83335894540675e-05,
+      "loss": 0.4671,
+      "step": 7931
+    },
+    {
+      "epoch": 1.414949603068415,
+      "grad_norm": 0.46683749556541443,
+      "learning_rate": 9.827789890310795e-05,
+      "loss": 0.547,
+      "step": 7932
+    },
+    {
+      "epoch": 1.4151279992864152,
+      "grad_norm": 0.5841799378395081,
+      "learning_rate": 9.822222026833703e-05,
+      "loss": 0.8423,
+      "step": 7933
+    },
+    {
+      "epoch": 1.4153063955044152,
+      "grad_norm": 0.520228922367096,
+      "learning_rate": 9.816655355412748e-05,
+      "loss": 0.6169,
+      "step": 7934
+    },
+    {
+      "epoch": 1.4154847917224154,
+      "grad_norm": 0.47443145513534546,
+      "learning_rate": 9.81108987648516e-05,
+      "loss": 0.725,
+      "step": 7935
+    },
+    {
+      "epoch": 1.4156631879404156,
+      "grad_norm": 0.5204434990882874,
+      "learning_rate": 9.805525590488037e-05,
+      "loss": 0.7623,
+      "step": 7936
+    },
+    {
+      "epoch": 1.4158415841584158,
+      "grad_norm": 0.4761342704296112,
+      "learning_rate": 9.799962497858387e-05,
+      "loss": 0.7241,
+      "step": 7937
+    },
+    {
+      "epoch": 1.416019980376416,
+      "grad_norm": 0.4414876401424408,
+      "learning_rate": 9.794400599033146e-05,
+      "loss": 0.5095,
+      "step": 7938
+    },
+    {
+      "epoch": 1.4161983765944162,
+      "grad_norm": 0.543932318687439,
+      "learning_rate": 9.788839894449134e-05,
+      "loss": 0.692,
+      "step": 7939
+    },
+    {
+      "epoch": 1.4163767728124164,
+      "grad_norm": 0.5136645436286926,
+      "learning_rate": 9.783280384543097e-05,
+      "loss": 0.6467,
+      "step": 7940
+    },
+    {
+      "epoch": 1.4165551690304166,
+      "grad_norm": 0.5013181567192078,
+      "learning_rate": 9.777722069751663e-05,
+      "loss": 0.6199,
+      "step": 7941
+    },
+    {
+      "epoch": 1.4167335652484168,
+      "grad_norm": 0.4954874813556671,
+      "learning_rate": 9.772164950511386e-05,
+      "loss": 0.4462,
+      "step": 7942
+    },
+    {
+      "epoch": 1.416911961466417,
+      "grad_norm": 0.42616045475006104,
+      "learning_rate": 9.76660902725873e-05,
+      "loss": 0.4236,
+      "step": 7943
+    },
+    {
+      "epoch": 1.4170903576844172,
+      "grad_norm": 0.4762613773345947,
+      "learning_rate": 9.761054300430036e-05,
+      "loss": 0.6666,
+      "step": 7944
+    },
+    {
+      "epoch": 1.4172687539024174,
+      "grad_norm": 0.43737682700157166,
+      "learning_rate": 9.75550077046159e-05,
+      "loss": 0.4711,
+      "step": 7945
+    },
+    {
+      "epoch": 1.4174471501204176,
+      "grad_norm": 0.514951229095459,
+      "learning_rate": 9.749948437789544e-05,
+      "loss": 0.5771,
+      "step": 7946
+    },
+    {
+      "epoch": 1.4176255463384178,
+      "grad_norm": 1.177257776260376,
+      "learning_rate": 9.744397302849995e-05,
+      "loss": 0.6234,
+      "step": 7947
+    },
+    {
+      "epoch": 1.4178039425564177,
+      "grad_norm": 0.4536794424057007,
+      "learning_rate": 9.738847366078912e-05,
+      "loss": 0.6181,
+      "step": 7948
+    },
+    {
+      "epoch": 1.417982338774418,
+      "grad_norm": 0.6050822138786316,
+      "learning_rate": 9.733298627912185e-05,
+      "loss": 0.6667,
+      "step": 7949
+    },
+    {
+      "epoch": 1.4181607349924181,
+      "grad_norm": 0.5121973752975464,
+      "learning_rate": 9.727751088785621e-05,
+      "loss": 0.6663,
+      "step": 7950
+    },
+    {
+      "epoch": 1.4183391312104183,
+      "grad_norm": 0.5665974020957947,
+      "learning_rate": 9.722204749134908e-05,
+      "loss": 0.634,
+      "step": 7951
+    },
+    {
+      "epoch": 1.4185175274284185,
+      "grad_norm": 0.5509992241859436,
+      "learning_rate": 9.716659609395665e-05,
+      "loss": 0.6952,
+      "step": 7952
+    },
+    {
+      "epoch": 1.4186959236464187,
+      "grad_norm": 0.4740985035896301,
+      "learning_rate": 9.711115670003393e-05,
+      "loss": 0.5432,
+      "step": 7953
+    },
+    {
+      "epoch": 1.418874319864419,
+      "grad_norm": 0.4883805513381958,
+      "learning_rate": 9.705572931393525e-05,
+      "loss": 0.5195,
+      "step": 7954
+    },
+    {
+      "epoch": 1.4190527160824191,
+      "grad_norm": 0.4700393080711365,
+      "learning_rate": 9.700031394001366e-05,
+      "loss": 0.5883,
+      "step": 7955
+    },
+    {
+      "epoch": 1.419231112300419,
+      "grad_norm": 0.46181434392929077,
+      "learning_rate": 9.694491058262162e-05,
+      "loss": 0.5271,
+      "step": 7956
+    },
+    {
+      "epoch": 1.4194095085184193,
+      "grad_norm": 0.48557132482528687,
+      "learning_rate": 9.688951924611048e-05,
+      "loss": 0.5342,
+      "step": 7957
+    },
+    {
+      "epoch": 1.4195879047364195,
+      "grad_norm": 0.4598628878593445,
+      "learning_rate": 9.683413993483053e-05,
+      "loss": 0.4381,
+      "step": 7958
+    },
+    {
+      "epoch": 1.4197663009544197,
+      "grad_norm": 0.7683700323104858,
+      "learning_rate": 9.677877265313143e-05,
+      "loss": 0.5352,
+      "step": 7959
+    },
+    {
+      "epoch": 1.41994469717242,
+      "grad_norm": 0.4990261197090149,
+      "learning_rate": 9.67234174053615e-05,
+      "loss": 0.5855,
+      "step": 7960
+    },
+    {
+      "epoch": 1.42012309339042,
+      "grad_norm": 0.46791011095046997,
+      "learning_rate": 9.666807419586849e-05,
+      "loss": 0.4852,
+      "step": 7961
+    },
+    {
+      "epoch": 1.4203014896084203,
+      "grad_norm": 0.5144142508506775,
+      "learning_rate": 9.661274302899891e-05,
+      "loss": 0.5664,
+      "step": 7962
+    },
+    {
+      "epoch": 1.4204798858264205,
+      "grad_norm": 0.4881432354450226,
+      "learning_rate": 9.655742390909845e-05,
+      "loss": 0.4927,
+      "step": 7963
+    },
+    {
+      "epoch": 1.4206582820444207,
+      "grad_norm": 0.4858154356479645,
+      "learning_rate": 9.650211684051193e-05,
+      "loss": 0.5022,
+      "step": 7964
+    },
+    {
+      "epoch": 1.4208366782624209,
+      "grad_norm": 0.4513051509857178,
+      "learning_rate": 9.644682182758304e-05,
+      "loss": 0.4369,
+      "step": 7965
+    },
+    {
+      "epoch": 1.421015074480421,
+      "grad_norm": 0.6125142574310303,
+      "learning_rate": 9.639153887465477e-05,
+      "loss": 0.6945,
+      "step": 7966
+    },
+    {
+      "epoch": 1.4211934706984213,
+      "grad_norm": 0.45510220527648926,
+      "learning_rate": 9.633626798606885e-05,
+      "loss": 0.5108,
+      "step": 7967
+    },
+    {
+      "epoch": 1.4213718669164215,
+      "grad_norm": 0.5278647541999817,
+      "learning_rate": 9.628100916616638e-05,
+      "loss": 0.8597,
+      "step": 7968
+    },
+    {
+      "epoch": 1.4215502631344217,
+      "grad_norm": 0.4560278654098511,
+      "learning_rate": 9.622576241928733e-05,
+      "loss": 0.5682,
+      "step": 7969
+    },
+    {
+      "epoch": 1.4217286593524217,
+      "grad_norm": 0.49285340309143066,
+      "learning_rate": 9.617052774977061e-05,
+      "loss": 0.522,
+      "step": 7970
+    },
+    {
+      "epoch": 1.4219070555704219,
+      "grad_norm": 0.4475046992301941,
+      "learning_rate": 9.611530516195454e-05,
+      "loss": 0.493,
+      "step": 7971
+    },
+    {
+      "epoch": 1.422085451788422,
+      "grad_norm": 0.46044278144836426,
+      "learning_rate": 9.606009466017602e-05,
+      "loss": 0.5749,
+      "step": 7972
+    },
+    {
+      "epoch": 1.4222638480064222,
+      "grad_norm": 0.53668612241745,
+      "learning_rate": 9.600489624877157e-05,
+      "loss": 0.562,
+      "step": 7973
+    },
+    {
+      "epoch": 1.4224422442244224,
+      "grad_norm": 0.4615837335586548,
+      "learning_rate": 9.59497099320763e-05,
+      "loss": 0.5402,
+      "step": 7974
+    },
+    {
+      "epoch": 1.4226206404424226,
+      "grad_norm": 0.531973659992218,
+      "learning_rate": 9.589453571442444e-05,
+      "loss": 0.7516,
+      "step": 7975
+    },
+    {
+      "epoch": 1.4227990366604228,
+      "grad_norm": 0.5027945637702942,
+      "learning_rate": 9.583937360014952e-05,
+      "loss": 0.6668,
+      "step": 7976
+    },
+    {
+      "epoch": 1.422977432878423,
+      "grad_norm": 0.49252915382385254,
+      "learning_rate": 9.578422359358377e-05,
+      "loss": 0.624,
+      "step": 7977
+    },
+    {
+      "epoch": 1.423155829096423,
+      "grad_norm": 0.4632330536842346,
+      "learning_rate": 9.572908569905883e-05,
+      "loss": 0.6206,
+      "step": 7978
+    },
+    {
+      "epoch": 1.4233342253144232,
+      "grad_norm": 0.46430516242980957,
+      "learning_rate": 9.5673959920905e-05,
+      "loss": 0.5605,
+      "step": 7979
+    },
+    {
+      "epoch": 1.4235126215324234,
+      "grad_norm": 0.5031800270080566,
+      "learning_rate": 9.561884626345205e-05,
+      "loss": 0.541,
+      "step": 7980
+    },
+    {
+      "epoch": 1.4236910177504236,
+      "grad_norm": 0.5108554363250732,
+      "learning_rate": 9.556374473102839e-05,
+      "loss": 0.5613,
+      "step": 7981
+    },
+    {
+      "epoch": 1.4238694139684238,
+      "grad_norm": 0.5381075143814087,
+      "learning_rate": 9.550865532796185e-05,
+      "loss": 0.635,
+      "step": 7982
+    },
+    {
+      "epoch": 1.424047810186424,
+      "grad_norm": 0.522135317325592,
+      "learning_rate": 9.545357805857901e-05,
+      "loss": 0.5884,
+      "step": 7983
+    },
+    {
+      "epoch": 1.4242262064044242,
+      "grad_norm": 0.5213260650634766,
+      "learning_rate": 9.539851292720562e-05,
+      "loss": 0.6214,
+      "step": 7984
+    },
+    {
+      "epoch": 1.4244046026224244,
+      "grad_norm": 0.4458264112472534,
+      "learning_rate": 9.53434599381665e-05,
+      "loss": 0.47,
+      "step": 7985
+    },
+    {
+      "epoch": 1.4245829988404246,
+      "grad_norm": 0.543368935585022,
+      "learning_rate": 9.528841909578545e-05,
+      "loss": 0.6638,
+      "step": 7986
+    },
+    {
+      "epoch": 1.4247613950584248,
+      "grad_norm": 0.5813699960708618,
+      "learning_rate": 9.523339040438536e-05,
+      "loss": 0.7685,
+      "step": 7987
+    },
+    {
+      "epoch": 1.424939791276425,
+      "grad_norm": 0.5484562516212463,
+      "learning_rate": 9.517837386828829e-05,
+      "loss": 0.6551,
+      "step": 7988
+    },
+    {
+      "epoch": 1.4251181874944252,
+      "grad_norm": 0.5792138576507568,
+      "learning_rate": 9.512336949181502e-05,
+      "loss": 0.6305,
+      "step": 7989
+    },
+    {
+      "epoch": 1.4252965837124254,
+      "grad_norm": 0.5417539477348328,
+      "learning_rate": 9.506837727928577e-05,
+      "loss": 0.7499,
+      "step": 7990
+    },
+    {
+      "epoch": 1.4254749799304256,
+      "grad_norm": 0.49307334423065186,
+      "learning_rate": 9.501339723501937e-05,
+      "loss": 0.5378,
+      "step": 7991
+    },
+    {
+      "epoch": 1.4256533761484256,
+      "grad_norm": 0.44642525911331177,
+      "learning_rate": 9.495842936333415e-05,
+      "loss": 0.4978,
+      "step": 7992
+    },
+    {
+      "epoch": 1.4258317723664258,
+      "grad_norm": 0.5179886817932129,
+      "learning_rate": 9.49034736685471e-05,
+      "loss": 0.6413,
+      "step": 7993
+    },
+    {
+      "epoch": 1.426010168584426,
+      "grad_norm": 0.48655420541763306,
+      "learning_rate": 9.484853015497458e-05,
+      "loss": 0.5178,
+      "step": 7994
+    },
+    {
+      "epoch": 1.4261885648024262,
+      "grad_norm": 0.4873103201389313,
+      "learning_rate": 9.47935988269317e-05,
+      "loss": 0.5396,
+      "step": 7995
+    },
+    {
+      "epoch": 1.4263669610204264,
+      "grad_norm": 0.49392032623291016,
+      "learning_rate": 9.47386796887327e-05,
+      "loss": 0.5208,
+      "step": 7996
+    },
+    {
+      "epoch": 1.4265453572384266,
+      "grad_norm": 0.5058830380439758,
+      "learning_rate": 9.468377274469109e-05,
+      "loss": 0.5961,
+      "step": 7997
+    },
+    {
+      "epoch": 1.4267237534564268,
+      "grad_norm": 0.4766574800014496,
+      "learning_rate": 9.462887799911904e-05,
+      "loss": 0.3617,
+      "step": 7998
+    },
+    {
+      "epoch": 1.426902149674427,
+      "grad_norm": 0.44295012950897217,
+      "learning_rate": 9.457399545632814e-05,
+      "loss": 0.5054,
+      "step": 7999
+    },
+    {
+      "epoch": 1.427080545892427,
+      "grad_norm": 0.5294760465621948,
+      "learning_rate": 9.451912512062863e-05,
+      "loss": 0.6995,
+      "step": 8000
+    },
+    {
+      "epoch": 1.4272589421104271,
+      "grad_norm": 0.46700701117515564,
+      "learning_rate": 9.446426699633023e-05,
+      "loss": 0.613,
+      "step": 8001
+    },
+    {
+      "epoch": 1.4274373383284273,
+      "grad_norm": 0.5137273669242859,
+      "learning_rate": 9.44094210877413e-05,
+      "loss": 0.7029,
+      "step": 8002
+    },
+    {
+      "epoch": 1.4276157345464275,
+      "grad_norm": 0.5057253837585449,
+      "learning_rate": 9.435458739916946e-05,
+      "loss": 0.6471,
+      "step": 8003
+    },
+    {
+      "epoch": 1.4277941307644277,
+      "grad_norm": 0.5789628028869629,
+      "learning_rate": 9.429976593492146e-05,
+      "loss": 0.7758,
+      "step": 8004
+    },
+    {
+      "epoch": 1.427972526982428,
+      "grad_norm": 0.5334276556968689,
+      "learning_rate": 9.424495669930272e-05,
+      "loss": 0.6382,
+      "step": 8005
+    },
+    {
+      "epoch": 1.4281509232004281,
+      "grad_norm": 0.5051296949386597,
+      "learning_rate": 9.419015969661814e-05,
+      "loss": 0.6057,
+      "step": 8006
+    },
+    {
+      "epoch": 1.4283293194184283,
+      "grad_norm": 0.5126160979270935,
+      "learning_rate": 9.413537493117142e-05,
+      "loss": 0.7181,
+      "step": 8007
+    },
+    {
+      "epoch": 1.4285077156364285,
+      "grad_norm": 0.5015937089920044,
+      "learning_rate": 9.408060240726515e-05,
+      "loss": 0.5093,
+      "step": 8008
+    },
+    {
+      "epoch": 1.4286861118544287,
+      "grad_norm": 0.5069268941879272,
+      "learning_rate": 9.402584212920134e-05,
+      "loss": 0.5187,
+      "step": 8009
+    },
+    {
+      "epoch": 1.428864508072429,
+      "grad_norm": 0.5026482343673706,
+      "learning_rate": 9.397109410128071e-05,
+      "loss": 0.5594,
+      "step": 8010
+    },
+    {
+      "epoch": 1.4290429042904291,
+      "grad_norm": 0.4498440623283386,
+      "learning_rate": 9.391635832780329e-05,
+      "loss": 0.4379,
+      "step": 8011
+    },
+    {
+      "epoch": 1.4292213005084293,
+      "grad_norm": 0.494722843170166,
+      "learning_rate": 9.386163481306784e-05,
+      "loss": 0.5101,
+      "step": 8012
+    },
+    {
+      "epoch": 1.4293996967264295,
+      "grad_norm": 0.4928201138973236,
+      "learning_rate": 9.380692356137247e-05,
+      "loss": 0.5785,
+      "step": 8013
+    },
+    {
+      "epoch": 1.4295780929444295,
+      "grad_norm": 0.49480581283569336,
+      "learning_rate": 9.375222457701401e-05,
+      "loss": 0.6379,
+      "step": 8014
+    },
+    {
+      "epoch": 1.4297564891624297,
+      "grad_norm": 0.46592390537261963,
+      "learning_rate": 9.369753786428869e-05,
+      "loss": 0.5308,
+      "step": 8015
+    },
+    {
+      "epoch": 1.42993488538043,
+      "grad_norm": 0.5155026912689209,
+      "learning_rate": 9.364286342749151e-05,
+      "loss": 0.6286,
+      "step": 8016
+    },
+    {
+      "epoch": 1.43011328159843,
+      "grad_norm": 0.4269189238548279,
+      "learning_rate": 9.358820127091636e-05,
+      "loss": 0.4702,
+      "step": 8017
+    },
+    {
+      "epoch": 1.4302916778164303,
+      "grad_norm": 0.4979037344455719,
+      "learning_rate": 9.353355139885672e-05,
+      "loss": 0.5917,
+      "step": 8018
+    },
+    {
+      "epoch": 1.4304700740344305,
+      "grad_norm": 0.43645283579826355,
+      "learning_rate": 9.347891381560455e-05,
+      "loss": 0.5333,
+      "step": 8019
+    },
+    {
+      "epoch": 1.4306484702524307,
+      "grad_norm": 0.47989028692245483,
+      "learning_rate": 9.342428852545123e-05,
+      "loss": 0.6838,
+      "step": 8020
+    },
+    {
+      "epoch": 1.430826866470431,
+      "grad_norm": 0.41572171449661255,
+      "learning_rate": 9.336967553268691e-05,
+      "loss": 0.3637,
+      "step": 8021
+    },
+    {
+      "epoch": 1.4310052626884309,
+      "grad_norm": 0.5728147029876709,
+      "learning_rate": 9.33150748416008e-05,
+      "loss": 0.7151,
+      "step": 8022
+    },
+    {
+      "epoch": 1.431183658906431,
+      "grad_norm": 0.48712360858917236,
+      "learning_rate": 9.326048645648134e-05,
+      "loss": 0.5196,
+      "step": 8023
+    },
+    {
+      "epoch": 1.4313620551244313,
+      "grad_norm": 0.44054681062698364,
+      "learning_rate": 9.320591038161574e-05,
+      "loss": 0.4682,
+      "step": 8024
+    },
+    {
+      "epoch": 1.4315404513424315,
+      "grad_norm": 0.48739486932754517,
+      "learning_rate": 9.315134662129058e-05,
+      "loss": 0.5445,
+      "step": 8025
+    },
+    {
+      "epoch": 1.4317188475604317,
+      "grad_norm": 0.44029682874679565,
+      "learning_rate": 9.309679517979102e-05,
+      "loss": 0.4334,
+      "step": 8026
+    },
+    {
+      "epoch": 1.4318972437784319,
+      "grad_norm": 0.47833380103111267,
+      "learning_rate": 9.304225606140176e-05,
+      "loss": 0.4728,
+      "step": 8027
+    },
+    {
+      "epoch": 1.432075639996432,
+      "grad_norm": 0.5727766156196594,
+      "learning_rate": 9.298772927040618e-05,
+      "loss": 0.7175,
+      "step": 8028
+    },
+    {
+      "epoch": 1.4322540362144323,
+      "grad_norm": 0.4518079161643982,
+      "learning_rate": 9.293321481108668e-05,
+      "loss": 0.5195,
+      "step": 8029
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 0.526343584060669,
+      "learning_rate": 9.2878712687725e-05,
+      "loss": 0.588,
+      "step": 8030
+    },
+    {
+      "epoch": 1.4326108286504327,
+      "grad_norm": 0.4975213408470154,
+      "learning_rate": 9.282422290460149e-05,
+      "loss": 0.6872,
+      "step": 8031
+    },
+    {
+      "epoch": 1.4327892248684329,
+      "grad_norm": 0.5089072585105896,
+      "learning_rate": 9.276974546599599e-05,
+      "loss": 0.6198,
+      "step": 8032
+    },
+    {
+      "epoch": 1.432967621086433,
+      "grad_norm": 0.4496549665927887,
+      "learning_rate": 9.27152803761869e-05,
+      "loss": 0.501,
+      "step": 8033
+    },
+    {
+      "epoch": 1.4331460173044333,
+      "grad_norm": 0.5993954539299011,
+      "learning_rate": 9.266082763945202e-05,
+      "loss": 0.7204,
+      "step": 8034
+    },
+    {
+      "epoch": 1.4333244135224334,
+      "grad_norm": 0.5009292960166931,
+      "learning_rate": 9.260638726006812e-05,
+      "loss": 0.5112,
+      "step": 8035
+    },
+    {
+      "epoch": 1.4335028097404334,
+      "grad_norm": 0.49264848232269287,
+      "learning_rate": 9.255195924231075e-05,
+      "loss": 0.5595,
+      "step": 8036
+    },
+    {
+      "epoch": 1.4336812059584336,
+      "grad_norm": 0.481372594833374,
+      "learning_rate": 9.249754359045484e-05,
+      "loss": 0.5221,
+      "step": 8037
+    },
+    {
+      "epoch": 1.4338596021764338,
+      "grad_norm": 0.5572468638420105,
+      "learning_rate": 9.244314030877398e-05,
+      "loss": 0.7088,
+      "step": 8038
+    },
+    {
+      "epoch": 1.434037998394434,
+      "grad_norm": 0.4595695436000824,
+      "learning_rate": 9.238874940154116e-05,
+      "loss": 0.4761,
+      "step": 8039
+    },
+    {
+      "epoch": 1.4342163946124342,
+      "grad_norm": 0.43319255113601685,
+      "learning_rate": 9.233437087302806e-05,
+      "loss": 0.5051,
+      "step": 8040
+    },
+    {
+      "epoch": 1.4343947908304344,
+      "grad_norm": 0.4756569266319275,
+      "learning_rate": 9.228000472750569e-05,
+      "loss": 0.5404,
+      "step": 8041
+    },
+    {
+      "epoch": 1.4345731870484346,
+      "grad_norm": 0.5660195350646973,
+      "learning_rate": 9.22256509692439e-05,
+      "loss": 0.6584,
+      "step": 8042
+    },
+    {
+      "epoch": 1.4347515832664348,
+      "grad_norm": 0.5442212224006653,
+      "learning_rate": 9.21713096025115e-05,
+      "loss": 0.7664,
+      "step": 8043
+    },
+    {
+      "epoch": 1.4349299794844348,
+      "grad_norm": 0.4902845323085785,
+      "learning_rate": 9.211698063157659e-05,
+      "loss": 0.6131,
+      "step": 8044
+    },
+    {
+      "epoch": 1.435108375702435,
+      "grad_norm": 0.47630754113197327,
+      "learning_rate": 9.206266406070601e-05,
+      "loss": 0.5899,
+      "step": 8045
+    },
+    {
+      "epoch": 1.4352867719204352,
+      "grad_norm": 0.4589233696460724,
+      "learning_rate": 9.200835989416589e-05,
+      "loss": 0.5258,
+      "step": 8046
+    },
+    {
+      "epoch": 1.4354651681384354,
+      "grad_norm": 0.5233177542686462,
+      "learning_rate": 9.195406813622115e-05,
+      "loss": 0.6126,
+      "step": 8047
+    },
+    {
+      "epoch": 1.4356435643564356,
+      "grad_norm": 0.475238561630249,
+      "learning_rate": 9.189978879113587e-05,
+      "loss": 0.6048,
+      "step": 8048
+    },
+    {
+      "epoch": 1.4358219605744358,
+      "grad_norm": 0.47865399718284607,
+      "learning_rate": 9.184552186317321e-05,
+      "loss": 0.6042,
+      "step": 8049
+    },
+    {
+      "epoch": 1.436000356792436,
+      "grad_norm": 0.49072253704071045,
+      "learning_rate": 9.179126735659513e-05,
+      "loss": 0.4852,
+      "step": 8050
+    },
+    {
+      "epoch": 1.4361787530104362,
+      "grad_norm": 0.47413280606269836,
+      "learning_rate": 9.173702527566292e-05,
+      "loss": 0.4936,
+      "step": 8051
+    },
+    {
+      "epoch": 1.4363571492284364,
+      "grad_norm": 0.5141640901565552,
+      "learning_rate": 9.168279562463655e-05,
+      "loss": 0.6562,
+      "step": 8052
+    },
+    {
+      "epoch": 1.4365355454464366,
+      "grad_norm": 0.5093448758125305,
+      "learning_rate": 9.162857840777535e-05,
+      "loss": 0.7238,
+      "step": 8053
+    },
+    {
+      "epoch": 1.4367139416644368,
+      "grad_norm": 0.45524322986602783,
+      "learning_rate": 9.157437362933749e-05,
+      "loss": 0.5362,
+      "step": 8054
+    },
+    {
+      "epoch": 1.436892337882437,
+      "grad_norm": 0.479604572057724,
+      "learning_rate": 9.152018129358003e-05,
+      "loss": 0.4765,
+      "step": 8055
+    },
+    {
+      "epoch": 1.4370707341004372,
+      "grad_norm": 0.4629632532596588,
+      "learning_rate": 9.146600140475944e-05,
+      "loss": 0.5064,
+      "step": 8056
+    },
+    {
+      "epoch": 1.4372491303184374,
+      "grad_norm": 0.595065176486969,
+      "learning_rate": 9.141183396713077e-05,
+      "loss": 0.7567,
+      "step": 8057
+    },
+    {
+      "epoch": 1.4374275265364373,
+      "grad_norm": 0.47794783115386963,
+      "learning_rate": 9.13576789849485e-05,
+      "loss": 0.4937,
+      "step": 8058
+    },
+    {
+      "epoch": 1.4376059227544375,
+      "grad_norm": 0.5082597732543945,
+      "learning_rate": 9.130353646246578e-05,
+      "loss": 0.7144,
+      "step": 8059
+    },
+    {
+      "epoch": 1.4377843189724377,
+      "grad_norm": 0.46265289187431335,
+      "learning_rate": 9.124940640393512e-05,
+      "loss": 0.5559,
+      "step": 8060
+    },
+    {
+      "epoch": 1.437962715190438,
+      "grad_norm": 0.5391396880149841,
+      "learning_rate": 9.119528881360764e-05,
+      "loss": 0.4127,
+      "step": 8061
+    },
+    {
+      "epoch": 1.4381411114084381,
+      "grad_norm": 0.4582943618297577,
+      "learning_rate": 9.114118369573393e-05,
+      "loss": 0.3894,
+      "step": 8062
+    },
+    {
+      "epoch": 1.4383195076264383,
+      "grad_norm": 0.460657000541687,
+      "learning_rate": 9.108709105456323e-05,
+      "loss": 0.4666,
+      "step": 8063
+    },
+    {
+      "epoch": 1.4384979038444385,
+      "grad_norm": 0.5155767798423767,
+      "learning_rate": 9.103301089434399e-05,
+      "loss": 0.673,
+      "step": 8064
+    },
+    {
+      "epoch": 1.4386763000624387,
+      "grad_norm": 0.41279855370521545,
+      "learning_rate": 9.097894321932377e-05,
+      "loss": 0.3958,
+      "step": 8065
+    },
+    {
+      "epoch": 1.4388546962804387,
+      "grad_norm": 0.4912382662296295,
+      "learning_rate": 9.09248880337489e-05,
+      "loss": 0.6629,
+      "step": 8066
+    },
+    {
+      "epoch": 1.439033092498439,
+      "grad_norm": 0.5284416675567627,
+      "learning_rate": 9.087084534186476e-05,
+      "loss": 0.6543,
+      "step": 8067
+    },
+    {
+      "epoch": 1.4392114887164391,
+      "grad_norm": 0.49840492010116577,
+      "learning_rate": 9.081681514791609e-05,
+      "loss": 0.65,
+      "step": 8068
+    },
+    {
+      "epoch": 1.4393898849344393,
+      "grad_norm": 0.4940994679927826,
+      "learning_rate": 9.076279745614613e-05,
+      "loss": 0.5973,
+      "step": 8069
+    },
+    {
+      "epoch": 1.4395682811524395,
+      "grad_norm": 0.49154722690582275,
+      "learning_rate": 9.070879227079765e-05,
+      "loss": 0.6159,
+      "step": 8070
+    },
+    {
+      "epoch": 1.4397466773704397,
+      "grad_norm": 0.4630894660949707,
+      "learning_rate": 9.065479959611194e-05,
+      "loss": 0.5687,
+      "step": 8071
+    },
+    {
+      "epoch": 1.43992507358844,
+      "grad_norm": 0.7224597930908203,
+      "learning_rate": 9.060081943632983e-05,
+      "loss": 0.499,
+      "step": 8072
+    },
+    {
+      "epoch": 1.44010346980644,
+      "grad_norm": 0.47265711426734924,
+      "learning_rate": 9.054685179569066e-05,
+      "loss": 0.5876,
+      "step": 8073
+    },
+    {
+      "epoch": 1.4402818660244403,
+      "grad_norm": 0.5103265643119812,
+      "learning_rate": 9.049289667843325e-05,
+      "loss": 0.7843,
+      "step": 8074
+    },
+    {
+      "epoch": 1.4404602622424405,
+      "grad_norm": 0.5575939416885376,
+      "learning_rate": 9.043895408879505e-05,
+      "loss": 0.7539,
+      "step": 8075
+    },
+    {
+      "epoch": 1.4406386584604407,
+      "grad_norm": 0.46837523579597473,
+      "learning_rate": 9.038502403101268e-05,
+      "loss": 0.5082,
+      "step": 8076
+    },
+    {
+      "epoch": 1.440817054678441,
+      "grad_norm": 0.5041645169258118,
+      "learning_rate": 9.033110650932188e-05,
+      "loss": 0.635,
+      "step": 8077
+    },
+    {
+      "epoch": 1.440995450896441,
+      "grad_norm": 0.4759989380836487,
+      "learning_rate": 9.027720152795721e-05,
+      "loss": 0.4989,
+      "step": 8078
+    },
+    {
+      "epoch": 1.4411738471144413,
+      "grad_norm": 0.49040788412094116,
+      "learning_rate": 9.022330909115239e-05,
+      "loss": 0.5498,
+      "step": 8079
+    },
+    {
+      "epoch": 1.4413522433324413,
+      "grad_norm": 0.44871506094932556,
+      "learning_rate": 9.01694292031402e-05,
+      "loss": 0.5627,
+      "step": 8080
+    },
+    {
+      "epoch": 1.4415306395504415,
+      "grad_norm": 0.45151621103286743,
+      "learning_rate": 9.011556186815217e-05,
+      "loss": 0.4521,
+      "step": 8081
+    },
+    {
+      "epoch": 1.4417090357684417,
+      "grad_norm": 0.5003564953804016,
+      "learning_rate": 9.006170709041922e-05,
+      "loss": 0.5973,
+      "step": 8082
+    },
+    {
+      "epoch": 1.4418874319864419,
+      "grad_norm": 0.559359073638916,
+      "learning_rate": 9.000786487417084e-05,
+      "loss": 0.65,
+      "step": 8083
+    },
+    {
+      "epoch": 1.442065828204442,
+      "grad_norm": 0.48790860176086426,
+      "learning_rate": 8.995403522363602e-05,
+      "loss": 0.5723,
+      "step": 8084
+    },
+    {
+      "epoch": 1.4422442244224423,
+      "grad_norm": 0.5341598391532898,
+      "learning_rate": 8.99002181430423e-05,
+      "loss": 0.7464,
+      "step": 8085
+    },
+    {
+      "epoch": 1.4424226206404425,
+      "grad_norm": 0.5069512128829956,
+      "learning_rate": 8.984641363661666e-05,
+      "loss": 0.7237,
+      "step": 8086
+    },
+    {
+      "epoch": 1.4426010168584427,
+      "grad_norm": 0.5093790888786316,
+      "learning_rate": 8.979262170858474e-05,
+      "loss": 0.6849,
+      "step": 8087
+    },
+    {
+      "epoch": 1.4427794130764429,
+      "grad_norm": 0.4905139207839966,
+      "learning_rate": 8.973884236317131e-05,
+      "loss": 0.6719,
+      "step": 8088
+    },
+    {
+      "epoch": 1.4429578092944428,
+      "grad_norm": 0.5070201754570007,
+      "learning_rate": 8.968507560460029e-05,
+      "loss": 0.6181,
+      "step": 8089
+    },
+    {
+      "epoch": 1.443136205512443,
+      "grad_norm": 0.47273147106170654,
+      "learning_rate": 8.963132143709437e-05,
+      "loss": 0.5333,
+      "step": 8090
+    },
+    {
+      "epoch": 1.4433146017304432,
+      "grad_norm": 0.5406144261360168,
+      "learning_rate": 8.957757986487556e-05,
+      "loss": 0.6466,
+      "step": 8091
+    },
+    {
+      "epoch": 1.4434929979484434,
+      "grad_norm": 0.506829023361206,
+      "learning_rate": 8.95238508921645e-05,
+      "loss": 0.5386,
+      "step": 8092
+    },
+    {
+      "epoch": 1.4436713941664436,
+      "grad_norm": 0.515042781829834,
+      "learning_rate": 8.94701345231812e-05,
+      "loss": 0.6152,
+      "step": 8093
+    },
+    {
+      "epoch": 1.4438497903844438,
+      "grad_norm": 0.48475679755210876,
+      "learning_rate": 8.941643076214436e-05,
+      "loss": 0.6374,
+      "step": 8094
+    },
+    {
+      "epoch": 1.444028186602444,
+      "grad_norm": 0.5149679780006409,
+      "learning_rate": 8.936273961327198e-05,
+      "loss": 0.6776,
+      "step": 8095
+    },
+    {
+      "epoch": 1.4442065828204442,
+      "grad_norm": 0.610506534576416,
+      "learning_rate": 8.930906108078096e-05,
+      "loss": 0.5515,
+      "step": 8096
+    },
+    {
+      "epoch": 1.4443849790384444,
+      "grad_norm": 0.4772319793701172,
+      "learning_rate": 8.925539516888706e-05,
+      "loss": 0.5777,
+      "step": 8097
+    },
+    {
+      "epoch": 1.4445633752564446,
+      "grad_norm": 0.42948904633522034,
+      "learning_rate": 8.920174188180533e-05,
+      "loss": 0.4874,
+      "step": 8098
+    },
+    {
+      "epoch": 1.4447417714744448,
+      "grad_norm": 0.44335153698921204,
+      "learning_rate": 8.91481012237495e-05,
+      "loss": 0.3909,
+      "step": 8099
+    },
+    {
+      "epoch": 1.444920167692445,
+      "grad_norm": 0.5608864426612854,
+      "learning_rate": 8.909447319893269e-05,
+      "loss": 0.7402,
+      "step": 8100
+    },
+    {
+      "epoch": 1.4450985639104452,
+      "grad_norm": 0.5135005116462708,
+      "learning_rate": 8.904085781156671e-05,
+      "loss": 0.6809,
+      "step": 8101
+    },
+    {
+      "epoch": 1.4452769601284452,
+      "grad_norm": 0.5145747661590576,
+      "learning_rate": 8.898725506586239e-05,
+      "loss": 0.6553,
+      "step": 8102
+    },
+    {
+      "epoch": 1.4454553563464454,
+      "grad_norm": 0.46137893199920654,
+      "learning_rate": 8.893366496602984e-05,
+      "loss": 0.4671,
+      "step": 8103
+    },
+    {
+      "epoch": 1.4456337525644456,
+      "grad_norm": 0.47708660364151,
+      "learning_rate": 8.888008751627788e-05,
+      "loss": 0.4231,
+      "step": 8104
+    },
+    {
+      "epoch": 1.4458121487824458,
+      "grad_norm": 0.5265963077545166,
+      "learning_rate": 8.882652272081457e-05,
+      "loss": 0.6085,
+      "step": 8105
+    },
+    {
+      "epoch": 1.445990545000446,
+      "grad_norm": 0.5075244307518005,
+      "learning_rate": 8.877297058384673e-05,
+      "loss": 0.53,
+      "step": 8106
+    },
+    {
+      "epoch": 1.4461689412184462,
+      "grad_norm": 0.504497230052948,
+      "learning_rate": 8.871943110958048e-05,
+      "loss": 0.5873,
+      "step": 8107
+    },
+    {
+      "epoch": 1.4463473374364464,
+      "grad_norm": 0.5384894013404846,
+      "learning_rate": 8.866590430222072e-05,
+      "loss": 0.542,
+      "step": 8108
+    },
+    {
+      "epoch": 1.4465257336544466,
+      "grad_norm": 0.4829096496105194,
+      "learning_rate": 8.861239016597123e-05,
+      "loss": 0.5085,
+      "step": 8109
+    },
+    {
+      "epoch": 1.4467041298724468,
+      "grad_norm": 0.5008237361907959,
+      "learning_rate": 8.855888870503535e-05,
+      "loss": 0.5474,
+      "step": 8110
+    },
+    {
+      "epoch": 1.4468825260904468,
+      "grad_norm": 0.4817045331001282,
+      "learning_rate": 8.850539992361475e-05,
+      "loss": 0.4752,
+      "step": 8111
+    },
+    {
+      "epoch": 1.447060922308447,
+      "grad_norm": 0.5164246559143066,
+      "learning_rate": 8.845192382591067e-05,
+      "loss": 0.6815,
+      "step": 8112
+    },
+    {
+      "epoch": 1.4472393185264472,
+      "grad_norm": 0.4688229262828827,
+      "learning_rate": 8.839846041612295e-05,
+      "loss": 0.5217,
+      "step": 8113
+    },
+    {
+      "epoch": 1.4474177147444474,
+      "grad_norm": 0.4715273082256317,
+      "learning_rate": 8.834500969845052e-05,
+      "loss": 0.5708,
+      "step": 8114
+    },
+    {
+      "epoch": 1.4475961109624476,
+      "grad_norm": 0.42476388812065125,
+      "learning_rate": 8.829157167709157e-05,
+      "loss": 0.4669,
+      "step": 8115
+    },
+    {
+      "epoch": 1.4477745071804478,
+      "grad_norm": 0.5047100782394409,
+      "learning_rate": 8.823814635624288e-05,
+      "loss": 0.7172,
+      "step": 8116
+    },
+    {
+      "epoch": 1.447952903398448,
+      "grad_norm": 0.49228885769844055,
+      "learning_rate": 8.81847337401007e-05,
+      "loss": 0.5684,
+      "step": 8117
+    },
+    {
+      "epoch": 1.4481312996164482,
+      "grad_norm": 0.5029509663581848,
+      "learning_rate": 8.813133383285977e-05,
+      "loss": 0.6608,
+      "step": 8118
+    },
+    {
+      "epoch": 1.4483096958344484,
+      "grad_norm": 0.5024793148040771,
+      "learning_rate": 8.807794663871429e-05,
+      "loss": 0.6344,
+      "step": 8119
+    },
+    {
+      "epoch": 1.4484880920524486,
+      "grad_norm": 0.49262115359306335,
+      "learning_rate": 8.802457216185717e-05,
+      "loss": 0.6377,
+      "step": 8120
+    },
+    {
+      "epoch": 1.4486664882704487,
+      "grad_norm": 0.43594643473625183,
+      "learning_rate": 8.797121040648049e-05,
+      "loss": 0.527,
+      "step": 8121
+    },
+    {
+      "epoch": 1.448844884488449,
+      "grad_norm": 0.4278947114944458,
+      "learning_rate": 8.791786137677524e-05,
+      "loss": 0.5052,
+      "step": 8122
+    },
+    {
+      "epoch": 1.4490232807064491,
+      "grad_norm": 0.4548480212688446,
+      "learning_rate": 8.78645250769313e-05,
+      "loss": 0.4939,
+      "step": 8123
+    },
+    {
+      "epoch": 1.4492016769244491,
+      "grad_norm": 0.5257616639137268,
+      "learning_rate": 8.781120151113788e-05,
+      "loss": 0.6517,
+      "step": 8124
+    },
+    {
+      "epoch": 1.4493800731424493,
+      "grad_norm": 0.5141300559043884,
+      "learning_rate": 8.775789068358283e-05,
+      "loss": 0.5612,
+      "step": 8125
+    },
+    {
+      "epoch": 1.4495584693604495,
+      "grad_norm": 0.464578241109848,
+      "learning_rate": 8.770459259845323e-05,
+      "loss": 0.4524,
+      "step": 8126
+    },
+    {
+      "epoch": 1.4497368655784497,
+      "grad_norm": 0.48500585556030273,
+      "learning_rate": 8.765130725993514e-05,
+      "loss": 0.6593,
+      "step": 8127
+    },
+    {
+      "epoch": 1.44991526179645,
+      "grad_norm": 0.4844876825809479,
+      "learning_rate": 8.759803467221348e-05,
+      "loss": 0.589,
+      "step": 8128
+    },
+    {
+      "epoch": 1.4500936580144501,
+      "grad_norm": 0.5171689391136169,
+      "learning_rate": 8.754477483947232e-05,
+      "loss": 0.6696,
+      "step": 8129
+    },
+    {
+      "epoch": 1.4502720542324503,
+      "grad_norm": 0.5518413186073303,
+      "learning_rate": 8.749152776589459e-05,
+      "loss": 0.7856,
+      "step": 8130
+    },
+    {
+      "epoch": 1.4504504504504505,
+      "grad_norm": 0.4898933470249176,
+      "learning_rate": 8.74382934556624e-05,
+      "loss": 0.6053,
+      "step": 8131
+    },
+    {
+      "epoch": 1.4506288466684507,
+      "grad_norm": 0.4357226490974426,
+      "learning_rate": 8.738507191295658e-05,
+      "loss": 0.5202,
+      "step": 8132
+    },
+    {
+      "epoch": 1.4508072428864507,
+      "grad_norm": 0.5392679572105408,
+      "learning_rate": 8.733186314195734e-05,
+      "loss": 0.6614,
+      "step": 8133
+    },
+    {
+      "epoch": 1.4509856391044509,
+      "grad_norm": 0.566256582736969,
+      "learning_rate": 8.72786671468436e-05,
+      "loss": 0.7098,
+      "step": 8134
+    },
+    {
+      "epoch": 1.451164035322451,
+      "grad_norm": 0.45817068219184875,
+      "learning_rate": 8.722548393179319e-05,
+      "loss": 0.5082,
+      "step": 8135
+    },
+    {
+      "epoch": 1.4513424315404513,
+      "grad_norm": 0.4491539001464844,
+      "learning_rate": 8.717231350098331e-05,
+      "loss": 0.4243,
+      "step": 8136
+    },
+    {
+      "epoch": 1.4515208277584515,
+      "grad_norm": 0.5039847493171692,
+      "learning_rate": 8.711915585858979e-05,
+      "loss": 0.5085,
+      "step": 8137
+    },
+    {
+      "epoch": 1.4516992239764517,
+      "grad_norm": 0.5082625150680542,
+      "learning_rate": 8.706601100878778e-05,
+      "loss": 0.5393,
+      "step": 8138
+    },
+    {
+      "epoch": 1.4518776201944519,
+      "grad_norm": 0.5001698732376099,
+      "learning_rate": 8.701287895575102e-05,
+      "loss": 0.6006,
+      "step": 8139
+    },
+    {
+      "epoch": 1.452056016412452,
+      "grad_norm": 0.4417283535003662,
+      "learning_rate": 8.695975970365264e-05,
+      "loss": 0.6028,
+      "step": 8140
+    },
+    {
+      "epoch": 1.4522344126304523,
+      "grad_norm": 0.46480488777160645,
+      "learning_rate": 8.690665325666463e-05,
+      "loss": 0.5018,
+      "step": 8141
+    },
+    {
+      "epoch": 1.4524128088484525,
+      "grad_norm": 0.5661883354187012,
+      "learning_rate": 8.685355961895783e-05,
+      "loss": 0.7788,
+      "step": 8142
+    },
+    {
+      "epoch": 1.4525912050664527,
+      "grad_norm": 0.46579429507255554,
+      "learning_rate": 8.680047879470233e-05,
+      "loss": 0.5646,
+      "step": 8143
+    },
+    {
+      "epoch": 1.4527696012844529,
+      "grad_norm": 0.4672081768512726,
+      "learning_rate": 8.67474107880669e-05,
+      "loss": 0.5924,
+      "step": 8144
+    },
+    {
+      "epoch": 1.452947997502453,
+      "grad_norm": 0.4720962941646576,
+      "learning_rate": 8.669435560321968e-05,
+      "loss": 0.4921,
+      "step": 8145
+    },
+    {
+      "epoch": 1.453126393720453,
+      "grad_norm": 0.5264710783958435,
+      "learning_rate": 8.664131324432745e-05,
+      "loss": 0.6339,
+      "step": 8146
+    },
+    {
+      "epoch": 1.4533047899384532,
+      "grad_norm": 0.4442679286003113,
+      "learning_rate": 8.658828371555613e-05,
+      "loss": 0.4405,
+      "step": 8147
+    },
+    {
+      "epoch": 1.4534831861564534,
+      "grad_norm": 0.4961048364639282,
+      "learning_rate": 8.653526702107075e-05,
+      "loss": 0.5394,
+      "step": 8148
+    },
+    {
+      "epoch": 1.4536615823744536,
+      "grad_norm": 0.5059047341346741,
+      "learning_rate": 8.64822631650351e-05,
+      "loss": 0.5187,
+      "step": 8149
+    },
+    {
+      "epoch": 1.4538399785924538,
+      "grad_norm": 0.6599757671356201,
+      "learning_rate": 8.64292721516122e-05,
+      "loss": 0.6328,
+      "step": 8150
+    },
+    {
+      "epoch": 1.454018374810454,
+      "grad_norm": 0.48593562841415405,
+      "learning_rate": 8.637629398496377e-05,
+      "loss": 0.5988,
+      "step": 8151
+    },
+    {
+      "epoch": 1.4541967710284542,
+      "grad_norm": 0.5314595103263855,
+      "learning_rate": 8.632332866925091e-05,
+      "loss": 0.5746,
+      "step": 8152
+    },
+    {
+      "epoch": 1.4543751672464544,
+      "grad_norm": 0.4576443135738373,
+      "learning_rate": 8.627037620863328e-05,
+      "loss": 0.4491,
+      "step": 8153
+    },
+    {
+      "epoch": 1.4545535634644546,
+      "grad_norm": 0.4969932734966278,
+      "learning_rate": 8.621743660726994e-05,
+      "loss": 0.6862,
+      "step": 8154
+    },
+    {
+      "epoch": 1.4547319596824546,
+      "grad_norm": 0.5143519043922424,
+      "learning_rate": 8.616450986931857e-05,
+      "loss": 0.4361,
+      "step": 8155
+    },
+    {
+      "epoch": 1.4549103559004548,
+      "grad_norm": 0.498677134513855,
+      "learning_rate": 8.611159599893609e-05,
+      "loss": 0.5173,
+      "step": 8156
+    },
+    {
+      "epoch": 1.455088752118455,
+      "grad_norm": 0.47454512119293213,
+      "learning_rate": 8.605869500027838e-05,
+      "loss": 0.6723,
+      "step": 8157
+    },
+    {
+      "epoch": 1.4552671483364552,
+      "grad_norm": 0.46952202916145325,
+      "learning_rate": 8.600580687750017e-05,
+      "loss": 0.4888,
+      "step": 8158
+    },
+    {
+      "epoch": 1.4554455445544554,
+      "grad_norm": 0.6815268993377686,
+      "learning_rate": 8.59529316347554e-05,
+      "loss": 0.5794,
+      "step": 8159
+    },
+    {
+      "epoch": 1.4556239407724556,
+      "grad_norm": 0.5202351808547974,
+      "learning_rate": 8.590006927619676e-05,
+      "loss": 0.6447,
+      "step": 8160
+    },
+    {
+      "epoch": 1.4558023369904558,
+      "grad_norm": 0.4786362051963806,
+      "learning_rate": 8.584721980597599e-05,
+      "loss": 0.5809,
+      "step": 8161
+    },
+    {
+      "epoch": 1.455980733208456,
+      "grad_norm": 0.5115570425987244,
+      "learning_rate": 8.579438322824403e-05,
+      "loss": 0.6302,
+      "step": 8162
+    },
+    {
+      "epoch": 1.4561591294264562,
+      "grad_norm": 0.4989685118198395,
+      "learning_rate": 8.574155954715047e-05,
+      "loss": 0.5402,
+      "step": 8163
+    },
+    {
+      "epoch": 1.4563375256444564,
+      "grad_norm": 0.5098851919174194,
+      "learning_rate": 8.568874876684418e-05,
+      "loss": 0.6773,
+      "step": 8164
+    },
+    {
+      "epoch": 1.4565159218624566,
+      "grad_norm": 0.5459485650062561,
+      "learning_rate": 8.56359508914728e-05,
+      "loss": 0.7716,
+      "step": 8165
+    },
+    {
+      "epoch": 1.4566943180804568,
+      "grad_norm": 0.5095139145851135,
+      "learning_rate": 8.55831659251832e-05,
+      "loss": 0.7628,
+      "step": 8166
+    },
+    {
+      "epoch": 1.456872714298457,
+      "grad_norm": 0.5003282427787781,
+      "learning_rate": 8.553039387212097e-05,
+      "loss": 0.5369,
+      "step": 8167
+    },
+    {
+      "epoch": 1.457051110516457,
+      "grad_norm": 0.5022057890892029,
+      "learning_rate": 8.547763473643074e-05,
+      "loss": 0.5863,
+      "step": 8168
+    },
+    {
+      "epoch": 1.4572295067344572,
+      "grad_norm": 0.4498676061630249,
+      "learning_rate": 8.542488852225638e-05,
+      "loss": 0.4954,
+      "step": 8169
+    },
+    {
+      "epoch": 1.4574079029524574,
+      "grad_norm": 0.6011500358581543,
+      "learning_rate": 8.537215523374037e-05,
+      "loss": 0.7241,
+      "step": 8170
+    },
+    {
+      "epoch": 1.4575862991704576,
+      "grad_norm": 0.48579999804496765,
+      "learning_rate": 8.531943487502445e-05,
+      "loss": 0.4106,
+      "step": 8171
+    },
+    {
+      "epoch": 1.4577646953884578,
+      "grad_norm": 0.5279344320297241,
+      "learning_rate": 8.52667274502493e-05,
+      "loss": 0.5385,
+      "step": 8172
+    },
+    {
+      "epoch": 1.457943091606458,
+      "grad_norm": 0.5536198616027832,
+      "learning_rate": 8.521403296355443e-05,
+      "loss": 0.7606,
+      "step": 8173
+    },
+    {
+      "epoch": 1.4581214878244582,
+      "grad_norm": 0.5520446300506592,
+      "learning_rate": 8.516135141907858e-05,
+      "loss": 0.5819,
+      "step": 8174
+    },
+    {
+      "epoch": 1.4582998840424584,
+      "grad_norm": 0.49602624773979187,
+      "learning_rate": 8.510868282095916e-05,
+      "loss": 0.6204,
+      "step": 8175
+    },
+    {
+      "epoch": 1.4584782802604586,
+      "grad_norm": 0.4096802771091461,
+      "learning_rate": 8.505602717333291e-05,
+      "loss": 0.3461,
+      "step": 8176
+    },
+    {
+      "epoch": 1.4586566764784585,
+      "grad_norm": 0.5201975107192993,
+      "learning_rate": 8.500338448033524e-05,
+      "loss": 0.5463,
+      "step": 8177
+    },
+    {
+      "epoch": 1.4588350726964587,
+      "grad_norm": 0.45560964941978455,
+      "learning_rate": 8.495075474610081e-05,
+      "loss": 0.4585,
+      "step": 8178
+    },
+    {
+      "epoch": 1.459013468914459,
+      "grad_norm": 0.4644407629966736,
+      "learning_rate": 8.489813797476303e-05,
+      "loss": 0.5049,
+      "step": 8179
+    },
+    {
+      "epoch": 1.4591918651324591,
+      "grad_norm": 0.4632987678050995,
+      "learning_rate": 8.484553417045448e-05,
+      "loss": 0.4836,
+      "step": 8180
+    },
+    {
+      "epoch": 1.4593702613504593,
+      "grad_norm": 0.5432924628257751,
+      "learning_rate": 8.479294333730664e-05,
+      "loss": 0.6128,
+      "step": 8181
+    },
+    {
+      "epoch": 1.4595486575684595,
+      "grad_norm": 0.4162793755531311,
+      "learning_rate": 8.474036547944985e-05,
+      "loss": 0.4007,
+      "step": 8182
+    },
+    {
+      "epoch": 1.4597270537864597,
+      "grad_norm": 0.49239131808280945,
+      "learning_rate": 8.468780060101372e-05,
+      "loss": 0.6786,
+      "step": 8183
+    },
+    {
+      "epoch": 1.45990545000446,
+      "grad_norm": 0.5322994589805603,
+      "learning_rate": 8.463524870612649e-05,
+      "loss": 0.6764,
+      "step": 8184
+    },
+    {
+      "epoch": 1.4600838462224601,
+      "grad_norm": 0.5249794125556946,
+      "learning_rate": 8.458270979891578e-05,
+      "loss": 0.5932,
+      "step": 8185
+    },
+    {
+      "epoch": 1.4602622424404603,
+      "grad_norm": 0.5151904821395874,
+      "learning_rate": 8.453018388350772e-05,
+      "loss": 0.5464,
+      "step": 8186
+    },
+    {
+      "epoch": 1.4604406386584605,
+      "grad_norm": 0.5483924746513367,
+      "learning_rate": 8.447767096402787e-05,
+      "loss": 0.6301,
+      "step": 8187
+    },
+    {
+      "epoch": 1.4606190348764607,
+      "grad_norm": 0.4792023301124573,
+      "learning_rate": 8.442517104460057e-05,
+      "loss": 0.5793,
+      "step": 8188
+    },
+    {
+      "epoch": 1.460797431094461,
+      "grad_norm": 0.5382273197174072,
+      "learning_rate": 8.437268412934898e-05,
+      "loss": 0.7128,
+      "step": 8189
+    },
+    {
+      "epoch": 1.460975827312461,
+      "grad_norm": 0.5009669065475464,
+      "learning_rate": 8.432021022239561e-05,
+      "loss": 0.5538,
+      "step": 8190
+    },
+    {
+      "epoch": 1.461154223530461,
+      "grad_norm": 0.4763016998767853,
+      "learning_rate": 8.426774932786154e-05,
+      "loss": 0.5062,
+      "step": 8191
+    },
+    {
+      "epoch": 1.4613326197484613,
+      "grad_norm": 0.49051937460899353,
+      "learning_rate": 8.421530144986722e-05,
+      "loss": 0.6787,
+      "step": 8192
+    },
+    {
+      "epoch": 1.4615110159664615,
+      "grad_norm": 0.48328647017478943,
+      "learning_rate": 8.416286659253178e-05,
+      "loss": 0.5788,
+      "step": 8193
+    },
+    {
+      "epoch": 1.4616894121844617,
+      "grad_norm": 0.48622655868530273,
+      "learning_rate": 8.411044475997331e-05,
+      "loss": 0.6081,
+      "step": 8194
+    },
+    {
+      "epoch": 1.4618678084024619,
+      "grad_norm": 0.45523685216903687,
+      "learning_rate": 8.405803595630926e-05,
+      "loss": 0.5096,
+      "step": 8195
+    },
+    {
+      "epoch": 1.462046204620462,
+      "grad_norm": 0.5314157009124756,
+      "learning_rate": 8.400564018565554e-05,
+      "loss": 0.6395,
+      "step": 8196
+    },
+    {
+      "epoch": 1.4622246008384623,
+      "grad_norm": 0.48079803586006165,
+      "learning_rate": 8.395325745212747e-05,
+      "loss": 0.5513,
+      "step": 8197
+    },
+    {
+      "epoch": 1.4624029970564625,
+      "grad_norm": 0.4904074966907501,
+      "learning_rate": 8.390088775983906e-05,
+      "loss": 0.566,
+      "step": 8198
+    },
+    {
+      "epoch": 1.4625813932744625,
+      "grad_norm": 0.48788025975227356,
+      "learning_rate": 8.384853111290352e-05,
+      "loss": 0.6323,
+      "step": 8199
+    },
+    {
+      "epoch": 1.4627597894924627,
+      "grad_norm": 0.455021470785141,
+      "learning_rate": 8.379618751543274e-05,
+      "loss": 0.5147,
+      "step": 8200
+    },
+    {
+      "epoch": 1.4629381857104629,
+      "grad_norm": 0.47679826617240906,
+      "learning_rate": 8.374385697153791e-05,
+      "loss": 0.6236,
+      "step": 8201
+    },
+    {
+      "epoch": 1.463116581928463,
+      "grad_norm": 0.5282472968101501,
+      "learning_rate": 8.369153948532907e-05,
+      "loss": 0.6563,
+      "step": 8202
+    },
+    {
+      "epoch": 1.4632949781464633,
+      "grad_norm": 0.5126857757568359,
+      "learning_rate": 8.363923506091506e-05,
+      "loss": 0.5623,
+      "step": 8203
+    },
+    {
+      "epoch": 1.4634733743644635,
+      "grad_norm": 0.476513147354126,
+      "learning_rate": 8.358694370240402e-05,
+      "loss": 0.443,
+      "step": 8204
+    },
+    {
+      "epoch": 1.4636517705824637,
+      "grad_norm": 0.5799571871757507,
+      "learning_rate": 8.353466541390273e-05,
+      "loss": 0.5796,
+      "step": 8205
+    },
+    {
+      "epoch": 1.4638301668004639,
+      "grad_norm": 0.41350480914115906,
+      "learning_rate": 8.348240019951728e-05,
+      "loss": 0.3979,
+      "step": 8206
+    },
+    {
+      "epoch": 1.464008563018464,
+      "grad_norm": 0.47292932868003845,
+      "learning_rate": 8.343014806335245e-05,
+      "loss": 0.4637,
+      "step": 8207
+    },
+    {
+      "epoch": 1.4641869592364642,
+      "grad_norm": 0.5536729693412781,
+      "learning_rate": 8.3377909009512e-05,
+      "loss": 0.6503,
+      "step": 8208
+    },
+    {
+      "epoch": 1.4643653554544644,
+      "grad_norm": 0.5005719065666199,
+      "learning_rate": 8.3325683042099e-05,
+      "loss": 0.4994,
+      "step": 8209
+    },
+    {
+      "epoch": 1.4645437516724646,
+      "grad_norm": 0.47958654165267944,
+      "learning_rate": 8.327347016521503e-05,
+      "loss": 0.5329,
+      "step": 8210
+    },
+    {
+      "epoch": 1.4647221478904648,
+      "grad_norm": 0.5664541125297546,
+      "learning_rate": 8.322127038296104e-05,
+      "loss": 0.766,
+      "step": 8211
+    },
+    {
+      "epoch": 1.4649005441084648,
+      "grad_norm": 0.5163717865943909,
+      "learning_rate": 8.316908369943663e-05,
+      "loss": 0.5823,
+      "step": 8212
+    },
+    {
+      "epoch": 1.465078940326465,
+      "grad_norm": 0.4811464548110962,
+      "learning_rate": 8.311691011874067e-05,
+      "loss": 0.55,
+      "step": 8213
+    },
+    {
+      "epoch": 1.4652573365444652,
+      "grad_norm": 0.4442557990550995,
+      "learning_rate": 8.306474964497076e-05,
+      "loss": 0.4014,
+      "step": 8214
+    },
+    {
+      "epoch": 1.4654357327624654,
+      "grad_norm": 0.5611222982406616,
+      "learning_rate": 8.301260228222351e-05,
+      "loss": 0.7519,
+      "step": 8215
+    },
+    {
+      "epoch": 1.4656141289804656,
+      "grad_norm": 0.4814377427101135,
+      "learning_rate": 8.29604680345947e-05,
+      "loss": 0.5494,
+      "step": 8216
+    },
+    {
+      "epoch": 1.4657925251984658,
+      "grad_norm": 0.42738616466522217,
+      "learning_rate": 8.290834690617868e-05,
+      "loss": 0.5058,
+      "step": 8217
+    },
+    {
+      "epoch": 1.465970921416466,
+      "grad_norm": 0.6096622347831726,
+      "learning_rate": 8.285623890106936e-05,
+      "loss": 0.573,
+      "step": 8218
+    },
+    {
+      "epoch": 1.4661493176344662,
+      "grad_norm": 0.43758609890937805,
+      "learning_rate": 8.280414402335909e-05,
+      "loss": 0.4614,
+      "step": 8219
+    },
+    {
+      "epoch": 1.4663277138524664,
+      "grad_norm": 0.48316100239753723,
+      "learning_rate": 8.275206227713936e-05,
+      "loss": 0.5601,
+      "step": 8220
+    },
+    {
+      "epoch": 1.4665061100704664,
+      "grad_norm": 0.5487798452377319,
+      "learning_rate": 8.269999366650071e-05,
+      "loss": 0.7356,
+      "step": 8221
+    },
+    {
+      "epoch": 1.4666845062884666,
+      "grad_norm": 0.4505595862865448,
+      "learning_rate": 8.264793819553252e-05,
+      "loss": 0.4967,
+      "step": 8222
+    },
+    {
+      "epoch": 1.4668629025064668,
+      "grad_norm": 0.46539705991744995,
+      "learning_rate": 8.259589586832331e-05,
+      "loss": 0.4694,
+      "step": 8223
+    },
+    {
+      "epoch": 1.467041298724467,
+      "grad_norm": 0.5119168758392334,
+      "learning_rate": 8.254386668896033e-05,
+      "loss": 0.5304,
+      "step": 8224
+    },
+    {
+      "epoch": 1.4672196949424672,
+      "grad_norm": 0.494151771068573,
+      "learning_rate": 8.249185066153006e-05,
+      "loss": 0.5928,
+      "step": 8225
+    },
+    {
+      "epoch": 1.4673980911604674,
+      "grad_norm": 0.4593886137008667,
+      "learning_rate": 8.243984779011779e-05,
+      "loss": 0.5024,
+      "step": 8226
+    },
+    {
+      "epoch": 1.4675764873784676,
+      "grad_norm": 0.46240317821502686,
+      "learning_rate": 8.238785807880767e-05,
+      "loss": 0.534,
+      "step": 8227
+    },
+    {
+      "epoch": 1.4677548835964678,
+      "grad_norm": 0.49743345379829407,
+      "learning_rate": 8.233588153168312e-05,
+      "loss": 0.5577,
+      "step": 8228
+    },
+    {
+      "epoch": 1.467933279814468,
+      "grad_norm": 0.43200600147247314,
+      "learning_rate": 8.228391815282619e-05,
+      "loss": 0.3676,
+      "step": 8229
+    },
+    {
+      "epoch": 1.4681116760324682,
+      "grad_norm": 0.46943768858909607,
+      "learning_rate": 8.223196794631826e-05,
+      "loss": 0.4637,
+      "step": 8230
+    },
+    {
+      "epoch": 1.4682900722504684,
+      "grad_norm": 0.5156434178352356,
+      "learning_rate": 8.218003091623927e-05,
+      "loss": 0.6567,
+      "step": 8231
+    },
+    {
+      "epoch": 1.4684684684684686,
+      "grad_norm": 0.46780145168304443,
+      "learning_rate": 8.212810706666846e-05,
+      "loss": 0.5733,
+      "step": 8232
+    },
+    {
+      "epoch": 1.4686468646864688,
+      "grad_norm": 0.5769681334495544,
+      "learning_rate": 8.207619640168393e-05,
+      "loss": 0.8275,
+      "step": 8233
+    },
+    {
+      "epoch": 1.4688252609044687,
+      "grad_norm": 0.4969267249107361,
+      "learning_rate": 8.202429892536261e-05,
+      "loss": 0.5933,
+      "step": 8234
+    },
+    {
+      "epoch": 1.469003657122469,
+      "grad_norm": 0.5537621378898621,
+      "learning_rate": 8.197241464178065e-05,
+      "loss": 0.7404,
+      "step": 8235
+    },
+    {
+      "epoch": 1.4691820533404691,
+      "grad_norm": 0.5195968747138977,
+      "learning_rate": 8.192054355501282e-05,
+      "loss": 0.6725,
+      "step": 8236
+    },
+    {
+      "epoch": 1.4693604495584693,
+      "grad_norm": 0.47820261120796204,
+      "learning_rate": 8.18686856691333e-05,
+      "loss": 0.6093,
+      "step": 8237
+    },
+    {
+      "epoch": 1.4695388457764695,
+      "grad_norm": 0.45387136936187744,
+      "learning_rate": 8.181684098821474e-05,
+      "loss": 0.552,
+      "step": 8238
+    },
+    {
+      "epoch": 1.4697172419944697,
+      "grad_norm": 0.49222466349601746,
+      "learning_rate": 8.17650095163292e-05,
+      "loss": 0.6341,
+      "step": 8239
+    },
+    {
+      "epoch": 1.46989563821247,
+      "grad_norm": 0.47406187653541565,
+      "learning_rate": 8.171319125754745e-05,
+      "loss": 0.5789,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4700740344304701,
+      "grad_norm": 0.49534159898757935,
+      "learning_rate": 8.16613862159391e-05,
+      "loss": 0.5961,
+      "step": 8241
+    },
+    {
+      "epoch": 1.4702524306484703,
+      "grad_norm": 0.5249592661857605,
+      "learning_rate": 8.160959439557316e-05,
+      "loss": 0.6454,
+      "step": 8242
+    },
+    {
+      "epoch": 1.4704308268664703,
+      "grad_norm": 0.4379402697086334,
+      "learning_rate": 8.155781580051714e-05,
+      "loss": 0.4289,
+      "step": 8243
+    },
+    {
+      "epoch": 1.4706092230844705,
+      "grad_norm": 0.4104413092136383,
+      "learning_rate": 8.150605043483783e-05,
+      "loss": 0.4201,
+      "step": 8244
+    },
+    {
+      "epoch": 1.4707876193024707,
+      "grad_norm": 0.5442726612091064,
+      "learning_rate": 8.145429830260073e-05,
+      "loss": 0.7512,
+      "step": 8245
+    },
+    {
+      "epoch": 1.470966015520471,
+      "grad_norm": 0.47776398062705994,
+      "learning_rate": 8.140255940787059e-05,
+      "loss": 0.5361,
+      "step": 8246
+    },
+    {
+      "epoch": 1.471144411738471,
+      "grad_norm": 0.49612924456596375,
+      "learning_rate": 8.13508337547108e-05,
+      "loss": 0.6154,
+      "step": 8247
+    },
+    {
+      "epoch": 1.4713228079564713,
+      "grad_norm": 0.45897966623306274,
+      "learning_rate": 8.129912134718398e-05,
+      "loss": 0.4906,
+      "step": 8248
+    },
+    {
+      "epoch": 1.4715012041744715,
+      "grad_norm": 0.5512779355049133,
+      "learning_rate": 8.124742218935164e-05,
+      "loss": 0.739,
+      "step": 8249
+    },
+    {
+      "epoch": 1.4716796003924717,
+      "grad_norm": 0.5072031617164612,
+      "learning_rate": 8.119573628527404e-05,
+      "loss": 0.5826,
+      "step": 8250
+    },
+    {
+      "epoch": 1.471857996610472,
+      "grad_norm": 0.47292739152908325,
+      "learning_rate": 8.114406363901078e-05,
+      "loss": 0.4861,
+      "step": 8251
+    },
+    {
+      "epoch": 1.472036392828472,
+      "grad_norm": 0.448589026927948,
+      "learning_rate": 8.109240425462008e-05,
+      "loss": 0.4831,
+      "step": 8252
+    },
+    {
+      "epoch": 1.4722147890464723,
+      "grad_norm": 0.4658273756504059,
+      "learning_rate": 8.104075813615918e-05,
+      "loss": 0.5328,
+      "step": 8253
+    },
+    {
+      "epoch": 1.4723931852644725,
+      "grad_norm": 0.5529447197914124,
+      "learning_rate": 8.098912528768452e-05,
+      "loss": 0.6661,
+      "step": 8254
+    },
+    {
+      "epoch": 1.4725715814824727,
+      "grad_norm": 0.46816307306289673,
+      "learning_rate": 8.093750571325112e-05,
+      "loss": 0.4844,
+      "step": 8255
+    },
+    {
+      "epoch": 1.4727499777004727,
+      "grad_norm": 0.5499048829078674,
+      "learning_rate": 8.088589941691338e-05,
+      "loss": 0.5624,
+      "step": 8256
+    },
+    {
+      "epoch": 1.4729283739184729,
+      "grad_norm": 0.5012791752815247,
+      "learning_rate": 8.083430640272424e-05,
+      "loss": 0.5128,
+      "step": 8257
+    },
+    {
+      "epoch": 1.473106770136473,
+      "grad_norm": 0.5420527458190918,
+      "learning_rate": 8.078272667473593e-05,
+      "loss": 0.5709,
+      "step": 8258
+    },
+    {
+      "epoch": 1.4732851663544733,
+      "grad_norm": 0.4874211549758911,
+      "learning_rate": 8.073116023699939e-05,
+      "loss": 0.5259,
+      "step": 8259
+    },
+    {
+      "epoch": 1.4734635625724735,
+      "grad_norm": 0.5653133988380432,
+      "learning_rate": 8.067960709356478e-05,
+      "loss": 0.6445,
+      "step": 8260
+    },
+    {
+      "epoch": 1.4736419587904737,
+      "grad_norm": 0.4977559745311737,
+      "learning_rate": 8.062806724848093e-05,
+      "loss": 0.5627,
+      "step": 8261
+    },
+    {
+      "epoch": 1.4738203550084739,
+      "grad_norm": 0.5231120586395264,
+      "learning_rate": 8.057654070579573e-05,
+      "loss": 0.5859,
+      "step": 8262
+    },
+    {
+      "epoch": 1.473998751226474,
+      "grad_norm": 0.5144016146659851,
+      "learning_rate": 8.052502746955612e-05,
+      "loss": 0.6625,
+      "step": 8263
+    },
+    {
+      "epoch": 1.4741771474444743,
+      "grad_norm": 0.5300917029380798,
+      "learning_rate": 8.04735275438079e-05,
+      "loss": 0.6414,
+      "step": 8264
+    },
+    {
+      "epoch": 1.4743555436624742,
+      "grad_norm": 0.506809651851654,
+      "learning_rate": 8.042204093259597e-05,
+      "loss": 0.5545,
+      "step": 8265
+    },
+    {
+      "epoch": 1.4745339398804744,
+      "grad_norm": 0.4742717742919922,
+      "learning_rate": 8.037056763996398e-05,
+      "loss": 0.5692,
+      "step": 8266
+    },
+    {
+      "epoch": 1.4747123360984746,
+      "grad_norm": 0.477117657661438,
+      "learning_rate": 8.031910766995451e-05,
+      "loss": 0.4557,
+      "step": 8267
+    },
+    {
+      "epoch": 1.4748907323164748,
+      "grad_norm": 0.4841883182525635,
+      "learning_rate": 8.02676610266094e-05,
+      "loss": 0.6635,
+      "step": 8268
+    },
+    {
+      "epoch": 1.475069128534475,
+      "grad_norm": 0.4822656810283661,
+      "learning_rate": 8.021622771396905e-05,
+      "loss": 0.557,
+      "step": 8269
+    },
+    {
+      "epoch": 1.4752475247524752,
+      "grad_norm": 0.49790987372398376,
+      "learning_rate": 8.01648077360732e-05,
+      "loss": 0.6972,
+      "step": 8270
+    },
+    {
+      "epoch": 1.4754259209704754,
+      "grad_norm": 0.5201764702796936,
+      "learning_rate": 8.01134010969602e-05,
+      "loss": 0.6559,
+      "step": 8271
+    },
+    {
+      "epoch": 1.4756043171884756,
+      "grad_norm": 0.4643043577671051,
+      "learning_rate": 8.006200780066763e-05,
+      "loss": 0.4562,
+      "step": 8272
+    },
+    {
+      "epoch": 1.4757827134064758,
+      "grad_norm": 0.4936201870441437,
+      "learning_rate": 8.001062785123184e-05,
+      "loss": 0.519,
+      "step": 8273
+    },
+    {
+      "epoch": 1.475961109624476,
+      "grad_norm": 0.5245823860168457,
+      "learning_rate": 7.995926125268813e-05,
+      "loss": 0.7527,
+      "step": 8274
+    },
+    {
+      "epoch": 1.4761395058424762,
+      "grad_norm": 0.45136356353759766,
+      "learning_rate": 7.99079080090709e-05,
+      "loss": 0.4744,
+      "step": 8275
+    },
+    {
+      "epoch": 1.4763179020604764,
+      "grad_norm": 0.5103075504302979,
+      "learning_rate": 7.985656812441336e-05,
+      "loss": 0.5456,
+      "step": 8276
+    },
+    {
+      "epoch": 1.4764962982784766,
+      "grad_norm": 0.5783809423446655,
+      "learning_rate": 7.980524160274776e-05,
+      "loss": 0.6718,
+      "step": 8277
+    },
+    {
+      "epoch": 1.4766746944964766,
+      "grad_norm": 0.41144779324531555,
+      "learning_rate": 7.975392844810523e-05,
+      "loss": 0.4706,
+      "step": 8278
+    },
+    {
+      "epoch": 1.4768530907144768,
+      "grad_norm": 0.5507053732872009,
+      "learning_rate": 7.970262866451583e-05,
+      "loss": 0.6462,
+      "step": 8279
+    },
+    {
+      "epoch": 1.477031486932477,
+      "grad_norm": 0.5084280967712402,
+      "learning_rate": 7.965134225600881e-05,
+      "loss": 0.5376,
+      "step": 8280
+    },
+    {
+      "epoch": 1.4772098831504772,
+      "grad_norm": 0.46495214104652405,
+      "learning_rate": 7.960006922661197e-05,
+      "loss": 0.4167,
+      "step": 8281
+    },
+    {
+      "epoch": 1.4773882793684774,
+      "grad_norm": 0.54592365026474,
+      "learning_rate": 7.954880958035245e-05,
+      "loss": 0.6447,
+      "step": 8282
+    },
+    {
+      "epoch": 1.4775666755864776,
+      "grad_norm": 0.588589608669281,
+      "learning_rate": 7.949756332125599e-05,
+      "loss": 0.7427,
+      "step": 8283
+    },
+    {
+      "epoch": 1.4777450718044778,
+      "grad_norm": 0.5428383350372314,
+      "learning_rate": 7.944633045334763e-05,
+      "loss": 0.628,
+      "step": 8284
+    },
+    {
+      "epoch": 1.477923468022478,
+      "grad_norm": 0.45825469493865967,
+      "learning_rate": 7.939511098065097e-05,
+      "loss": 0.4774,
+      "step": 8285
+    },
+    {
+      "epoch": 1.4781018642404782,
+      "grad_norm": 0.591995358467102,
+      "learning_rate": 7.934390490718898e-05,
+      "loss": 0.5751,
+      "step": 8286
+    },
+    {
+      "epoch": 1.4782802604584782,
+      "grad_norm": 0.5302790403366089,
+      "learning_rate": 7.929271223698326e-05,
+      "loss": 0.5787,
+      "step": 8287
+    },
+    {
+      "epoch": 1.4784586566764784,
+      "grad_norm": 0.5035462379455566,
+      "learning_rate": 7.924153297405437e-05,
+      "loss": 0.5414,
+      "step": 8288
+    },
+    {
+      "epoch": 1.4786370528944786,
+      "grad_norm": 0.4451696574687958,
+      "learning_rate": 7.919036712242205e-05,
+      "loss": 0.4648,
+      "step": 8289
+    },
+    {
+      "epoch": 1.4788154491124788,
+      "grad_norm": 0.5414475798606873,
+      "learning_rate": 7.913921468610477e-05,
+      "loss": 0.7727,
+      "step": 8290
+    },
+    {
+      "epoch": 1.478993845330479,
+      "grad_norm": 0.44653692841529846,
+      "learning_rate": 7.908807566912007e-05,
+      "loss": 0.5023,
+      "step": 8291
+    },
+    {
+      "epoch": 1.4791722415484791,
+      "grad_norm": 0.4529421925544739,
+      "learning_rate": 7.903695007548432e-05,
+      "loss": 0.534,
+      "step": 8292
+    },
+    {
+      "epoch": 1.4793506377664793,
+      "grad_norm": 0.5018486380577087,
+      "learning_rate": 7.898583790921291e-05,
+      "loss": 0.514,
+      "step": 8293
+    },
+    {
+      "epoch": 1.4795290339844795,
+      "grad_norm": 0.4705956280231476,
+      "learning_rate": 7.893473917432029e-05,
+      "loss": 0.5485,
+      "step": 8294
+    },
+    {
+      "epoch": 1.4797074302024797,
+      "grad_norm": 0.4740748107433319,
+      "learning_rate": 7.888365387481955e-05,
+      "loss": 0.5393,
+      "step": 8295
+    },
+    {
+      "epoch": 1.47988582642048,
+      "grad_norm": 0.5045658349990845,
+      "learning_rate": 7.88325820147231e-05,
+      "loss": 0.5581,
+      "step": 8296
+    },
+    {
+      "epoch": 1.4800642226384801,
+      "grad_norm": 0.48949524760246277,
+      "learning_rate": 7.87815235980419e-05,
+      "loss": 0.5983,
+      "step": 8297
+    },
+    {
+      "epoch": 1.4802426188564803,
+      "grad_norm": 0.47559529542922974,
+      "learning_rate": 7.873047862878624e-05,
+      "loss": 0.566,
+      "step": 8298
+    },
+    {
+      "epoch": 1.4804210150744805,
+      "grad_norm": 0.49631282687187195,
+      "learning_rate": 7.867944711096508e-05,
+      "loss": 0.5443,
+      "step": 8299
+    },
+    {
+      "epoch": 1.4805994112924805,
+      "grad_norm": 0.508677065372467,
+      "learning_rate": 7.862842904858633e-05,
+      "loss": 0.602,
+      "step": 8300
+    },
+    {
+      "epoch": 1.4807778075104807,
+      "grad_norm": 0.5043870806694031,
+      "learning_rate": 7.857742444565713e-05,
+      "loss": 0.5086,
+      "step": 8301
+    },
+    {
+      "epoch": 1.480956203728481,
+      "grad_norm": 0.5479679107666016,
+      "learning_rate": 7.852643330618314e-05,
+      "loss": 0.6697,
+      "step": 8302
+    },
+    {
+      "epoch": 1.4811345999464811,
+      "grad_norm": 0.5437744855880737,
+      "learning_rate": 7.847545563416936e-05,
+      "loss": 0.6367,
+      "step": 8303
+    },
+    {
+      "epoch": 1.4813129961644813,
+      "grad_norm": 0.5325936675071716,
+      "learning_rate": 7.842449143361943e-05,
+      "loss": 0.6722,
+      "step": 8304
+    },
+    {
+      "epoch": 1.4814913923824815,
+      "grad_norm": 0.48976537585258484,
+      "learning_rate": 7.837354070853616e-05,
+      "loss": 0.6325,
+      "step": 8305
+    },
+    {
+      "epoch": 1.4816697886004817,
+      "grad_norm": 0.5160368084907532,
+      "learning_rate": 7.832260346292117e-05,
+      "loss": 0.5519,
+      "step": 8306
+    },
+    {
+      "epoch": 1.481848184818482,
+      "grad_norm": 0.49977096915245056,
+      "learning_rate": 7.827167970077492e-05,
+      "loss": 0.5907,
+      "step": 8307
+    },
+    {
+      "epoch": 1.482026581036482,
+      "grad_norm": 0.47831645607948303,
+      "learning_rate": 7.822076942609707e-05,
+      "loss": 0.5587,
+      "step": 8308
+    },
+    {
+      "epoch": 1.482204977254482,
+      "grad_norm": 0.496364951133728,
+      "learning_rate": 7.816987264288606e-05,
+      "loss": 0.5853,
+      "step": 8309
+    },
+    {
+      "epoch": 1.4823833734724823,
+      "grad_norm": 0.45871299505233765,
+      "learning_rate": 7.811898935513936e-05,
+      "loss": 0.4607,
+      "step": 8310
+    },
+    {
+      "epoch": 1.4825617696904825,
+      "grad_norm": 0.4363771677017212,
+      "learning_rate": 7.806811956685331e-05,
+      "loss": 0.5019,
+      "step": 8311
+    },
+    {
+      "epoch": 1.4827401659084827,
+      "grad_norm": 0.5848436951637268,
+      "learning_rate": 7.801726328202305e-05,
+      "loss": 0.6789,
+      "step": 8312
+    },
+    {
+      "epoch": 1.4829185621264829,
+      "grad_norm": 0.4620121121406555,
+      "learning_rate": 7.796642050464303e-05,
+      "loss": 0.5479,
+      "step": 8313
+    },
+    {
+      "epoch": 1.483096958344483,
+      "grad_norm": 0.5338596701622009,
+      "learning_rate": 7.79155912387062e-05,
+      "loss": 0.5664,
+      "step": 8314
+    },
+    {
+      "epoch": 1.4832753545624833,
+      "grad_norm": 0.4988861382007599,
+      "learning_rate": 7.786477548820489e-05,
+      "loss": 0.5782,
+      "step": 8315
+    },
+    {
+      "epoch": 1.4834537507804835,
+      "grad_norm": 0.6170352101325989,
+      "learning_rate": 7.781397325712994e-05,
+      "loss": 0.5044,
+      "step": 8316
+    },
+    {
+      "epoch": 1.4836321469984837,
+      "grad_norm": 0.5399267077445984,
+      "learning_rate": 7.776318454947154e-05,
+      "loss": 0.6209,
+      "step": 8317
+    },
+    {
+      "epoch": 1.4838105432164839,
+      "grad_norm": 0.5562114715576172,
+      "learning_rate": 7.771240936921839e-05,
+      "loss": 0.7107,
+      "step": 8318
+    },
+    {
+      "epoch": 1.483988939434484,
+      "grad_norm": 0.46948572993278503,
+      "learning_rate": 7.766164772035855e-05,
+      "loss": 0.5113,
+      "step": 8319
+    },
+    {
+      "epoch": 1.4841673356524843,
+      "grad_norm": 0.5978415012359619,
+      "learning_rate": 7.761089960687876e-05,
+      "loss": 0.6767,
+      "step": 8320
+    },
+    {
+      "epoch": 1.4843457318704845,
+      "grad_norm": 0.4891349673271179,
+      "learning_rate": 7.756016503276464e-05,
+      "loss": 0.5338,
+      "step": 8321
+    },
+    {
+      "epoch": 1.4845241280884847,
+      "grad_norm": 0.45725369453430176,
+      "learning_rate": 7.750944400200102e-05,
+      "loss": 0.4976,
+      "step": 8322
+    },
+    {
+      "epoch": 1.4847025243064846,
+      "grad_norm": 0.42975348234176636,
+      "learning_rate": 7.745873651857138e-05,
+      "loss": 0.4602,
+      "step": 8323
+    },
+    {
+      "epoch": 1.4848809205244848,
+      "grad_norm": 0.4848777949810028,
+      "learning_rate": 7.740804258645831e-05,
+      "loss": 0.6199,
+      "step": 8324
+    },
+    {
+      "epoch": 1.485059316742485,
+      "grad_norm": 0.4301760792732239,
+      "learning_rate": 7.735736220964337e-05,
+      "loss": 0.4882,
+      "step": 8325
+    },
+    {
+      "epoch": 1.4852377129604852,
+      "grad_norm": 0.4948813021183014,
+      "learning_rate": 7.730669539210686e-05,
+      "loss": 0.6615,
+      "step": 8326
+    },
+    {
+      "epoch": 1.4854161091784854,
+      "grad_norm": 0.4712623059749603,
+      "learning_rate": 7.725604213782824e-05,
+      "loss": 0.5958,
+      "step": 8327
+    },
+    {
+      "epoch": 1.4855945053964856,
+      "grad_norm": 0.484377384185791,
+      "learning_rate": 7.720540245078567e-05,
+      "loss": 0.6291,
+      "step": 8328
+    },
+    {
+      "epoch": 1.4857729016144858,
+      "grad_norm": 0.5532923340797424,
+      "learning_rate": 7.715477633495649e-05,
+      "loss": 0.712,
+      "step": 8329
+    },
+    {
+      "epoch": 1.485951297832486,
+      "grad_norm": 0.4761124849319458,
+      "learning_rate": 7.71041637943167e-05,
+      "loss": 0.4219,
+      "step": 8330
+    },
+    {
+      "epoch": 1.486129694050486,
+      "grad_norm": 0.4717256724834442,
+      "learning_rate": 7.70535648328416e-05,
+      "loss": 0.5188,
+      "step": 8331
+    },
+    {
+      "epoch": 1.4863080902684862,
+      "grad_norm": 0.4593075215816498,
+      "learning_rate": 7.700297945450507e-05,
+      "loss": 0.4863,
+      "step": 8332
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 0.4873063266277313,
+      "learning_rate": 7.695240766328002e-05,
+      "loss": 0.5419,
+      "step": 8333
+    },
+    {
+      "epoch": 1.4866648827044866,
+      "grad_norm": 0.5233901143074036,
+      "learning_rate": 7.690184946313846e-05,
+      "loss": 0.5865,
+      "step": 8334
+    },
+    {
+      "epoch": 1.4868432789224868,
+      "grad_norm": 0.5119859576225281,
+      "learning_rate": 7.685130485805112e-05,
+      "loss": 0.5873,
+      "step": 8335
+    },
+    {
+      "epoch": 1.487021675140487,
+      "grad_norm": 0.4452389180660248,
+      "learning_rate": 7.680077385198783e-05,
+      "loss": 0.3941,
+      "step": 8336
+    },
+    {
+      "epoch": 1.4872000713584872,
+      "grad_norm": 0.49985215067863464,
+      "learning_rate": 7.675025644891714e-05,
+      "loss": 0.586,
+      "step": 8337
+    },
+    {
+      "epoch": 1.4873784675764874,
+      "grad_norm": 0.5451793074607849,
+      "learning_rate": 7.669975265280688e-05,
+      "loss": 0.597,
+      "step": 8338
+    },
+    {
+      "epoch": 1.4875568637944876,
+      "grad_norm": 0.47509869933128357,
+      "learning_rate": 7.664926246762335e-05,
+      "loss": 0.5232,
+      "step": 8339
+    },
+    {
+      "epoch": 1.4877352600124878,
+      "grad_norm": 0.44898343086242676,
+      "learning_rate": 7.659878589733216e-05,
+      "loss": 0.4391,
+      "step": 8340
+    },
+    {
+      "epoch": 1.487913656230488,
+      "grad_norm": 0.5049835443496704,
+      "learning_rate": 7.654832294589775e-05,
+      "loss": 0.5392,
+      "step": 8341
+    },
+    {
+      "epoch": 1.4880920524484882,
+      "grad_norm": 0.5432078242301941,
+      "learning_rate": 7.649787361728338e-05,
+      "loss": 0.694,
+      "step": 8342
+    },
+    {
+      "epoch": 1.4882704486664884,
+      "grad_norm": 0.4939239025115967,
+      "learning_rate": 7.644743791545141e-05,
+      "loss": 0.4921,
+      "step": 8343
+    },
+    {
+      "epoch": 1.4884488448844886,
+      "grad_norm": 0.43726587295532227,
+      "learning_rate": 7.639701584436292e-05,
+      "loss": 0.4466,
+      "step": 8344
+    },
+    {
+      "epoch": 1.4886272411024886,
+      "grad_norm": 0.48199325799942017,
+      "learning_rate": 7.634660740797817e-05,
+      "loss": 0.5736,
+      "step": 8345
+    },
+    {
+      "epoch": 1.4888056373204888,
+      "grad_norm": 0.4477761387825012,
+      "learning_rate": 7.629621261025613e-05,
+      "loss": 0.4244,
+      "step": 8346
+    },
+    {
+      "epoch": 1.488984033538489,
+      "grad_norm": 0.564702570438385,
+      "learning_rate": 7.624583145515474e-05,
+      "loss": 0.7607,
+      "step": 8347
+    },
+    {
+      "epoch": 1.4891624297564892,
+      "grad_norm": 0.4603510797023773,
+      "learning_rate": 7.619546394663104e-05,
+      "loss": 0.4789,
+      "step": 8348
+    },
+    {
+      "epoch": 1.4893408259744894,
+      "grad_norm": 0.4973684549331665,
+      "learning_rate": 7.614511008864073e-05,
+      "loss": 0.5349,
+      "step": 8349
+    },
+    {
+      "epoch": 1.4895192221924896,
+      "grad_norm": 0.4872196316719055,
+      "learning_rate": 7.609476988513875e-05,
+      "loss": 0.5265,
+      "step": 8350
+    },
+    {
+      "epoch": 1.4896976184104898,
+      "grad_norm": 0.5209683775901794,
+      "learning_rate": 7.604444334007862e-05,
+      "loss": 0.6186,
+      "step": 8351
+    },
+    {
+      "epoch": 1.48987601462849,
+      "grad_norm": 0.48873645067214966,
+      "learning_rate": 7.599413045741313e-05,
+      "loss": 0.5037,
+      "step": 8352
+    },
+    {
+      "epoch": 1.49005441084649,
+      "grad_norm": 0.45188388228416443,
+      "learning_rate": 7.594383124109375e-05,
+      "loss": 0.4037,
+      "step": 8353
+    },
+    {
+      "epoch": 1.4902328070644901,
+      "grad_norm": 0.5492420196533203,
+      "learning_rate": 7.589354569507081e-05,
+      "loss": 0.589,
+      "step": 8354
+    },
+    {
+      "epoch": 1.4904112032824903,
+      "grad_norm": 0.532741904258728,
+      "learning_rate": 7.584327382329401e-05,
+      "loss": 0.6725,
+      "step": 8355
+    },
+    {
+      "epoch": 1.4905895995004905,
+      "grad_norm": 0.5740191340446472,
+      "learning_rate": 7.579301562971147e-05,
+      "loss": 0.6417,
+      "step": 8356
+    },
+    {
+      "epoch": 1.4907679957184907,
+      "grad_norm": 0.4454120099544525,
+      "learning_rate": 7.57427711182706e-05,
+      "loss": 0.397,
+      "step": 8357
+    },
+    {
+      "epoch": 1.490946391936491,
+      "grad_norm": 0.49578696489334106,
+      "learning_rate": 7.56925402929175e-05,
+      "loss": 0.4834,
+      "step": 8358
+    },
+    {
+      "epoch": 1.4911247881544911,
+      "grad_norm": 0.5603286623954773,
+      "learning_rate": 7.564232315759718e-05,
+      "loss": 0.4489,
+      "step": 8359
+    },
+    {
+      "epoch": 1.4913031843724913,
+      "grad_norm": 0.5061764717102051,
+      "learning_rate": 7.559211971625385e-05,
+      "loss": 0.5227,
+      "step": 8360
+    },
+    {
+      "epoch": 1.4914815805904915,
+      "grad_norm": 0.48216429352760315,
+      "learning_rate": 7.554192997283033e-05,
+      "loss": 0.605,
+      "step": 8361
+    },
+    {
+      "epoch": 1.4916599768084917,
+      "grad_norm": 0.47625431418418884,
+      "learning_rate": 7.549175393126861e-05,
+      "loss": 0.4631,
+      "step": 8362
+    },
+    {
+      "epoch": 1.491838373026492,
+      "grad_norm": 0.5075172781944275,
+      "learning_rate": 7.544159159550937e-05,
+      "loss": 0.5409,
+      "step": 8363
+    },
+    {
+      "epoch": 1.4920167692444921,
+      "grad_norm": 0.5858179330825806,
+      "learning_rate": 7.539144296949246e-05,
+      "loss": 0.6648,
+      "step": 8364
+    },
+    {
+      "epoch": 1.4921951654624923,
+      "grad_norm": 0.47895708680152893,
+      "learning_rate": 7.534130805715644e-05,
+      "loss": 0.5248,
+      "step": 8365
+    },
+    {
+      "epoch": 1.4923735616804925,
+      "grad_norm": 0.4535577893257141,
+      "learning_rate": 7.529118686243897e-05,
+      "loss": 0.5158,
+      "step": 8366
+    },
+    {
+      "epoch": 1.4925519578984925,
+      "grad_norm": 0.4920910894870758,
+      "learning_rate": 7.524107938927652e-05,
+      "loss": 0.5183,
+      "step": 8367
+    },
+    {
+      "epoch": 1.4927303541164927,
+      "grad_norm": 0.49673569202423096,
+      "learning_rate": 7.51909856416044e-05,
+      "loss": 0.5476,
+      "step": 8368
+    },
+    {
+      "epoch": 1.4929087503344929,
+      "grad_norm": 0.5008490681648254,
+      "learning_rate": 7.514090562335712e-05,
+      "loss": 0.5373,
+      "step": 8369
+    },
+    {
+      "epoch": 1.493087146552493,
+      "grad_norm": 0.46472302079200745,
+      "learning_rate": 7.50908393384678e-05,
+      "loss": 0.4453,
+      "step": 8370
+    },
+    {
+      "epoch": 1.4932655427704933,
+      "grad_norm": 0.4681701362133026,
+      "learning_rate": 7.504078679086868e-05,
+      "loss": 0.5433,
+      "step": 8371
+    },
+    {
+      "epoch": 1.4934439389884935,
+      "grad_norm": 0.48634573817253113,
+      "learning_rate": 7.499074798449095e-05,
+      "loss": 0.4632,
+      "step": 8372
+    },
+    {
+      "epoch": 1.4936223352064937,
+      "grad_norm": 0.5008024573326111,
+      "learning_rate": 7.494072292326448e-05,
+      "loss": 0.5638,
+      "step": 8373
+    },
+    {
+      "epoch": 1.4938007314244939,
+      "grad_norm": 0.5322297811508179,
+      "learning_rate": 7.48907116111184e-05,
+      "loss": 0.642,
+      "step": 8374
+    },
+    {
+      "epoch": 1.4939791276424939,
+      "grad_norm": 0.5250205993652344,
+      "learning_rate": 7.484071405198037e-05,
+      "loss": 0.6071,
+      "step": 8375
+    },
+    {
+      "epoch": 1.494157523860494,
+      "grad_norm": 0.5197263360023499,
+      "learning_rate": 7.479073024977736e-05,
+      "loss": 0.6804,
+      "step": 8376
+    },
+    {
+      "epoch": 1.4943359200784943,
+      "grad_norm": 0.4836525022983551,
+      "learning_rate": 7.474076020843496e-05,
+      "loss": 0.5058,
+      "step": 8377
+    },
+    {
+      "epoch": 1.4945143162964944,
+      "grad_norm": 0.5223329663276672,
+      "learning_rate": 7.469080393187786e-05,
+      "loss": 0.6614,
+      "step": 8378
+    },
+    {
+      "epoch": 1.4946927125144946,
+      "grad_norm": 0.4857519865036011,
+      "learning_rate": 7.464086142402959e-05,
+      "loss": 0.5892,
+      "step": 8379
+    },
+    {
+      "epoch": 1.4948711087324948,
+      "grad_norm": 0.4820669889450073,
+      "learning_rate": 7.459093268881254e-05,
+      "loss": 0.5444,
+      "step": 8380
+    },
+    {
+      "epoch": 1.495049504950495,
+      "grad_norm": 0.4609326720237732,
+      "learning_rate": 7.45410177301482e-05,
+      "loss": 0.4923,
+      "step": 8381
+    },
+    {
+      "epoch": 1.4952279011684952,
+      "grad_norm": 0.47613245248794556,
+      "learning_rate": 7.449111655195678e-05,
+      "loss": 0.5083,
+      "step": 8382
+    },
+    {
+      "epoch": 1.4954062973864954,
+      "grad_norm": 0.496552973985672,
+      "learning_rate": 7.444122915815759e-05,
+      "loss": 0.5721,
+      "step": 8383
+    },
+    {
+      "epoch": 1.4955846936044956,
+      "grad_norm": 0.493670254945755,
+      "learning_rate": 7.439135555266866e-05,
+      "loss": 0.461,
+      "step": 8384
+    },
+    {
+      "epoch": 1.4957630898224958,
+      "grad_norm": 0.5373759269714355,
+      "learning_rate": 7.434149573940707e-05,
+      "loss": 0.5685,
+      "step": 8385
+    },
+    {
+      "epoch": 1.495941486040496,
+      "grad_norm": 0.5208349227905273,
+      "learning_rate": 7.429164972228891e-05,
+      "loss": 0.5965,
+      "step": 8386
+    },
+    {
+      "epoch": 1.4961198822584962,
+      "grad_norm": 0.5484915971755981,
+      "learning_rate": 7.424181750522887e-05,
+      "loss": 0.5944,
+      "step": 8387
+    },
+    {
+      "epoch": 1.4962982784764964,
+      "grad_norm": 0.5299747586250305,
+      "learning_rate": 7.419199909214095e-05,
+      "loss": 0.5318,
+      "step": 8388
+    },
+    {
+      "epoch": 1.4964766746944964,
+      "grad_norm": 0.5350375175476074,
+      "learning_rate": 7.414219448693769e-05,
+      "loss": 0.6534,
+      "step": 8389
+    },
+    {
+      "epoch": 1.4966550709124966,
+      "grad_norm": 0.5854753255844116,
+      "learning_rate": 7.409240369353084e-05,
+      "loss": 0.7582,
+      "step": 8390
+    },
+    {
+      "epoch": 1.4968334671304968,
+      "grad_norm": 0.5602213144302368,
+      "learning_rate": 7.404262671583092e-05,
+      "loss": 0.583,
+      "step": 8391
+    },
+    {
+      "epoch": 1.497011863348497,
+      "grad_norm": 0.4914671778678894,
+      "learning_rate": 7.399286355774732e-05,
+      "loss": 0.5397,
+      "step": 8392
+    },
+    {
+      "epoch": 1.4971902595664972,
+      "grad_norm": 0.47588059306144714,
+      "learning_rate": 7.394311422318853e-05,
+      "loss": 0.4886,
+      "step": 8393
+    },
+    {
+      "epoch": 1.4973686557844974,
+      "grad_norm": 0.5675820708274841,
+      "learning_rate": 7.389337871606172e-05,
+      "loss": 0.6568,
+      "step": 8394
+    },
+    {
+      "epoch": 1.4975470520024976,
+      "grad_norm": 0.508738100528717,
+      "learning_rate": 7.384365704027321e-05,
+      "loss": 0.5592,
+      "step": 8395
+    },
+    {
+      "epoch": 1.4977254482204978,
+      "grad_norm": 0.4801078140735626,
+      "learning_rate": 7.379394919972804e-05,
+      "loss": 0.4698,
+      "step": 8396
+    },
+    {
+      "epoch": 1.4979038444384978,
+      "grad_norm": 0.48387011885643005,
+      "learning_rate": 7.374425519833031e-05,
+      "loss": 0.551,
+      "step": 8397
+    },
+    {
+      "epoch": 1.498082240656498,
+      "grad_norm": 0.5585319399833679,
+      "learning_rate": 7.369457503998286e-05,
+      "loss": 0.7281,
+      "step": 8398
+    },
+    {
+      "epoch": 1.4982606368744982,
+      "grad_norm": 0.5239479541778564,
+      "learning_rate": 7.36449087285877e-05,
+      "loss": 0.6676,
+      "step": 8399
+    },
+    {
+      "epoch": 1.4984390330924984,
+      "grad_norm": 0.550262451171875,
+      "learning_rate": 7.359525626804544e-05,
+      "loss": 0.7478,
+      "step": 8400
+    },
+    {
+      "epoch": 1.4986174293104986,
+      "grad_norm": 0.45570600032806396,
+      "learning_rate": 7.354561766225584e-05,
+      "loss": 0.5162,
+      "step": 8401
+    },
+    {
+      "epoch": 1.4987958255284988,
+      "grad_norm": 0.43634846806526184,
+      "learning_rate": 7.349599291511757e-05,
+      "loss": 0.4103,
+      "step": 8402
+    },
+    {
+      "epoch": 1.498974221746499,
+      "grad_norm": 0.5149981379508972,
+      "learning_rate": 7.344638203052798e-05,
+      "loss": 0.594,
+      "step": 8403
+    },
+    {
+      "epoch": 1.4991526179644992,
+      "grad_norm": 0.4976111054420471,
+      "learning_rate": 7.339678501238364e-05,
+      "loss": 0.5622,
+      "step": 8404
+    },
+    {
+      "epoch": 1.4993310141824994,
+      "grad_norm": 0.5391610860824585,
+      "learning_rate": 7.334720186457982e-05,
+      "loss": 0.6742,
+      "step": 8405
+    },
+    {
+      "epoch": 1.4995094104004996,
+      "grad_norm": 0.4595610499382019,
+      "learning_rate": 7.329763259101069e-05,
+      "loss": 0.4807,
+      "step": 8406
+    },
+    {
+      "epoch": 1.4996878066184998,
+      "grad_norm": 0.49623727798461914,
+      "learning_rate": 7.32480771955695e-05,
+      "loss": 0.6094,
+      "step": 8407
+    },
+    {
+      "epoch": 1.4998662028365,
+      "grad_norm": 0.47364911437034607,
+      "learning_rate": 7.319853568214818e-05,
+      "loss": 0.5401,
+      "step": 8408
+    },
+    {
+      "epoch": 1.5000445990545002,
+      "grad_norm": 0.4699852764606476,
+      "learning_rate": 7.314900805463789e-05,
+      "loss": 0.4449,
+      "step": 8409
+    },
+    {
+      "epoch": 1.5002229952725004,
+      "grad_norm": 0.4462581276893616,
+      "learning_rate": 7.30994943169283e-05,
+      "loss": 0.4451,
+      "step": 8410
+    },
+    {
+      "epoch": 1.5004013914905006,
+      "grad_norm": 0.5866194367408752,
+      "learning_rate": 7.304999447290838e-05,
+      "loss": 0.6961,
+      "step": 8411
+    },
+    {
+      "epoch": 1.5005797877085005,
+      "grad_norm": 0.4362695515155792,
+      "learning_rate": 7.300050852646578e-05,
+      "loss": 0.3869,
+      "step": 8412
+    },
+    {
+      "epoch": 1.5007581839265007,
+      "grad_norm": 0.6117103099822998,
+      "learning_rate": 7.295103648148697e-05,
+      "loss": 0.6624,
+      "step": 8413
+    },
+    {
+      "epoch": 1.500936580144501,
+      "grad_norm": 0.49935221672058105,
+      "learning_rate": 7.290157834185763e-05,
+      "loss": 0.4923,
+      "step": 8414
+    },
+    {
+      "epoch": 1.5011149763625011,
+      "grad_norm": 0.5208079218864441,
+      "learning_rate": 7.285213411146205e-05,
+      "loss": 0.4946,
+      "step": 8415
+    },
+    {
+      "epoch": 1.5012933725805013,
+      "grad_norm": 0.6199221611022949,
+      "learning_rate": 7.280270379418363e-05,
+      "loss": 0.6581,
+      "step": 8416
+    },
+    {
+      "epoch": 1.5014717687985015,
+      "grad_norm": 0.49500852823257446,
+      "learning_rate": 7.275328739390465e-05,
+      "loss": 0.4407,
+      "step": 8417
+    },
+    {
+      "epoch": 1.5016501650165015,
+      "grad_norm": 0.4843553602695465,
+      "learning_rate": 7.270388491450616e-05,
+      "loss": 0.5689,
+      "step": 8418
+    },
+    {
+      "epoch": 1.5018285612345017,
+      "grad_norm": 0.5777263045310974,
+      "learning_rate": 7.265449635986831e-05,
+      "loss": 0.5386,
+      "step": 8419
+    },
+    {
+      "epoch": 1.502006957452502,
+      "grad_norm": 0.5656114220619202,
+      "learning_rate": 7.260512173386993e-05,
+      "loss": 0.6281,
+      "step": 8420
+    },
+    {
+      "epoch": 1.502185353670502,
+      "grad_norm": 0.48054200410842896,
+      "learning_rate": 7.255576104038902e-05,
+      "loss": 0.5009,
+      "step": 8421
+    },
+    {
+      "epoch": 1.5023637498885023,
+      "grad_norm": 0.47279128432273865,
+      "learning_rate": 7.25064142833022e-05,
+      "loss": 0.4902,
+      "step": 8422
+    },
+    {
+      "epoch": 1.5025421461065025,
+      "grad_norm": 0.5290525555610657,
+      "learning_rate": 7.24570814664853e-05,
+      "loss": 0.4631,
+      "step": 8423
+    },
+    {
+      "epoch": 1.5027205423245027,
+      "grad_norm": 0.46618005633354187,
+      "learning_rate": 7.240776259381276e-05,
+      "loss": 0.5708,
+      "step": 8424
+    },
+    {
+      "epoch": 1.502898938542503,
+      "grad_norm": 0.4757915139198303,
+      "learning_rate": 7.235845766915819e-05,
+      "loss": 0.499,
+      "step": 8425
+    },
+    {
+      "epoch": 1.503077334760503,
+      "grad_norm": 0.4191453754901886,
+      "learning_rate": 7.23091666963939e-05,
+      "loss": 0.3913,
+      "step": 8426
+    },
+    {
+      "epoch": 1.5032557309785033,
+      "grad_norm": 0.4610666036605835,
+      "learning_rate": 7.225988967939113e-05,
+      "loss": 0.3548,
+      "step": 8427
+    },
+    {
+      "epoch": 1.5034341271965035,
+      "grad_norm": 0.488862544298172,
+      "learning_rate": 7.221062662202018e-05,
+      "loss": 0.5241,
+      "step": 8428
+    },
+    {
+      "epoch": 1.5036125234145037,
+      "grad_norm": 0.5716550946235657,
+      "learning_rate": 7.216137752815005e-05,
+      "loss": 0.7528,
+      "step": 8429
+    },
+    {
+      "epoch": 1.5037909196325039,
+      "grad_norm": 0.5017661452293396,
+      "learning_rate": 7.211214240164887e-05,
+      "loss": 0.5129,
+      "step": 8430
+    },
+    {
+      "epoch": 1.503969315850504,
+      "grad_norm": 0.4955231547355652,
+      "learning_rate": 7.206292124638342e-05,
+      "loss": 0.5957,
+      "step": 8431
+    },
+    {
+      "epoch": 1.5041477120685043,
+      "grad_norm": 0.5126091837882996,
+      "learning_rate": 7.201371406621954e-05,
+      "loss": 0.6606,
+      "step": 8432
+    },
+    {
+      "epoch": 1.5043261082865045,
+      "grad_norm": 0.4719708561897278,
+      "learning_rate": 7.196452086502206e-05,
+      "loss": 0.4889,
+      "step": 8433
+    },
+    {
+      "epoch": 1.5045045045045045,
+      "grad_norm": 0.4715104401111603,
+      "learning_rate": 7.191534164665439e-05,
+      "loss": 0.4824,
+      "step": 8434
+    },
+    {
+      "epoch": 1.5046829007225047,
+      "grad_norm": 0.49331334233283997,
+      "learning_rate": 7.186617641497926e-05,
+      "loss": 0.5115,
+      "step": 8435
+    },
+    {
+      "epoch": 1.5048612969405049,
+      "grad_norm": 0.4969932436943054,
+      "learning_rate": 7.181702517385788e-05,
+      "loss": 0.5298,
+      "step": 8436
+    },
+    {
+      "epoch": 1.505039693158505,
+      "grad_norm": 0.5154744386672974,
+      "learning_rate": 7.176788792715074e-05,
+      "loss": 0.5574,
+      "step": 8437
+    },
+    {
+      "epoch": 1.5052180893765053,
+      "grad_norm": 0.5841989517211914,
+      "learning_rate": 7.171876467871699e-05,
+      "loss": 0.5517,
+      "step": 8438
+    },
+    {
+      "epoch": 1.5053964855945055,
+      "grad_norm": 0.5319039225578308,
+      "learning_rate": 7.166965543241466e-05,
+      "loss": 0.7089,
+      "step": 8439
+    },
+    {
+      "epoch": 1.5055748818125054,
+      "grad_norm": 0.5384335517883301,
+      "learning_rate": 7.162056019210095e-05,
+      "loss": 0.604,
+      "step": 8440
+    },
+    {
+      "epoch": 1.5057532780305056,
+      "grad_norm": 0.48112642765045166,
+      "learning_rate": 7.157147896163157e-05,
+      "loss": 0.464,
+      "step": 8441
+    },
+    {
+      "epoch": 1.5059316742485058,
+      "grad_norm": 0.519540011882782,
+      "learning_rate": 7.152241174486154e-05,
+      "loss": 0.5771,
+      "step": 8442
+    },
+    {
+      "epoch": 1.506110070466506,
+      "grad_norm": 0.49812787771224976,
+      "learning_rate": 7.147335854564444e-05,
+      "loss": 0.5615,
+      "step": 8443
+    },
+    {
+      "epoch": 1.5062884666845062,
+      "grad_norm": 0.4924754500389099,
+      "learning_rate": 7.142431936783297e-05,
+      "loss": 0.4846,
+      "step": 8444
+    },
+    {
+      "epoch": 1.5064668629025064,
+      "grad_norm": 0.5007926225662231,
+      "learning_rate": 7.137529421527852e-05,
+      "loss": 0.5509,
+      "step": 8445
+    },
+    {
+      "epoch": 1.5066452591205066,
+      "grad_norm": 0.4786141514778137,
+      "learning_rate": 7.132628309183165e-05,
+      "loss": 0.4272,
+      "step": 8446
+    },
+    {
+      "epoch": 1.5068236553385068,
+      "grad_norm": 0.4895467758178711,
+      "learning_rate": 7.127728600134164e-05,
+      "loss": 0.4545,
+      "step": 8447
+    },
+    {
+      "epoch": 1.507002051556507,
+      "grad_norm": 0.4716068506240845,
+      "learning_rate": 7.122830294765664e-05,
+      "loss": 0.4106,
+      "step": 8448
+    },
+    {
+      "epoch": 1.5071804477745072,
+      "grad_norm": 0.5841825008392334,
+      "learning_rate": 7.117933393462384e-05,
+      "loss": 0.7319,
+      "step": 8449
+    },
+    {
+      "epoch": 1.5073588439925074,
+      "grad_norm": 0.4899286925792694,
+      "learning_rate": 7.113037896608923e-05,
+      "loss": 0.4271,
+      "step": 8450
+    },
+    {
+      "epoch": 1.5075372402105076,
+      "grad_norm": 0.5296789407730103,
+      "learning_rate": 7.108143804589759e-05,
+      "loss": 0.5624,
+      "step": 8451
+    },
+    {
+      "epoch": 1.5077156364285078,
+      "grad_norm": 0.48504510521888733,
+      "learning_rate": 7.10325111778929e-05,
+      "loss": 0.5338,
+      "step": 8452
+    },
+    {
+      "epoch": 1.507894032646508,
+      "grad_norm": 0.5016950368881226,
+      "learning_rate": 7.098359836591764e-05,
+      "loss": 0.5561,
+      "step": 8453
+    },
+    {
+      "epoch": 1.5080724288645082,
+      "grad_norm": 0.5088602900505066,
+      "learning_rate": 7.093469961381365e-05,
+      "loss": 0.6215,
+      "step": 8454
+    },
+    {
+      "epoch": 1.5082508250825084,
+      "grad_norm": 0.6125203371047974,
+      "learning_rate": 7.088581492542121e-05,
+      "loss": 0.6588,
+      "step": 8455
+    },
+    {
+      "epoch": 1.5084292213005084,
+      "grad_norm": 0.8469102382659912,
+      "learning_rate": 7.083694430457988e-05,
+      "loss": 0.5289,
+      "step": 8456
+    },
+    {
+      "epoch": 1.5086076175185086,
+      "grad_norm": 0.49744218587875366,
+      "learning_rate": 7.078808775512774e-05,
+      "loss": 0.6164,
+      "step": 8457
+    },
+    {
+      "epoch": 1.5087860137365088,
+      "grad_norm": 0.5175665616989136,
+      "learning_rate": 7.073924528090214e-05,
+      "loss": 0.5147,
+      "step": 8458
+    },
+    {
+      "epoch": 1.508964409954509,
+      "grad_norm": 0.492880254983902,
+      "learning_rate": 7.06904168857391e-05,
+      "loss": 0.5535,
+      "step": 8459
+    },
+    {
+      "epoch": 1.5091428061725092,
+      "grad_norm": 0.5100347399711609,
+      "learning_rate": 7.064160257347346e-05,
+      "loss": 0.6079,
+      "step": 8460
+    },
+    {
+      "epoch": 1.5093212023905094,
+      "grad_norm": 0.4987751543521881,
+      "learning_rate": 7.059280234793927e-05,
+      "loss": 0.5412,
+      "step": 8461
+    },
+    {
+      "epoch": 1.5094995986085094,
+      "grad_norm": 0.4200221002101898,
+      "learning_rate": 7.054401621296899e-05,
+      "loss": 0.3935,
+      "step": 8462
+    },
+    {
+      "epoch": 1.5096779948265096,
+      "grad_norm": 0.5284053087234497,
+      "learning_rate": 7.049524417239465e-05,
+      "loss": 0.6021,
+      "step": 8463
+    },
+    {
+      "epoch": 1.5098563910445097,
+      "grad_norm": 0.5351274013519287,
+      "learning_rate": 7.044648623004654e-05,
+      "loss": 0.6754,
+      "step": 8464
+    },
+    {
+      "epoch": 1.51003478726251,
+      "grad_norm": 0.5108939409255981,
+      "learning_rate": 7.039774238975408e-05,
+      "loss": 0.6219,
+      "step": 8465
+    },
+    {
+      "epoch": 1.5102131834805101,
+      "grad_norm": 0.5333287119865417,
+      "learning_rate": 7.034901265534571e-05,
+      "loss": 0.581,
+      "step": 8466
+    },
+    {
+      "epoch": 1.5103915796985103,
+      "grad_norm": 0.6006919145584106,
+      "learning_rate": 7.030029703064849e-05,
+      "loss": 0.6058,
+      "step": 8467
+    },
+    {
+      "epoch": 1.5105699759165105,
+      "grad_norm": 0.4753846526145935,
+      "learning_rate": 7.02515955194887e-05,
+      "loss": 0.4262,
+      "step": 8468
+    },
+    {
+      "epoch": 1.5107483721345107,
+      "grad_norm": 0.48568055033683777,
+      "learning_rate": 7.020290812569119e-05,
+      "loss": 0.4365,
+      "step": 8469
+    },
+    {
+      "epoch": 1.510926768352511,
+      "grad_norm": 0.488115519285202,
+      "learning_rate": 7.015423485307996e-05,
+      "loss": 0.5195,
+      "step": 8470
+    },
+    {
+      "epoch": 1.5111051645705111,
+      "grad_norm": 0.5295347571372986,
+      "learning_rate": 7.010557570547774e-05,
+      "loss": 0.5334,
+      "step": 8471
+    },
+    {
+      "epoch": 1.5112835607885113,
+      "grad_norm": 0.49440649151802063,
+      "learning_rate": 7.00569306867061e-05,
+      "loss": 0.546,
+      "step": 8472
+    },
+    {
+      "epoch": 1.5114619570065115,
+      "grad_norm": 0.5544924736022949,
+      "learning_rate": 7.000829980058576e-05,
+      "loss": 0.5707,
+      "step": 8473
+    },
+    {
+      "epoch": 1.5116403532245117,
+      "grad_norm": 0.6297951936721802,
+      "learning_rate": 6.995968305093603e-05,
+      "loss": 0.5151,
+      "step": 8474
+    },
+    {
+      "epoch": 1.511818749442512,
+      "grad_norm": 0.4706231653690338,
+      "learning_rate": 6.991108044157537e-05,
+      "loss": 0.5719,
+      "step": 8475
+    },
+    {
+      "epoch": 1.5119971456605121,
+      "grad_norm": 0.5109243988990784,
+      "learning_rate": 6.986249197632092e-05,
+      "loss": 0.6104,
+      "step": 8476
+    },
+    {
+      "epoch": 1.5121755418785123,
+      "grad_norm": 0.5036676526069641,
+      "learning_rate": 6.981391765898881e-05,
+      "loss": 0.6026,
+      "step": 8477
+    },
+    {
+      "epoch": 1.5123539380965123,
+      "grad_norm": 0.5482192635536194,
+      "learning_rate": 6.976535749339413e-05,
+      "loss": 0.5634,
+      "step": 8478
+    },
+    {
+      "epoch": 1.5125323343145125,
+      "grad_norm": 0.47447332739830017,
+      "learning_rate": 6.971681148335066e-05,
+      "loss": 0.4998,
+      "step": 8479
+    },
+    {
+      "epoch": 1.5127107305325127,
+      "grad_norm": 0.46494314074516296,
+      "learning_rate": 6.96682796326713e-05,
+      "loss": 0.4863,
+      "step": 8480
+    },
+    {
+      "epoch": 1.512889126750513,
+      "grad_norm": 0.525435745716095,
+      "learning_rate": 6.961976194516759e-05,
+      "loss": 0.5581,
+      "step": 8481
+    },
+    {
+      "epoch": 1.513067522968513,
+      "grad_norm": 0.5048510432243347,
+      "learning_rate": 6.95712584246502e-05,
+      "loss": 0.5747,
+      "step": 8482
+    },
+    {
+      "epoch": 1.5132459191865133,
+      "grad_norm": 0.4954485297203064,
+      "learning_rate": 6.952276907492846e-05,
+      "loss": 0.504,
+      "step": 8483
+    },
+    {
+      "epoch": 1.5134243154045133,
+      "grad_norm": 0.5358535051345825,
+      "learning_rate": 6.947429389981085e-05,
+      "loss": 0.6579,
+      "step": 8484
+    },
+    {
+      "epoch": 1.5136027116225135,
+      "grad_norm": 0.4358263611793518,
+      "learning_rate": 6.942583290310453e-05,
+      "loss": 0.4034,
+      "step": 8485
+    },
+    {
+      "epoch": 1.5137811078405137,
+      "grad_norm": 0.43710726499557495,
+      "learning_rate": 6.937738608861552e-05,
+      "loss": 0.4148,
+      "step": 8486
+    },
+    {
+      "epoch": 1.5139595040585139,
+      "grad_norm": 0.4802948832511902,
+      "learning_rate": 6.932895346014893e-05,
+      "loss": 0.4326,
+      "step": 8487
+    },
+    {
+      "epoch": 1.514137900276514,
+      "grad_norm": 0.515270471572876,
+      "learning_rate": 6.928053502150849e-05,
+      "loss": 0.4963,
+      "step": 8488
+    },
+    {
+      "epoch": 1.5143162964945143,
+      "grad_norm": 0.5241411924362183,
+      "learning_rate": 6.923213077649718e-05,
+      "loss": 0.5903,
+      "step": 8489
+    },
+    {
+      "epoch": 1.5144946927125145,
+      "grad_norm": 0.5037186741828918,
+      "learning_rate": 6.918374072891643e-05,
+      "loss": 0.6109,
+      "step": 8490
+    },
+    {
+      "epoch": 1.5146730889305147,
+      "grad_norm": 0.586534857749939,
+      "learning_rate": 6.913536488256695e-05,
+      "loss": 0.6432,
+      "step": 8491
+    },
+    {
+      "epoch": 1.5148514851485149,
+      "grad_norm": 0.49035558104515076,
+      "learning_rate": 6.9087003241248e-05,
+      "loss": 0.4847,
+      "step": 8492
+    },
+    {
+      "epoch": 1.515029881366515,
+      "grad_norm": 0.5941229462623596,
+      "learning_rate": 6.903865580875795e-05,
+      "loss": 0.7171,
+      "step": 8493
+    },
+    {
+      "epoch": 1.5152082775845153,
+      "grad_norm": 0.4788690209388733,
+      "learning_rate": 6.899032258889409e-05,
+      "loss": 0.5057,
+      "step": 8494
+    },
+    {
+      "epoch": 1.5153866738025155,
+      "grad_norm": 0.6357868909835815,
+      "learning_rate": 6.894200358545233e-05,
+      "loss": 0.7302,
+      "step": 8495
+    },
+    {
+      "epoch": 1.5155650700205157,
+      "grad_norm": 0.5005565881729126,
+      "learning_rate": 6.889369880222776e-05,
+      "loss": 0.4858,
+      "step": 8496
+    },
+    {
+      "epoch": 1.5157434662385159,
+      "grad_norm": 0.5777773857116699,
+      "learning_rate": 6.884540824301416e-05,
+      "loss": 0.6461,
+      "step": 8497
+    },
+    {
+      "epoch": 1.515921862456516,
+      "grad_norm": 0.4599872827529907,
+      "learning_rate": 6.879713191160417e-05,
+      "loss": 0.5051,
+      "step": 8498
+    },
+    {
+      "epoch": 1.5161002586745163,
+      "grad_norm": 0.44417500495910645,
+      "learning_rate": 6.874886981178952e-05,
+      "loss": 0.4822,
+      "step": 8499
+    },
+    {
+      "epoch": 1.5162786548925162,
+      "grad_norm": 0.46541309356689453,
+      "learning_rate": 6.870062194736057e-05,
+      "loss": 0.4517,
+      "step": 8500
+    },
+    {
+      "epoch": 1.5164570511105164,
+      "grad_norm": 0.6577228903770447,
+      "learning_rate": 6.865238832210682e-05,
+      "loss": 0.5325,
+      "step": 8501
+    },
+    {
+      "epoch": 1.5166354473285166,
+      "grad_norm": 0.43800440430641174,
+      "learning_rate": 6.860416893981638e-05,
+      "loss": 0.4597,
+      "step": 8502
+    },
+    {
+      "epoch": 1.5168138435465168,
+      "grad_norm": 0.46744707226753235,
+      "learning_rate": 6.855596380427651e-05,
+      "loss": 0.5437,
+      "step": 8503
+    },
+    {
+      "epoch": 1.516992239764517,
+      "grad_norm": 0.48990002274513245,
+      "learning_rate": 6.85077729192731e-05,
+      "loss": 0.5202,
+      "step": 8504
+    },
+    {
+      "epoch": 1.5171706359825172,
+      "grad_norm": 0.49149778485298157,
+      "learning_rate": 6.845959628859119e-05,
+      "loss": 0.5856,
+      "step": 8505
+    },
+    {
+      "epoch": 1.5173490322005172,
+      "grad_norm": 0.5475661158561707,
+      "learning_rate": 6.841143391601445e-05,
+      "loss": 0.7378,
+      "step": 8506
+    },
+    {
+      "epoch": 1.5175274284185174,
+      "grad_norm": 0.5574973821640015,
+      "learning_rate": 6.836328580532547e-05,
+      "loss": 0.5967,
+      "step": 8507
+    },
+    {
+      "epoch": 1.5177058246365176,
+      "grad_norm": 0.522175133228302,
+      "learning_rate": 6.831515196030588e-05,
+      "loss": 0.6041,
+      "step": 8508
+    },
+    {
+      "epoch": 1.5178842208545178,
+      "grad_norm": 0.5970126390457153,
+      "learning_rate": 6.826703238473605e-05,
+      "loss": 0.8074,
+      "step": 8509
+    },
+    {
+      "epoch": 1.518062617072518,
+      "grad_norm": 0.5063888430595398,
+      "learning_rate": 6.821892708239535e-05,
+      "loss": 0.5915,
+      "step": 8510
+    },
+    {
+      "epoch": 1.5182410132905182,
+      "grad_norm": 0.5437260866165161,
+      "learning_rate": 6.817083605706193e-05,
+      "loss": 0.6038,
+      "step": 8511
+    },
+    {
+      "epoch": 1.5184194095085184,
+      "grad_norm": 0.49606233835220337,
+      "learning_rate": 6.812275931251268e-05,
+      "loss": 0.5352,
+      "step": 8512
+    },
+    {
+      "epoch": 1.5185978057265186,
+      "grad_norm": 0.509730339050293,
+      "learning_rate": 6.807469685252376e-05,
+      "loss": 0.5743,
+      "step": 8513
+    },
+    {
+      "epoch": 1.5187762019445188,
+      "grad_norm": 0.5043073296546936,
+      "learning_rate": 6.802664868086978e-05,
+      "loss": 0.496,
+      "step": 8514
+    },
+    {
+      "epoch": 1.518954598162519,
+      "grad_norm": 0.5181002020835876,
+      "learning_rate": 6.797861480132456e-05,
+      "loss": 0.5839,
+      "step": 8515
+    },
+    {
+      "epoch": 1.5191329943805192,
+      "grad_norm": 0.5447046756744385,
+      "learning_rate": 6.793059521766054e-05,
+      "loss": 0.6502,
+      "step": 8516
+    },
+    {
+      "epoch": 1.5193113905985194,
+      "grad_norm": 0.5003826022148132,
+      "learning_rate": 6.788258993364929e-05,
+      "loss": 0.5199,
+      "step": 8517
+    },
+    {
+      "epoch": 1.5194897868165196,
+      "grad_norm": 0.5281699299812317,
+      "learning_rate": 6.783459895306107e-05,
+      "loss": 0.6103,
+      "step": 8518
+    },
+    {
+      "epoch": 1.5196681830345198,
+      "grad_norm": 0.4539282023906708,
+      "learning_rate": 6.778662227966495e-05,
+      "loss": 0.4092,
+      "step": 8519
+    },
+    {
+      "epoch": 1.51984657925252,
+      "grad_norm": 0.49163827300071716,
+      "learning_rate": 6.773865991722921e-05,
+      "loss": 0.4445,
+      "step": 8520
+    },
+    {
+      "epoch": 1.5200249754705202,
+      "grad_norm": 0.47441521286964417,
+      "learning_rate": 6.76907118695206e-05,
+      "loss": 0.4945,
+      "step": 8521
+    },
+    {
+      "epoch": 1.5202033716885202,
+      "grad_norm": 0.6085272431373596,
+      "learning_rate": 6.76427781403051e-05,
+      "loss": 0.7252,
+      "step": 8522
+    },
+    {
+      "epoch": 1.5203817679065204,
+      "grad_norm": 0.5102150440216064,
+      "learning_rate": 6.759485873334725e-05,
+      "loss": 0.4936,
+      "step": 8523
+    },
+    {
+      "epoch": 1.5205601641245206,
+      "grad_norm": 0.48524045944213867,
+      "learning_rate": 6.754695365241071e-05,
+      "loss": 0.5631,
+      "step": 8524
+    },
+    {
+      "epoch": 1.5207385603425208,
+      "grad_norm": 0.433088481426239,
+      "learning_rate": 6.749906290125799e-05,
+      "loss": 0.4493,
+      "step": 8525
+    },
+    {
+      "epoch": 1.520916956560521,
+      "grad_norm": 0.5861080288887024,
+      "learning_rate": 6.745118648365026e-05,
+      "loss": 0.6544,
+      "step": 8526
+    },
+    {
+      "epoch": 1.5210953527785211,
+      "grad_norm": 0.4654527008533478,
+      "learning_rate": 6.740332440334784e-05,
+      "loss": 0.4399,
+      "step": 8527
+    },
+    {
+      "epoch": 1.5212737489965211,
+      "grad_norm": 0.5457956194877625,
+      "learning_rate": 6.735547666410968e-05,
+      "loss": 0.6563,
+      "step": 8528
+    },
+    {
+      "epoch": 1.5214521452145213,
+      "grad_norm": 0.6288897395133972,
+      "learning_rate": 6.730764326969388e-05,
+      "loss": 0.7027,
+      "step": 8529
+    },
+    {
+      "epoch": 1.5216305414325215,
+      "grad_norm": 0.5368294715881348,
+      "learning_rate": 6.725982422385715e-05,
+      "loss": 0.6109,
+      "step": 8530
+    },
+    {
+      "epoch": 1.5218089376505217,
+      "grad_norm": 0.5222885012626648,
+      "learning_rate": 6.721201953035511e-05,
+      "loss": 0.5362,
+      "step": 8531
+    },
+    {
+      "epoch": 1.521987333868522,
+      "grad_norm": 0.49758774042129517,
+      "learning_rate": 6.716422919294247e-05,
+      "loss": 0.4813,
+      "step": 8532
+    },
+    {
+      "epoch": 1.5221657300865221,
+      "grad_norm": 0.5573477745056152,
+      "learning_rate": 6.71164532153725e-05,
+      "loss": 0.7868,
+      "step": 8533
+    },
+    {
+      "epoch": 1.5223441263045223,
+      "grad_norm": 0.46816423535346985,
+      "learning_rate": 6.706869160139767e-05,
+      "loss": 0.4346,
+      "step": 8534
+    },
+    {
+      "epoch": 1.5225225225225225,
+      "grad_norm": 0.521041750907898,
+      "learning_rate": 6.702094435476902e-05,
+      "loss": 0.5432,
+      "step": 8535
+    },
+    {
+      "epoch": 1.5227009187405227,
+      "grad_norm": 0.49728789925575256,
+      "learning_rate": 6.697321147923671e-05,
+      "loss": 0.6175,
+      "step": 8536
+    },
+    {
+      "epoch": 1.522879314958523,
+      "grad_norm": 0.5513471364974976,
+      "learning_rate": 6.692549297854956e-05,
+      "loss": 0.6744,
+      "step": 8537
+    },
+    {
+      "epoch": 1.523057711176523,
+      "grad_norm": 1.1942397356033325,
+      "learning_rate": 6.68777888564554e-05,
+      "loss": 0.6059,
+      "step": 8538
+    },
+    {
+      "epoch": 1.5232361073945233,
+      "grad_norm": 0.5422153472900391,
+      "learning_rate": 6.683009911670095e-05,
+      "loss": 0.5448,
+      "step": 8539
+    },
+    {
+      "epoch": 1.5234145036125235,
+      "grad_norm": 0.5084717869758606,
+      "learning_rate": 6.678242376303165e-05,
+      "loss": 0.5318,
+      "step": 8540
+    },
+    {
+      "epoch": 1.5235928998305237,
+      "grad_norm": 0.4946412444114685,
+      "learning_rate": 6.673476279919202e-05,
+      "loss": 0.4886,
+      "step": 8541
+    },
+    {
+      "epoch": 1.523771296048524,
+      "grad_norm": 0.5950092673301697,
+      "learning_rate": 6.668711622892515e-05,
+      "loss": 0.7057,
+      "step": 8542
+    },
+    {
+      "epoch": 1.523949692266524,
+      "grad_norm": 0.45473888516426086,
+      "learning_rate": 6.663948405597339e-05,
+      "loss": 0.4004,
+      "step": 8543
+    },
+    {
+      "epoch": 1.524128088484524,
+      "grad_norm": 0.5750323534011841,
+      "learning_rate": 6.659186628407762e-05,
+      "loss": 0.6046,
+      "step": 8544
+    },
+    {
+      "epoch": 1.5243064847025243,
+      "grad_norm": 0.5322027802467346,
+      "learning_rate": 6.654426291697768e-05,
+      "loss": 0.5181,
+      "step": 8545
+    },
+    {
+      "epoch": 1.5244848809205245,
+      "grad_norm": 0.4996216893196106,
+      "learning_rate": 6.649667395841247e-05,
+      "loss": 0.5104,
+      "step": 8546
+    },
+    {
+      "epoch": 1.5246632771385247,
+      "grad_norm": 0.5934137105941772,
+      "learning_rate": 6.644909941211943e-05,
+      "loss": 0.7318,
+      "step": 8547
+    },
+    {
+      "epoch": 1.5248416733565249,
+      "grad_norm": 0.4300279915332794,
+      "learning_rate": 6.640153928183523e-05,
+      "loss": 0.3769,
+      "step": 8548
+    },
+    {
+      "epoch": 1.525020069574525,
+      "grad_norm": 0.5084002614021301,
+      "learning_rate": 6.635399357129501e-05,
+      "loss": 0.4071,
+      "step": 8549
+    },
+    {
+      "epoch": 1.525198465792525,
+      "grad_norm": 0.47641390562057495,
+      "learning_rate": 6.630646228423323e-05,
+      "loss": 0.6074,
+      "step": 8550
+    },
+    {
+      "epoch": 1.5253768620105252,
+      "grad_norm": 0.5083361268043518,
+      "learning_rate": 6.625894542438283e-05,
+      "loss": 0.657,
+      "step": 8551
+    },
+    {
+      "epoch": 1.5255552582285254,
+      "grad_norm": 0.5362805128097534,
+      "learning_rate": 6.621144299547572e-05,
+      "loss": 0.6543,
+      "step": 8552
+    },
+    {
+      "epoch": 1.5257336544465256,
+      "grad_norm": 0.46762344241142273,
+      "learning_rate": 6.616395500124276e-05,
+      "loss": 0.4705,
+      "step": 8553
+    },
+    {
+      "epoch": 1.5259120506645258,
+      "grad_norm": 0.41276419162750244,
+      "learning_rate": 6.611648144541369e-05,
+      "loss": 0.4374,
+      "step": 8554
+    },
+    {
+      "epoch": 1.526090446882526,
+      "grad_norm": 0.5008599758148193,
+      "learning_rate": 6.60690223317171e-05,
+      "loss": 0.5399,
+      "step": 8555
+    },
+    {
+      "epoch": 1.5262688431005262,
+      "grad_norm": 0.8646233677864075,
+      "learning_rate": 6.602157766388034e-05,
+      "loss": 0.4534,
+      "step": 8556
+    },
+    {
+      "epoch": 1.5264472393185264,
+      "grad_norm": 0.4763386845588684,
+      "learning_rate": 6.597414744562963e-05,
+      "loss": 0.5278,
+      "step": 8557
+    },
+    {
+      "epoch": 1.5266256355365266,
+      "grad_norm": 0.5339080691337585,
+      "learning_rate": 6.592673168069027e-05,
+      "loss": 0.555,
+      "step": 8558
+    },
+    {
+      "epoch": 1.5268040317545268,
+      "grad_norm": 0.5438008904457092,
+      "learning_rate": 6.587933037278609e-05,
+      "loss": 0.6286,
+      "step": 8559
+    },
+    {
+      "epoch": 1.526982427972527,
+      "grad_norm": 0.4510871469974518,
+      "learning_rate": 6.583194352564017e-05,
+      "loss": 0.4141,
+      "step": 8560
+    },
+    {
+      "epoch": 1.5271608241905272,
+      "grad_norm": 0.5599578619003296,
+      "learning_rate": 6.578457114297407e-05,
+      "loss": 0.619,
+      "step": 8561
+    },
+    {
+      "epoch": 1.5273392204085274,
+      "grad_norm": 0.5268929600715637,
+      "learning_rate": 6.573721322850854e-05,
+      "loss": 0.6176,
+      "step": 8562
+    },
+    {
+      "epoch": 1.5275176166265276,
+      "grad_norm": 0.5147603154182434,
+      "learning_rate": 6.568986978596291e-05,
+      "loss": 0.5238,
+      "step": 8563
+    },
+    {
+      "epoch": 1.5276960128445278,
+      "grad_norm": 0.5306607484817505,
+      "learning_rate": 6.564254081905571e-05,
+      "loss": 0.5801,
+      "step": 8564
+    },
+    {
+      "epoch": 1.527874409062528,
+      "grad_norm": 0.5111916065216064,
+      "learning_rate": 6.559522633150397e-05,
+      "loss": 0.5071,
+      "step": 8565
+    },
+    {
+      "epoch": 1.528052805280528,
+      "grad_norm": 0.48746898770332336,
+      "learning_rate": 6.554792632702376e-05,
+      "loss": 0.5161,
+      "step": 8566
+    },
+    {
+      "epoch": 1.5282312014985282,
+      "grad_norm": 0.48005223274230957,
+      "learning_rate": 6.55006408093301e-05,
+      "loss": 0.4149,
+      "step": 8567
+    },
+    {
+      "epoch": 1.5284095977165284,
+      "grad_norm": 0.5667544007301331,
+      "learning_rate": 6.545336978213664e-05,
+      "loss": 0.6735,
+      "step": 8568
+    },
+    {
+      "epoch": 1.5285879939345286,
+      "grad_norm": 0.5892688035964966,
+      "learning_rate": 6.54061132491561e-05,
+      "loss": 0.5779,
+      "step": 8569
+    },
+    {
+      "epoch": 1.5287663901525288,
+      "grad_norm": 0.4722636342048645,
+      "learning_rate": 6.535887121410006e-05,
+      "loss": 0.4264,
+      "step": 8570
+    },
+    {
+      "epoch": 1.528944786370529,
+      "grad_norm": 0.5367898941040039,
+      "learning_rate": 6.531164368067874e-05,
+      "loss": 0.5761,
+      "step": 8571
+    },
+    {
+      "epoch": 1.529123182588529,
+      "grad_norm": 0.5178304314613342,
+      "learning_rate": 6.526443065260154e-05,
+      "loss": 0.5354,
+      "step": 8572
+    },
+    {
+      "epoch": 1.5293015788065292,
+      "grad_norm": 0.4545149803161621,
+      "learning_rate": 6.521723213357635e-05,
+      "loss": 0.3896,
+      "step": 8573
+    },
+    {
+      "epoch": 1.5294799750245294,
+      "grad_norm": 0.5091972947120667,
+      "learning_rate": 6.517004812731034e-05,
+      "loss": 0.6283,
+      "step": 8574
+    },
+    {
+      "epoch": 1.5296583712425296,
+      "grad_norm": 0.5601230263710022,
+      "learning_rate": 6.512287863750912e-05,
+      "loss": 0.6376,
+      "step": 8575
+    },
+    {
+      "epoch": 1.5298367674605298,
+      "grad_norm": 0.44082170724868774,
+      "learning_rate": 6.507572366787753e-05,
+      "loss": 0.412,
+      "step": 8576
+    },
+    {
+      "epoch": 1.53001516367853,
+      "grad_norm": 0.6141606569290161,
+      "learning_rate": 6.502858322211902e-05,
+      "loss": 0.7171,
+      "step": 8577
+    },
+    {
+      "epoch": 1.5301935598965302,
+      "grad_norm": 0.5599898099899292,
+      "learning_rate": 6.498145730393592e-05,
+      "loss": 0.6881,
+      "step": 8578
+    },
+    {
+      "epoch": 1.5303719561145304,
+      "grad_norm": 0.5843207836151123,
+      "learning_rate": 6.493434591702962e-05,
+      "loss": 0.7688,
+      "step": 8579
+    },
+    {
+      "epoch": 1.5305503523325306,
+      "grad_norm": 0.4394163191318512,
+      "learning_rate": 6.488724906510008e-05,
+      "loss": 0.4879,
+      "step": 8580
+    },
+    {
+      "epoch": 1.5307287485505308,
+      "grad_norm": 0.49832600355148315,
+      "learning_rate": 6.484016675184639e-05,
+      "loss": 0.5856,
+      "step": 8581
+    },
+    {
+      "epoch": 1.530907144768531,
+      "grad_norm": 0.5916674137115479,
+      "learning_rate": 6.479309898096627e-05,
+      "loss": 0.5649,
+      "step": 8582
+    },
+    {
+      "epoch": 1.5310855409865312,
+      "grad_norm": 0.4655032455921173,
+      "learning_rate": 6.474604575615653e-05,
+      "loss": 0.5289,
+      "step": 8583
+    },
+    {
+      "epoch": 1.5312639372045314,
+      "grad_norm": 0.48161542415618896,
+      "learning_rate": 6.469900708111254e-05,
+      "loss": 0.4909,
+      "step": 8584
+    },
+    {
+      "epoch": 1.5314423334225316,
+      "grad_norm": 0.564775288105011,
+      "learning_rate": 6.465198295952881e-05,
+      "loss": 0.5441,
+      "step": 8585
+    },
+    {
+      "epoch": 1.5316207296405318,
+      "grad_norm": 0.4698559641838074,
+      "learning_rate": 6.460497339509864e-05,
+      "loss": 0.487,
+      "step": 8586
+    },
+    {
+      "epoch": 1.531799125858532,
+      "grad_norm": 0.45102497935295105,
+      "learning_rate": 6.4557978391514e-05,
+      "loss": 0.5289,
+      "step": 8587
+    },
+    {
+      "epoch": 1.531977522076532,
+      "grad_norm": 0.5399623513221741,
+      "learning_rate": 6.451099795246604e-05,
+      "loss": 0.7627,
+      "step": 8588
+    },
+    {
+      "epoch": 1.5321559182945321,
+      "grad_norm": 0.5812605023384094,
+      "learning_rate": 6.446403208164436e-05,
+      "loss": 0.7045,
+      "step": 8589
+    },
+    {
+      "epoch": 1.5323343145125323,
+      "grad_norm": 0.5008137822151184,
+      "learning_rate": 6.441708078273787e-05,
+      "loss": 0.4697,
+      "step": 8590
+    },
+    {
+      "epoch": 1.5325127107305325,
+      "grad_norm": 0.5091111660003662,
+      "learning_rate": 6.437014405943397e-05,
+      "loss": 0.4132,
+      "step": 8591
+    },
+    {
+      "epoch": 1.5326911069485327,
+      "grad_norm": 0.6219770312309265,
+      "learning_rate": 6.432322191541901e-05,
+      "loss": 0.5969,
+      "step": 8592
+    },
+    {
+      "epoch": 1.532869503166533,
+      "grad_norm": 0.5382397770881653,
+      "learning_rate": 6.427631435437836e-05,
+      "loss": 0.5745,
+      "step": 8593
+    },
+    {
+      "epoch": 1.533047899384533,
+      "grad_norm": 0.5575965642929077,
+      "learning_rate": 6.422942137999598e-05,
+      "loss": 0.4856,
+      "step": 8594
+    },
+    {
+      "epoch": 1.533226295602533,
+      "grad_norm": 0.5452744960784912,
+      "learning_rate": 6.418254299595499e-05,
+      "loss": 0.5125,
+      "step": 8595
+    },
+    {
+      "epoch": 1.5334046918205333,
+      "grad_norm": 0.5527557134628296,
+      "learning_rate": 6.413567920593705e-05,
+      "loss": 0.6013,
+      "step": 8596
+    },
+    {
+      "epoch": 1.5335830880385335,
+      "grad_norm": 0.47276362776756287,
+      "learning_rate": 6.408883001362292e-05,
+      "loss": 0.4906,
+      "step": 8597
+    },
+    {
+      "epoch": 1.5337614842565337,
+      "grad_norm": 0.5150408148765564,
+      "learning_rate": 6.404199542269213e-05,
+      "loss": 0.5248,
+      "step": 8598
+    },
+    {
+      "epoch": 1.533939880474534,
+      "grad_norm": 0.4958198070526123,
+      "learning_rate": 6.399517543682278e-05,
+      "loss": 0.5716,
+      "step": 8599
+    },
+    {
+      "epoch": 1.534118276692534,
+      "grad_norm": 0.49585381150245667,
+      "learning_rate": 6.39483700596925e-05,
+      "loss": 0.6034,
+      "step": 8600
+    },
+    {
+      "epoch": 1.5342966729105343,
+      "grad_norm": 0.5567946434020996,
+      "learning_rate": 6.390157929497708e-05,
+      "loss": 0.6809,
+      "step": 8601
+    },
+    {
+      "epoch": 1.5344750691285345,
+      "grad_norm": 0.541259765625,
+      "learning_rate": 6.385480314635162e-05,
+      "loss": 0.6461,
+      "step": 8602
+    },
+    {
+      "epoch": 1.5346534653465347,
+      "grad_norm": 0.5411567091941833,
+      "learning_rate": 6.380804161748982e-05,
+      "loss": 0.6509,
+      "step": 8603
+    },
+    {
+      "epoch": 1.5348318615645349,
+      "grad_norm": 0.4732004404067993,
+      "learning_rate": 6.376129471206422e-05,
+      "loss": 0.4766,
+      "step": 8604
+    },
+    {
+      "epoch": 1.535010257782535,
+      "grad_norm": 0.45597603917121887,
+      "learning_rate": 6.371456243374646e-05,
+      "loss": 0.5052,
+      "step": 8605
+    },
+    {
+      "epoch": 1.5351886540005353,
+      "grad_norm": 0.49953046441078186,
+      "learning_rate": 6.366784478620674e-05,
+      "loss": 0.6363,
+      "step": 8606
+    },
+    {
+      "epoch": 1.5353670502185355,
+      "grad_norm": 0.4946024715900421,
+      "learning_rate": 6.362114177311437e-05,
+      "loss": 0.5352,
+      "step": 8607
+    },
+    {
+      "epoch": 1.5355454464365357,
+      "grad_norm": 0.4868670403957367,
+      "learning_rate": 6.357445339813726e-05,
+      "loss": 0.4525,
+      "step": 8608
+    },
+    {
+      "epoch": 1.5357238426545359,
+      "grad_norm": 0.5404412150382996,
+      "learning_rate": 6.352777966494242e-05,
+      "loss": 0.6858,
+      "step": 8609
+    },
+    {
+      "epoch": 1.5359022388725359,
+      "grad_norm": 0.5547458529472351,
+      "learning_rate": 6.348112057719551e-05,
+      "loss": 0.6672,
+      "step": 8610
+    },
+    {
+      "epoch": 1.536080635090536,
+      "grad_norm": 0.5338977575302124,
+      "learning_rate": 6.343447613856108e-05,
+      "loss": 0.694,
+      "step": 8611
+    },
+    {
+      "epoch": 1.5362590313085362,
+      "grad_norm": 0.5286486744880676,
+      "learning_rate": 6.338784635270264e-05,
+      "loss": 0.6343,
+      "step": 8612
+    },
+    {
+      "epoch": 1.5364374275265364,
+      "grad_norm": 0.49329471588134766,
+      "learning_rate": 6.334123122328239e-05,
+      "loss": 0.5588,
+      "step": 8613
+    },
+    {
+      "epoch": 1.5366158237445366,
+      "grad_norm": 0.5088777542114258,
+      "learning_rate": 6.32946307539616e-05,
+      "loss": 0.5553,
+      "step": 8614
+    },
+    {
+      "epoch": 1.5367942199625368,
+      "grad_norm": 0.627946674823761,
+      "learning_rate": 6.324804494840008e-05,
+      "loss": 0.4839,
+      "step": 8615
+    },
+    {
+      "epoch": 1.5369726161805368,
+      "grad_norm": 0.5603138208389282,
+      "learning_rate": 6.320147381025673e-05,
+      "loss": 0.7505,
+      "step": 8616
+    },
+    {
+      "epoch": 1.537151012398537,
+      "grad_norm": 0.4293627440929413,
+      "learning_rate": 6.315491734318934e-05,
+      "loss": 0.3965,
+      "step": 8617
+    },
+    {
+      "epoch": 1.5373294086165372,
+      "grad_norm": 0.5595969557762146,
+      "learning_rate": 6.310837555085424e-05,
+      "loss": 0.716,
+      "step": 8618
+    },
+    {
+      "epoch": 1.5375078048345374,
+      "grad_norm": 0.4659646153450012,
+      "learning_rate": 6.306184843690699e-05,
+      "loss": 0.4463,
+      "step": 8619
+    },
+    {
+      "epoch": 1.5376862010525376,
+      "grad_norm": 0.5158681869506836,
+      "learning_rate": 6.301533600500165e-05,
+      "loss": 0.5118,
+      "step": 8620
+    },
+    {
+      "epoch": 1.5378645972705378,
+      "grad_norm": 0.4700511693954468,
+      "learning_rate": 6.296883825879141e-05,
+      "loss": 0.4758,
+      "step": 8621
+    },
+    {
+      "epoch": 1.538042993488538,
+      "grad_norm": 0.5160353183746338,
+      "learning_rate": 6.292235520192807e-05,
+      "loss": 0.6518,
+      "step": 8622
+    },
+    {
+      "epoch": 1.5382213897065382,
+      "grad_norm": 0.4958648085594177,
+      "learning_rate": 6.28758868380625e-05,
+      "loss": 0.4795,
+      "step": 8623
+    },
+    {
+      "epoch": 1.5383997859245384,
+      "grad_norm": 0.42343762516975403,
+      "learning_rate": 6.282943317084428e-05,
+      "loss": 0.3622,
+      "step": 8624
+    },
+    {
+      "epoch": 1.5385781821425386,
+      "grad_norm": 0.5087177753448486,
+      "learning_rate": 6.278299420392173e-05,
+      "loss": 0.4692,
+      "step": 8625
+    },
+    {
+      "epoch": 1.5387565783605388,
+      "grad_norm": 0.44395479559898376,
+      "learning_rate": 6.273656994094232e-05,
+      "loss": 0.4136,
+      "step": 8626
+    },
+    {
+      "epoch": 1.538934974578539,
+      "grad_norm": 0.5368765592575073,
+      "learning_rate": 6.269016038555206e-05,
+      "loss": 0.648,
+      "step": 8627
+    },
+    {
+      "epoch": 1.5391133707965392,
+      "grad_norm": 0.6212941408157349,
+      "learning_rate": 6.264376554139608e-05,
+      "loss": 0.7711,
+      "step": 8628
+    },
+    {
+      "epoch": 1.5392917670145394,
+      "grad_norm": 0.5054884552955627,
+      "learning_rate": 6.259738541211804e-05,
+      "loss": 0.5336,
+      "step": 8629
+    },
+    {
+      "epoch": 1.5394701632325396,
+      "grad_norm": 0.5286135673522949,
+      "learning_rate": 6.255102000136073e-05,
+      "loss": 0.5894,
+      "step": 8630
+    },
+    {
+      "epoch": 1.5396485594505398,
+      "grad_norm": 0.581322193145752,
+      "learning_rate": 6.250466931276569e-05,
+      "loss": 0.732,
+      "step": 8631
+    },
+    {
+      "epoch": 1.5398269556685398,
+      "grad_norm": 0.492714524269104,
+      "learning_rate": 6.245833334997317e-05,
+      "loss": 0.583,
+      "step": 8632
+    },
+    {
+      "epoch": 1.54000535188654,
+      "grad_norm": 0.49149906635284424,
+      "learning_rate": 6.241201211662254e-05,
+      "loss": 0.496,
+      "step": 8633
+    },
+    {
+      "epoch": 1.5401837481045402,
+      "grad_norm": 0.45817720890045166,
+      "learning_rate": 6.236570561635163e-05,
+      "loss": 0.5408,
+      "step": 8634
+    },
+    {
+      "epoch": 1.5403621443225404,
+      "grad_norm": 0.5533788204193115,
+      "learning_rate": 6.231941385279757e-05,
+      "loss": 0.6682,
+      "step": 8635
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 0.5434373021125793,
+      "learning_rate": 6.227313682959596e-05,
+      "loss": 0.6231,
+      "step": 8636
+    },
+    {
+      "epoch": 1.5407189367585408,
+      "grad_norm": 0.45322054624557495,
+      "learning_rate": 6.222687455038134e-05,
+      "loss": 0.3839,
+      "step": 8637
+    },
+    {
+      "epoch": 1.5408973329765407,
+      "grad_norm": 0.5391322374343872,
+      "learning_rate": 6.218062701878724e-05,
+      "loss": 0.4676,
+      "step": 8638
+    },
+    {
+      "epoch": 1.541075729194541,
+      "grad_norm": 0.4535292685031891,
+      "learning_rate": 6.213439423844583e-05,
+      "loss": 0.4573,
+      "step": 8639
+    },
+    {
+      "epoch": 1.5412541254125411,
+      "grad_norm": 0.46426570415496826,
+      "learning_rate": 6.208817621298829e-05,
+      "loss": 0.4326,
+      "step": 8640
+    },
+    {
+      "epoch": 1.5414325216305413,
+      "grad_norm": 0.6217679977416992,
+      "learning_rate": 6.204197294604446e-05,
+      "loss": 0.6326,
+      "step": 8641
+    },
+    {
+      "epoch": 1.5416109178485415,
+      "grad_norm": 0.4808856248855591,
+      "learning_rate": 6.199578444124329e-05,
+      "loss": 0.4614,
+      "step": 8642
+    },
+    {
+      "epoch": 1.5417893140665417,
+      "grad_norm": 0.5089523792266846,
+      "learning_rate": 6.194961070221219e-05,
+      "loss": 0.5047,
+      "step": 8643
+    },
+    {
+      "epoch": 1.541967710284542,
+      "grad_norm": 0.43264392018318176,
+      "learning_rate": 6.190345173257786e-05,
+      "loss": 0.3809,
+      "step": 8644
+    },
+    {
+      "epoch": 1.5421461065025421,
+      "grad_norm": 0.450720876455307,
+      "learning_rate": 6.185730753596539e-05,
+      "loss": 0.4279,
+      "step": 8645
+    },
+    {
+      "epoch": 1.5423245027205423,
+      "grad_norm": 0.4698851704597473,
+      "learning_rate": 6.181117811599901e-05,
+      "loss": 0.4536,
+      "step": 8646
+    },
+    {
+      "epoch": 1.5425028989385425,
+      "grad_norm": 0.46005532145500183,
+      "learning_rate": 6.176506347630181e-05,
+      "loss": 0.4388,
+      "step": 8647
+    },
+    {
+      "epoch": 1.5426812951565427,
+      "grad_norm": 0.5724323391914368,
+      "learning_rate": 6.171896362049542e-05,
+      "loss": 0.6823,
+      "step": 8648
+    },
+    {
+      "epoch": 1.542859691374543,
+      "grad_norm": 0.47760656476020813,
+      "learning_rate": 6.167287855220072e-05,
+      "loss": 0.4862,
+      "step": 8649
+    },
+    {
+      "epoch": 1.5430380875925431,
+      "grad_norm": 0.533450186252594,
+      "learning_rate": 6.162680827503705e-05,
+      "loss": 0.606,
+      "step": 8650
+    },
+    {
+      "epoch": 1.5432164838105433,
+      "grad_norm": 0.5100364089012146,
+      "learning_rate": 6.158075279262273e-05,
+      "loss": 0.5607,
+      "step": 8651
+    },
+    {
+      "epoch": 1.5433948800285435,
+      "grad_norm": 0.5601518154144287,
+      "learning_rate": 6.153471210857511e-05,
+      "loss": 0.6058,
+      "step": 8652
+    },
+    {
+      "epoch": 1.5435732762465437,
+      "grad_norm": 0.4649224579334259,
+      "learning_rate": 6.148868622650999e-05,
+      "loss": 0.461,
+      "step": 8653
+    },
+    {
+      "epoch": 1.5437516724645437,
+      "grad_norm": 0.510765016078949,
+      "learning_rate": 6.144267515004243e-05,
+      "loss": 0.5787,
+      "step": 8654
+    },
+    {
+      "epoch": 1.543930068682544,
+      "grad_norm": 0.5434867143630981,
+      "learning_rate": 6.139667888278594e-05,
+      "loss": 0.5901,
+      "step": 8655
+    },
+    {
+      "epoch": 1.544108464900544,
+      "grad_norm": 0.5194826722145081,
+      "learning_rate": 6.13506974283532e-05,
+      "loss": 0.5809,
+      "step": 8656
+    },
+    {
+      "epoch": 1.5442868611185443,
+      "grad_norm": 0.42716190218925476,
+      "learning_rate": 6.130473079035548e-05,
+      "loss": 0.4636,
+      "step": 8657
+    },
+    {
+      "epoch": 1.5444652573365445,
+      "grad_norm": 0.4728177785873413,
+      "learning_rate": 6.125877897240295e-05,
+      "loss": 0.4347,
+      "step": 8658
+    },
+    {
+      "epoch": 1.5446436535545447,
+      "grad_norm": 0.4928957223892212,
+      "learning_rate": 6.121284197810476e-05,
+      "loss": 0.5965,
+      "step": 8659
+    },
+    {
+      "epoch": 1.5448220497725447,
+      "grad_norm": 0.5190236568450928,
+      "learning_rate": 6.116691981106868e-05,
+      "loss": 0.5734,
+      "step": 8660
+    },
+    {
+      "epoch": 1.5450004459905449,
+      "grad_norm": 0.4811258018016815,
+      "learning_rate": 6.11210124749014e-05,
+      "loss": 0.5823,
+      "step": 8661
+    },
+    {
+      "epoch": 1.545178842208545,
+      "grad_norm": 0.5223371982574463,
+      "learning_rate": 6.107511997320863e-05,
+      "loss": 0.52,
+      "step": 8662
+    },
+    {
+      "epoch": 1.5453572384265453,
+      "grad_norm": 0.5431803464889526,
+      "learning_rate": 6.102924230959456e-05,
+      "loss": 0.6178,
+      "step": 8663
+    },
+    {
+      "epoch": 1.5455356346445455,
+      "grad_norm": 0.5262665152549744,
+      "learning_rate": 6.0983379487662555e-05,
+      "loss": 0.5756,
+      "step": 8664
+    },
+    {
+      "epoch": 1.5457140308625457,
+      "grad_norm": 0.5077388882637024,
+      "learning_rate": 6.09375315110145e-05,
+      "loss": 0.5741,
+      "step": 8665
+    },
+    {
+      "epoch": 1.5458924270805459,
+      "grad_norm": 0.5087509155273438,
+      "learning_rate": 6.089169838325143e-05,
+      "loss": 0.5392,
+      "step": 8666
+    },
+    {
+      "epoch": 1.546070823298546,
+      "grad_norm": 0.4039164185523987,
+      "learning_rate": 6.084588010797293e-05,
+      "loss": 0.3906,
+      "step": 8667
+    },
+    {
+      "epoch": 1.5462492195165463,
+      "grad_norm": 0.4832989573478699,
+      "learning_rate": 6.0800076688777684e-05,
+      "loss": 0.5393,
+      "step": 8668
+    },
+    {
+      "epoch": 1.5464276157345465,
+      "grad_norm": 0.4995843470096588,
+      "learning_rate": 6.0754288129262895e-05,
+      "loss": 0.6855,
+      "step": 8669
+    },
+    {
+      "epoch": 1.5466060119525467,
+      "grad_norm": 0.47281551361083984,
+      "learning_rate": 6.070851443302497e-05,
+      "loss": 0.489,
+      "step": 8670
+    },
+    {
+      "epoch": 1.5467844081705469,
+      "grad_norm": 0.4991108477115631,
+      "learning_rate": 6.066275560365886e-05,
+      "loss": 0.5685,
+      "step": 8671
+    },
+    {
+      "epoch": 1.546962804388547,
+      "grad_norm": 0.5392157435417175,
+      "learning_rate": 6.0617011644758385e-05,
+      "loss": 0.7395,
+      "step": 8672
+    },
+    {
+      "epoch": 1.5471412006065473,
+      "grad_norm": 0.6251504421234131,
+      "learning_rate": 6.057128255991637e-05,
+      "loss": 0.777,
+      "step": 8673
+    },
+    {
+      "epoch": 1.5473195968245474,
+      "grad_norm": 0.4212150573730469,
+      "learning_rate": 6.052556835272424e-05,
+      "loss": 0.3672,
+      "step": 8674
+    },
+    {
+      "epoch": 1.5474979930425476,
+      "grad_norm": 0.3978899419307709,
+      "learning_rate": 6.047986902677252e-05,
+      "loss": 0.352,
+      "step": 8675
+    },
+    {
+      "epoch": 1.5476763892605476,
+      "grad_norm": 0.5088796615600586,
+      "learning_rate": 6.0434184585650256e-05,
+      "loss": 0.5676,
+      "step": 8676
+    },
+    {
+      "epoch": 1.5478547854785478,
+      "grad_norm": 0.42953309416770935,
+      "learning_rate": 6.038851503294554e-05,
+      "loss": 0.3769,
+      "step": 8677
+    },
+    {
+      "epoch": 1.548033181696548,
+      "grad_norm": 0.49997836351394653,
+      "learning_rate": 6.0342860372245344e-05,
+      "loss": 0.6421,
+      "step": 8678
+    },
+    {
+      "epoch": 1.5482115779145482,
+      "grad_norm": 0.5206395387649536,
+      "learning_rate": 6.029722060713519e-05,
+      "loss": 0.521,
+      "step": 8679
+    },
+    {
+      "epoch": 1.5483899741325484,
+      "grad_norm": 0.4601840376853943,
+      "learning_rate": 6.025159574119979e-05,
+      "loss": 0.5562,
+      "step": 8680
+    },
+    {
+      "epoch": 1.5485683703505486,
+      "grad_norm": 0.5508368015289307,
+      "learning_rate": 6.0205985778022305e-05,
+      "loss": 0.5545,
+      "step": 8681
+    },
+    {
+      "epoch": 1.5487467665685486,
+      "grad_norm": 0.5534743070602417,
+      "learning_rate": 6.016039072118512e-05,
+      "loss": 0.5463,
+      "step": 8682
+    },
+    {
+      "epoch": 1.5489251627865488,
+      "grad_norm": 0.4522656798362732,
+      "learning_rate": 6.011481057426915e-05,
+      "loss": 0.4339,
+      "step": 8683
+    },
+    {
+      "epoch": 1.549103559004549,
+      "grad_norm": 0.44534674286842346,
+      "learning_rate": 6.006924534085414e-05,
+      "loss": 0.4096,
+      "step": 8684
+    },
+    {
+      "epoch": 1.5492819552225492,
+      "grad_norm": 0.4928872883319855,
+      "learning_rate": 6.002369502451899e-05,
+      "loss": 0.5002,
+      "step": 8685
+    },
+    {
+      "epoch": 1.5494603514405494,
+      "grad_norm": 0.509283185005188,
+      "learning_rate": 5.997815962884098e-05,
+      "loss": 0.4871,
+      "step": 8686
+    },
+    {
+      "epoch": 1.5496387476585496,
+      "grad_norm": 0.5575897097587585,
+      "learning_rate": 5.993263915739661e-05,
+      "loss": 0.4546,
+      "step": 8687
+    },
+    {
+      "epoch": 1.5498171438765498,
+      "grad_norm": 0.49267205595970154,
+      "learning_rate": 5.988713361376089e-05,
+      "loss": 0.5014,
+      "step": 8688
+    },
+    {
+      "epoch": 1.54999554009455,
+      "grad_norm": 0.4820170998573303,
+      "learning_rate": 5.984164300150796e-05,
+      "loss": 0.4648,
+      "step": 8689
+    },
+    {
+      "epoch": 1.5501739363125502,
+      "grad_norm": 0.5354559421539307,
+      "learning_rate": 5.9796167324210505e-05,
+      "loss": 0.6559,
+      "step": 8690
+    },
+    {
+      "epoch": 1.5503523325305504,
+      "grad_norm": 0.5515713095664978,
+      "learning_rate": 5.975070658544021e-05,
+      "loss": 0.5613,
+      "step": 8691
+    },
+    {
+      "epoch": 1.5505307287485506,
+      "grad_norm": 0.5345812439918518,
+      "learning_rate": 5.9705260788767594e-05,
+      "loss": 0.6878,
+      "step": 8692
+    },
+    {
+      "epoch": 1.5507091249665508,
+      "grad_norm": 0.48849523067474365,
+      "learning_rate": 5.9659829937761865e-05,
+      "loss": 0.4398,
+      "step": 8693
+    },
+    {
+      "epoch": 1.550887521184551,
+      "grad_norm": 0.5398678183555603,
+      "learning_rate": 5.9614414035991244e-05,
+      "loss": 0.6064,
+      "step": 8694
+    },
+    {
+      "epoch": 1.5510659174025512,
+      "grad_norm": 0.47150692343711853,
+      "learning_rate": 5.9569013087022614e-05,
+      "loss": 0.5028,
+      "step": 8695
+    },
+    {
+      "epoch": 1.5512443136205514,
+      "grad_norm": 0.4753498435020447,
+      "learning_rate": 5.9523627094421664e-05,
+      "loss": 0.4533,
+      "step": 8696
+    },
+    {
+      "epoch": 1.5514227098385516,
+      "grad_norm": 0.5937864184379578,
+      "learning_rate": 5.947825606175317e-05,
+      "loss": 0.6236,
+      "step": 8697
+    },
+    {
+      "epoch": 1.5516011060565515,
+      "grad_norm": 0.5065819025039673,
+      "learning_rate": 5.943289999258036e-05,
+      "loss": 0.5351,
+      "step": 8698
+    },
+    {
+      "epoch": 1.5517795022745517,
+      "grad_norm": 0.5110275149345398,
+      "learning_rate": 5.938755889046565e-05,
+      "loss": 0.5605,
+      "step": 8699
+    },
+    {
+      "epoch": 1.551957898492552,
+      "grad_norm": 0.5337404608726501,
+      "learning_rate": 5.934223275896999e-05,
+      "loss": 0.6949,
+      "step": 8700
+    },
+    {
+      "epoch": 1.5521362947105521,
+      "grad_norm": 0.5709247589111328,
+      "learning_rate": 5.92969216016534e-05,
+      "loss": 0.7602,
+      "step": 8701
+    },
+    {
+      "epoch": 1.5523146909285523,
+      "grad_norm": 0.4845641255378723,
+      "learning_rate": 5.9251625422074406e-05,
+      "loss": 0.4694,
+      "step": 8702
+    },
+    {
+      "epoch": 1.5524930871465525,
+      "grad_norm": 0.5533928871154785,
+      "learning_rate": 5.920634422379079e-05,
+      "loss": 0.6047,
+      "step": 8703
+    },
+    {
+      "epoch": 1.5526714833645525,
+      "grad_norm": 0.5261991024017334,
+      "learning_rate": 5.916107801035875e-05,
+      "loss": 0.6173,
+      "step": 8704
+    },
+    {
+      "epoch": 1.5528498795825527,
+      "grad_norm": 0.5172926783561707,
+      "learning_rate": 5.9115826785333473e-05,
+      "loss": 0.5984,
+      "step": 8705
+    },
+    {
+      "epoch": 1.553028275800553,
+      "grad_norm": 0.47893017530441284,
+      "learning_rate": 5.907059055226907e-05,
+      "loss": 0.4444,
+      "step": 8706
+    },
+    {
+      "epoch": 1.5532066720185531,
+      "grad_norm": 0.5027623772621155,
+      "learning_rate": 5.902536931471819e-05,
+      "loss": 0.4835,
+      "step": 8707
+    },
+    {
+      "epoch": 1.5533850682365533,
+      "grad_norm": 0.566230297088623,
+      "learning_rate": 5.898016307623275e-05,
+      "loss": 0.5732,
+      "step": 8708
+    },
+    {
+      "epoch": 1.5535634644545535,
+      "grad_norm": 0.49636390805244446,
+      "learning_rate": 5.8934971840363116e-05,
+      "loss": 0.5499,
+      "step": 8709
+    },
+    {
+      "epoch": 1.5537418606725537,
+      "grad_norm": 0.5405829548835754,
+      "learning_rate": 5.888979561065849e-05,
+      "loss": 0.5984,
+      "step": 8710
+    },
+    {
+      "epoch": 1.553920256890554,
+      "grad_norm": 0.4457230865955353,
+      "learning_rate": 5.8844634390667176e-05,
+      "loss": 0.4734,
+      "step": 8711
+    },
+    {
+      "epoch": 1.554098653108554,
+      "grad_norm": 0.5478752851486206,
+      "learning_rate": 5.879948818393591e-05,
+      "loss": 0.5332,
+      "step": 8712
+    },
+    {
+      "epoch": 1.5542770493265543,
+      "grad_norm": 0.49561017751693726,
+      "learning_rate": 5.8754356994010634e-05,
+      "loss": 0.5383,
+      "step": 8713
+    },
+    {
+      "epoch": 1.5544554455445545,
+      "grad_norm": 0.571239709854126,
+      "learning_rate": 5.8709240824435795e-05,
+      "loss": 0.4595,
+      "step": 8714
+    },
+    {
+      "epoch": 1.5546338417625547,
+      "grad_norm": 0.48616328835487366,
+      "learning_rate": 5.8664139678754944e-05,
+      "loss": 0.5232,
+      "step": 8715
+    },
+    {
+      "epoch": 1.554812237980555,
+      "grad_norm": 0.5562140345573425,
+      "learning_rate": 5.861905356051023e-05,
+      "loss": 0.6471,
+      "step": 8716
+    },
+    {
+      "epoch": 1.554990634198555,
+      "grad_norm": 0.516631007194519,
+      "learning_rate": 5.857398247324261e-05,
+      "loss": 0.572,
+      "step": 8717
+    },
+    {
+      "epoch": 1.5551690304165553,
+      "grad_norm": 0.4869072139263153,
+      "learning_rate": 5.85289264204921e-05,
+      "loss": 0.5075,
+      "step": 8718
+    },
+    {
+      "epoch": 1.5553474266345555,
+      "grad_norm": 0.5700942873954773,
+      "learning_rate": 5.8483885405797246e-05,
+      "loss": 0.6769,
+      "step": 8719
+    },
+    {
+      "epoch": 1.5555258228525555,
+      "grad_norm": 1.086318850517273,
+      "learning_rate": 5.843885943269567e-05,
+      "loss": 0.4651,
+      "step": 8720
+    },
+    {
+      "epoch": 1.5557042190705557,
+      "grad_norm": 0.45219793915748596,
+      "learning_rate": 5.83938485047236e-05,
+      "loss": 0.4308,
+      "step": 8721
+    },
+    {
+      "epoch": 1.5558826152885559,
+      "grad_norm": 0.5690205097198486,
+      "learning_rate": 5.834885262541617e-05,
+      "loss": 0.628,
+      "step": 8722
+    },
+    {
+      "epoch": 1.556061011506556,
+      "grad_norm": 0.5013719797134399,
+      "learning_rate": 5.830387179830748e-05,
+      "loss": 0.5243,
+      "step": 8723
+    },
+    {
+      "epoch": 1.5562394077245563,
+      "grad_norm": 0.4831756055355072,
+      "learning_rate": 5.825890602693013e-05,
+      "loss": 0.4328,
+      "step": 8724
+    },
+    {
+      "epoch": 1.5564178039425565,
+      "grad_norm": 0.511838436126709,
+      "learning_rate": 5.8213955314815853e-05,
+      "loss": 0.4934,
+      "step": 8725
+    },
+    {
+      "epoch": 1.5565962001605564,
+      "grad_norm": 0.5139269232749939,
+      "learning_rate": 5.816901966549495e-05,
+      "loss": 0.4621,
+      "step": 8726
+    },
+    {
+      "epoch": 1.5567745963785566,
+      "grad_norm": 0.4796489477157593,
+      "learning_rate": 5.8124099082496745e-05,
+      "loss": 0.5047,
+      "step": 8727
+    },
+    {
+      "epoch": 1.5569529925965568,
+      "grad_norm": 0.47193264961242676,
+      "learning_rate": 5.807919356934915e-05,
+      "loss": 0.467,
+      "step": 8728
+    },
+    {
+      "epoch": 1.557131388814557,
+      "grad_norm": 0.570898711681366,
+      "learning_rate": 5.8034303129579164e-05,
+      "loss": 0.7563,
+      "step": 8729
+    },
+    {
+      "epoch": 1.5573097850325572,
+      "grad_norm": 0.4879952073097229,
+      "learning_rate": 5.798942776671243e-05,
+      "loss": 0.4973,
+      "step": 8730
+    },
+    {
+      "epoch": 1.5574881812505574,
+      "grad_norm": 0.44711223244667053,
+      "learning_rate": 5.794456748427332e-05,
+      "loss": 0.3708,
+      "step": 8731
+    },
+    {
+      "epoch": 1.5576665774685576,
+      "grad_norm": 0.6115127801895142,
+      "learning_rate": 5.78997222857853e-05,
+      "loss": 0.6516,
+      "step": 8732
+    },
+    {
+      "epoch": 1.5578449736865578,
+      "grad_norm": 0.5244799852371216,
+      "learning_rate": 5.785489217477036e-05,
+      "loss": 0.507,
+      "step": 8733
+    },
+    {
+      "epoch": 1.558023369904558,
+      "grad_norm": 0.5613646507263184,
+      "learning_rate": 5.7810077154749566e-05,
+      "loss": 0.6944,
+      "step": 8734
+    },
+    {
+      "epoch": 1.5582017661225582,
+      "grad_norm": 0.49775010347366333,
+      "learning_rate": 5.7765277229242546e-05,
+      "loss": 0.4435,
+      "step": 8735
+    },
+    {
+      "epoch": 1.5583801623405584,
+      "grad_norm": 0.46436411142349243,
+      "learning_rate": 5.772049240176799e-05,
+      "loss": 0.4435,
+      "step": 8736
+    },
+    {
+      "epoch": 1.5585585585585586,
+      "grad_norm": 0.5458922982215881,
+      "learning_rate": 5.7675722675843144e-05,
+      "loss": 0.5705,
+      "step": 8737
+    },
+    {
+      "epoch": 1.5587369547765588,
+      "grad_norm": 0.5422109365463257,
+      "learning_rate": 5.763096805498427e-05,
+      "loss": 0.6135,
+      "step": 8738
+    },
+    {
+      "epoch": 1.558915350994559,
+      "grad_norm": 0.574105441570282,
+      "learning_rate": 5.758622854270648e-05,
+      "loss": 0.7073,
+      "step": 8739
+    },
+    {
+      "epoch": 1.5590937472125592,
+      "grad_norm": 0.5504136085510254,
+      "learning_rate": 5.7541504142523406e-05,
+      "loss": 0.6011,
+      "step": 8740
+    },
+    {
+      "epoch": 1.5592721434305594,
+      "grad_norm": 0.5172324180603027,
+      "learning_rate": 5.7496794857947846e-05,
+      "loss": 0.4303,
+      "step": 8741
+    },
+    {
+      "epoch": 1.5594505396485594,
+      "grad_norm": 0.5464860200881958,
+      "learning_rate": 5.745210069249118e-05,
+      "loss": 0.6365,
+      "step": 8742
+    },
+    {
+      "epoch": 1.5596289358665596,
+      "grad_norm": 0.47427821159362793,
+      "learning_rate": 5.740742164966362e-05,
+      "loss": 0.4361,
+      "step": 8743
+    },
+    {
+      "epoch": 1.5598073320845598,
+      "grad_norm": 1.116687536239624,
+      "learning_rate": 5.736275773297431e-05,
+      "loss": 0.699,
+      "step": 8744
+    },
+    {
+      "epoch": 1.55998572830256,
+      "grad_norm": 0.4698363244533539,
+      "learning_rate": 5.731810894593106e-05,
+      "loss": 0.5028,
+      "step": 8745
+    },
+    {
+      "epoch": 1.5601641245205602,
+      "grad_norm": 0.47105473279953003,
+      "learning_rate": 5.727347529204069e-05,
+      "loss": 0.5059,
+      "step": 8746
+    },
+    {
+      "epoch": 1.5603425207385604,
+      "grad_norm": 0.5197162628173828,
+      "learning_rate": 5.722885677480857e-05,
+      "loss": 0.5229,
+      "step": 8747
+    },
+    {
+      "epoch": 1.5605209169565604,
+      "grad_norm": 0.5358749032020569,
+      "learning_rate": 5.718425339773914e-05,
+      "loss": 0.5431,
+      "step": 8748
+    },
+    {
+      "epoch": 1.5606993131745606,
+      "grad_norm": 0.6025269627571106,
+      "learning_rate": 5.713966516433541e-05,
+      "loss": 0.791,
+      "step": 8749
+    },
+    {
+      "epoch": 1.5608777093925608,
+      "grad_norm": 0.502884030342102,
+      "learning_rate": 5.709509207809946e-05,
+      "loss": 0.5499,
+      "step": 8750
+    },
+    {
+      "epoch": 1.561056105610561,
+      "grad_norm": 0.49932992458343506,
+      "learning_rate": 5.705053414253195e-05,
+      "loss": 0.5878,
+      "step": 8751
+    },
+    {
+      "epoch": 1.5612345018285612,
+      "grad_norm": 0.4696747362613678,
+      "learning_rate": 5.700599136113238e-05,
+      "loss": 0.4294,
+      "step": 8752
+    },
+    {
+      "epoch": 1.5614128980465614,
+      "grad_norm": 0.4963851273059845,
+      "learning_rate": 5.6961463737399215e-05,
+      "loss": 0.5904,
+      "step": 8753
+    },
+    {
+      "epoch": 1.5615912942645616,
+      "grad_norm": 0.542914867401123,
+      "learning_rate": 5.691695127482968e-05,
+      "loss": 0.7059,
+      "step": 8754
+    },
+    {
+      "epoch": 1.5617696904825618,
+      "grad_norm": 0.44281649589538574,
+      "learning_rate": 5.687245397691962e-05,
+      "loss": 0.416,
+      "step": 8755
+    },
+    {
+      "epoch": 1.561948086700562,
+      "grad_norm": 0.48408737778663635,
+      "learning_rate": 5.682797184716401e-05,
+      "loss": 0.5626,
+      "step": 8756
+    },
+    {
+      "epoch": 1.5621264829185622,
+      "grad_norm": 0.4773915410041809,
+      "learning_rate": 5.6783504889056285e-05,
+      "loss": 0.5477,
+      "step": 8757
+    },
+    {
+      "epoch": 1.5623048791365624,
+      "grad_norm": 0.5725618600845337,
+      "learning_rate": 5.6739053106088984e-05,
+      "loss": 0.5617,
+      "step": 8758
+    },
+    {
+      "epoch": 1.5624832753545626,
+      "grad_norm": 0.4656837582588196,
+      "learning_rate": 5.6694616501753256e-05,
+      "loss": 0.4798,
+      "step": 8759
+    },
+    {
+      "epoch": 1.5626616715725627,
+      "grad_norm": 0.4770435094833374,
+      "learning_rate": 5.6650195079539194e-05,
+      "loss": 0.5398,
+      "step": 8760
+    },
+    {
+      "epoch": 1.562840067790563,
+      "grad_norm": 0.47299638390541077,
+      "learning_rate": 5.6605788842935544e-05,
+      "loss": 0.471,
+      "step": 8761
+    },
+    {
+      "epoch": 1.5630184640085631,
+      "grad_norm": 0.4295462369918823,
+      "learning_rate": 5.6561397795430096e-05,
+      "loss": 0.4549,
+      "step": 8762
+    },
+    {
+      "epoch": 1.5631968602265633,
+      "grad_norm": 0.5424062609672546,
+      "learning_rate": 5.6517021940509225e-05,
+      "loss": 0.649,
+      "step": 8763
+    },
+    {
+      "epoch": 1.5633752564445633,
+      "grad_norm": 0.4859166145324707,
+      "learning_rate": 5.6472661281658125e-05,
+      "loss": 0.5242,
+      "step": 8764
+    },
+    {
+      "epoch": 1.5635536526625635,
+      "grad_norm": 0.5274991393089294,
+      "learning_rate": 5.642831582236096e-05,
+      "loss": 0.5842,
+      "step": 8765
+    },
+    {
+      "epoch": 1.5637320488805637,
+      "grad_norm": 0.5063260197639465,
+      "learning_rate": 5.6383985566100525e-05,
+      "loss": 0.4953,
+      "step": 8766
+    },
+    {
+      "epoch": 1.563910445098564,
+      "grad_norm": 0.5385729074478149,
+      "learning_rate": 5.6339670516358633e-05,
+      "loss": 0.6145,
+      "step": 8767
+    },
+    {
+      "epoch": 1.5640888413165641,
+      "grad_norm": 0.429574579000473,
+      "learning_rate": 5.6295370676615584e-05,
+      "loss": 0.3678,
+      "step": 8768
+    },
+    {
+      "epoch": 1.5642672375345643,
+      "grad_norm": 0.45450228452682495,
+      "learning_rate": 5.625108605035076e-05,
+      "loss": 0.4379,
+      "step": 8769
+    },
+    {
+      "epoch": 1.5644456337525643,
+      "grad_norm": 0.5811430215835571,
+      "learning_rate": 5.620681664104235e-05,
+      "loss": 0.6731,
+      "step": 8770
+    },
+    {
+      "epoch": 1.5646240299705645,
+      "grad_norm": 0.5694655179977417,
+      "learning_rate": 5.6162562452167085e-05,
+      "loss": 0.6028,
+      "step": 8771
+    },
+    {
+      "epoch": 1.5648024261885647,
+      "grad_norm": 0.5351256728172302,
+      "learning_rate": 5.6118323487200806e-05,
+      "loss": 0.4978,
+      "step": 8772
+    },
+    {
+      "epoch": 1.5649808224065649,
+      "grad_norm": 0.43898749351501465,
+      "learning_rate": 5.6074099749617914e-05,
+      "loss": 0.4938,
+      "step": 8773
+    },
+    {
+      "epoch": 1.565159218624565,
+      "grad_norm": 0.5240178108215332,
+      "learning_rate": 5.602989124289185e-05,
+      "loss": 0.4156,
+      "step": 8774
+    },
+    {
+      "epoch": 1.5653376148425653,
+      "grad_norm": 0.4986781179904938,
+      "learning_rate": 5.598569797049466e-05,
+      "loss": 0.4061,
+      "step": 8775
+    },
+    {
+      "epoch": 1.5655160110605655,
+      "grad_norm": 0.6685777306556702,
+      "learning_rate": 5.5941519935897164e-05,
+      "loss": 0.8649,
+      "step": 8776
+    },
+    {
+      "epoch": 1.5656944072785657,
+      "grad_norm": 0.46365004777908325,
+      "learning_rate": 5.589735714256927e-05,
+      "loss": 0.4269,
+      "step": 8777
+    },
+    {
+      "epoch": 1.5658728034965659,
+      "grad_norm": 0.5603156089782715,
+      "learning_rate": 5.5853209593979354e-05,
+      "loss": 0.7361,
+      "step": 8778
+    },
+    {
+      "epoch": 1.566051199714566,
+      "grad_norm": 0.5568613409996033,
+      "learning_rate": 5.580907729359486e-05,
+      "loss": 0.5341,
+      "step": 8779
+    },
+    {
+      "epoch": 1.5662295959325663,
+      "grad_norm": 0.5259753465652466,
+      "learning_rate": 5.5764960244881815e-05,
+      "loss": 0.511,
+      "step": 8780
+    },
+    {
+      "epoch": 1.5664079921505665,
+      "grad_norm": 0.5009199976921082,
+      "learning_rate": 5.5720858451305255e-05,
+      "loss": 0.5754,
+      "step": 8781
+    },
+    {
+      "epoch": 1.5665863883685667,
+      "grad_norm": 0.48167309165000916,
+      "learning_rate": 5.567677191632883e-05,
+      "loss": 0.4789,
+      "step": 8782
+    },
+    {
+      "epoch": 1.5667647845865669,
+      "grad_norm": 0.5015170574188232,
+      "learning_rate": 5.563270064341508e-05,
+      "loss": 0.5136,
+      "step": 8783
+    },
+    {
+      "epoch": 1.566943180804567,
+      "grad_norm": 0.5342589616775513,
+      "learning_rate": 5.558864463602548e-05,
+      "loss": 0.6359,
+      "step": 8784
+    },
+    {
+      "epoch": 1.5671215770225673,
+      "grad_norm": 0.5268881916999817,
+      "learning_rate": 5.5544603897619976e-05,
+      "loss": 0.5012,
+      "step": 8785
+    },
+    {
+      "epoch": 1.5672999732405672,
+      "grad_norm": 0.5006667375564575,
+      "learning_rate": 5.5500578431657675e-05,
+      "loss": 0.5217,
+      "step": 8786
+    },
+    {
+      "epoch": 1.5674783694585674,
+      "grad_norm": 0.49677345156669617,
+      "learning_rate": 5.545656824159617e-05,
+      "loss": 0.5327,
+      "step": 8787
+    },
+    {
+      "epoch": 1.5676567656765676,
+      "grad_norm": 0.5768446326255798,
+      "learning_rate": 5.5412573330892165e-05,
+      "loss": 0.6573,
+      "step": 8788
+    },
+    {
+      "epoch": 1.5678351618945678,
+      "grad_norm": 0.48240020871162415,
+      "learning_rate": 5.53685937030009e-05,
+      "loss": 0.5009,
+      "step": 8789
+    },
+    {
+      "epoch": 1.568013558112568,
+      "grad_norm": 0.4898053705692291,
+      "learning_rate": 5.532462936137647e-05,
+      "loss": 0.5342,
+      "step": 8790
+    },
+    {
+      "epoch": 1.5681919543305682,
+      "grad_norm": 0.5353338122367859,
+      "learning_rate": 5.528068030947192e-05,
+      "loss": 0.5924,
+      "step": 8791
+    },
+    {
+      "epoch": 1.5683703505485682,
+      "grad_norm": 0.47058504819869995,
+      "learning_rate": 5.52367465507389e-05,
+      "loss": 0.4101,
+      "step": 8792
+    },
+    {
+      "epoch": 1.5685487467665684,
+      "grad_norm": 0.47491228580474854,
+      "learning_rate": 5.519282808862805e-05,
+      "loss": 0.4776,
+      "step": 8793
+    },
+    {
+      "epoch": 1.5687271429845686,
+      "grad_norm": 0.47631198167800903,
+      "learning_rate": 5.5148924926588574e-05,
+      "loss": 0.4497,
+      "step": 8794
+    },
+    {
+      "epoch": 1.5689055392025688,
+      "grad_norm": 0.4845799207687378,
+      "learning_rate": 5.510503706806877e-05,
+      "loss": 0.5219,
+      "step": 8795
+    },
+    {
+      "epoch": 1.569083935420569,
+      "grad_norm": 0.5517118573188782,
+      "learning_rate": 5.506116451651547e-05,
+      "loss": 0.5573,
+      "step": 8796
+    },
+    {
+      "epoch": 1.5692623316385692,
+      "grad_norm": 0.529208242893219,
+      "learning_rate": 5.501730727537435e-05,
+      "loss": 0.5191,
+      "step": 8797
+    },
+    {
+      "epoch": 1.5694407278565694,
+      "grad_norm": 0.48718565702438354,
+      "learning_rate": 5.497346534809011e-05,
+      "loss": 0.4791,
+      "step": 8798
+    },
+    {
+      "epoch": 1.5696191240745696,
+      "grad_norm": 0.5780670642852783,
+      "learning_rate": 5.4929638738105805e-05,
+      "loss": 0.5571,
+      "step": 8799
+    },
+    {
+      "epoch": 1.5697975202925698,
+      "grad_norm": 0.5497144460678101,
+      "learning_rate": 5.488582744886386e-05,
+      "loss": 0.6058,
+      "step": 8800
+    },
+    {
+      "epoch": 1.56997591651057,
+      "grad_norm": 0.5026035904884338,
+      "learning_rate": 5.484203148380509e-05,
+      "loss": 0.436,
+      "step": 8801
+    },
+    {
+      "epoch": 1.5701543127285702,
+      "grad_norm": 0.4526662528514862,
+      "learning_rate": 5.479825084636911e-05,
+      "loss": 0.4857,
+      "step": 8802
+    },
+    {
+      "epoch": 1.5703327089465704,
+      "grad_norm": 0.5049575567245483,
+      "learning_rate": 5.475448553999454e-05,
+      "loss": 0.5069,
+      "step": 8803
+    },
+    {
+      "epoch": 1.5705111051645706,
+      "grad_norm": 0.503604531288147,
+      "learning_rate": 5.471073556811862e-05,
+      "loss": 0.5236,
+      "step": 8804
+    },
+    {
+      "epoch": 1.5706895013825708,
+      "grad_norm": 0.5270532369613647,
+      "learning_rate": 5.4667000934177564e-05,
+      "loss": 0.5193,
+      "step": 8805
+    },
+    {
+      "epoch": 1.570867897600571,
+      "grad_norm": 0.49302002787590027,
+      "learning_rate": 5.4623281641606096e-05,
+      "loss": 0.4832,
+      "step": 8806
+    },
+    {
+      "epoch": 1.5710462938185712,
+      "grad_norm": 0.4588510990142822,
+      "learning_rate": 5.457957769383812e-05,
+      "loss": 0.4937,
+      "step": 8807
+    },
+    {
+      "epoch": 1.5712246900365712,
+      "grad_norm": 0.4889686703681946,
+      "learning_rate": 5.453588909430593e-05,
+      "loss": 0.4553,
+      "step": 8808
+    },
+    {
+      "epoch": 1.5714030862545714,
+      "grad_norm": 0.536016047000885,
+      "learning_rate": 5.4492215846440953e-05,
+      "loss": 0.6425,
+      "step": 8809
+    },
+    {
+      "epoch": 1.5715814824725716,
+      "grad_norm": 0.4554092288017273,
+      "learning_rate": 5.4448557953673204e-05,
+      "loss": 0.434,
+      "step": 8810
+    },
+    {
+      "epoch": 1.5717598786905718,
+      "grad_norm": 0.4832884669303894,
+      "learning_rate": 5.440491541943152e-05,
+      "loss": 0.4417,
+      "step": 8811
+    },
+    {
+      "epoch": 1.571938274908572,
+      "grad_norm": 0.5221408009529114,
+      "learning_rate": 5.4361288247143646e-05,
+      "loss": 0.5256,
+      "step": 8812
+    },
+    {
+      "epoch": 1.5721166711265722,
+      "grad_norm": 0.4202393889427185,
+      "learning_rate": 5.4317676440235967e-05,
+      "loss": 0.3462,
+      "step": 8813
+    },
+    {
+      "epoch": 1.5722950673445721,
+      "grad_norm": 0.4951157867908478,
+      "learning_rate": 5.427408000213374e-05,
+      "loss": 0.4871,
+      "step": 8814
+    },
+    {
+      "epoch": 1.5724734635625723,
+      "grad_norm": 0.4902489483356476,
+      "learning_rate": 5.423049893626114e-05,
+      "loss": 0.5418,
+      "step": 8815
+    },
+    {
+      "epoch": 1.5726518597805725,
+      "grad_norm": 0.5259934067726135,
+      "learning_rate": 5.418693324604082e-05,
+      "loss": 0.5426,
+      "step": 8816
+    },
+    {
+      "epoch": 1.5728302559985727,
+      "grad_norm": 0.5680667757987976,
+      "learning_rate": 5.414338293489457e-05,
+      "loss": 0.6104,
+      "step": 8817
+    },
+    {
+      "epoch": 1.573008652216573,
+      "grad_norm": 0.5012747645378113,
+      "learning_rate": 5.409984800624265e-05,
+      "loss": 0.5192,
+      "step": 8818
+    },
+    {
+      "epoch": 1.5731870484345731,
+      "grad_norm": 0.48108309507369995,
+      "learning_rate": 5.4056328463504475e-05,
+      "loss": 0.5373,
+      "step": 8819
+    },
+    {
+      "epoch": 1.5733654446525733,
+      "grad_norm": 0.5479810237884521,
+      "learning_rate": 5.401282431009785e-05,
+      "loss": 0.6286,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5735438408705735,
+      "grad_norm": 0.5330629944801331,
+      "learning_rate": 5.3969335549439726e-05,
+      "loss": 0.5358,
+      "step": 8821
+    },
+    {
+      "epoch": 1.5737222370885737,
+      "grad_norm": 0.5190926194190979,
+      "learning_rate": 5.392586218494563e-05,
+      "loss": 0.6252,
+      "step": 8822
+    },
+    {
+      "epoch": 1.573900633306574,
+      "grad_norm": 0.5403081178665161,
+      "learning_rate": 5.388240422002991e-05,
+      "loss": 0.5921,
+      "step": 8823
+    },
+    {
+      "epoch": 1.5740790295245741,
+      "grad_norm": 0.5377727746963501,
+      "learning_rate": 5.383896165810578e-05,
+      "loss": 0.4856,
+      "step": 8824
+    },
+    {
+      "epoch": 1.5742574257425743,
+      "grad_norm": 0.513603925704956,
+      "learning_rate": 5.3795534502585146e-05,
+      "loss": 0.5591,
+      "step": 8825
+    },
+    {
+      "epoch": 1.5744358219605745,
+      "grad_norm": 0.5131911039352417,
+      "learning_rate": 5.375212275687888e-05,
+      "loss": 0.4422,
+      "step": 8826
+    },
+    {
+      "epoch": 1.5746142181785747,
+      "grad_norm": 0.519425630569458,
+      "learning_rate": 5.3708726424396365e-05,
+      "loss": 0.5824,
+      "step": 8827
+    },
+    {
+      "epoch": 1.574792614396575,
+      "grad_norm": 0.47717246413230896,
+      "learning_rate": 5.366534550854607e-05,
+      "loss": 0.4575,
+      "step": 8828
+    },
+    {
+      "epoch": 1.5749710106145751,
+      "grad_norm": 0.563754677772522,
+      "learning_rate": 5.3621980012734965e-05,
+      "loss": 0.5664,
+      "step": 8829
+    },
+    {
+      "epoch": 1.575149406832575,
+      "grad_norm": 0.5223338603973389,
+      "learning_rate": 5.357862994036905e-05,
+      "loss": 0.498,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5753278030505753,
+      "grad_norm": 0.5984678268432617,
+      "learning_rate": 5.35352952948531e-05,
+      "loss": 0.6791,
+      "step": 8831
+    },
+    {
+      "epoch": 1.5755061992685755,
+      "grad_norm": 0.554014265537262,
+      "learning_rate": 5.349197607959042e-05,
+      "loss": 0.6039,
+      "step": 8832
+    },
+    {
+      "epoch": 1.5756845954865757,
+      "grad_norm": 0.561259388923645,
+      "learning_rate": 5.3448672297983445e-05,
+      "loss": 0.5535,
+      "step": 8833
+    },
+    {
+      "epoch": 1.5758629917045759,
+      "grad_norm": 0.49919480085372925,
+      "learning_rate": 5.3405383953433196e-05,
+      "loss": 0.4131,
+      "step": 8834
+    },
+    {
+      "epoch": 1.576041387922576,
+      "grad_norm": 0.5846698880195618,
+      "learning_rate": 5.336211104933938e-05,
+      "loss": 0.6158,
+      "step": 8835
+    },
+    {
+      "epoch": 1.576219784140576,
+      "grad_norm": 0.5364250540733337,
+      "learning_rate": 5.3318853589100824e-05,
+      "loss": 0.6105,
+      "step": 8836
+    },
+    {
+      "epoch": 1.5763981803585763,
+      "grad_norm": 0.5680550932884216,
+      "learning_rate": 5.3275611576114825e-05,
+      "loss": 0.7472,
+      "step": 8837
+    },
+    {
+      "epoch": 1.5765765765765765,
+      "grad_norm": 0.4823981821537018,
+      "learning_rate": 5.323238501377767e-05,
+      "loss": 0.4895,
+      "step": 8838
+    },
+    {
+      "epoch": 1.5767549727945767,
+      "grad_norm": 0.6050514578819275,
+      "learning_rate": 5.318917390548428e-05,
+      "loss": 0.7538,
+      "step": 8839
+    },
+    {
+      "epoch": 1.5769333690125769,
+      "grad_norm": 0.4770009517669678,
+      "learning_rate": 5.314597825462852e-05,
+      "loss": 0.4793,
+      "step": 8840
+    },
+    {
+      "epoch": 1.577111765230577,
+      "grad_norm": 0.4747209846973419,
+      "learning_rate": 5.310279806460286e-05,
+      "loss": 0.4359,
+      "step": 8841
+    },
+    {
+      "epoch": 1.5772901614485773,
+      "grad_norm": 0.44378408789634705,
+      "learning_rate": 5.305963333879879e-05,
+      "loss": 0.3823,
+      "step": 8842
+    },
+    {
+      "epoch": 1.5774685576665775,
+      "grad_norm": 0.4758981466293335,
+      "learning_rate": 5.301648408060633e-05,
+      "loss": 0.6459,
+      "step": 8843
+    },
+    {
+      "epoch": 1.5776469538845777,
+      "grad_norm": 0.5262596011161804,
+      "learning_rate": 5.297335029341433e-05,
+      "loss": 0.5195,
+      "step": 8844
+    },
+    {
+      "epoch": 1.5778253501025779,
+      "grad_norm": 0.4246023893356323,
+      "learning_rate": 5.2930231980610756e-05,
+      "loss": 0.3224,
+      "step": 8845
+    },
+    {
+      "epoch": 1.578003746320578,
+      "grad_norm": 0.502919614315033,
+      "learning_rate": 5.288712914558189e-05,
+      "loss": 0.5362,
+      "step": 8846
+    },
+    {
+      "epoch": 1.5781821425385782,
+      "grad_norm": 0.5143377184867859,
+      "learning_rate": 5.284404179171312e-05,
+      "loss": 0.5843,
+      "step": 8847
+    },
+    {
+      "epoch": 1.5783605387565784,
+      "grad_norm": 0.48670274019241333,
+      "learning_rate": 5.2800969922388474e-05,
+      "loss": 0.4592,
+      "step": 8848
+    },
+    {
+      "epoch": 1.5785389349745786,
+      "grad_norm": 0.45652881264686584,
+      "learning_rate": 5.2757913540990715e-05,
+      "loss": 0.42,
+      "step": 8849
+    },
+    {
+      "epoch": 1.5787173311925788,
+      "grad_norm": 0.5632809996604919,
+      "learning_rate": 5.271487265090164e-05,
+      "loss": 0.7437,
+      "step": 8850
+    },
+    {
+      "epoch": 1.578895727410579,
+      "grad_norm": 0.5193389654159546,
+      "learning_rate": 5.267184725550148e-05,
+      "loss": 0.6235,
+      "step": 8851
+    },
+    {
+      "epoch": 1.5790741236285792,
+      "grad_norm": 0.45982155203819275,
+      "learning_rate": 5.262883735816959e-05,
+      "loss": 0.4623,
+      "step": 8852
+    },
+    {
+      "epoch": 1.5792525198465792,
+      "grad_norm": 0.5154650211334229,
+      "learning_rate": 5.25858429622838e-05,
+      "loss": 0.537,
+      "step": 8853
+    },
+    {
+      "epoch": 1.5794309160645794,
+      "grad_norm": 0.46093055605888367,
+      "learning_rate": 5.254286407122103e-05,
+      "loss": 0.4579,
+      "step": 8854
+    },
+    {
+      "epoch": 1.5796093122825796,
+      "grad_norm": 0.4196965992450714,
+      "learning_rate": 5.249990068835675e-05,
+      "loss": 0.3861,
+      "step": 8855
+    },
+    {
+      "epoch": 1.5797877085005798,
+      "grad_norm": 0.5842517614364624,
+      "learning_rate": 5.24569528170652e-05,
+      "loss": 0.68,
+      "step": 8856
+    },
+    {
+      "epoch": 1.57996610471858,
+      "grad_norm": 0.5660757422447205,
+      "learning_rate": 5.2414020460719636e-05,
+      "loss": 0.5804,
+      "step": 8857
+    },
+    {
+      "epoch": 1.58014450093658,
+      "grad_norm": 0.49591630697250366,
+      "learning_rate": 5.237110362269182e-05,
+      "loss": 0.5518,
+      "step": 8858
+    },
+    {
+      "epoch": 1.5803228971545802,
+      "grad_norm": 0.5173941850662231,
+      "learning_rate": 5.232820230635255e-05,
+      "loss": 0.4721,
+      "step": 8859
+    },
+    {
+      "epoch": 1.5805012933725804,
+      "grad_norm": 0.538420557975769,
+      "learning_rate": 5.228531651507112e-05,
+      "loss": 0.5329,
+      "step": 8860
+    },
+    {
+      "epoch": 1.5806796895905806,
+      "grad_norm": 0.5405678749084473,
+      "learning_rate": 5.2242446252215856e-05,
+      "loss": 0.652,
+      "step": 8861
+    },
+    {
+      "epoch": 1.5808580858085808,
+      "grad_norm": 0.5772570967674255,
+      "learning_rate": 5.219959152115381e-05,
+      "loss": 0.459,
+      "step": 8862
+    },
+    {
+      "epoch": 1.581036482026581,
+      "grad_norm": 0.5803507566452026,
+      "learning_rate": 5.215675232525069e-05,
+      "loss": 0.6247,
+      "step": 8863
+    },
+    {
+      "epoch": 1.5812148782445812,
+      "grad_norm": 0.49514850974082947,
+      "learning_rate": 5.211392866787115e-05,
+      "loss": 0.4549,
+      "step": 8864
+    },
+    {
+      "epoch": 1.5813932744625814,
+      "grad_norm": 0.5379515290260315,
+      "learning_rate": 5.207112055237842e-05,
+      "loss": 0.6696,
+      "step": 8865
+    },
+    {
+      "epoch": 1.5815716706805816,
+      "grad_norm": 0.5178011059761047,
+      "learning_rate": 5.20283279821348e-05,
+      "loss": 0.5623,
+      "step": 8866
+    },
+    {
+      "epoch": 1.5817500668985818,
+      "grad_norm": 0.4703742563724518,
+      "learning_rate": 5.198555096050103e-05,
+      "loss": 0.5029,
+      "step": 8867
+    },
+    {
+      "epoch": 1.581928463116582,
+      "grad_norm": 0.5060920119285583,
+      "learning_rate": 5.194278949083695e-05,
+      "loss": 0.5898,
+      "step": 8868
+    },
+    {
+      "epoch": 1.5821068593345822,
+      "grad_norm": 0.4721207618713379,
+      "learning_rate": 5.1900043576500936e-05,
+      "loss": 0.3526,
+      "step": 8869
+    },
+    {
+      "epoch": 1.5822852555525824,
+      "grad_norm": 0.46023473143577576,
+      "learning_rate": 5.185731322085019e-05,
+      "loss": 0.5206,
+      "step": 8870
+    },
+    {
+      "epoch": 1.5824636517705826,
+      "grad_norm": 0.46701812744140625,
+      "learning_rate": 5.181459842724087e-05,
+      "loss": 0.4646,
+      "step": 8871
+    },
+    {
+      "epoch": 1.5826420479885828,
+      "grad_norm": 0.5196385383605957,
+      "learning_rate": 5.177189919902761e-05,
+      "loss": 0.4874,
+      "step": 8872
+    },
+    {
+      "epoch": 1.582820444206583,
+      "grad_norm": 0.5126696228981018,
+      "learning_rate": 5.172921553956417e-05,
+      "loss": 0.525,
+      "step": 8873
+    },
+    {
+      "epoch": 1.5829988404245832,
+      "grad_norm": 0.5253236889839172,
+      "learning_rate": 5.168654745220275e-05,
+      "loss": 0.5603,
+      "step": 8874
+    },
+    {
+      "epoch": 1.5831772366425831,
+      "grad_norm": 0.5543179512023926,
+      "learning_rate": 5.164389494029456e-05,
+      "loss": 0.6045,
+      "step": 8875
+    },
+    {
+      "epoch": 1.5833556328605833,
+      "grad_norm": 0.4786223769187927,
+      "learning_rate": 5.160125800718956e-05,
+      "loss": 0.4731,
+      "step": 8876
+    },
+    {
+      "epoch": 1.5835340290785835,
+      "grad_norm": 0.5591602921485901,
+      "learning_rate": 5.1558636656236306e-05,
+      "loss": 0.6183,
+      "step": 8877
+    },
+    {
+      "epoch": 1.5837124252965837,
+      "grad_norm": 0.6518511772155762,
+      "learning_rate": 5.151603089078241e-05,
+      "loss": 0.5388,
+      "step": 8878
+    },
+    {
+      "epoch": 1.583890821514584,
+      "grad_norm": 0.5040208101272583,
+      "learning_rate": 5.147344071417398e-05,
+      "loss": 0.4481,
+      "step": 8879
+    },
+    {
+      "epoch": 1.584069217732584,
+      "grad_norm": 0.533556342124939,
+      "learning_rate": 5.143086612975611e-05,
+      "loss": 0.5437,
+      "step": 8880
+    },
+    {
+      "epoch": 1.584247613950584,
+      "grad_norm": 0.4733051657676697,
+      "learning_rate": 5.1388307140872614e-05,
+      "loss": 0.51,
+      "step": 8881
+    },
+    {
+      "epoch": 1.5844260101685843,
+      "grad_norm": 0.5339987277984619,
+      "learning_rate": 5.134576375086591e-05,
+      "loss": 0.5314,
+      "step": 8882
+    },
+    {
+      "epoch": 1.5846044063865845,
+      "grad_norm": 0.477483332157135,
+      "learning_rate": 5.130323596307751e-05,
+      "loss": 0.3984,
+      "step": 8883
+    },
+    {
+      "epoch": 1.5847828026045847,
+      "grad_norm": 0.509962260723114,
+      "learning_rate": 5.126072378084737e-05,
+      "loss": 0.461,
+      "step": 8884
+    },
+    {
+      "epoch": 1.584961198822585,
+      "grad_norm": 0.5546766519546509,
+      "learning_rate": 5.121822720751454e-05,
+      "loss": 0.6239,
+      "step": 8885
+    },
+    {
+      "epoch": 1.585139595040585,
+      "grad_norm": 0.5097823739051819,
+      "learning_rate": 5.117574624641652e-05,
+      "loss": 0.4881,
+      "step": 8886
+    },
+    {
+      "epoch": 1.5853179912585853,
+      "grad_norm": 0.43369320034980774,
+      "learning_rate": 5.1133280900889924e-05,
+      "loss": 0.3753,
+      "step": 8887
+    },
+    {
+      "epoch": 1.5854963874765855,
+      "grad_norm": 0.523571252822876,
+      "learning_rate": 5.109083117426977e-05,
+      "loss": 0.6094,
+      "step": 8888
+    },
+    {
+      "epoch": 1.5856747836945857,
+      "grad_norm": 0.5631095170974731,
+      "learning_rate": 5.104839706989023e-05,
+      "loss": 0.5408,
+      "step": 8889
+    },
+    {
+      "epoch": 1.585853179912586,
+      "grad_norm": 0.4018095135688782,
+      "learning_rate": 5.100597859108388e-05,
+      "loss": 0.3599,
+      "step": 8890
+    },
+    {
+      "epoch": 1.586031576130586,
+      "grad_norm": 0.5422604084014893,
+      "learning_rate": 5.096357574118235e-05,
+      "loss": 0.4818,
+      "step": 8891
+    },
+    {
+      "epoch": 1.5862099723485863,
+      "grad_norm": 0.4900394082069397,
+      "learning_rate": 5.092118852351599e-05,
+      "loss": 0.4453,
+      "step": 8892
+    },
+    {
+      "epoch": 1.5863883685665865,
+      "grad_norm": 0.5061340928077698,
+      "learning_rate": 5.0878816941413744e-05,
+      "loss": 0.5332,
+      "step": 8893
+    },
+    {
+      "epoch": 1.5865667647845867,
+      "grad_norm": 0.4389035999774933,
+      "learning_rate": 5.0836460998203606e-05,
+      "loss": 0.361,
+      "step": 8894
+    },
+    {
+      "epoch": 1.586745161002587,
+      "grad_norm": 0.533852756023407,
+      "learning_rate": 5.0794120697212094e-05,
+      "loss": 0.5116,
+      "step": 8895
+    },
+    {
+      "epoch": 1.586923557220587,
+      "grad_norm": 0.5132311582565308,
+      "learning_rate": 5.075179604176458e-05,
+      "loss": 0.5874,
+      "step": 8896
+    },
+    {
+      "epoch": 1.587101953438587,
+      "grad_norm": 0.49478766322135925,
+      "learning_rate": 5.07094870351853e-05,
+      "loss": 0.4595,
+      "step": 8897
+    },
+    {
+      "epoch": 1.5872803496565873,
+      "grad_norm": 0.5392327904701233,
+      "learning_rate": 5.066719368079708e-05,
+      "loss": 0.7202,
+      "step": 8898
+    },
+    {
+      "epoch": 1.5874587458745875,
+      "grad_norm": 0.5095738768577576,
+      "learning_rate": 5.062491598192179e-05,
+      "loss": 0.5124,
+      "step": 8899
+    },
+    {
+      "epoch": 1.5876371420925877,
+      "grad_norm": 0.5276995301246643,
+      "learning_rate": 5.058265394187969e-05,
+      "loss": 0.6014,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5878155383105879,
+      "grad_norm": 0.5482483506202698,
+      "learning_rate": 5.054040756399023e-05,
+      "loss": 0.7651,
+      "step": 8901
+    },
+    {
+      "epoch": 1.5879939345285878,
+      "grad_norm": 0.46668532490730286,
+      "learning_rate": 5.049817685157132e-05,
+      "loss": 0.4828,
+      "step": 8902
+    },
+    {
+      "epoch": 1.588172330746588,
+      "grad_norm": 0.6453546285629272,
+      "learning_rate": 5.045596180793968e-05,
+      "loss": 0.7255,
+      "step": 8903
+    },
+    {
+      "epoch": 1.5883507269645882,
+      "grad_norm": 0.47187116742134094,
+      "learning_rate": 5.041376243641099e-05,
+      "loss": 0.4884,
+      "step": 8904
+    },
+    {
+      "epoch": 1.5885291231825884,
+      "grad_norm": 0.4838643968105316,
+      "learning_rate": 5.037157874029946e-05,
+      "loss": 0.5405,
+      "step": 8905
+    },
+    {
+      "epoch": 1.5887075194005886,
+      "grad_norm": 0.5167852640151978,
+      "learning_rate": 5.032941072291822e-05,
+      "loss": 0.6482,
+      "step": 8906
+    },
+    {
+      "epoch": 1.5888859156185888,
+      "grad_norm": 0.44085460901260376,
+      "learning_rate": 5.028725838757919e-05,
+      "loss": 0.4037,
+      "step": 8907
+    },
+    {
+      "epoch": 1.589064311836589,
+      "grad_norm": 0.5068758726119995,
+      "learning_rate": 5.024512173759288e-05,
+      "loss": 0.5228,
+      "step": 8908
+    },
+    {
+      "epoch": 1.5892427080545892,
+      "grad_norm": 0.4763888418674469,
+      "learning_rate": 5.020300077626883e-05,
+      "loss": 0.4412,
+      "step": 8909
+    },
+    {
+      "epoch": 1.5894211042725894,
+      "grad_norm": 0.5120984315872192,
+      "learning_rate": 5.016089550691505e-05,
+      "loss": 0.5497,
+      "step": 8910
+    },
+    {
+      "epoch": 1.5895995004905896,
+      "grad_norm": 0.5709746479988098,
+      "learning_rate": 5.0118805932838604e-05,
+      "loss": 0.5747,
+      "step": 8911
+    },
+    {
+      "epoch": 1.5897778967085898,
+      "grad_norm": 0.5391095280647278,
+      "learning_rate": 5.0076732057345034e-05,
+      "loss": 0.547,
+      "step": 8912
+    },
+    {
+      "epoch": 1.58995629292659,
+      "grad_norm": 0.5848898887634277,
+      "learning_rate": 5.0034673883738974e-05,
+      "loss": 0.5088,
+      "step": 8913
+    },
+    {
+      "epoch": 1.5901346891445902,
+      "grad_norm": 0.4863363206386566,
+      "learning_rate": 4.999263141532359e-05,
+      "loss": 0.4024,
+      "step": 8914
+    },
+    {
+      "epoch": 1.5903130853625904,
+      "grad_norm": 0.5417057871818542,
+      "learning_rate": 4.99506046554008e-05,
+      "loss": 0.5766,
+      "step": 8915
+    },
+    {
+      "epoch": 1.5904914815805906,
+      "grad_norm": 0.5208100080490112,
+      "learning_rate": 4.990859360727148e-05,
+      "loss": 0.617,
+      "step": 8916
+    },
+    {
+      "epoch": 1.5906698777985908,
+      "grad_norm": 0.5628762245178223,
+      "learning_rate": 4.9866598274235064e-05,
+      "loss": 0.6443,
+      "step": 8917
+    },
+    {
+      "epoch": 1.590848274016591,
+      "grad_norm": 0.5478858351707458,
+      "learning_rate": 4.9824618659589957e-05,
+      "loss": 0.7009,
+      "step": 8918
+    },
+    {
+      "epoch": 1.591026670234591,
+      "grad_norm": 0.5053479075431824,
+      "learning_rate": 4.9782654766633076e-05,
+      "loss": 0.4551,
+      "step": 8919
+    },
+    {
+      "epoch": 1.5912050664525912,
+      "grad_norm": 0.46800491213798523,
+      "learning_rate": 4.9740706598660396e-05,
+      "loss": 0.4591,
+      "step": 8920
+    },
+    {
+      "epoch": 1.5913834626705914,
+      "grad_norm": 0.47018712759017944,
+      "learning_rate": 4.9698774158966395e-05,
+      "loss": 0.4372,
+      "step": 8921
+    },
+    {
+      "epoch": 1.5915618588885916,
+      "grad_norm": 0.4378306567668915,
+      "learning_rate": 4.965685745084447e-05,
+      "loss": 0.4064,
+      "step": 8922
+    },
+    {
+      "epoch": 1.5917402551065918,
+      "grad_norm": 0.4751421809196472,
+      "learning_rate": 4.961495647758679e-05,
+      "loss": 0.5386,
+      "step": 8923
+    },
+    {
+      "epoch": 1.5919186513245918,
+      "grad_norm": 0.4474140703678131,
+      "learning_rate": 4.957307124248417e-05,
+      "loss": 0.5067,
+      "step": 8924
+    },
+    {
+      "epoch": 1.592097047542592,
+      "grad_norm": 0.5166023969650269,
+      "learning_rate": 4.9531201748826335e-05,
+      "loss": 0.5903,
+      "step": 8925
+    },
+    {
+      "epoch": 1.5922754437605922,
+      "grad_norm": 0.5033789277076721,
+      "learning_rate": 4.9489347999901567e-05,
+      "loss": 0.5299,
+      "step": 8926
+    },
+    {
+      "epoch": 1.5924538399785924,
+      "grad_norm": 0.4894052743911743,
+      "learning_rate": 4.944750999899719e-05,
+      "loss": 0.4912,
+      "step": 8927
+    },
+    {
+      "epoch": 1.5926322361965926,
+      "grad_norm": 0.484478235244751,
+      "learning_rate": 4.9405687749399076e-05,
+      "loss": 0.5635,
+      "step": 8928
+    },
+    {
+      "epoch": 1.5928106324145928,
+      "grad_norm": 0.47258055210113525,
+      "learning_rate": 4.936388125439184e-05,
+      "loss": 0.4479,
+      "step": 8929
+    },
+    {
+      "epoch": 1.592989028632593,
+      "grad_norm": 0.4957214593887329,
+      "learning_rate": 4.932209051725914e-05,
+      "loss": 0.554,
+      "step": 8930
+    },
+    {
+      "epoch": 1.5931674248505932,
+      "grad_norm": 0.5196431279182434,
+      "learning_rate": 4.9280315541282985e-05,
+      "loss": 0.5595,
+      "step": 8931
+    },
+    {
+      "epoch": 1.5933458210685933,
+      "grad_norm": 0.4598284065723419,
+      "learning_rate": 4.923855632974455e-05,
+      "loss": 0.421,
+      "step": 8932
+    },
+    {
+      "epoch": 1.5935242172865935,
+      "grad_norm": 0.42510855197906494,
+      "learning_rate": 4.919681288592345e-05,
+      "loss": 0.3646,
+      "step": 8933
+    },
+    {
+      "epoch": 1.5937026135045937,
+      "grad_norm": 0.5351304411888123,
+      "learning_rate": 4.91550852130983e-05,
+      "loss": 0.5159,
+      "step": 8934
+    },
+    {
+      "epoch": 1.593881009722594,
+      "grad_norm": 0.499664306640625,
+      "learning_rate": 4.911337331454635e-05,
+      "loss": 0.4688,
+      "step": 8935
+    },
+    {
+      "epoch": 1.5940594059405941,
+      "grad_norm": 0.47697949409484863,
+      "learning_rate": 4.907167719354347e-05,
+      "loss": 0.4619,
+      "step": 8936
+    },
+    {
+      "epoch": 1.5942378021585943,
+      "grad_norm": 0.48516783118247986,
+      "learning_rate": 4.9029996853364734e-05,
+      "loss": 0.5222,
+      "step": 8937
+    },
+    {
+      "epoch": 1.5944161983765945,
+      "grad_norm": 0.4567139446735382,
+      "learning_rate": 4.898833229728347e-05,
+      "loss": 0.4139,
+      "step": 8938
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 0.5564625859260559,
+      "learning_rate": 4.894668352857218e-05,
+      "loss": 0.5797,
+      "step": 8939
+    },
+    {
+      "epoch": 1.594772990812595,
+      "grad_norm": 0.5010836124420166,
+      "learning_rate": 4.890505055050182e-05,
+      "loss": 0.5232,
+      "step": 8940
+    },
+    {
+      "epoch": 1.594951387030595,
+      "grad_norm": 0.5170525908470154,
+      "learning_rate": 4.886343336634222e-05,
+      "loss": 0.5822,
+      "step": 8941
+    },
+    {
+      "epoch": 1.5951297832485951,
+      "grad_norm": 0.5084660649299622,
+      "learning_rate": 4.8821831979362044e-05,
+      "loss": 0.5954,
+      "step": 8942
+    },
+    {
+      "epoch": 1.5953081794665953,
+      "grad_norm": 0.49296459555625916,
+      "learning_rate": 4.878024639282855e-05,
+      "loss": 0.5069,
+      "step": 8943
+    },
+    {
+      "epoch": 1.5954865756845955,
+      "grad_norm": 0.515155017375946,
+      "learning_rate": 4.8738676610008e-05,
+      "loss": 0.5101,
+      "step": 8944
+    },
+    {
+      "epoch": 1.5956649719025957,
+      "grad_norm": 0.3873237073421478,
+      "learning_rate": 4.869712263416509e-05,
+      "loss": 0.2959,
+      "step": 8945
+    },
+    {
+      "epoch": 1.5958433681205957,
+      "grad_norm": 0.4798237681388855,
+      "learning_rate": 4.865558446856361e-05,
+      "loss": 0.5004,
+      "step": 8946
+    },
+    {
+      "epoch": 1.5960217643385959,
+      "grad_norm": 0.5282852053642273,
+      "learning_rate": 4.8614062116465826e-05,
+      "loss": 0.5694,
+      "step": 8947
+    },
+    {
+      "epoch": 1.596200160556596,
+      "grad_norm": 0.43830955028533936,
+      "learning_rate": 4.8572555581132995e-05,
+      "loss": 0.4124,
+      "step": 8948
+    },
+    {
+      "epoch": 1.5963785567745963,
+      "grad_norm": 0.5239266753196716,
+      "learning_rate": 4.853106486582498e-05,
+      "loss": 0.6721,
+      "step": 8949
+    },
+    {
+      "epoch": 1.5965569529925965,
+      "grad_norm": 0.46557673811912537,
+      "learning_rate": 4.8489589973800346e-05,
+      "loss": 0.4868,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5967353492105967,
+      "grad_norm": 0.5202271342277527,
+      "learning_rate": 4.844813090831668e-05,
+      "loss": 0.734,
+      "step": 8951
+    },
+    {
+      "epoch": 1.5969137454285969,
+      "grad_norm": 0.4756893515586853,
+      "learning_rate": 4.840668767262993e-05,
+      "loss": 0.5069,
+      "step": 8952
+    },
+    {
+      "epoch": 1.597092141646597,
+      "grad_norm": 0.40755748748779297,
+      "learning_rate": 4.836526026999532e-05,
+      "loss": 0.3434,
+      "step": 8953
+    },
+    {
+      "epoch": 1.5972705378645973,
+      "grad_norm": 0.5153047442436218,
+      "learning_rate": 4.8323848703666405e-05,
+      "loss": 0.4607,
+      "step": 8954
+    },
+    {
+      "epoch": 1.5974489340825975,
+      "grad_norm": 0.5465795397758484,
+      "learning_rate": 4.8282452976895566e-05,
+      "loss": 0.5989,
+      "step": 8955
+    },
+    {
+      "epoch": 1.5976273303005977,
+      "grad_norm": 0.47764524817466736,
+      "learning_rate": 4.8241073092934104e-05,
+      "loss": 0.4355,
+      "step": 8956
+    },
+    {
+      "epoch": 1.5978057265185979,
+      "grad_norm": 0.5183848142623901,
+      "learning_rate": 4.8199709055031876e-05,
+      "loss": 0.4979,
+      "step": 8957
+    },
+    {
+      "epoch": 1.597984122736598,
+      "grad_norm": 0.55250483751297,
+      "learning_rate": 4.815836086643774e-05,
+      "loss": 0.5273,
+      "step": 8958
+    },
+    {
+      "epoch": 1.5981625189545983,
+      "grad_norm": 0.4400150775909424,
+      "learning_rate": 4.811702853039901e-05,
+      "loss": 0.4164,
+      "step": 8959
+    },
+    {
+      "epoch": 1.5983409151725985,
+      "grad_norm": 0.5529983639717102,
+      "learning_rate": 4.807571205016206e-05,
+      "loss": 0.6202,
+      "step": 8960
+    },
+    {
+      "epoch": 1.5985193113905987,
+      "grad_norm": 0.5263171792030334,
+      "learning_rate": 4.803441142897178e-05,
+      "loss": 0.5444,
+      "step": 8961
+    },
+    {
+      "epoch": 1.5986977076085989,
+      "grad_norm": 0.4600481688976288,
+      "learning_rate": 4.799312667007183e-05,
+      "loss": 0.4249,
+      "step": 8962
+    },
+    {
+      "epoch": 1.5988761038265988,
+      "grad_norm": 0.5363233089447021,
+      "learning_rate": 4.795185777670485e-05,
+      "loss": 0.4851,
+      "step": 8963
+    },
+    {
+      "epoch": 1.599054500044599,
+      "grad_norm": 0.478712797164917,
+      "learning_rate": 4.791060475211198e-05,
+      "loss": 0.3852,
+      "step": 8964
+    },
+    {
+      "epoch": 1.5992328962625992,
+      "grad_norm": 0.5279228687286377,
+      "learning_rate": 4.7869367599533284e-05,
+      "loss": 0.6256,
+      "step": 8965
+    },
+    {
+      "epoch": 1.5994112924805994,
+      "grad_norm": 0.4477089047431946,
+      "learning_rate": 4.782814632220742e-05,
+      "loss": 0.4195,
+      "step": 8966
+    },
+    {
+      "epoch": 1.5995896886985996,
+      "grad_norm": 0.5009279847145081,
+      "learning_rate": 4.778694092337194e-05,
+      "loss": 0.5,
+      "step": 8967
+    },
+    {
+      "epoch": 1.5997680849165996,
+      "grad_norm": 0.47871533036231995,
+      "learning_rate": 4.7745751406263163e-05,
+      "loss": 0.5258,
+      "step": 8968
+    },
+    {
+      "epoch": 1.5999464811345998,
+      "grad_norm": 0.47330227494239807,
+      "learning_rate": 4.770457777411597e-05,
+      "loss": 0.4042,
+      "step": 8969
+    },
+    {
+      "epoch": 1.6001248773526,
+      "grad_norm": 0.4700087308883667,
+      "learning_rate": 4.766342003016424e-05,
+      "loss": 0.462,
+      "step": 8970
+    },
+    {
+      "epoch": 1.6003032735706002,
+      "grad_norm": 0.5046184659004211,
+      "learning_rate": 4.7622278177640366e-05,
+      "loss": 0.4632,
+      "step": 8971
+    },
+    {
+      "epoch": 1.6004816697886004,
+      "grad_norm": 0.5220001935958862,
+      "learning_rate": 4.758115221977574e-05,
+      "loss": 0.531,
+      "step": 8972
+    },
+    {
+      "epoch": 1.6006600660066006,
+      "grad_norm": 0.5016428232192993,
+      "learning_rate": 4.7540042159800264e-05,
+      "loss": 0.5008,
+      "step": 8973
+    },
+    {
+      "epoch": 1.6008384622246008,
+      "grad_norm": 0.4888211786746979,
+      "learning_rate": 4.7498948000942814e-05,
+      "loss": 0.4499,
+      "step": 8974
+    },
+    {
+      "epoch": 1.601016858442601,
+      "grad_norm": 0.5536308288574219,
+      "learning_rate": 4.745786974643082e-05,
+      "loss": 0.5715,
+      "step": 8975
+    },
+    {
+      "epoch": 1.6011952546606012,
+      "grad_norm": 0.48591530323028564,
+      "learning_rate": 4.741680739949053e-05,
+      "loss": 0.4089,
+      "step": 8976
+    },
+    {
+      "epoch": 1.6013736508786014,
+      "grad_norm": 0.6232868432998657,
+      "learning_rate": 4.7375760963347056e-05,
+      "loss": 0.5869,
+      "step": 8977
+    },
+    {
+      "epoch": 1.6015520470966016,
+      "grad_norm": 0.5616008043289185,
+      "learning_rate": 4.733473044122408e-05,
+      "loss": 0.5639,
+      "step": 8978
+    },
+    {
+      "epoch": 1.6017304433146018,
+      "grad_norm": 0.5132565498352051,
+      "learning_rate": 4.72937158363442e-05,
+      "loss": 0.53,
+      "step": 8979
+    },
+    {
+      "epoch": 1.601908839532602,
+      "grad_norm": 0.5778293609619141,
+      "learning_rate": 4.725271715192861e-05,
+      "loss": 0.7178,
+      "step": 8980
+    },
+    {
+      "epoch": 1.6020872357506022,
+      "grad_norm": 0.5367584228515625,
+      "learning_rate": 4.721173439119742e-05,
+      "loss": 0.5345,
+      "step": 8981
+    },
+    {
+      "epoch": 1.6022656319686024,
+      "grad_norm": 0.4915335476398468,
+      "learning_rate": 4.7170767557369264e-05,
+      "loss": 0.5284,
+      "step": 8982
+    },
+    {
+      "epoch": 1.6024440281866026,
+      "grad_norm": 0.524190366268158,
+      "learning_rate": 4.712981665366176e-05,
+      "loss": 0.5688,
+      "step": 8983
+    },
+    {
+      "epoch": 1.6026224244046028,
+      "grad_norm": 0.5471283793449402,
+      "learning_rate": 4.70888816832912e-05,
+      "loss": 0.5799,
+      "step": 8984
+    },
+    {
+      "epoch": 1.6028008206226028,
+      "grad_norm": 0.5151530504226685,
+      "learning_rate": 4.7047962649472504e-05,
+      "loss": 0.5424,
+      "step": 8985
+    },
+    {
+      "epoch": 1.602979216840603,
+      "grad_norm": 0.5410858988761902,
+      "learning_rate": 4.7007059555419535e-05,
+      "loss": 0.624,
+      "step": 8986
+    },
+    {
+      "epoch": 1.6031576130586032,
+      "grad_norm": 0.4905013144016266,
+      "learning_rate": 4.696617240434475e-05,
+      "loss": 0.4588,
+      "step": 8987
+    },
+    {
+      "epoch": 1.6033360092766034,
+      "grad_norm": 0.4888882637023926,
+      "learning_rate": 4.692530119945937e-05,
+      "loss": 0.4577,
+      "step": 8988
+    },
+    {
+      "epoch": 1.6035144054946036,
+      "grad_norm": 0.4958007335662842,
+      "learning_rate": 4.688444594397351e-05,
+      "loss": 0.519,
+      "step": 8989
+    },
+    {
+      "epoch": 1.6036928017126035,
+      "grad_norm": 0.5347129106521606,
+      "learning_rate": 4.684360664109577e-05,
+      "loss": 0.687,
+      "step": 8990
+    },
+    {
+      "epoch": 1.6038711979306037,
+      "grad_norm": 0.5105026364326477,
+      "learning_rate": 4.680278329403381e-05,
+      "loss": 0.5317,
+      "step": 8991
+    },
+    {
+      "epoch": 1.604049594148604,
+      "grad_norm": 0.4922098219394684,
+      "learning_rate": 4.676197590599377e-05,
+      "loss": 0.482,
+      "step": 8992
+    },
+    {
+      "epoch": 1.6042279903666041,
+      "grad_norm": 0.44692763686180115,
+      "learning_rate": 4.672118448018073e-05,
+      "loss": 0.4144,
+      "step": 8993
+    },
+    {
+      "epoch": 1.6044063865846043,
+      "grad_norm": 0.5374802947044373,
+      "learning_rate": 4.6680409019798364e-05,
+      "loss": 0.6081,
+      "step": 8994
+    },
+    {
+      "epoch": 1.6045847828026045,
+      "grad_norm": 0.54610675573349,
+      "learning_rate": 4.6639649528049135e-05,
+      "loss": 0.6023,
+      "step": 8995
+    },
+    {
+      "epoch": 1.6047631790206047,
+      "grad_norm": 0.4369504451751709,
+      "learning_rate": 4.659890600813438e-05,
+      "loss": 0.357,
+      "step": 8996
+    },
+    {
+      "epoch": 1.604941575238605,
+      "grad_norm": 0.47287872433662415,
+      "learning_rate": 4.6558178463253944e-05,
+      "loss": 0.4559,
+      "step": 8997
+    },
+    {
+      "epoch": 1.6051199714566051,
+      "grad_norm": 0.4916176497936249,
+      "learning_rate": 4.651746689660663e-05,
+      "loss": 0.5313,
+      "step": 8998
+    },
+    {
+      "epoch": 1.6052983676746053,
+      "grad_norm": 0.5396872758865356,
+      "learning_rate": 4.647677131138997e-05,
+      "loss": 0.5313,
+      "step": 8999
+    },
+    {
+      "epoch": 1.6054767638926055,
+      "grad_norm": 0.5110781788825989,
+      "learning_rate": 4.643609171080001e-05,
+      "loss": 0.5488,
+      "step": 9000
+    },
+    {
+      "epoch": 1.6056551601106057,
+      "grad_norm": 0.5322538018226624,
+      "learning_rate": 4.63954280980319e-05,
+      "loss": 0.5717,
+      "step": 9001
+    },
+    {
+      "epoch": 1.605833556328606,
+      "grad_norm": 0.49910300970077515,
+      "learning_rate": 4.63547804762792e-05,
+      "loss": 0.5186,
+      "step": 9002
+    },
+    {
+      "epoch": 1.6060119525466061,
+      "grad_norm": 0.48461076617240906,
+      "learning_rate": 4.631414884873444e-05,
+      "loss": 0.5344,
+      "step": 9003
+    },
+    {
+      "epoch": 1.6061903487646063,
+      "grad_norm": 0.626358151435852,
+      "learning_rate": 4.627353321858874e-05,
+      "loss": 0.7774,
+      "step": 9004
+    },
+    {
+      "epoch": 1.6063687449826065,
+      "grad_norm": 0.6041339039802551,
+      "learning_rate": 4.6232933589032105e-05,
+      "loss": 0.6707,
+      "step": 9005
+    },
+    {
+      "epoch": 1.6065471412006067,
+      "grad_norm": 0.4728602468967438,
+      "learning_rate": 4.619234996325314e-05,
+      "loss": 0.4132,
+      "step": 9006
+    },
+    {
+      "epoch": 1.6067255374186067,
+      "grad_norm": 0.5530493259429932,
+      "learning_rate": 4.6151782344439366e-05,
+      "loss": 0.5413,
+      "step": 9007
+    },
+    {
+      "epoch": 1.6069039336366069,
+      "grad_norm": 0.5874399542808533,
+      "learning_rate": 4.611123073577686e-05,
+      "loss": 0.7141,
+      "step": 9008
+    },
+    {
+      "epoch": 1.607082329854607,
+      "grad_norm": 0.5239754319190979,
+      "learning_rate": 4.607069514045051e-05,
+      "loss": 0.6276,
+      "step": 9009
+    },
+    {
+      "epoch": 1.6072607260726073,
+      "grad_norm": 0.5062006115913391,
+      "learning_rate": 4.603017556164407e-05,
+      "loss": 0.4942,
+      "step": 9010
+    },
+    {
+      "epoch": 1.6074391222906075,
+      "grad_norm": 0.49783483147621155,
+      "learning_rate": 4.5989672002539785e-05,
+      "loss": 0.4906,
+      "step": 9011
+    },
+    {
+      "epoch": 1.6076175185086075,
+      "grad_norm": 0.5406367182731628,
+      "learning_rate": 4.594918446631896e-05,
+      "loss": 0.6172,
+      "step": 9012
+    },
+    {
+      "epoch": 1.6077959147266077,
+      "grad_norm": 0.48416727781295776,
+      "learning_rate": 4.590871295616128e-05,
+      "loss": 0.5055,
+      "step": 9013
+    },
+    {
+      "epoch": 1.6079743109446079,
+      "grad_norm": 0.5911693572998047,
+      "learning_rate": 4.586825747524548e-05,
+      "loss": 0.7576,
+      "step": 9014
+    },
+    {
+      "epoch": 1.608152707162608,
+      "grad_norm": 0.5087306499481201,
+      "learning_rate": 4.582781802674896e-05,
+      "loss": 0.4797,
+      "step": 9015
+    },
+    {
+      "epoch": 1.6083311033806083,
+      "grad_norm": 0.4570036232471466,
+      "learning_rate": 4.578739461384765e-05,
+      "loss": 0.5285,
+      "step": 9016
+    },
+    {
+      "epoch": 1.6085094995986084,
+      "grad_norm": 0.46012383699417114,
+      "learning_rate": 4.57469872397166e-05,
+      "loss": 0.4309,
+      "step": 9017
+    },
+    {
+      "epoch": 1.6086878958166086,
+      "grad_norm": 0.4837871789932251,
+      "learning_rate": 4.570659590752918e-05,
+      "loss": 0.5151,
+      "step": 9018
+    },
+    {
+      "epoch": 1.6088662920346088,
+      "grad_norm": 0.5085347890853882,
+      "learning_rate": 4.566622062045786e-05,
+      "loss": 0.6172,
+      "step": 9019
+    },
+    {
+      "epoch": 1.609044688252609,
+      "grad_norm": 0.4480245113372803,
+      "learning_rate": 4.562586138167368e-05,
+      "loss": 0.3621,
+      "step": 9020
+    },
+    {
+      "epoch": 1.6092230844706092,
+      "grad_norm": 0.5031635761260986,
+      "learning_rate": 4.558551819434631e-05,
+      "loss": 0.5342,
+      "step": 9021
+    },
+    {
+      "epoch": 1.6094014806886094,
+      "grad_norm": 0.5782181024551392,
+      "learning_rate": 4.554519106164442e-05,
+      "loss": 0.551,
+      "step": 9022
+    },
+    {
+      "epoch": 1.6095798769066096,
+      "grad_norm": 0.5325855612754822,
+      "learning_rate": 4.550487998673519e-05,
+      "loss": 0.5286,
+      "step": 9023
+    },
+    {
+      "epoch": 1.6097582731246098,
+      "grad_norm": 0.4554942548274994,
+      "learning_rate": 4.5464584972784774e-05,
+      "loss": 0.453,
+      "step": 9024
+    },
+    {
+      "epoch": 1.60993666934261,
+      "grad_norm": 0.5228267312049866,
+      "learning_rate": 4.5424306022957745e-05,
+      "loss": 0.6083,
+      "step": 9025
+    },
+    {
+      "epoch": 1.6101150655606102,
+      "grad_norm": 0.5161296725273132,
+      "learning_rate": 4.538404314041775e-05,
+      "loss": 0.5022,
+      "step": 9026
+    },
+    {
+      "epoch": 1.6102934617786104,
+      "grad_norm": 0.5143990516662598,
+      "learning_rate": 4.534379632832691e-05,
+      "loss": 0.4937,
+      "step": 9027
+    },
+    {
+      "epoch": 1.6104718579966106,
+      "grad_norm": 0.483243852853775,
+      "learning_rate": 4.530356558984622e-05,
+      "loss": 0.4821,
+      "step": 9028
+    },
+    {
+      "epoch": 1.6106502542146106,
+      "grad_norm": 0.5501946210861206,
+      "learning_rate": 4.5263350928135465e-05,
+      "loss": 0.6569,
+      "step": 9029
+    },
+    {
+      "epoch": 1.6108286504326108,
+      "grad_norm": 0.5732588171958923,
+      "learning_rate": 4.5223152346352964e-05,
+      "loss": 0.7411,
+      "step": 9030
+    },
+    {
+      "epoch": 1.611007046650611,
+      "grad_norm": 0.4644021987915039,
+      "learning_rate": 4.518296984765599e-05,
+      "loss": 0.506,
+      "step": 9031
+    },
+    {
+      "epoch": 1.6111854428686112,
+      "grad_norm": 0.5346475839614868,
+      "learning_rate": 4.514280343520041e-05,
+      "loss": 0.5926,
+      "step": 9032
+    },
+    {
+      "epoch": 1.6113638390866114,
+      "grad_norm": 0.4996803104877472,
+      "learning_rate": 4.510265311214093e-05,
+      "loss": 0.4819,
+      "step": 9033
+    },
+    {
+      "epoch": 1.6115422353046114,
+      "grad_norm": 0.5373480319976807,
+      "learning_rate": 4.50625188816309e-05,
+      "loss": 0.5144,
+      "step": 9034
+    },
+    {
+      "epoch": 1.6117206315226116,
+      "grad_norm": 0.5674717426300049,
+      "learning_rate": 4.5022400746822374e-05,
+      "loss": 0.5336,
+      "step": 9035
+    },
+    {
+      "epoch": 1.6118990277406118,
+      "grad_norm": 0.531956672668457,
+      "learning_rate": 4.498229871086637e-05,
+      "loss": 0.4986,
+      "step": 9036
+    },
+    {
+      "epoch": 1.612077423958612,
+      "grad_norm": 0.5014679431915283,
+      "learning_rate": 4.4942212776912325e-05,
+      "loss": 0.5051,
+      "step": 9037
+    },
+    {
+      "epoch": 1.6122558201766122,
+      "grad_norm": 0.5485700368881226,
+      "learning_rate": 4.4902142948108684e-05,
+      "loss": 0.502,
+      "step": 9038
+    },
+    {
+      "epoch": 1.6124342163946124,
+      "grad_norm": 0.5426976084709167,
+      "learning_rate": 4.486208922760243e-05,
+      "loss": 0.5571,
+      "step": 9039
+    },
+    {
+      "epoch": 1.6126126126126126,
+      "grad_norm": 0.49919578433036804,
+      "learning_rate": 4.48220516185395e-05,
+      "loss": 0.4595,
+      "step": 9040
+    },
+    {
+      "epoch": 1.6127910088306128,
+      "grad_norm": 0.43158578872680664,
+      "learning_rate": 4.4782030124064314e-05,
+      "loss": 0.3117,
+      "step": 9041
+    },
+    {
+      "epoch": 1.612969405048613,
+      "grad_norm": 0.5657439231872559,
+      "learning_rate": 4.474202474732011e-05,
+      "loss": 0.5668,
+      "step": 9042
+    },
+    {
+      "epoch": 1.6131478012666132,
+      "grad_norm": 0.5453231334686279,
+      "learning_rate": 4.470203549144902e-05,
+      "loss": 0.6741,
+      "step": 9043
+    },
+    {
+      "epoch": 1.6133261974846134,
+      "grad_norm": 0.5499635338783264,
+      "learning_rate": 4.4662062359591585e-05,
+      "loss": 0.719,
+      "step": 9044
+    },
+    {
+      "epoch": 1.6135045937026136,
+      "grad_norm": 0.43844422698020935,
+      "learning_rate": 4.4622105354887534e-05,
+      "loss": 0.3209,
+      "step": 9045
+    },
+    {
+      "epoch": 1.6136829899206138,
+      "grad_norm": 0.5634275078773499,
+      "learning_rate": 4.458216448047494e-05,
+      "loss": 0.5132,
+      "step": 9046
+    },
+    {
+      "epoch": 1.613861386138614,
+      "grad_norm": 0.4690277874469757,
+      "learning_rate": 4.4542239739490705e-05,
+      "loss": 0.3584,
+      "step": 9047
+    },
+    {
+      "epoch": 1.6140397823566142,
+      "grad_norm": 0.4696029722690582,
+      "learning_rate": 4.45023311350706e-05,
+      "loss": 0.4606,
+      "step": 9048
+    },
+    {
+      "epoch": 1.6142181785746144,
+      "grad_norm": 0.4961758553981781,
+      "learning_rate": 4.446243867034891e-05,
+      "loss": 0.5177,
+      "step": 9049
+    },
+    {
+      "epoch": 1.6143965747926146,
+      "grad_norm": 0.5290566682815552,
+      "learning_rate": 4.44225623484589e-05,
+      "loss": 0.5366,
+      "step": 9050
+    },
+    {
+      "epoch": 1.6145749710106145,
+      "grad_norm": 0.5504355430603027,
+      "learning_rate": 4.438270217253232e-05,
+      "loss": 0.6723,
+      "step": 9051
+    },
+    {
+      "epoch": 1.6147533672286147,
+      "grad_norm": 0.49992886185646057,
+      "learning_rate": 4.434285814569988e-05,
+      "loss": 0.5295,
+      "step": 9052
+    },
+    {
+      "epoch": 1.614931763446615,
+      "grad_norm": 0.5225392580032349,
+      "learning_rate": 4.430303027109081e-05,
+      "loss": 0.575,
+      "step": 9053
+    },
+    {
+      "epoch": 1.6151101596646151,
+      "grad_norm": 0.5753360390663147,
+      "learning_rate": 4.4263218551833294e-05,
+      "loss": 0.6629,
+      "step": 9054
+    },
+    {
+      "epoch": 1.6152885558826153,
+      "grad_norm": 0.4710274040699005,
+      "learning_rate": 4.422342299105403e-05,
+      "loss": 0.5181,
+      "step": 9055
+    },
+    {
+      "epoch": 1.6154669521006153,
+      "grad_norm": 0.5554249286651611,
+      "learning_rate": 4.4183643591878515e-05,
+      "loss": 0.6106,
+      "step": 9056
+    },
+    {
+      "epoch": 1.6156453483186155,
+      "grad_norm": 0.5206986665725708,
+      "learning_rate": 4.414388035743114e-05,
+      "loss": 0.5476,
+      "step": 9057
+    },
+    {
+      "epoch": 1.6158237445366157,
+      "grad_norm": 0.5251834392547607,
+      "learning_rate": 4.410413329083473e-05,
+      "loss": 0.6463,
+      "step": 9058
+    },
+    {
+      "epoch": 1.616002140754616,
+      "grad_norm": 0.49586114287376404,
+      "learning_rate": 4.4064402395211116e-05,
+      "loss": 0.427,
+      "step": 9059
+    },
+    {
+      "epoch": 1.616180536972616,
+      "grad_norm": 0.4815883934497833,
+      "learning_rate": 4.402468767368076e-05,
+      "loss": 0.4655,
+      "step": 9060
+    },
+    {
+      "epoch": 1.6163589331906163,
+      "grad_norm": 0.5105904936790466,
+      "learning_rate": 4.3984989129362744e-05,
+      "loss": 0.4924,
+      "step": 9061
+    },
+    {
+      "epoch": 1.6165373294086165,
+      "grad_norm": 0.47116348147392273,
+      "learning_rate": 4.3945306765375086e-05,
+      "loss": 0.4197,
+      "step": 9062
+    },
+    {
+      "epoch": 1.6167157256266167,
+      "grad_norm": 0.516294538974762,
+      "learning_rate": 4.390564058483429e-05,
+      "loss": 0.5377,
+      "step": 9063
+    },
+    {
+      "epoch": 1.616894121844617,
+      "grad_norm": 0.4575921595096588,
+      "learning_rate": 4.3865990590855885e-05,
+      "loss": 0.5038,
+      "step": 9064
+    },
+    {
+      "epoch": 1.617072518062617,
+      "grad_norm": 0.5077314376831055,
+      "learning_rate": 4.3826356786553776e-05,
+      "loss": 0.4385,
+      "step": 9065
+    },
+    {
+      "epoch": 1.6172509142806173,
+      "grad_norm": 0.4583306908607483,
+      "learning_rate": 4.378673917504094e-05,
+      "loss": 0.4456,
+      "step": 9066
+    },
+    {
+      "epoch": 1.6174293104986175,
+      "grad_norm": 0.5505391955375671,
+      "learning_rate": 4.37471377594289e-05,
+      "loss": 0.5579,
+      "step": 9067
+    },
+    {
+      "epoch": 1.6176077067166177,
+      "grad_norm": 0.501794159412384,
+      "learning_rate": 4.3707552542827824e-05,
+      "loss": 0.4491,
+      "step": 9068
+    },
+    {
+      "epoch": 1.6177861029346179,
+      "grad_norm": 0.4820462465286255,
+      "learning_rate": 4.366798352834686e-05,
+      "loss": 0.4494,
+      "step": 9069
+    },
+    {
+      "epoch": 1.617964499152618,
+      "grad_norm": 0.6349626183509827,
+      "learning_rate": 4.3628430719093614e-05,
+      "loss": 0.6564,
+      "step": 9070
+    },
+    {
+      "epoch": 1.6181428953706183,
+      "grad_norm": 0.577896773815155,
+      "learning_rate": 4.3588894118174685e-05,
+      "loss": 0.6188,
+      "step": 9071
+    },
+    {
+      "epoch": 1.6183212915886185,
+      "grad_norm": 0.5295639634132385,
+      "learning_rate": 4.3549373728695105e-05,
+      "loss": 0.5548,
+      "step": 9072
+    },
+    {
+      "epoch": 1.6184996878066185,
+      "grad_norm": 0.5311222076416016,
+      "learning_rate": 4.350986955375893e-05,
+      "loss": 0.4302,
+      "step": 9073
+    },
+    {
+      "epoch": 1.6186780840246187,
+      "grad_norm": 0.4720657467842102,
+      "learning_rate": 4.3470381596468714e-05,
+      "loss": 0.4666,
+      "step": 9074
+    },
+    {
+      "epoch": 1.6188564802426189,
+      "grad_norm": 0.4858959913253784,
+      "learning_rate": 4.3430909859925814e-05,
+      "loss": 0.5341,
+      "step": 9075
+    },
+    {
+      "epoch": 1.619034876460619,
+      "grad_norm": 0.5353506207466125,
+      "learning_rate": 4.339145434723044e-05,
+      "loss": 0.5534,
+      "step": 9076
+    },
+    {
+      "epoch": 1.6192132726786193,
+      "grad_norm": 0.4330075681209564,
+      "learning_rate": 4.335201506148126e-05,
+      "loss": 0.363,
+      "step": 9077
+    },
+    {
+      "epoch": 1.6193916688966192,
+      "grad_norm": 0.5149824023246765,
+      "learning_rate": 4.3312592005775946e-05,
+      "loss": 0.5245,
+      "step": 9078
+    },
+    {
+      "epoch": 1.6195700651146194,
+      "grad_norm": 0.49177101254463196,
+      "learning_rate": 4.327318518321074e-05,
+      "loss": 0.5551,
+      "step": 9079
+    },
+    {
+      "epoch": 1.6197484613326196,
+      "grad_norm": 0.47272542119026184,
+      "learning_rate": 4.323379459688051e-05,
+      "loss": 0.4035,
+      "step": 9080
+    },
+    {
+      "epoch": 1.6199268575506198,
+      "grad_norm": 0.509996235370636,
+      "learning_rate": 4.319442024987916e-05,
+      "loss": 0.5317,
+      "step": 9081
+    },
+    {
+      "epoch": 1.62010525376862,
+      "grad_norm": 0.5650163888931274,
+      "learning_rate": 4.315506214529899e-05,
+      "loss": 0.7105,
+      "step": 9082
+    },
+    {
+      "epoch": 1.6202836499866202,
+      "grad_norm": 0.5097303986549377,
+      "learning_rate": 4.3115720286231257e-05,
+      "loss": 0.4972,
+      "step": 9083
+    },
+    {
+      "epoch": 1.6204620462046204,
+      "grad_norm": 0.5151503086090088,
+      "learning_rate": 4.3076394675765796e-05,
+      "loss": 0.5673,
+      "step": 9084
+    },
+    {
+      "epoch": 1.6206404424226206,
+      "grad_norm": 0.5100242495536804,
+      "learning_rate": 4.30370853169913e-05,
+      "loss": 0.4747,
+      "step": 9085
+    },
+    {
+      "epoch": 1.6208188386406208,
+      "grad_norm": 0.5815212726593018,
+      "learning_rate": 4.299779221299499e-05,
+      "loss": 0.6796,
+      "step": 9086
+    },
+    {
+      "epoch": 1.620997234858621,
+      "grad_norm": 0.5294939875602722,
+      "learning_rate": 4.2958515366863075e-05,
+      "loss": 0.5649,
+      "step": 9087
+    },
+    {
+      "epoch": 1.6211756310766212,
+      "grad_norm": 0.45899614691734314,
+      "learning_rate": 4.291925478168024e-05,
+      "loss": 0.4118,
+      "step": 9088
+    },
+    {
+      "epoch": 1.6213540272946214,
+      "grad_norm": 0.5030704736709595,
+      "learning_rate": 4.288001046052992e-05,
+      "loss": 0.5717,
+      "step": 9089
+    },
+    {
+      "epoch": 1.6215324235126216,
+      "grad_norm": 0.4515380561351776,
+      "learning_rate": 4.284078240649458e-05,
+      "loss": 0.4641,
+      "step": 9090
+    },
+    {
+      "epoch": 1.6217108197306218,
+      "grad_norm": 0.5059680938720703,
+      "learning_rate": 4.280157062265497e-05,
+      "loss": 0.4881,
+      "step": 9091
+    },
+    {
+      "epoch": 1.621889215948622,
+      "grad_norm": 0.4840905964374542,
+      "learning_rate": 4.2762375112090886e-05,
+      "loss": 0.4164,
+      "step": 9092
+    },
+    {
+      "epoch": 1.6220676121666222,
+      "grad_norm": 0.5307683348655701,
+      "learning_rate": 4.2723195877880706e-05,
+      "loss": 0.5169,
+      "step": 9093
+    },
+    {
+      "epoch": 1.6222460083846224,
+      "grad_norm": 0.6119624376296997,
+      "learning_rate": 4.268403292310144e-05,
+      "loss": 0.4207,
+      "step": 9094
+    },
+    {
+      "epoch": 1.6224244046026224,
+      "grad_norm": 0.4793750047683716,
+      "learning_rate": 4.264488625082907e-05,
+      "loss": 0.4304,
+      "step": 9095
+    },
+    {
+      "epoch": 1.6226028008206226,
+      "grad_norm": 0.5105946660041809,
+      "learning_rate": 4.260575586413806e-05,
+      "loss": 0.4994,
+      "step": 9096
+    },
+    {
+      "epoch": 1.6227811970386228,
+      "grad_norm": 0.5425340533256531,
+      "learning_rate": 4.256664176610178e-05,
+      "loss": 0.5135,
+      "step": 9097
+    },
+    {
+      "epoch": 1.622959593256623,
+      "grad_norm": 0.5028733611106873,
+      "learning_rate": 4.252754395979216e-05,
+      "loss": 0.4335,
+      "step": 9098
+    },
+    {
+      "epoch": 1.6231379894746232,
+      "grad_norm": 0.48356130719184875,
+      "learning_rate": 4.2488462448280005e-05,
+      "loss": 0.3819,
+      "step": 9099
+    },
+    {
+      "epoch": 1.6233163856926232,
+      "grad_norm": 0.5290387868881226,
+      "learning_rate": 4.244939723463467e-05,
+      "loss": 0.4932,
+      "step": 9100
+    },
+    {
+      "epoch": 1.6234947819106234,
+      "grad_norm": 0.4586488604545593,
+      "learning_rate": 4.241034832192434e-05,
+      "loss": 0.3835,
+      "step": 9101
+    },
+    {
+      "epoch": 1.6236731781286236,
+      "grad_norm": 0.5234144926071167,
+      "learning_rate": 4.237131571321598e-05,
+      "loss": 0.5544,
+      "step": 9102
+    },
+    {
+      "epoch": 1.6238515743466237,
+      "grad_norm": 0.46121421456336975,
+      "learning_rate": 4.233229941157504e-05,
+      "loss": 0.3899,
+      "step": 9103
+    },
+    {
+      "epoch": 1.624029970564624,
+      "grad_norm": 0.500857412815094,
+      "learning_rate": 4.229329942006604e-05,
+      "loss": 0.6343,
+      "step": 9104
+    },
+    {
+      "epoch": 1.6242083667826241,
+      "grad_norm": 0.47724878787994385,
+      "learning_rate": 4.225431574175184e-05,
+      "loss": 0.4782,
+      "step": 9105
+    },
+    {
+      "epoch": 1.6243867630006243,
+      "grad_norm": 0.46206358075141907,
+      "learning_rate": 4.221534837969429e-05,
+      "loss": 0.44,
+      "step": 9106
+    },
+    {
+      "epoch": 1.6245651592186245,
+      "grad_norm": 0.5365794897079468,
+      "learning_rate": 4.217639733695391e-05,
+      "loss": 0.5337,
+      "step": 9107
+    },
+    {
+      "epoch": 1.6247435554366247,
+      "grad_norm": 0.49346688389778137,
+      "learning_rate": 4.21374626165898e-05,
+      "loss": 0.4771,
+      "step": 9108
+    },
+    {
+      "epoch": 1.624921951654625,
+      "grad_norm": 0.5585914254188538,
+      "learning_rate": 4.209854422165998e-05,
+      "loss": 0.5715,
+      "step": 9109
+    },
+    {
+      "epoch": 1.6251003478726251,
+      "grad_norm": 0.49180254340171814,
+      "learning_rate": 4.205964215522096e-05,
+      "loss": 0.5383,
+      "step": 9110
+    },
+    {
+      "epoch": 1.6252787440906253,
+      "grad_norm": 0.5028634667396545,
+      "learning_rate": 4.202075642032824e-05,
+      "loss": 0.5635,
+      "step": 9111
+    },
+    {
+      "epoch": 1.6254571403086255,
+      "grad_norm": 0.583336353302002,
+      "learning_rate": 4.198188702003575e-05,
+      "loss": 0.6608,
+      "step": 9112
+    },
+    {
+      "epoch": 1.6256355365266257,
+      "grad_norm": 0.5647164583206177,
+      "learning_rate": 4.194303395739638e-05,
+      "loss": 0.7343,
+      "step": 9113
+    },
+    {
+      "epoch": 1.625813932744626,
+      "grad_norm": 0.4452681839466095,
+      "learning_rate": 4.19041972354616e-05,
+      "loss": 0.4388,
+      "step": 9114
+    },
+    {
+      "epoch": 1.6259923289626261,
+      "grad_norm": 0.3759744167327881,
+      "learning_rate": 4.186537685728156e-05,
+      "loss": 0.2556,
+      "step": 9115
+    },
+    {
+      "epoch": 1.6261707251806263,
+      "grad_norm": 0.6420390605926514,
+      "learning_rate": 4.1826572825905296e-05,
+      "loss": 0.8147,
+      "step": 9116
+    },
+    {
+      "epoch": 1.6263491213986263,
+      "grad_norm": 0.4719489812850952,
+      "learning_rate": 4.178778514438036e-05,
+      "loss": 0.501,
+      "step": 9117
+    },
+    {
+      "epoch": 1.6265275176166265,
+      "grad_norm": 0.5943500995635986,
+      "learning_rate": 4.174901381575327e-05,
+      "loss": 0.5954,
+      "step": 9118
+    },
+    {
+      "epoch": 1.6267059138346267,
+      "grad_norm": 0.45145153999328613,
+      "learning_rate": 4.171025884306892e-05,
+      "loss": 0.3752,
+      "step": 9119
+    },
+    {
+      "epoch": 1.626884310052627,
+      "grad_norm": 0.49264243245124817,
+      "learning_rate": 4.1671520229371234e-05,
+      "loss": 0.3958,
+      "step": 9120
+    },
+    {
+      "epoch": 1.627062706270627,
+      "grad_norm": 0.5612894296646118,
+      "learning_rate": 4.163279797770275e-05,
+      "loss": 0.6676,
+      "step": 9121
+    },
+    {
+      "epoch": 1.627241102488627,
+      "grad_norm": 0.6251209378242493,
+      "learning_rate": 4.1594092091104594e-05,
+      "loss": 0.7134,
+      "step": 9122
+    },
+    {
+      "epoch": 1.6274194987066273,
+      "grad_norm": 0.49562308192253113,
+      "learning_rate": 4.155540257261681e-05,
+      "loss": 0.5082,
+      "step": 9123
+    },
+    {
+      "epoch": 1.6275978949246275,
+      "grad_norm": 0.6067000031471252,
+      "learning_rate": 4.1516729425277924e-05,
+      "loss": 0.6371,
+      "step": 9124
+    },
+    {
+      "epoch": 1.6277762911426277,
+      "grad_norm": 0.46219685673713684,
+      "learning_rate": 4.14780726521255e-05,
+      "loss": 0.4726,
+      "step": 9125
+    },
+    {
+      "epoch": 1.6279546873606279,
+      "grad_norm": 0.5683661103248596,
+      "learning_rate": 4.143943225619548e-05,
+      "loss": 0.6561,
+      "step": 9126
+    },
+    {
+      "epoch": 1.628133083578628,
+      "grad_norm": 0.5201235413551331,
+      "learning_rate": 4.140080824052264e-05,
+      "loss": 0.6107,
+      "step": 9127
+    },
+    {
+      "epoch": 1.6283114797966283,
+      "grad_norm": 0.5444372296333313,
+      "learning_rate": 4.1362200608140635e-05,
+      "loss": 0.6235,
+      "step": 9128
+    },
+    {
+      "epoch": 1.6284898760146285,
+      "grad_norm": 0.5139498114585876,
+      "learning_rate": 4.132360936208154e-05,
+      "loss": 0.4917,
+      "step": 9129
+    },
+    {
+      "epoch": 1.6286682722326287,
+      "grad_norm": 0.43743810057640076,
+      "learning_rate": 4.1285034505376436e-05,
+      "loss": 0.4096,
+      "step": 9130
+    },
+    {
+      "epoch": 1.6288466684506289,
+      "grad_norm": 0.5461446046829224,
+      "learning_rate": 4.124647604105483e-05,
+      "loss": 0.6073,
+      "step": 9131
+    },
+    {
+      "epoch": 1.629025064668629,
+      "grad_norm": 0.5800282955169678,
+      "learning_rate": 4.120793397214523e-05,
+      "loss": 0.6349,
+      "step": 9132
+    },
+    {
+      "epoch": 1.6292034608866293,
+      "grad_norm": 0.6056262850761414,
+      "learning_rate": 4.1169408301674566e-05,
+      "loss": 0.5903,
+      "step": 9133
+    },
+    {
+      "epoch": 1.6293818571046295,
+      "grad_norm": 0.47324472665786743,
+      "learning_rate": 4.113089903266879e-05,
+      "loss": 0.4385,
+      "step": 9134
+    },
+    {
+      "epoch": 1.6295602533226297,
+      "grad_norm": 0.5741338133811951,
+      "learning_rate": 4.109240616815227e-05,
+      "loss": 0.6246,
+      "step": 9135
+    },
+    {
+      "epoch": 1.6297386495406299,
+      "grad_norm": 0.5018966794013977,
+      "learning_rate": 4.105392971114824e-05,
+      "loss": 0.5335,
+      "step": 9136
+    },
+    {
+      "epoch": 1.62991704575863,
+      "grad_norm": 0.5375863909721375,
+      "learning_rate": 4.101546966467873e-05,
+      "loss": 0.6031,
+      "step": 9137
+    },
+    {
+      "epoch": 1.6300954419766303,
+      "grad_norm": 0.5858981013298035,
+      "learning_rate": 4.0977026031764286e-05,
+      "loss": 0.4568,
+      "step": 9138
+    },
+    {
+      "epoch": 1.6302738381946302,
+      "grad_norm": 0.43643826246261597,
+      "learning_rate": 4.093859881542422e-05,
+      "loss": 0.3848,
+      "step": 9139
+    },
+    {
+      "epoch": 1.6304522344126304,
+      "grad_norm": 0.5104368329048157,
+      "learning_rate": 4.09001880186767e-05,
+      "loss": 0.5939,
+      "step": 9140
+    },
+    {
+      "epoch": 1.6306306306306306,
+      "grad_norm": 0.5159505605697632,
+      "learning_rate": 4.0861793644538374e-05,
+      "loss": 0.55,
+      "step": 9141
+    },
+    {
+      "epoch": 1.6308090268486308,
+      "grad_norm": 0.5293989777565002,
+      "learning_rate": 4.082341569602482e-05,
+      "loss": 0.5429,
+      "step": 9142
+    },
+    {
+      "epoch": 1.630987423066631,
+      "grad_norm": 0.42517781257629395,
+      "learning_rate": 4.0785054176150135e-05,
+      "loss": 0.4012,
+      "step": 9143
+    },
+    {
+      "epoch": 1.631165819284631,
+      "grad_norm": 0.5382733941078186,
+      "learning_rate": 4.07467090879273e-05,
+      "loss": 0.5455,
+      "step": 9144
+    },
+    {
+      "epoch": 1.6313442155026312,
+      "grad_norm": 0.503835141658783,
+      "learning_rate": 4.0708380434367864e-05,
+      "loss": 0.5749,
+      "step": 9145
+    },
+    {
+      "epoch": 1.6315226117206314,
+      "grad_norm": 0.44363367557525635,
+      "learning_rate": 4.067006821848218e-05,
+      "loss": 0.4532,
+      "step": 9146
+    },
+    {
+      "epoch": 1.6317010079386316,
+      "grad_norm": 0.4841151833534241,
+      "learning_rate": 4.063177244327929e-05,
+      "loss": 0.4812,
+      "step": 9147
+    },
+    {
+      "epoch": 1.6318794041566318,
+      "grad_norm": 0.4346253573894501,
+      "learning_rate": 4.059349311176683e-05,
+      "loss": 0.4208,
+      "step": 9148
+    },
+    {
+      "epoch": 1.632057800374632,
+      "grad_norm": 0.46320968866348267,
+      "learning_rate": 4.055523022695135e-05,
+      "loss": 0.5005,
+      "step": 9149
+    },
+    {
+      "epoch": 1.6322361965926322,
+      "grad_norm": 0.4925728440284729,
+      "learning_rate": 4.051698379183791e-05,
+      "loss": 0.5086,
+      "step": 9150
+    },
+    {
+      "epoch": 1.6324145928106324,
+      "grad_norm": 0.4870889484882355,
+      "learning_rate": 4.047875380943039e-05,
+      "loss": 0.4466,
+      "step": 9151
+    },
+    {
+      "epoch": 1.6325929890286326,
+      "grad_norm": 0.4800761342048645,
+      "learning_rate": 4.0440540282731476e-05,
+      "loss": 0.4516,
+      "step": 9152
+    },
+    {
+      "epoch": 1.6327713852466328,
+      "grad_norm": 0.5413798689842224,
+      "learning_rate": 4.040234321474226e-05,
+      "loss": 0.455,
+      "step": 9153
+    },
+    {
+      "epoch": 1.632949781464633,
+      "grad_norm": 0.4921496510505676,
+      "learning_rate": 4.0364162608462904e-05,
+      "loss": 0.4303,
+      "step": 9154
+    },
+    {
+      "epoch": 1.6331281776826332,
+      "grad_norm": 0.5225081443786621,
+      "learning_rate": 4.0325998466891914e-05,
+      "loss": 0.4553,
+      "step": 9155
+    },
+    {
+      "epoch": 1.6333065739006334,
+      "grad_norm": 0.4533441364765167,
+      "learning_rate": 4.0287850793026825e-05,
+      "loss": 0.4079,
+      "step": 9156
+    },
+    {
+      "epoch": 1.6334849701186336,
+      "grad_norm": 0.521223247051239,
+      "learning_rate": 4.024971958986365e-05,
+      "loss": 0.5097,
+      "step": 9157
+    },
+    {
+      "epoch": 1.6336633663366338,
+      "grad_norm": 0.4792344272136688,
+      "learning_rate": 4.0211604860397295e-05,
+      "loss": 0.4567,
+      "step": 9158
+    },
+    {
+      "epoch": 1.633841762554634,
+      "grad_norm": 0.5261356830596924,
+      "learning_rate": 4.0173506607621227e-05,
+      "loss": 0.5597,
+      "step": 9159
+    },
+    {
+      "epoch": 1.6340201587726342,
+      "grad_norm": 0.5096288919448853,
+      "learning_rate": 4.013542483452759e-05,
+      "loss": 0.3727,
+      "step": 9160
+    },
+    {
+      "epoch": 1.6341985549906342,
+      "grad_norm": 0.5453589558601379,
+      "learning_rate": 4.0097359544107424e-05,
+      "loss": 0.5976,
+      "step": 9161
+    },
+    {
+      "epoch": 1.6343769512086344,
+      "grad_norm": 0.5257166028022766,
+      "learning_rate": 4.005931073935024e-05,
+      "loss": 0.4802,
+      "step": 9162
+    },
+    {
+      "epoch": 1.6345553474266346,
+      "grad_norm": 0.4476458728313446,
+      "learning_rate": 4.002127842324452e-05,
+      "loss": 0.4914,
+      "step": 9163
+    },
+    {
+      "epoch": 1.6347337436446348,
+      "grad_norm": 0.4358140230178833,
+      "learning_rate": 3.998326259877716e-05,
+      "loss": 0.3802,
+      "step": 9164
+    },
+    {
+      "epoch": 1.634912139862635,
+      "grad_norm": 0.4720216989517212,
+      "learning_rate": 3.994526326893405e-05,
+      "loss": 0.4643,
+      "step": 9165
+    },
+    {
+      "epoch": 1.635090536080635,
+      "grad_norm": 0.5170959234237671,
+      "learning_rate": 3.990728043669953e-05,
+      "loss": 0.6022,
+      "step": 9166
+    },
+    {
+      "epoch": 1.6352689322986351,
+      "grad_norm": 0.49744245409965515,
+      "learning_rate": 3.986931410505676e-05,
+      "loss": 0.4991,
+      "step": 9167
+    },
+    {
+      "epoch": 1.6354473285166353,
+      "grad_norm": 0.5377562642097473,
+      "learning_rate": 3.9831364276987717e-05,
+      "loss": 0.5476,
+      "step": 9168
+    },
+    {
+      "epoch": 1.6356257247346355,
+      "grad_norm": 0.5449107885360718,
+      "learning_rate": 3.97934309554728e-05,
+      "loss": 0.5108,
+      "step": 9169
+    },
+    {
+      "epoch": 1.6358041209526357,
+      "grad_norm": 0.49150487780570984,
+      "learning_rate": 3.9755514143491434e-05,
+      "loss": 0.3777,
+      "step": 9170
+    },
+    {
+      "epoch": 1.635982517170636,
+      "grad_norm": 0.5345403552055359,
+      "learning_rate": 3.971761384402145e-05,
+      "loss": 0.5278,
+      "step": 9171
+    },
+    {
+      "epoch": 1.6361609133886361,
+      "grad_norm": 0.5451551079750061,
+      "learning_rate": 3.9679730060039634e-05,
+      "loss": 0.4898,
+      "step": 9172
+    },
+    {
+      "epoch": 1.6363393096066363,
+      "grad_norm": 0.5874769687652588,
+      "learning_rate": 3.964186279452131e-05,
+      "loss": 0.6324,
+      "step": 9173
+    },
+    {
+      "epoch": 1.6365177058246365,
+      "grad_norm": 0.47333958745002747,
+      "learning_rate": 3.960401205044051e-05,
+      "loss": 0.4932,
+      "step": 9174
+    },
+    {
+      "epoch": 1.6366961020426367,
+      "grad_norm": 0.4722568094730377,
+      "learning_rate": 3.95661778307701e-05,
+      "loss": 0.3842,
+      "step": 9175
+    },
+    {
+      "epoch": 1.636874498260637,
+      "grad_norm": 0.55815190076828,
+      "learning_rate": 3.952836013848149e-05,
+      "loss": 0.668,
+      "step": 9176
+    },
+    {
+      "epoch": 1.637052894478637,
+      "grad_norm": 0.5843459963798523,
+      "learning_rate": 3.9490558976544965e-05,
+      "loss": 0.6534,
+      "step": 9177
+    },
+    {
+      "epoch": 1.6372312906966373,
+      "grad_norm": 0.44996100664138794,
+      "learning_rate": 3.9452774347929264e-05,
+      "loss": 0.3837,
+      "step": 9178
+    },
+    {
+      "epoch": 1.6374096869146375,
+      "grad_norm": 0.46350741386413574,
+      "learning_rate": 3.9415006255602123e-05,
+      "loss": 0.4841,
+      "step": 9179
+    },
+    {
+      "epoch": 1.6375880831326377,
+      "grad_norm": 0.5201796889305115,
+      "learning_rate": 3.9377254702529784e-05,
+      "loss": 0.5497,
+      "step": 9180
+    },
+    {
+      "epoch": 1.637766479350638,
+      "grad_norm": 0.5969486236572266,
+      "learning_rate": 3.933951969167709e-05,
+      "loss": 0.6576,
+      "step": 9181
+    },
+    {
+      "epoch": 1.637944875568638,
+      "grad_norm": 0.5387462377548218,
+      "learning_rate": 3.9301801226008014e-05,
+      "loss": 0.7601,
+      "step": 9182
+    },
+    {
+      "epoch": 1.638123271786638,
+      "grad_norm": 0.6476037502288818,
+      "learning_rate": 3.926409930848471e-05,
+      "loss": 0.7146,
+      "step": 9183
+    },
+    {
+      "epoch": 1.6383016680046383,
+      "grad_norm": 0.48097339272499084,
+      "learning_rate": 3.922641394206844e-05,
+      "loss": 0.4778,
+      "step": 9184
+    },
+    {
+      "epoch": 1.6384800642226385,
+      "grad_norm": 0.46164533495903015,
+      "learning_rate": 3.91887451297189e-05,
+      "loss": 0.3921,
+      "step": 9185
+    },
+    {
+      "epoch": 1.6386584604406387,
+      "grad_norm": 0.4835575520992279,
+      "learning_rate": 3.915109287439453e-05,
+      "loss": 0.4748,
+      "step": 9186
+    },
+    {
+      "epoch": 1.6388368566586389,
+      "grad_norm": 0.48624852299690247,
+      "learning_rate": 3.911345717905268e-05,
+      "loss": 0.4619,
+      "step": 9187
+    },
+    {
+      "epoch": 1.6390152528766389,
+      "grad_norm": 0.5269492864608765,
+      "learning_rate": 3.907583804664908e-05,
+      "loss": 0.4905,
+      "step": 9188
+    },
+    {
+      "epoch": 1.639193649094639,
+      "grad_norm": 0.538076639175415,
+      "learning_rate": 3.9038235480138435e-05,
+      "loss": 0.6448,
+      "step": 9189
+    },
+    {
+      "epoch": 1.6393720453126392,
+      "grad_norm": 0.5943297147750854,
+      "learning_rate": 3.9000649482473946e-05,
+      "loss": 0.6616,
+      "step": 9190
+    },
+    {
+      "epoch": 1.6395504415306394,
+      "grad_norm": 0.509810745716095,
+      "learning_rate": 3.8963080056607705e-05,
+      "loss": 0.5005,
+      "step": 9191
+    },
+    {
+      "epoch": 1.6397288377486396,
+      "grad_norm": 0.5434547662734985,
+      "learning_rate": 3.892552720549028e-05,
+      "loss": 0.5415,
+      "step": 9192
+    },
+    {
+      "epoch": 1.6399072339666398,
+      "grad_norm": 0.47341305017471313,
+      "learning_rate": 3.8887990932071156e-05,
+      "loss": 0.4871,
+      "step": 9193
+    },
+    {
+      "epoch": 1.64008563018464,
+      "grad_norm": 0.5444488525390625,
+      "learning_rate": 3.885047123929841e-05,
+      "loss": 0.6654,
+      "step": 9194
+    },
+    {
+      "epoch": 1.6402640264026402,
+      "grad_norm": 0.49865496158599854,
+      "learning_rate": 3.8812968130118696e-05,
+      "loss": 0.5161,
+      "step": 9195
+    },
+    {
+      "epoch": 1.6404424226206404,
+      "grad_norm": 0.6822000741958618,
+      "learning_rate": 3.877548160747768e-05,
+      "loss": 0.7265,
+      "step": 9196
+    },
+    {
+      "epoch": 1.6406208188386406,
+      "grad_norm": 0.5185505151748657,
+      "learning_rate": 3.873801167431928e-05,
+      "loss": 0.4423,
+      "step": 9197
+    },
+    {
+      "epoch": 1.6407992150566408,
+      "grad_norm": 0.5268929600715637,
+      "learning_rate": 3.8700558333586686e-05,
+      "loss": 0.563,
+      "step": 9198
+    },
+    {
+      "epoch": 1.640977611274641,
+      "grad_norm": 0.5534122586250305,
+      "learning_rate": 3.86631215882213e-05,
+      "loss": 0.6052,
+      "step": 9199
+    },
+    {
+      "epoch": 1.6411560074926412,
+      "grad_norm": 0.4893801212310791,
+      "learning_rate": 3.862570144116334e-05,
+      "loss": 0.5551,
+      "step": 9200
+    },
+    {
+      "epoch": 1.6413344037106414,
+      "grad_norm": 0.5151475667953491,
+      "learning_rate": 3.858829789535187e-05,
+      "loss": 0.5263,
+      "step": 9201
+    },
+    {
+      "epoch": 1.6415127999286416,
+      "grad_norm": 0.5644826292991638,
+      "learning_rate": 3.8550910953724456e-05,
+      "loss": 0.5625,
+      "step": 9202
+    },
+    {
+      "epoch": 1.6416911961466418,
+      "grad_norm": 0.5256550312042236,
+      "learning_rate": 3.851354061921758e-05,
+      "loss": 0.6462,
+      "step": 9203
+    },
+    {
+      "epoch": 1.641869592364642,
+      "grad_norm": 0.47919219732284546,
+      "learning_rate": 3.847618689476612e-05,
+      "loss": 0.4722,
+      "step": 9204
+    },
+    {
+      "epoch": 1.642047988582642,
+      "grad_norm": 0.5013111233711243,
+      "learning_rate": 3.8438849783304e-05,
+      "loss": 0.5176,
+      "step": 9205
+    },
+    {
+      "epoch": 1.6422263848006422,
+      "grad_norm": 0.5357193350791931,
+      "learning_rate": 3.840152928776358e-05,
+      "loss": 0.6217,
+      "step": 9206
+    },
+    {
+      "epoch": 1.6424047810186424,
+      "grad_norm": 0.5302641987800598,
+      "learning_rate": 3.836422541107593e-05,
+      "loss": 0.5448,
+      "step": 9207
+    },
+    {
+      "epoch": 1.6425831772366426,
+      "grad_norm": 0.5999414324760437,
+      "learning_rate": 3.832693815617097e-05,
+      "loss": 0.5155,
+      "step": 9208
+    },
+    {
+      "epoch": 1.6427615734546428,
+      "grad_norm": 0.5149564146995544,
+      "learning_rate": 3.828966752597718e-05,
+      "loss": 0.5502,
+      "step": 9209
+    },
+    {
+      "epoch": 1.6429399696726428,
+      "grad_norm": 0.46984735131263733,
+      "learning_rate": 3.8252413523421816e-05,
+      "loss": 0.3823,
+      "step": 9210
+    },
+    {
+      "epoch": 1.643118365890643,
+      "grad_norm": 0.4093347191810608,
+      "learning_rate": 3.821517615143075e-05,
+      "loss": 0.3818,
+      "step": 9211
+    },
+    {
+      "epoch": 1.6432967621086432,
+      "grad_norm": 0.5009578466415405,
+      "learning_rate": 3.817795541292859e-05,
+      "loss": 0.5433,
+      "step": 9212
+    },
+    {
+      "epoch": 1.6434751583266434,
+      "grad_norm": 0.5088503360748291,
+      "learning_rate": 3.8140751310838715e-05,
+      "loss": 0.4853,
+      "step": 9213
+    },
+    {
+      "epoch": 1.6436535545446436,
+      "grad_norm": 0.42338132858276367,
+      "learning_rate": 3.810356384808303e-05,
+      "loss": 0.3726,
+      "step": 9214
+    },
+    {
+      "epoch": 1.6438319507626438,
+      "grad_norm": 0.5196266770362854,
+      "learning_rate": 3.806639302758227e-05,
+      "loss": 0.4542,
+      "step": 9215
+    },
+    {
+      "epoch": 1.644010346980644,
+      "grad_norm": 0.5105109214782715,
+      "learning_rate": 3.802923885225576e-05,
+      "loss": 0.4412,
+      "step": 9216
+    },
+    {
+      "epoch": 1.6441887431986442,
+      "grad_norm": 0.5296480655670166,
+      "learning_rate": 3.7992101325021674e-05,
+      "loss": 0.5869,
+      "step": 9217
+    },
+    {
+      "epoch": 1.6443671394166444,
+      "grad_norm": 0.5003146529197693,
+      "learning_rate": 3.7954980448796724e-05,
+      "loss": 0.4731,
+      "step": 9218
+    },
+    {
+      "epoch": 1.6445455356346446,
+      "grad_norm": 0.5123296976089478,
+      "learning_rate": 3.7917876226496284e-05,
+      "loss": 0.5295,
+      "step": 9219
+    },
+    {
+      "epoch": 1.6447239318526448,
+      "grad_norm": 0.45916229486465454,
+      "learning_rate": 3.788078866103467e-05,
+      "loss": 0.4446,
+      "step": 9220
+    },
+    {
+      "epoch": 1.644902328070645,
+      "grad_norm": 0.5705180168151855,
+      "learning_rate": 3.7843717755324525e-05,
+      "loss": 0.5378,
+      "step": 9221
+    },
+    {
+      "epoch": 1.6450807242886452,
+      "grad_norm": 0.5570235848426819,
+      "learning_rate": 3.78066635122776e-05,
+      "loss": 0.5277,
+      "step": 9222
+    },
+    {
+      "epoch": 1.6452591205066454,
+      "grad_norm": 0.5336431860923767,
+      "learning_rate": 3.7769625934803904e-05,
+      "loss": 0.5871,
+      "step": 9223
+    },
+    {
+      "epoch": 1.6454375167246456,
+      "grad_norm": 0.4929081201553345,
+      "learning_rate": 3.773260502581255e-05,
+      "loss": 0.5053,
+      "step": 9224
+    },
+    {
+      "epoch": 1.6456159129426458,
+      "grad_norm": 0.4710312783718109,
+      "learning_rate": 3.7695600788210967e-05,
+      "loss": 0.5122,
+      "step": 9225
+    },
+    {
+      "epoch": 1.645794309160646,
+      "grad_norm": 0.44707542657852173,
+      "learning_rate": 3.7658613224905606e-05,
+      "loss": 0.3666,
+      "step": 9226
+    },
+    {
+      "epoch": 1.645972705378646,
+      "grad_norm": 0.6010311245918274,
+      "learning_rate": 3.7621642338801335e-05,
+      "loss": 0.5016,
+      "step": 9227
+    },
+    {
+      "epoch": 1.6461511015966461,
+      "grad_norm": 0.5563137531280518,
+      "learning_rate": 3.758468813280186e-05,
+      "loss": 0.5181,
+      "step": 9228
+    },
+    {
+      "epoch": 1.6463294978146463,
+      "grad_norm": 0.4832812547683716,
+      "learning_rate": 3.754775060980964e-05,
+      "loss": 0.5077,
+      "step": 9229
+    },
+    {
+      "epoch": 1.6465078940326465,
+      "grad_norm": 0.5060386657714844,
+      "learning_rate": 3.75108297727256e-05,
+      "loss": 0.5059,
+      "step": 9230
+    },
+    {
+      "epoch": 1.6466862902506467,
+      "grad_norm": 0.4577668309211731,
+      "learning_rate": 3.7473925624449625e-05,
+      "loss": 0.4391,
+      "step": 9231
+    },
+    {
+      "epoch": 1.6468646864686467,
+      "grad_norm": 0.5205204486846924,
+      "learning_rate": 3.743703816788005e-05,
+      "loss": 0.614,
+      "step": 9232
+    },
+    {
+      "epoch": 1.647043082686647,
+      "grad_norm": 0.550428569316864,
+      "learning_rate": 3.740016740591398e-05,
+      "loss": 0.558,
+      "step": 9233
+    },
+    {
+      "epoch": 1.647221478904647,
+      "grad_norm": 0.5250325798988342,
+      "learning_rate": 3.736331334144733e-05,
+      "loss": 0.4774,
+      "step": 9234
+    },
+    {
+      "epoch": 1.6473998751226473,
+      "grad_norm": 0.5304805040359497,
+      "learning_rate": 3.7326475977374486e-05,
+      "loss": 0.5603,
+      "step": 9235
+    },
+    {
+      "epoch": 1.6475782713406475,
+      "grad_norm": 0.5167137384414673,
+      "learning_rate": 3.728965531658876e-05,
+      "loss": 0.5417,
+      "step": 9236
+    },
+    {
+      "epoch": 1.6477566675586477,
+      "grad_norm": 0.4665099084377289,
+      "learning_rate": 3.725285136198189e-05,
+      "loss": 0.5014,
+      "step": 9237
+    },
+    {
+      "epoch": 1.647935063776648,
+      "grad_norm": 0.5300410985946655,
+      "learning_rate": 3.72160641164446e-05,
+      "loss": 0.5706,
+      "step": 9238
+    },
+    {
+      "epoch": 1.648113459994648,
+      "grad_norm": 0.5009768009185791,
+      "learning_rate": 3.7179293582866064e-05,
+      "loss": 0.4469,
+      "step": 9239
+    },
+    {
+      "epoch": 1.6482918562126483,
+      "grad_norm": 0.4268481433391571,
+      "learning_rate": 3.714253976413418e-05,
+      "loss": 0.3472,
+      "step": 9240
+    },
+    {
+      "epoch": 1.6484702524306485,
+      "grad_norm": 0.5363745093345642,
+      "learning_rate": 3.710580266313565e-05,
+      "loss": 0.6022,
+      "step": 9241
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 0.5669705271720886,
+      "learning_rate": 3.706908228275571e-05,
+      "loss": 0.6687,
+      "step": 9242
+    },
+    {
+      "epoch": 1.6488270448666489,
+      "grad_norm": 0.5033018589019775,
+      "learning_rate": 3.703237862587844e-05,
+      "loss": 0.4968,
+      "step": 9243
+    },
+    {
+      "epoch": 1.649005441084649,
+      "grad_norm": 0.491277813911438,
+      "learning_rate": 3.699569169538655e-05,
+      "loss": 0.4729,
+      "step": 9244
+    },
+    {
+      "epoch": 1.6491838373026493,
+      "grad_norm": 0.43952322006225586,
+      "learning_rate": 3.695902149416133e-05,
+      "loss": 0.3538,
+      "step": 9245
+    },
+    {
+      "epoch": 1.6493622335206495,
+      "grad_norm": 0.6005352735519409,
+      "learning_rate": 3.692236802508292e-05,
+      "loss": 0.6145,
+      "step": 9246
+    },
+    {
+      "epoch": 1.6495406297386497,
+      "grad_norm": 0.566465437412262,
+      "learning_rate": 3.688573129102999e-05,
+      "loss": 0.5943,
+      "step": 9247
+    },
+    {
+      "epoch": 1.6497190259566499,
+      "grad_norm": 0.4471827745437622,
+      "learning_rate": 3.6849111294880056e-05,
+      "loss": 0.4295,
+      "step": 9248
+    },
+    {
+      "epoch": 1.6498974221746499,
+      "grad_norm": 0.4851609766483307,
+      "learning_rate": 3.681250803950914e-05,
+      "loss": 0.4421,
+      "step": 9249
+    },
+    {
+      "epoch": 1.65007581839265,
+      "grad_norm": 0.5344951152801514,
+      "learning_rate": 3.6775921527792164e-05,
+      "loss": 0.483,
+      "step": 9250
+    },
+    {
+      "epoch": 1.6502542146106502,
+      "grad_norm": 0.5168811082839966,
+      "learning_rate": 3.673935176260249e-05,
+      "loss": 0.4603,
+      "step": 9251
+    },
+    {
+      "epoch": 1.6504326108286504,
+      "grad_norm": 0.5323466062545776,
+      "learning_rate": 3.67027987468124e-05,
+      "loss": 0.5537,
+      "step": 9252
+    },
+    {
+      "epoch": 1.6506110070466506,
+      "grad_norm": 0.46962469816207886,
+      "learning_rate": 3.6666262483292715e-05,
+      "loss": 0.3963,
+      "step": 9253
+    },
+    {
+      "epoch": 1.6507894032646506,
+      "grad_norm": 0.5902261734008789,
+      "learning_rate": 3.662974297491292e-05,
+      "loss": 0.6236,
+      "step": 9254
+    },
+    {
+      "epoch": 1.6509677994826508,
+      "grad_norm": 0.5006431937217712,
+      "learning_rate": 3.6593240224541357e-05,
+      "loss": 0.4853,
+      "step": 9255
+    },
+    {
+      "epoch": 1.651146195700651,
+      "grad_norm": 0.6096543669700623,
+      "learning_rate": 3.6556754235044815e-05,
+      "loss": 0.7908,
+      "step": 9256
+    },
+    {
+      "epoch": 1.6513245919186512,
+      "grad_norm": 0.4698275625705719,
+      "learning_rate": 3.6520285009289e-05,
+      "loss": 0.4004,
+      "step": 9257
+    },
+    {
+      "epoch": 1.6515029881366514,
+      "grad_norm": 0.49396297335624695,
+      "learning_rate": 3.648383255013804e-05,
+      "loss": 0.5341,
+      "step": 9258
+    },
+    {
+      "epoch": 1.6516813843546516,
+      "grad_norm": 0.49175316095352173,
+      "learning_rate": 3.644739686045503e-05,
+      "loss": 0.4549,
+      "step": 9259
+    },
+    {
+      "epoch": 1.6518597805726518,
+      "grad_norm": 0.533931314945221,
+      "learning_rate": 3.6410977943101606e-05,
+      "loss": 0.5214,
+      "step": 9260
+    },
+    {
+      "epoch": 1.652038176790652,
+      "grad_norm": 0.5131736397743225,
+      "learning_rate": 3.6374575800938004e-05,
+      "loss": 0.546,
+      "step": 9261
+    },
+    {
+      "epoch": 1.6522165730086522,
+      "grad_norm": 0.4640941321849823,
+      "learning_rate": 3.633819043682338e-05,
+      "loss": 0.4511,
+      "step": 9262
+    },
+    {
+      "epoch": 1.6523949692266524,
+      "grad_norm": 0.506545901298523,
+      "learning_rate": 3.6301821853615216e-05,
+      "loss": 0.4355,
+      "step": 9263
+    },
+    {
+      "epoch": 1.6525733654446526,
+      "grad_norm": 0.5474511981010437,
+      "learning_rate": 3.6265470054170107e-05,
+      "loss": 0.5829,
+      "step": 9264
+    },
+    {
+      "epoch": 1.6527517616626528,
+      "grad_norm": 0.5812618732452393,
+      "learning_rate": 3.622913504134298e-05,
+      "loss": 0.5626,
+      "step": 9265
+    },
+    {
+      "epoch": 1.652930157880653,
+      "grad_norm": 0.5066866874694824,
+      "learning_rate": 3.619281681798756e-05,
+      "loss": 0.5741,
+      "step": 9266
+    },
+    {
+      "epoch": 1.6531085540986532,
+      "grad_norm": 0.48909926414489746,
+      "learning_rate": 3.615651538695633e-05,
+      "loss": 0.4439,
+      "step": 9267
+    },
+    {
+      "epoch": 1.6532869503166534,
+      "grad_norm": 0.5382171869277954,
+      "learning_rate": 3.6120230751100295e-05,
+      "loss": 0.5164,
+      "step": 9268
+    },
+    {
+      "epoch": 1.6534653465346536,
+      "grad_norm": 0.5178452730178833,
+      "learning_rate": 3.608396291326938e-05,
+      "loss": 0.5316,
+      "step": 9269
+    },
+    {
+      "epoch": 1.6536437427526538,
+      "grad_norm": 0.5068187713623047,
+      "learning_rate": 3.6047711876311895e-05,
+      "loss": 0.4757,
+      "step": 9270
+    },
+    {
+      "epoch": 1.6538221389706538,
+      "grad_norm": 0.5174327492713928,
+      "learning_rate": 3.601147764307511e-05,
+      "loss": 0.4613,
+      "step": 9271
+    },
+    {
+      "epoch": 1.654000535188654,
+      "grad_norm": 0.4983610212802887,
+      "learning_rate": 3.597526021640471e-05,
+      "loss": 0.4841,
+      "step": 9272
+    },
+    {
+      "epoch": 1.6541789314066542,
+      "grad_norm": 0.5520349740982056,
+      "learning_rate": 3.593905959914528e-05,
+      "loss": 0.6076,
+      "step": 9273
+    },
+    {
+      "epoch": 1.6543573276246544,
+      "grad_norm": 0.5307412147521973,
+      "learning_rate": 3.590287579414006e-05,
+      "loss": 0.5639,
+      "step": 9274
+    },
+    {
+      "epoch": 1.6545357238426546,
+      "grad_norm": 0.5179523825645447,
+      "learning_rate": 3.586670880423079e-05,
+      "loss": 0.5217,
+      "step": 9275
+    },
+    {
+      "epoch": 1.6547141200606548,
+      "grad_norm": 0.5160864591598511,
+      "learning_rate": 3.5830558632258095e-05,
+      "loss": 0.597,
+      "step": 9276
+    },
+    {
+      "epoch": 1.6548925162786547,
+      "grad_norm": 0.5735654830932617,
+      "learning_rate": 3.579442528106111e-05,
+      "loss": 0.6853,
+      "step": 9277
+    },
+    {
+      "epoch": 1.655070912496655,
+      "grad_norm": 0.5181892514228821,
+      "learning_rate": 3.5758308753477855e-05,
+      "loss": 0.4761,
+      "step": 9278
+    },
+    {
+      "epoch": 1.6552493087146551,
+      "grad_norm": 0.4839540421962738,
+      "learning_rate": 3.5722209052344826e-05,
+      "loss": 0.444,
+      "step": 9279
+    },
+    {
+      "epoch": 1.6554277049326553,
+      "grad_norm": 0.4774402678012848,
+      "learning_rate": 3.5686126180497214e-05,
+      "loss": 0.4561,
+      "step": 9280
+    },
+    {
+      "epoch": 1.6556061011506555,
+      "grad_norm": 0.5063000321388245,
+      "learning_rate": 3.56500601407691e-05,
+      "loss": 0.4304,
+      "step": 9281
+    },
+    {
+      "epoch": 1.6557844973686557,
+      "grad_norm": 0.39081576466560364,
+      "learning_rate": 3.561401093599295e-05,
+      "loss": 0.318,
+      "step": 9282
+    },
+    {
+      "epoch": 1.655962893586656,
+      "grad_norm": 0.5337099432945251,
+      "learning_rate": 3.557797856900022e-05,
+      "loss": 0.3636,
+      "step": 9283
+    },
+    {
+      "epoch": 1.6561412898046561,
+      "grad_norm": 0.5275738835334778,
+      "learning_rate": 3.554196304262067e-05,
+      "loss": 0.4887,
+      "step": 9284
+    },
+    {
+      "epoch": 1.6563196860226563,
+      "grad_norm": 0.5422784686088562,
+      "learning_rate": 3.5505964359683146e-05,
+      "loss": 0.5377,
+      "step": 9285
+    },
+    {
+      "epoch": 1.6564980822406565,
+      "grad_norm": 0.48426032066345215,
+      "learning_rate": 3.546998252301487e-05,
+      "loss": 0.4597,
+      "step": 9286
+    },
+    {
+      "epoch": 1.6566764784586567,
+      "grad_norm": 0.5601605176925659,
+      "learning_rate": 3.543401753544179e-05,
+      "loss": 0.6027,
+      "step": 9287
+    },
+    {
+      "epoch": 1.656854874676657,
+      "grad_norm": 0.4989088773727417,
+      "learning_rate": 3.539806939978868e-05,
+      "loss": 0.4151,
+      "step": 9288
+    },
+    {
+      "epoch": 1.6570332708946571,
+      "grad_norm": 0.5130565762519836,
+      "learning_rate": 3.536213811887876e-05,
+      "loss": 0.4985,
+      "step": 9289
+    },
+    {
+      "epoch": 1.6572116671126573,
+      "grad_norm": 0.4882428050041199,
+      "learning_rate": 3.532622369553423e-05,
+      "loss": 0.5041,
+      "step": 9290
+    },
+    {
+      "epoch": 1.6573900633306575,
+      "grad_norm": 0.5201172828674316,
+      "learning_rate": 3.529032613257574e-05,
+      "loss": 0.5467,
+      "step": 9291
+    },
+    {
+      "epoch": 1.6575684595486577,
+      "grad_norm": 0.5011335015296936,
+      "learning_rate": 3.525444543282255e-05,
+      "loss": 0.4953,
+      "step": 9292
+    },
+    {
+      "epoch": 1.6577468557666577,
+      "grad_norm": 0.4961317777633667,
+      "learning_rate": 3.521858159909289e-05,
+      "loss": 0.4933,
+      "step": 9293
+    },
+    {
+      "epoch": 1.657925251984658,
+      "grad_norm": 0.5875723958015442,
+      "learning_rate": 3.518273463420332e-05,
+      "loss": 0.5181,
+      "step": 9294
+    },
+    {
+      "epoch": 1.658103648202658,
+      "grad_norm": 0.4927757680416107,
+      "learning_rate": 3.5146904540969414e-05,
+      "loss": 0.3826,
+      "step": 9295
+    },
+    {
+      "epoch": 1.6582820444206583,
+      "grad_norm": 0.5020849108695984,
+      "learning_rate": 3.511109132220508e-05,
+      "loss": 0.4191,
+      "step": 9296
+    },
+    {
+      "epoch": 1.6584604406386585,
+      "grad_norm": 0.44826123118400574,
+      "learning_rate": 3.507529498072323e-05,
+      "loss": 0.3812,
+      "step": 9297
+    },
+    {
+      "epoch": 1.6586388368566587,
+      "grad_norm": 0.5366451144218445,
+      "learning_rate": 3.5039515519335236e-05,
+      "loss": 0.4824,
+      "step": 9298
+    },
+    {
+      "epoch": 1.6588172330746587,
+      "grad_norm": 0.5516796112060547,
+      "learning_rate": 3.500375294085112e-05,
+      "loss": 0.5474,
+      "step": 9299
+    },
+    {
+      "epoch": 1.6589956292926589,
+      "grad_norm": 0.5130575895309448,
+      "learning_rate": 3.4968007248079776e-05,
+      "loss": 0.4346,
+      "step": 9300
+    },
+    {
+      "epoch": 1.659174025510659,
+      "grad_norm": 0.6171131134033203,
+      "learning_rate": 3.493227844382857e-05,
+      "loss": 0.6627,
+      "step": 9301
+    },
+    {
+      "epoch": 1.6593524217286593,
+      "grad_norm": 0.5802091956138611,
+      "learning_rate": 3.48965665309037e-05,
+      "loss": 0.5415,
+      "step": 9302
+    },
+    {
+      "epoch": 1.6595308179466595,
+      "grad_norm": 0.5305059552192688,
+      "learning_rate": 3.48608715121099e-05,
+      "loss": 0.5481,
+      "step": 9303
+    },
+    {
+      "epoch": 1.6597092141646597,
+      "grad_norm": 0.5003054738044739,
+      "learning_rate": 3.4825193390250645e-05,
+      "loss": 0.511,
+      "step": 9304
+    },
+    {
+      "epoch": 1.6598876103826599,
+      "grad_norm": 0.5035673379898071,
+      "learning_rate": 3.478953216812816e-05,
+      "loss": 0.5759,
+      "step": 9305
+    },
+    {
+      "epoch": 1.66006600660066,
+      "grad_norm": 0.5683543682098389,
+      "learning_rate": 3.4753887848543163e-05,
+      "loss": 0.602,
+      "step": 9306
+    },
+    {
+      "epoch": 1.6602444028186603,
+      "grad_norm": 0.4327681362628937,
+      "learning_rate": 3.471826043429524e-05,
+      "loss": 0.3409,
+      "step": 9307
+    },
+    {
+      "epoch": 1.6604227990366605,
+      "grad_norm": 0.5470651388168335,
+      "learning_rate": 3.468264992818246e-05,
+      "loss": 0.5559,
+      "step": 9308
+    },
+    {
+      "epoch": 1.6606011952546607,
+      "grad_norm": 0.49540919065475464,
+      "learning_rate": 3.464705633300172e-05,
+      "loss": 0.4914,
+      "step": 9309
+    },
+    {
+      "epoch": 1.6607795914726609,
+      "grad_norm": 0.45050835609436035,
+      "learning_rate": 3.461147965154846e-05,
+      "loss": 0.4139,
+      "step": 9310
+    },
+    {
+      "epoch": 1.660957987690661,
+      "grad_norm": 0.5284720659255981,
+      "learning_rate": 3.457591988661696e-05,
+      "loss": 0.4919,
+      "step": 9311
+    },
+    {
+      "epoch": 1.6611363839086613,
+      "grad_norm": 0.4902994930744171,
+      "learning_rate": 3.4540377040999995e-05,
+      "loss": 0.4287,
+      "step": 9312
+    },
+    {
+      "epoch": 1.6613147801266615,
+      "grad_norm": 0.4877963364124298,
+      "learning_rate": 3.450485111748905e-05,
+      "loss": 0.4539,
+      "step": 9313
+    },
+    {
+      "epoch": 1.6614931763446616,
+      "grad_norm": 0.5532664656639099,
+      "learning_rate": 3.446934211887443e-05,
+      "loss": 0.4228,
+      "step": 9314
+    },
+    {
+      "epoch": 1.6616715725626616,
+      "grad_norm": 0.5521690249443054,
+      "learning_rate": 3.443385004794486e-05,
+      "loss": 0.5662,
+      "step": 9315
+    },
+    {
+      "epoch": 1.6618499687806618,
+      "grad_norm": 0.6033298969268799,
+      "learning_rate": 3.439837490748798e-05,
+      "loss": 0.6267,
+      "step": 9316
+    },
+    {
+      "epoch": 1.662028364998662,
+      "grad_norm": 0.5335450172424316,
+      "learning_rate": 3.436291670028993e-05,
+      "loss": 0.5272,
+      "step": 9317
+    },
+    {
+      "epoch": 1.6622067612166622,
+      "grad_norm": 0.5540075302124023,
+      "learning_rate": 3.432747542913564e-05,
+      "loss": 0.4624,
+      "step": 9318
+    },
+    {
+      "epoch": 1.6623851574346624,
+      "grad_norm": 0.5262585282325745,
+      "learning_rate": 3.429205109680858e-05,
+      "loss": 0.526,
+      "step": 9319
+    },
+    {
+      "epoch": 1.6625635536526626,
+      "grad_norm": 0.5248980522155762,
+      "learning_rate": 3.425664370609099e-05,
+      "loss": 0.5116,
+      "step": 9320
+    },
+    {
+      "epoch": 1.6627419498706626,
+      "grad_norm": 0.547280490398407,
+      "learning_rate": 3.422125325976383e-05,
+      "loss": 0.561,
+      "step": 9321
+    },
+    {
+      "epoch": 1.6629203460886628,
+      "grad_norm": 0.4935222268104553,
+      "learning_rate": 3.418587976060653e-05,
+      "loss": 0.4185,
+      "step": 9322
+    },
+    {
+      "epoch": 1.663098742306663,
+      "grad_norm": 0.5188893675804138,
+      "learning_rate": 3.415052321139739e-05,
+      "loss": 0.5019,
+      "step": 9323
+    },
+    {
+      "epoch": 1.6632771385246632,
+      "grad_norm": 0.4804520905017853,
+      "learning_rate": 3.411518361491328e-05,
+      "loss": 0.4503,
+      "step": 9324
+    },
+    {
+      "epoch": 1.6634555347426634,
+      "grad_norm": 0.5874333381652832,
+      "learning_rate": 3.407986097392971e-05,
+      "loss": 0.5774,
+      "step": 9325
+    },
+    {
+      "epoch": 1.6636339309606636,
+      "grad_norm": 0.49579691886901855,
+      "learning_rate": 3.404455529122097e-05,
+      "loss": 0.4188,
+      "step": 9326
+    },
+    {
+      "epoch": 1.6638123271786638,
+      "grad_norm": 0.45815184712409973,
+      "learning_rate": 3.400926656955988e-05,
+      "loss": 0.4121,
+      "step": 9327
+    },
+    {
+      "epoch": 1.663990723396664,
+      "grad_norm": 0.4780590534210205,
+      "learning_rate": 3.397399481171812e-05,
+      "loss": 0.4392,
+      "step": 9328
+    },
+    {
+      "epoch": 1.6641691196146642,
+      "grad_norm": 0.4702582359313965,
+      "learning_rate": 3.393874002046576e-05,
+      "loss": 0.4654,
+      "step": 9329
+    },
+    {
+      "epoch": 1.6643475158326644,
+      "grad_norm": 0.4759620428085327,
+      "learning_rate": 3.3903502198571855e-05,
+      "loss": 0.4035,
+      "step": 9330
+    },
+    {
+      "epoch": 1.6645259120506646,
+      "grad_norm": 0.5115009546279907,
+      "learning_rate": 3.386828134880382e-05,
+      "loss": 0.5284,
+      "step": 9331
+    },
+    {
+      "epoch": 1.6647043082686648,
+      "grad_norm": 0.5846856832504272,
+      "learning_rate": 3.383307747392802e-05,
+      "loss": 0.6325,
+      "step": 9332
+    },
+    {
+      "epoch": 1.664882704486665,
+      "grad_norm": 0.6280512809753418,
+      "learning_rate": 3.379789057670929e-05,
+      "loss": 0.5222,
+      "step": 9333
+    },
+    {
+      "epoch": 1.6650611007046652,
+      "grad_norm": 0.5682896375656128,
+      "learning_rate": 3.376272065991115e-05,
+      "loss": 0.5462,
+      "step": 9334
+    },
+    {
+      "epoch": 1.6652394969226654,
+      "grad_norm": 0.5903425216674805,
+      "learning_rate": 3.372756772629587e-05,
+      "loss": 0.8133,
+      "step": 9335
+    },
+    {
+      "epoch": 1.6654178931406656,
+      "grad_norm": 0.5490643978118896,
+      "learning_rate": 3.369243177862436e-05,
+      "loss": 0.6145,
+      "step": 9336
+    },
+    {
+      "epoch": 1.6655962893586655,
+      "grad_norm": 0.5370810031890869,
+      "learning_rate": 3.3657312819656226e-05,
+      "loss": 0.5671,
+      "step": 9337
+    },
+    {
+      "epoch": 1.6657746855766657,
+      "grad_norm": 0.5224997997283936,
+      "learning_rate": 3.362221085214964e-05,
+      "loss": 0.5352,
+      "step": 9338
+    },
+    {
+      "epoch": 1.665953081794666,
+      "grad_norm": 0.5418823957443237,
+      "learning_rate": 3.358712587886143e-05,
+      "loss": 0.5853,
+      "step": 9339
+    },
+    {
+      "epoch": 1.6661314780126661,
+      "grad_norm": 0.5162733197212219,
+      "learning_rate": 3.3552057902547286e-05,
+      "loss": 0.4823,
+      "step": 9340
+    },
+    {
+      "epoch": 1.6663098742306663,
+      "grad_norm": 0.4687744081020355,
+      "learning_rate": 3.351700692596132e-05,
+      "loss": 0.4868,
+      "step": 9341
+    },
+    {
+      "epoch": 1.6664882704486665,
+      "grad_norm": 0.5440129637718201,
+      "learning_rate": 3.348197295185654e-05,
+      "loss": 0.5426,
+      "step": 9342
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.4801231920719147,
+      "learning_rate": 3.344695598298436e-05,
+      "loss": 0.5322,
+      "step": 9343
+    },
+    {
+      "epoch": 1.6668450628846667,
+      "grad_norm": 0.4614768624305725,
+      "learning_rate": 3.341195602209512e-05,
+      "loss": 0.3709,
+      "step": 9344
+    },
+    {
+      "epoch": 1.667023459102667,
+      "grad_norm": 0.4808447062969208,
+      "learning_rate": 3.3376973071937656e-05,
+      "loss": 0.4343,
+      "step": 9345
+    },
+    {
+      "epoch": 1.6672018553206671,
+      "grad_norm": 0.42641395330429077,
+      "learning_rate": 3.3342007135259425e-05,
+      "loss": 0.442,
+      "step": 9346
+    },
+    {
+      "epoch": 1.6673802515386673,
+      "grad_norm": 0.4673963189125061,
+      "learning_rate": 3.3307058214806814e-05,
+      "loss": 0.4625,
+      "step": 9347
+    },
+    {
+      "epoch": 1.6675586477566675,
+      "grad_norm": 0.4446796178817749,
+      "learning_rate": 3.327212631332452e-05,
+      "loss": 0.3887,
+      "step": 9348
+    },
+    {
+      "epoch": 1.6677370439746677,
+      "grad_norm": 0.42202529311180115,
+      "learning_rate": 3.323721143355621e-05,
+      "loss": 0.3426,
+      "step": 9349
+    },
+    {
+      "epoch": 1.667915440192668,
+      "grad_norm": 0.5200551748275757,
+      "learning_rate": 3.320231357824399e-05,
+      "loss": 0.5275,
+      "step": 9350
+    },
+    {
+      "epoch": 1.668093836410668,
+      "grad_norm": 0.4585854411125183,
+      "learning_rate": 3.3167432750128764e-05,
+      "loss": 0.4564,
+      "step": 9351
+    },
+    {
+      "epoch": 1.6682722326286683,
+      "grad_norm": 0.5158877968788147,
+      "learning_rate": 3.313256895195013e-05,
+      "loss": 0.4869,
+      "step": 9352
+    },
+    {
+      "epoch": 1.6684506288466685,
+      "grad_norm": 0.4725387692451477,
+      "learning_rate": 3.3097722186446135e-05,
+      "loss": 0.5002,
+      "step": 9353
+    },
+    {
+      "epoch": 1.6686290250646687,
+      "grad_norm": 0.44371485710144043,
+      "learning_rate": 3.306289245635374e-05,
+      "loss": 0.398,
+      "step": 9354
+    },
+    {
+      "epoch": 1.668807421282669,
+      "grad_norm": 0.5774705410003662,
+      "learning_rate": 3.3028079764408386e-05,
+      "loss": 0.5725,
+      "step": 9355
+    },
+    {
+      "epoch": 1.668985817500669,
+      "grad_norm": 0.5633199214935303,
+      "learning_rate": 3.2993284113344315e-05,
+      "loss": 0.5249,
+      "step": 9356
+    },
+    {
+      "epoch": 1.6691642137186693,
+      "grad_norm": 0.4772298336029053,
+      "learning_rate": 3.295850550589427e-05,
+      "loss": 0.3945,
+      "step": 9357
+    },
+    {
+      "epoch": 1.6693426099366695,
+      "grad_norm": 0.4757455587387085,
+      "learning_rate": 3.292374394478986e-05,
+      "loss": 0.4001,
+      "step": 9358
+    },
+    {
+      "epoch": 1.6695210061546695,
+      "grad_norm": 0.49446776509284973,
+      "learning_rate": 3.288899943276119e-05,
+      "loss": 0.3805,
+      "step": 9359
+    },
+    {
+      "epoch": 1.6696994023726697,
+      "grad_norm": 0.4915049970149994,
+      "learning_rate": 3.285427197253704e-05,
+      "loss": 0.4954,
+      "step": 9360
+    },
+    {
+      "epoch": 1.6698777985906699,
+      "grad_norm": 0.5779103636741638,
+      "learning_rate": 3.281956156684496e-05,
+      "loss": 0.514,
+      "step": 9361
+    },
+    {
+      "epoch": 1.67005619480867,
+      "grad_norm": 0.5662655830383301,
+      "learning_rate": 3.278486821841098e-05,
+      "loss": 0.5022,
+      "step": 9362
+    },
+    {
+      "epoch": 1.6702345910266703,
+      "grad_norm": 0.5742835998535156,
+      "learning_rate": 3.275019192996004e-05,
+      "loss": 0.6153,
+      "step": 9363
+    },
+    {
+      "epoch": 1.6704129872446705,
+      "grad_norm": 0.517959713935852,
+      "learning_rate": 3.271553270421551e-05,
+      "loss": 0.4259,
+      "step": 9364
+    },
+    {
+      "epoch": 1.6705913834626704,
+      "grad_norm": 0.5644903779029846,
+      "learning_rate": 3.26808905438995e-05,
+      "loss": 0.6748,
+      "step": 9365
+    },
+    {
+      "epoch": 1.6707697796806706,
+      "grad_norm": 0.5745575428009033,
+      "learning_rate": 3.264626545173291e-05,
+      "loss": 0.6036,
+      "step": 9366
+    },
+    {
+      "epoch": 1.6709481758986708,
+      "grad_norm": 0.5247329473495483,
+      "learning_rate": 3.261165743043501e-05,
+      "loss": 0.5762,
+      "step": 9367
+    },
+    {
+      "epoch": 1.671126572116671,
+      "grad_norm": 0.5291442275047302,
+      "learning_rate": 3.2577066482724074e-05,
+      "loss": 0.5571,
+      "step": 9368
+    },
+    {
+      "epoch": 1.6713049683346712,
+      "grad_norm": 0.5165200233459473,
+      "learning_rate": 3.2542492611316696e-05,
+      "loss": 0.6014,
+      "step": 9369
+    },
+    {
+      "epoch": 1.6714833645526714,
+      "grad_norm": 0.5730751156806946,
+      "learning_rate": 3.250793581892844e-05,
+      "loss": 0.7261,
+      "step": 9370
+    },
+    {
+      "epoch": 1.6716617607706716,
+      "grad_norm": 0.5317413806915283,
+      "learning_rate": 3.2473396108273297e-05,
+      "loss": 0.4659,
+      "step": 9371
+    },
+    {
+      "epoch": 1.6718401569886718,
+      "grad_norm": 0.5281661152839661,
+      "learning_rate": 3.243887348206395e-05,
+      "loss": 0.5902,
+      "step": 9372
+    },
+    {
+      "epoch": 1.672018553206672,
+      "grad_norm": 0.4783737361431122,
+      "learning_rate": 3.240436794301194e-05,
+      "loss": 0.5077,
+      "step": 9373
+    },
+    {
+      "epoch": 1.6721969494246722,
+      "grad_norm": 0.5345422029495239,
+      "learning_rate": 3.2369879493827167e-05,
+      "loss": 0.6897,
+      "step": 9374
+    },
+    {
+      "epoch": 1.6723753456426724,
+      "grad_norm": 0.4720827639102936,
+      "learning_rate": 3.233540813721844e-05,
+      "loss": 0.3754,
+      "step": 9375
+    },
+    {
+      "epoch": 1.6725537418606726,
+      "grad_norm": 0.4707276225090027,
+      "learning_rate": 3.2300953875893046e-05,
+      "loss": 0.4113,
+      "step": 9376
+    },
+    {
+      "epoch": 1.6727321380786728,
+      "grad_norm": 0.5764208436012268,
+      "learning_rate": 3.226651671255712e-05,
+      "loss": 0.5561,
+      "step": 9377
+    },
+    {
+      "epoch": 1.672910534296673,
+      "grad_norm": 0.4676361680030823,
+      "learning_rate": 3.2232096649915196e-05,
+      "loss": 0.4113,
+      "step": 9378
+    },
+    {
+      "epoch": 1.6730889305146732,
+      "grad_norm": 0.5252630710601807,
+      "learning_rate": 3.219769369067077e-05,
+      "loss": 0.5728,
+      "step": 9379
+    },
+    {
+      "epoch": 1.6732673267326734,
+      "grad_norm": 0.4781394302845001,
+      "learning_rate": 3.216330783752569e-05,
+      "loss": 0.2909,
+      "step": 9380
+    },
+    {
+      "epoch": 1.6734457229506734,
+      "grad_norm": 0.47112560272216797,
+      "learning_rate": 3.2128939093180655e-05,
+      "loss": 0.4191,
+      "step": 9381
+    },
+    {
+      "epoch": 1.6736241191686736,
+      "grad_norm": 0.5391256809234619,
+      "learning_rate": 3.209458746033506e-05,
+      "loss": 0.7091,
+      "step": 9382
+    },
+    {
+      "epoch": 1.6738025153866738,
+      "grad_norm": 0.49783357977867126,
+      "learning_rate": 3.206025294168677e-05,
+      "loss": 0.5378,
+      "step": 9383
+    },
+    {
+      "epoch": 1.673980911604674,
+      "grad_norm": 0.5017059445381165,
+      "learning_rate": 3.202593553993238e-05,
+      "loss": 0.4537,
+      "step": 9384
+    },
+    {
+      "epoch": 1.6741593078226742,
+      "grad_norm": 0.5273604989051819,
+      "learning_rate": 3.1991635257767274e-05,
+      "loss": 0.4771,
+      "step": 9385
+    },
+    {
+      "epoch": 1.6743377040406744,
+      "grad_norm": 0.5304795503616333,
+      "learning_rate": 3.195735209788528e-05,
+      "loss": 0.3931,
+      "step": 9386
+    },
+    {
+      "epoch": 1.6745161002586744,
+      "grad_norm": 0.5918444395065308,
+      "learning_rate": 3.1923086062979056e-05,
+      "loss": 0.7699,
+      "step": 9387
+    },
+    {
+      "epoch": 1.6746944964766746,
+      "grad_norm": 0.5880241990089417,
+      "learning_rate": 3.188883715573976e-05,
+      "loss": 0.6155,
+      "step": 9388
+    },
+    {
+      "epoch": 1.6748728926946748,
+      "grad_norm": 0.4606783092021942,
+      "learning_rate": 3.185460537885737e-05,
+      "loss": 0.4407,
+      "step": 9389
+    },
+    {
+      "epoch": 1.675051288912675,
+      "grad_norm": 0.44340142607688904,
+      "learning_rate": 3.182039073502035e-05,
+      "loss": 0.3536,
+      "step": 9390
+    },
+    {
+      "epoch": 1.6752296851306752,
+      "grad_norm": 0.5192257761955261,
+      "learning_rate": 3.1786193226916005e-05,
+      "loss": 0.5362,
+      "step": 9391
+    },
+    {
+      "epoch": 1.6754080813486754,
+      "grad_norm": 0.6925177574157715,
+      "learning_rate": 3.175201285723017e-05,
+      "loss": 0.7919,
+      "step": 9392
+    },
+    {
+      "epoch": 1.6755864775666756,
+      "grad_norm": 0.557155966758728,
+      "learning_rate": 3.171784962864724e-05,
+      "loss": 0.4573,
+      "step": 9393
+    },
+    {
+      "epoch": 1.6757648737846758,
+      "grad_norm": 0.5083150267601013,
+      "learning_rate": 3.1683703543850526e-05,
+      "loss": 0.5892,
+      "step": 9394
+    },
+    {
+      "epoch": 1.675943270002676,
+      "grad_norm": 0.46054404973983765,
+      "learning_rate": 3.164957460552173e-05,
+      "loss": 0.3862,
+      "step": 9395
+    },
+    {
+      "epoch": 1.6761216662206762,
+      "grad_norm": 0.5175045728683472,
+      "learning_rate": 3.16154628163414e-05,
+      "loss": 0.5591,
+      "step": 9396
+    },
+    {
+      "epoch": 1.6763000624386764,
+      "grad_norm": 0.4765566885471344,
+      "learning_rate": 3.1581368178988654e-05,
+      "loss": 0.405,
+      "step": 9397
+    },
+    {
+      "epoch": 1.6764784586566766,
+      "grad_norm": 0.5212094783782959,
+      "learning_rate": 3.154729069614123e-05,
+      "loss": 0.4395,
+      "step": 9398
+    },
+    {
+      "epoch": 1.6766568548746767,
+      "grad_norm": 0.5103222727775574,
+      "learning_rate": 3.1513230370475654e-05,
+      "loss": 0.4392,
+      "step": 9399
+    },
+    {
+      "epoch": 1.676835251092677,
+      "grad_norm": 0.46660804748535156,
+      "learning_rate": 3.147918720466689e-05,
+      "loss": 0.4229,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6770136473106771,
+      "grad_norm": 0.500568687915802,
+      "learning_rate": 3.1445161201388766e-05,
+      "loss": 0.4259,
+      "step": 9401
+    },
+    {
+      "epoch": 1.6771920435286773,
+      "grad_norm": 0.5917816162109375,
+      "learning_rate": 3.141115236331357e-05,
+      "loss": 0.7201,
+      "step": 9402
+    },
+    {
+      "epoch": 1.6773704397466773,
+      "grad_norm": 0.49264368414878845,
+      "learning_rate": 3.137716069311247e-05,
+      "loss": 0.4191,
+      "step": 9403
+    },
+    {
+      "epoch": 1.6775488359646775,
+      "grad_norm": 0.5691975355148315,
+      "learning_rate": 3.134318619345508e-05,
+      "loss": 0.6493,
+      "step": 9404
+    },
+    {
+      "epoch": 1.6777272321826777,
+      "grad_norm": 0.5436373353004456,
+      "learning_rate": 3.130922886700968e-05,
+      "loss": 0.5603,
+      "step": 9405
+    },
+    {
+      "epoch": 1.677905628400678,
+      "grad_norm": 0.47983697056770325,
+      "learning_rate": 3.127528871644342e-05,
+      "loss": 0.4834,
+      "step": 9406
+    },
+    {
+      "epoch": 1.6780840246186781,
+      "grad_norm": 0.5476971864700317,
+      "learning_rate": 3.12413657444218e-05,
+      "loss": 0.5386,
+      "step": 9407
+    },
+    {
+      "epoch": 1.6782624208366783,
+      "grad_norm": 0.5388808846473694,
+      "learning_rate": 3.120745995360921e-05,
+      "loss": 0.483,
+      "step": 9408
+    },
+    {
+      "epoch": 1.6784408170546783,
+      "grad_norm": 0.5438225865364075,
+      "learning_rate": 3.117357134666851e-05,
+      "loss": 0.5662,
+      "step": 9409
+    },
+    {
+      "epoch": 1.6786192132726785,
+      "grad_norm": 0.5238845348358154,
+      "learning_rate": 3.113969992626142e-05,
+      "loss": 0.5429,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6787976094906787,
+      "grad_norm": 0.5313870906829834,
+      "learning_rate": 3.110584569504804e-05,
+      "loss": 0.5065,
+      "step": 9411
+    },
+    {
+      "epoch": 1.6789760057086789,
+      "grad_norm": 0.5582557916641235,
+      "learning_rate": 3.1072008655687376e-05,
+      "loss": 0.5529,
+      "step": 9412
+    },
+    {
+      "epoch": 1.679154401926679,
+      "grad_norm": 0.552903950214386,
+      "learning_rate": 3.103818881083695e-05,
+      "loss": 0.553,
+      "step": 9413
+    },
+    {
+      "epoch": 1.6793327981446793,
+      "grad_norm": 0.5650268793106079,
+      "learning_rate": 3.100438616315293e-05,
+      "loss": 0.6105,
+      "step": 9414
+    },
+    {
+      "epoch": 1.6795111943626795,
+      "grad_norm": 0.4516279995441437,
+      "learning_rate": 3.0970600715290204e-05,
+      "loss": 0.3535,
+      "step": 9415
+    },
+    {
+      "epoch": 1.6796895905806797,
+      "grad_norm": 0.4835292100906372,
+      "learning_rate": 3.0936832469902226e-05,
+      "loss": 0.4125,
+      "step": 9416
+    },
+    {
+      "epoch": 1.6798679867986799,
+      "grad_norm": 0.6102882623672485,
+      "learning_rate": 3.0903081429641186e-05,
+      "loss": 0.5305,
+      "step": 9417
+    },
+    {
+      "epoch": 1.68004638301668,
+      "grad_norm": 0.485422283411026,
+      "learning_rate": 3.086934759715784e-05,
+      "loss": 0.4507,
+      "step": 9418
+    },
+    {
+      "epoch": 1.6802247792346803,
+      "grad_norm": 0.4775411784648895,
+      "learning_rate": 3.0835630975101587e-05,
+      "loss": 0.4746,
+      "step": 9419
+    },
+    {
+      "epoch": 1.6804031754526805,
+      "grad_norm": 0.4786146283149719,
+      "learning_rate": 3.08019315661206e-05,
+      "loss": 0.5359,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6805815716706807,
+      "grad_norm": 0.5550956130027771,
+      "learning_rate": 3.076824937286155e-05,
+      "loss": 0.4877,
+      "step": 9421
+    },
+    {
+      "epoch": 1.6807599678886809,
+      "grad_norm": 0.5382834076881409,
+      "learning_rate": 3.073458439796989e-05,
+      "loss": 0.6299,
+      "step": 9422
+    },
+    {
+      "epoch": 1.680938364106681,
+      "grad_norm": 0.5830110311508179,
+      "learning_rate": 3.0700936644089574e-05,
+      "loss": 0.7122,
+      "step": 9423
+    },
+    {
+      "epoch": 1.6811167603246813,
+      "grad_norm": 0.506211519241333,
+      "learning_rate": 3.0667306113863366e-05,
+      "loss": 0.3911,
+      "step": 9424
+    },
+    {
+      "epoch": 1.6812951565426812,
+      "grad_norm": 0.46619656682014465,
+      "learning_rate": 3.0633692809932554e-05,
+      "loss": 0.4391,
+      "step": 9425
+    },
+    {
+      "epoch": 1.6814735527606814,
+      "grad_norm": 0.4801982343196869,
+      "learning_rate": 3.060009673493702e-05,
+      "loss": 0.4515,
+      "step": 9426
+    },
+    {
+      "epoch": 1.6816519489786816,
+      "grad_norm": 0.5707588791847229,
+      "learning_rate": 3.0566517891515546e-05,
+      "loss": 0.645,
+      "step": 9427
+    },
+    {
+      "epoch": 1.6818303451966818,
+      "grad_norm": 0.5557790994644165,
+      "learning_rate": 3.0532956282305294e-05,
+      "loss": 0.6412,
+      "step": 9428
+    },
+    {
+      "epoch": 1.682008741414682,
+      "grad_norm": 0.5159487128257751,
+      "learning_rate": 3.0499411909942265e-05,
+      "loss": 0.5476,
+      "step": 9429
+    },
+    {
+      "epoch": 1.6821871376326822,
+      "grad_norm": 0.4673673212528229,
+      "learning_rate": 3.0465884777060943e-05,
+      "loss": 0.4216,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6823655338506822,
+      "grad_norm": 0.484874963760376,
+      "learning_rate": 3.0432374886294523e-05,
+      "loss": 0.5434,
+      "step": 9431
+    },
+    {
+      "epoch": 1.6825439300686824,
+      "grad_norm": 0.7181788682937622,
+      "learning_rate": 3.0398882240274955e-05,
+      "loss": 0.8602,
+      "step": 9432
+    },
+    {
+      "epoch": 1.6827223262866826,
+      "grad_norm": 0.4546528458595276,
+      "learning_rate": 3.036540684163261e-05,
+      "loss": 0.3598,
+      "step": 9433
+    },
+    {
+      "epoch": 1.6829007225046828,
+      "grad_norm": 0.4792789816856384,
+      "learning_rate": 3.033194869299674e-05,
+      "loss": 0.4134,
+      "step": 9434
+    },
+    {
+      "epoch": 1.683079118722683,
+      "grad_norm": 0.4788531959056854,
+      "learning_rate": 3.0298507796995056e-05,
+      "loss": 0.4332,
+      "step": 9435
+    },
+    {
+      "epoch": 1.6832575149406832,
+      "grad_norm": 0.6334043145179749,
+      "learning_rate": 3.0265084156254064e-05,
+      "loss": 0.4675,
+      "step": 9436
+    },
+    {
+      "epoch": 1.6834359111586834,
+      "grad_norm": 0.48785632848739624,
+      "learning_rate": 3.0231677773398748e-05,
+      "loss": 0.4613,
+      "step": 9437
+    },
+    {
+      "epoch": 1.6836143073766836,
+      "grad_norm": 0.5417238473892212,
+      "learning_rate": 3.019828865105295e-05,
+      "loss": 0.6667,
+      "step": 9438
+    },
+    {
+      "epoch": 1.6837927035946838,
+      "grad_norm": 0.46144363284111023,
+      "learning_rate": 3.016491679183897e-05,
+      "loss": 0.3897,
+      "step": 9439
+    },
+    {
+      "epoch": 1.683971099812684,
+      "grad_norm": 0.4795248508453369,
+      "learning_rate": 3.013156219837776e-05,
+      "loss": 0.4471,
+      "step": 9440
+    },
+    {
+      "epoch": 1.6841494960306842,
+      "grad_norm": 0.5149518251419067,
+      "learning_rate": 3.0098224873289086e-05,
+      "loss": 0.5248,
+      "step": 9441
+    },
+    {
+      "epoch": 1.6843278922486844,
+      "grad_norm": 0.5572359561920166,
+      "learning_rate": 3.0064904819191162e-05,
+      "loss": 0.6678,
+      "step": 9442
+    },
+    {
+      "epoch": 1.6845062884666846,
+      "grad_norm": 0.5138469934463501,
+      "learning_rate": 3.003160203870095e-05,
+      "loss": 0.5152,
+      "step": 9443
+    },
+    {
+      "epoch": 1.6846846846846848,
+      "grad_norm": 0.5161378383636475,
+      "learning_rate": 2.9998316534434135e-05,
+      "loss": 0.4903,
+      "step": 9444
+    },
+    {
+      "epoch": 1.684863080902685,
+      "grad_norm": 0.49434447288513184,
+      "learning_rate": 2.996504830900479e-05,
+      "loss": 0.5285,
+      "step": 9445
+    },
+    {
+      "epoch": 1.6850414771206852,
+      "grad_norm": 0.5236574411392212,
+      "learning_rate": 2.9931797365025937e-05,
+      "loss": 0.5159,
+      "step": 9446
+    },
+    {
+      "epoch": 1.6852198733386852,
+      "grad_norm": 0.544776201248169,
+      "learning_rate": 2.9898563705108932e-05,
+      "loss": 0.6266,
+      "step": 9447
+    },
+    {
+      "epoch": 1.6853982695566854,
+      "grad_norm": 0.5385255217552185,
+      "learning_rate": 2.9865347331864106e-05,
+      "loss": 0.5478,
+      "step": 9448
+    },
+    {
+      "epoch": 1.6855766657746856,
+      "grad_norm": 0.5094165205955505,
+      "learning_rate": 2.9832148247900092e-05,
+      "loss": 0.5402,
+      "step": 9449
+    },
+    {
+      "epoch": 1.6857550619926858,
+      "grad_norm": 0.5184796452522278,
+      "learning_rate": 2.9798966455824473e-05,
+      "loss": 0.4279,
+      "step": 9450
+    },
+    {
+      "epoch": 1.685933458210686,
+      "grad_norm": 0.5272186398506165,
+      "learning_rate": 2.9765801958243245e-05,
+      "loss": 0.4309,
+      "step": 9451
+    },
+    {
+      "epoch": 1.6861118544286862,
+      "grad_norm": 0.49219128489494324,
+      "learning_rate": 2.973265475776113e-05,
+      "loss": 0.4853,
+      "step": 9452
+    },
+    {
+      "epoch": 1.6862902506466861,
+      "grad_norm": 0.46345192193984985,
+      "learning_rate": 2.9699524856981574e-05,
+      "loss": 0.5184,
+      "step": 9453
+    },
+    {
+      "epoch": 1.6864686468646863,
+      "grad_norm": 0.528198778629303,
+      "learning_rate": 2.966641225850647e-05,
+      "loss": 0.572,
+      "step": 9454
+    },
+    {
+      "epoch": 1.6866470430826865,
+      "grad_norm": 0.5292263627052307,
+      "learning_rate": 2.963331696493657e-05,
+      "loss": 0.4926,
+      "step": 9455
+    },
+    {
+      "epoch": 1.6868254393006867,
+      "grad_norm": 0.48872697353363037,
+      "learning_rate": 2.9600238978871097e-05,
+      "loss": 0.3942,
+      "step": 9456
+    },
+    {
+      "epoch": 1.687003835518687,
+      "grad_norm": 0.5367798805236816,
+      "learning_rate": 2.9567178302908005e-05,
+      "loss": 0.5765,
+      "step": 9457
+    },
+    {
+      "epoch": 1.6871822317366871,
+      "grad_norm": 0.48006823658943176,
+      "learning_rate": 2.953413493964391e-05,
+      "loss": 0.4903,
+      "step": 9458
+    },
+    {
+      "epoch": 1.6873606279546873,
+      "grad_norm": 0.4655410349369049,
+      "learning_rate": 2.9501108891673928e-05,
+      "loss": 0.3782,
+      "step": 9459
+    },
+    {
+      "epoch": 1.6875390241726875,
+      "grad_norm": 0.49426302313804626,
+      "learning_rate": 2.9468100161592043e-05,
+      "loss": 0.5576,
+      "step": 9460
+    },
+    {
+      "epoch": 1.6877174203906877,
+      "grad_norm": 0.4561796188354492,
+      "learning_rate": 2.9435108751990596e-05,
+      "loss": 0.4239,
+      "step": 9461
+    },
+    {
+      "epoch": 1.687895816608688,
+      "grad_norm": 0.4534534513950348,
+      "learning_rate": 2.940213466546085e-05,
+      "loss": 0.3932,
+      "step": 9462
+    },
+    {
+      "epoch": 1.6880742128266881,
+      "grad_norm": 0.5681625604629517,
+      "learning_rate": 2.936917790459251e-05,
+      "loss": 0.5814,
+      "step": 9463
+    },
+    {
+      "epoch": 1.6882526090446883,
+      "grad_norm": 0.5786123275756836,
+      "learning_rate": 2.9336238471973947e-05,
+      "loss": 0.7368,
+      "step": 9464
+    },
+    {
+      "epoch": 1.6884310052626885,
+      "grad_norm": 0.6435004472732544,
+      "learning_rate": 2.930331637019229e-05,
+      "loss": 0.6626,
+      "step": 9465
+    },
+    {
+      "epoch": 1.6886094014806887,
+      "grad_norm": 0.4304559528827667,
+      "learning_rate": 2.9270411601833162e-05,
+      "loss": 0.3071,
+      "step": 9466
+    },
+    {
+      "epoch": 1.688787797698689,
+      "grad_norm": 0.5037026405334473,
+      "learning_rate": 2.9237524169480974e-05,
+      "loss": 0.4809,
+      "step": 9467
+    },
+    {
+      "epoch": 1.6889661939166891,
+      "grad_norm": 0.5338348150253296,
+      "learning_rate": 2.9204654075718568e-05,
+      "loss": 0.5474,
+      "step": 9468
+    },
+    {
+      "epoch": 1.689144590134689,
+      "grad_norm": 0.44183140993118286,
+      "learning_rate": 2.917180132312766e-05,
+      "loss": 0.3869,
+      "step": 9469
+    },
+    {
+      "epoch": 1.6893229863526893,
+      "grad_norm": 0.5616282224655151,
+      "learning_rate": 2.913896591428841e-05,
+      "loss": 0.5424,
+      "step": 9470
+    },
+    {
+      "epoch": 1.6895013825706895,
+      "grad_norm": 0.5086612701416016,
+      "learning_rate": 2.9106147851779785e-05,
+      "loss": 0.5467,
+      "step": 9471
+    },
+    {
+      "epoch": 1.6896797787886897,
+      "grad_norm": 0.44916772842407227,
+      "learning_rate": 2.907334713817919e-05,
+      "loss": 0.3491,
+      "step": 9472
+    },
+    {
+      "epoch": 1.6898581750066899,
+      "grad_norm": 0.5420513153076172,
+      "learning_rate": 2.9040563776062845e-05,
+      "loss": 0.5353,
+      "step": 9473
+    },
+    {
+      "epoch": 1.69003657122469,
+      "grad_norm": 0.4886205196380615,
+      "learning_rate": 2.9007797768005606e-05,
+      "loss": 0.4478,
+      "step": 9474
+    },
+    {
+      "epoch": 1.69021496744269,
+      "grad_norm": 0.47686657309532166,
+      "learning_rate": 2.897504911658075e-05,
+      "loss": 0.4001,
+      "step": 9475
+    },
+    {
+      "epoch": 1.6903933636606903,
+      "grad_norm": 0.5385986566543579,
+      "learning_rate": 2.8942317824360493e-05,
+      "loss": 0.6524,
+      "step": 9476
+    },
+    {
+      "epoch": 1.6905717598786905,
+      "grad_norm": 0.5010984539985657,
+      "learning_rate": 2.890960389391545e-05,
+      "loss": 0.4801,
+      "step": 9477
+    },
+    {
+      "epoch": 1.6907501560966907,
+      "grad_norm": 0.6029075384140015,
+      "learning_rate": 2.887690732781492e-05,
+      "loss": 0.6388,
+      "step": 9478
+    },
+    {
+      "epoch": 1.6909285523146909,
+      "grad_norm": 0.47980138659477234,
+      "learning_rate": 2.8844228128627e-05,
+      "loss": 0.4344,
+      "step": 9479
+    },
+    {
+      "epoch": 1.691106948532691,
+      "grad_norm": 0.5094255208969116,
+      "learning_rate": 2.881156629891815e-05,
+      "loss": 0.5046,
+      "step": 9480
+    },
+    {
+      "epoch": 1.6912853447506913,
+      "grad_norm": 0.4894050061702728,
+      "learning_rate": 2.8778921841253774e-05,
+      "loss": 0.4561,
+      "step": 9481
+    },
+    {
+      "epoch": 1.6914637409686915,
+      "grad_norm": 0.531238317489624,
+      "learning_rate": 2.8746294758197622e-05,
+      "loss": 0.5447,
+      "step": 9482
+    },
+    {
+      "epoch": 1.6916421371866917,
+      "grad_norm": 0.5232047438621521,
+      "learning_rate": 2.871368505231234e-05,
+      "loss": 0.474,
+      "step": 9483
+    },
+    {
+      "epoch": 1.6918205334046919,
+      "grad_norm": 0.4739258885383606,
+      "learning_rate": 2.8681092726158957e-05,
+      "loss": 0.4283,
+      "step": 9484
+    },
+    {
+      "epoch": 1.691998929622692,
+      "grad_norm": 0.4757033884525299,
+      "learning_rate": 2.8648517782297294e-05,
+      "loss": 0.3738,
+      "step": 9485
+    },
+    {
+      "epoch": 1.6921773258406922,
+      "grad_norm": 0.5776329636573792,
+      "learning_rate": 2.861596022328583e-05,
+      "loss": 0.7618,
+      "step": 9486
+    },
+    {
+      "epoch": 1.6923557220586924,
+      "grad_norm": 0.5738844275474548,
+      "learning_rate": 2.8583420051681545e-05,
+      "loss": 0.678,
+      "step": 9487
+    },
+    {
+      "epoch": 1.6925341182766926,
+      "grad_norm": 0.5281713604927063,
+      "learning_rate": 2.8550897270040148e-05,
+      "loss": 0.4907,
+      "step": 9488
+    },
+    {
+      "epoch": 1.6927125144946928,
+      "grad_norm": 0.5512872338294983,
+      "learning_rate": 2.851839188091604e-05,
+      "loss": 0.6453,
+      "step": 9489
+    },
+    {
+      "epoch": 1.692890910712693,
+      "grad_norm": 0.5240409970283508,
+      "learning_rate": 2.8485903886862093e-05,
+      "loss": 0.443,
+      "step": 9490
+    },
+    {
+      "epoch": 1.693069306930693,
+      "grad_norm": 0.6053004860877991,
+      "learning_rate": 2.845343329042996e-05,
+      "loss": 0.568,
+      "step": 9491
+    },
+    {
+      "epoch": 1.6932477031486932,
+      "grad_norm": 0.6063686013221741,
+      "learning_rate": 2.842098009416977e-05,
+      "loss": 0.5555,
+      "step": 9492
+    },
+    {
+      "epoch": 1.6934260993666934,
+      "grad_norm": 0.5201789736747742,
+      "learning_rate": 2.8388544300630538e-05,
+      "loss": 0.515,
+      "step": 9493
+    },
+    {
+      "epoch": 1.6936044955846936,
+      "grad_norm": 0.47415268421173096,
+      "learning_rate": 2.8356125912359587e-05,
+      "loss": 0.3263,
+      "step": 9494
+    },
+    {
+      "epoch": 1.6937828918026938,
+      "grad_norm": 0.4585654139518738,
+      "learning_rate": 2.832372493190319e-05,
+      "loss": 0.3953,
+      "step": 9495
+    },
+    {
+      "epoch": 1.693961288020694,
+      "grad_norm": 0.552780032157898,
+      "learning_rate": 2.8291341361805973e-05,
+      "loss": 0.5802,
+      "step": 9496
+    },
+    {
+      "epoch": 1.694139684238694,
+      "grad_norm": 0.5424774885177612,
+      "learning_rate": 2.8258975204611488e-05,
+      "loss": 0.5196,
+      "step": 9497
+    },
+    {
+      "epoch": 1.6943180804566942,
+      "grad_norm": 0.5058227777481079,
+      "learning_rate": 2.8226626462861645e-05,
+      "loss": 0.4721,
+      "step": 9498
+    },
+    {
+      "epoch": 1.6944964766746944,
+      "grad_norm": 0.4840177297592163,
+      "learning_rate": 2.819429513909705e-05,
+      "loss": 0.5075,
+      "step": 9499
+    },
+    {
+      "epoch": 1.6946748728926946,
+      "grad_norm": 0.47189828753471375,
+      "learning_rate": 2.816198123585714e-05,
+      "loss": 0.3598,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6948532691106948,
+      "grad_norm": 0.46200263500213623,
+      "learning_rate": 2.81296847556797e-05,
+      "loss": 0.4265,
+      "step": 9501
+    },
+    {
+      "epoch": 1.695031665328695,
+      "grad_norm": 0.5357261300086975,
+      "learning_rate": 2.809740570110142e-05,
+      "loss": 0.4951,
+      "step": 9502
+    },
+    {
+      "epoch": 1.6952100615466952,
+      "grad_norm": 0.5782269835472107,
+      "learning_rate": 2.8065144074657322e-05,
+      "loss": 0.586,
+      "step": 9503
+    },
+    {
+      "epoch": 1.6953884577646954,
+      "grad_norm": 0.5674153566360474,
+      "learning_rate": 2.8032899878881303e-05,
+      "loss": 0.5702,
+      "step": 9504
+    },
+    {
+      "epoch": 1.6955668539826956,
+      "grad_norm": 0.5033431649208069,
+      "learning_rate": 2.8000673116305857e-05,
+      "loss": 0.4803,
+      "step": 9505
+    },
+    {
+      "epoch": 1.6957452502006958,
+      "grad_norm": 0.5172526836395264,
+      "learning_rate": 2.7968463789461968e-05,
+      "loss": 0.5727,
+      "step": 9506
+    },
+    {
+      "epoch": 1.695923646418696,
+      "grad_norm": 0.5547386407852173,
+      "learning_rate": 2.793627190087944e-05,
+      "loss": 0.5522,
+      "step": 9507
+    },
+    {
+      "epoch": 1.6961020426366962,
+      "grad_norm": 0.5478581786155701,
+      "learning_rate": 2.7904097453086503e-05,
+      "loss": 0.5495,
+      "step": 9508
+    },
+    {
+      "epoch": 1.6962804388546964,
+      "grad_norm": 0.5385379791259766,
+      "learning_rate": 2.7871940448610244e-05,
+      "loss": 0.4659,
+      "step": 9509
+    },
+    {
+      "epoch": 1.6964588350726966,
+      "grad_norm": 0.4548676609992981,
+      "learning_rate": 2.7839800889976173e-05,
+      "loss": 0.4397,
+      "step": 9510
+    },
+    {
+      "epoch": 1.6966372312906968,
+      "grad_norm": 0.5475732684135437,
+      "learning_rate": 2.7807678779708466e-05,
+      "loss": 0.6873,
+      "step": 9511
+    },
+    {
+      "epoch": 1.696815627508697,
+      "grad_norm": 0.49754053354263306,
+      "learning_rate": 2.7775574120330132e-05,
+      "loss": 0.4393,
+      "step": 9512
+    },
+    {
+      "epoch": 1.696994023726697,
+      "grad_norm": 0.5146554708480835,
+      "learning_rate": 2.7743486914362513e-05,
+      "loss": 0.4761,
+      "step": 9513
+    },
+    {
+      "epoch": 1.6971724199446971,
+      "grad_norm": 0.5417225956916809,
+      "learning_rate": 2.771141716432585e-05,
+      "loss": 0.5846,
+      "step": 9514
+    },
+    {
+      "epoch": 1.6973508161626973,
+      "grad_norm": 0.5172662734985352,
+      "learning_rate": 2.7679364872738753e-05,
+      "loss": 0.4245,
+      "step": 9515
+    },
+    {
+      "epoch": 1.6975292123806975,
+      "grad_norm": 0.5381408929824829,
+      "learning_rate": 2.764733004211875e-05,
+      "loss": 0.4776,
+      "step": 9516
+    },
+    {
+      "epoch": 1.6977076085986977,
+      "grad_norm": 0.4274336099624634,
+      "learning_rate": 2.7615312674981686e-05,
+      "loss": 0.3263,
+      "step": 9517
+    },
+    {
+      "epoch": 1.697886004816698,
+      "grad_norm": 0.5179285407066345,
+      "learning_rate": 2.7583312773842267e-05,
+      "loss": 0.5205,
+      "step": 9518
+    },
+    {
+      "epoch": 1.698064401034698,
+      "grad_norm": 0.5593109726905823,
+      "learning_rate": 2.7551330341213793e-05,
+      "loss": 0.568,
+      "step": 9519
+    },
+    {
+      "epoch": 1.698242797252698,
+      "grad_norm": 0.6157004833221436,
+      "learning_rate": 2.7519365379608058e-05,
+      "loss": 0.7443,
+      "step": 9520
+    },
+    {
+      "epoch": 1.6984211934706983,
+      "grad_norm": 0.5052160620689392,
+      "learning_rate": 2.7487417891535693e-05,
+      "loss": 0.4443,
+      "step": 9521
+    },
+    {
+      "epoch": 1.6985995896886985,
+      "grad_norm": 0.569677472114563,
+      "learning_rate": 2.7455487879505747e-05,
+      "loss": 0.5609,
+      "step": 9522
+    },
+    {
+      "epoch": 1.6987779859066987,
+      "grad_norm": 0.45186012983322144,
+      "learning_rate": 2.742357534602599e-05,
+      "loss": 0.3359,
+      "step": 9523
+    },
+    {
+      "epoch": 1.698956382124699,
+      "grad_norm": 0.5323873162269592,
+      "learning_rate": 2.7391680293602866e-05,
+      "loss": 0.5906,
+      "step": 9524
+    },
+    {
+      "epoch": 1.699134778342699,
+      "grad_norm": 0.47437727451324463,
+      "learning_rate": 2.7359802724741367e-05,
+      "loss": 0.4004,
+      "step": 9525
+    },
+    {
+      "epoch": 1.6993131745606993,
+      "grad_norm": 0.47239458560943604,
+      "learning_rate": 2.7327942641945157e-05,
+      "loss": 0.4497,
+      "step": 9526
+    },
+    {
+      "epoch": 1.6994915707786995,
+      "grad_norm": 0.5352083444595337,
+      "learning_rate": 2.7296100047716488e-05,
+      "loss": 0.5579,
+      "step": 9527
+    },
+    {
+      "epoch": 1.6996699669966997,
+      "grad_norm": 0.530881404876709,
+      "learning_rate": 2.7264274944556328e-05,
+      "loss": 0.5424,
+      "step": 9528
+    },
+    {
+      "epoch": 1.6998483632147,
+      "grad_norm": 0.5889961123466492,
+      "learning_rate": 2.7232467334964095e-05,
+      "loss": 0.7217,
+      "step": 9529
+    },
+    {
+      "epoch": 1.7000267594327,
+      "grad_norm": 0.5244048833847046,
+      "learning_rate": 2.7200677221438098e-05,
+      "loss": 0.5383,
+      "step": 9530
+    },
+    {
+      "epoch": 1.7002051556507003,
+      "grad_norm": 0.5565990209579468,
+      "learning_rate": 2.7168904606475005e-05,
+      "loss": 0.642,
+      "step": 9531
+    },
+    {
+      "epoch": 1.7003835518687005,
+      "grad_norm": 0.4610227644443512,
+      "learning_rate": 2.7137149492570207e-05,
+      "loss": 0.3604,
+      "step": 9532
+    },
+    {
+      "epoch": 1.7005619480867007,
+      "grad_norm": 0.4633643329143524,
+      "learning_rate": 2.7105411882217852e-05,
+      "loss": 0.3937,
+      "step": 9533
+    },
+    {
+      "epoch": 1.700740344304701,
+      "grad_norm": 0.5488377213478088,
+      "learning_rate": 2.7073691777910387e-05,
+      "loss": 0.5148,
+      "step": 9534
+    },
+    {
+      "epoch": 1.7009187405227009,
+      "grad_norm": 0.5138922929763794,
+      "learning_rate": 2.7041989182139375e-05,
+      "loss": 0.4672,
+      "step": 9535
+    },
+    {
+      "epoch": 1.701097136740701,
+      "grad_norm": 0.5196233987808228,
+      "learning_rate": 2.7010304097394578e-05,
+      "loss": 0.4599,
+      "step": 9536
+    },
+    {
+      "epoch": 1.7012755329587013,
+      "grad_norm": 0.45630475878715515,
+      "learning_rate": 2.6978636526164474e-05,
+      "loss": 0.3845,
+      "step": 9537
+    },
+    {
+      "epoch": 1.7014539291767015,
+      "grad_norm": 0.5754191279411316,
+      "learning_rate": 2.6946986470936352e-05,
+      "loss": 0.7097,
+      "step": 9538
+    },
+    {
+      "epoch": 1.7016323253947017,
+      "grad_norm": 0.5594825148582458,
+      "learning_rate": 2.6915353934195864e-05,
+      "loss": 0.6275,
+      "step": 9539
+    },
+    {
+      "epoch": 1.7018107216127019,
+      "grad_norm": 0.5201573967933655,
+      "learning_rate": 2.688373891842752e-05,
+      "loss": 0.5501,
+      "step": 9540
+    },
+    {
+      "epoch": 1.7019891178307018,
+      "grad_norm": 0.5017893314361572,
+      "learning_rate": 2.685214142611428e-05,
+      "loss": 0.4429,
+      "step": 9541
+    },
+    {
+      "epoch": 1.702167514048702,
+      "grad_norm": 0.5738409161567688,
+      "learning_rate": 2.682056145973785e-05,
+      "loss": 0.6522,
+      "step": 9542
+    },
+    {
+      "epoch": 1.7023459102667022,
+      "grad_norm": 0.5462552905082703,
+      "learning_rate": 2.6788999021778506e-05,
+      "loss": 0.7407,
+      "step": 9543
+    },
+    {
+      "epoch": 1.7025243064847024,
+      "grad_norm": 0.5017895102500916,
+      "learning_rate": 2.6757454114715058e-05,
+      "loss": 0.4998,
+      "step": 9544
+    },
+    {
+      "epoch": 1.7027027027027026,
+      "grad_norm": 0.47052887082099915,
+      "learning_rate": 2.6725926741025143e-05,
+      "loss": 0.3548,
+      "step": 9545
+    },
+    {
+      "epoch": 1.7028810989207028,
+      "grad_norm": 0.5245179533958435,
+      "learning_rate": 2.6694416903184805e-05,
+      "loss": 0.5281,
+      "step": 9546
+    },
+    {
+      "epoch": 1.703059495138703,
+      "grad_norm": 0.4874159097671509,
+      "learning_rate": 2.6662924603668927e-05,
+      "loss": 0.443,
+      "step": 9547
+    },
+    {
+      "epoch": 1.7032378913567032,
+      "grad_norm": 0.5207834243774414,
+      "learning_rate": 2.6631449844950806e-05,
+      "loss": 0.4949,
+      "step": 9548
+    },
+    {
+      "epoch": 1.7034162875747034,
+      "grad_norm": 0.43551138043403625,
+      "learning_rate": 2.6599992629502466e-05,
+      "loss": 0.3538,
+      "step": 9549
+    },
+    {
+      "epoch": 1.7035946837927036,
+      "grad_norm": 0.49214231967926025,
+      "learning_rate": 2.656855295979463e-05,
+      "loss": 0.4541,
+      "step": 9550
+    },
+    {
+      "epoch": 1.7037730800107038,
+      "grad_norm": 0.4991733729839325,
+      "learning_rate": 2.6537130838296452e-05,
+      "loss": 0.3988,
+      "step": 9551
+    },
+    {
+      "epoch": 1.703951476228704,
+      "grad_norm": 0.5545747876167297,
+      "learning_rate": 2.650572626747588e-05,
+      "loss": 0.5933,
+      "step": 9552
+    },
+    {
+      "epoch": 1.7041298724467042,
+      "grad_norm": 0.542052686214447,
+      "learning_rate": 2.6474339249799362e-05,
+      "loss": 0.5359,
+      "step": 9553
+    },
+    {
+      "epoch": 1.7043082686647044,
+      "grad_norm": 0.5299772620201111,
+      "learning_rate": 2.6442969787732085e-05,
+      "loss": 0.5399,
+      "step": 9554
+    },
+    {
+      "epoch": 1.7044866648827046,
+      "grad_norm": 0.490618497133255,
+      "learning_rate": 2.641161788373772e-05,
+      "loss": 0.4699,
+      "step": 9555
+    },
+    {
+      "epoch": 1.7046650611007048,
+      "grad_norm": 0.5627853870391846,
+      "learning_rate": 2.6380283540278714e-05,
+      "loss": 0.6102,
+      "step": 9556
+    },
+    {
+      "epoch": 1.7048434573187048,
+      "grad_norm": 0.6248000860214233,
+      "learning_rate": 2.634896675981599e-05,
+      "loss": 0.5612,
+      "step": 9557
+    },
+    {
+      "epoch": 1.705021853536705,
+      "grad_norm": 0.5680274367332458,
+      "learning_rate": 2.6317667544809132e-05,
+      "loss": 0.5528,
+      "step": 9558
+    },
+    {
+      "epoch": 1.7052002497547052,
+      "grad_norm": 0.5207681059837341,
+      "learning_rate": 2.628638589771648e-05,
+      "loss": 0.526,
+      "step": 9559
+    },
+    {
+      "epoch": 1.7053786459727054,
+      "grad_norm": 0.4698508381843567,
+      "learning_rate": 2.6255121820994737e-05,
+      "loss": 0.4154,
+      "step": 9560
+    },
+    {
+      "epoch": 1.7055570421907056,
+      "grad_norm": 0.45155656337738037,
+      "learning_rate": 2.6223875317099492e-05,
+      "loss": 0.4199,
+      "step": 9561
+    },
+    {
+      "epoch": 1.7057354384087058,
+      "grad_norm": 0.5513625741004944,
+      "learning_rate": 2.6192646388484732e-05,
+      "loss": 0.5584,
+      "step": 9562
+    },
+    {
+      "epoch": 1.7059138346267058,
+      "grad_norm": 0.4611358642578125,
+      "learning_rate": 2.6161435037603264e-05,
+      "loss": 0.3435,
+      "step": 9563
+    },
+    {
+      "epoch": 1.706092230844706,
+      "grad_norm": 0.5059071183204651,
+      "learning_rate": 2.61302412669063e-05,
+      "loss": 0.4466,
+      "step": 9564
+    },
+    {
+      "epoch": 1.7062706270627062,
+      "grad_norm": 0.5149518251419067,
+      "learning_rate": 2.609906507884388e-05,
+      "loss": 0.4098,
+      "step": 9565
+    },
+    {
+      "epoch": 1.7064490232807064,
+      "grad_norm": 0.5157696604728699,
+      "learning_rate": 2.606790647586457e-05,
+      "loss": 0.4633,
+      "step": 9566
+    },
+    {
+      "epoch": 1.7066274194987066,
+      "grad_norm": 0.47338274121284485,
+      "learning_rate": 2.6036765460415447e-05,
+      "loss": 0.4056,
+      "step": 9567
+    },
+    {
+      "epoch": 1.7068058157167068,
+      "grad_norm": 0.49794062972068787,
+      "learning_rate": 2.6005642034942463e-05,
+      "loss": 0.5416,
+      "step": 9568
+    },
+    {
+      "epoch": 1.706984211934707,
+      "grad_norm": 0.6232036352157593,
+      "learning_rate": 2.597453620188997e-05,
+      "loss": 0.6416,
+      "step": 9569
+    },
+    {
+      "epoch": 1.7071626081527072,
+      "grad_norm": 0.5632308125495911,
+      "learning_rate": 2.5943447963700933e-05,
+      "loss": 0.5564,
+      "step": 9570
+    },
+    {
+      "epoch": 1.7073410043707073,
+      "grad_norm": 0.5408989191055298,
+      "learning_rate": 2.591237732281712e-05,
+      "loss": 0.5595,
+      "step": 9571
+    },
+    {
+      "epoch": 1.7075194005887075,
+      "grad_norm": 0.5331925749778748,
+      "learning_rate": 2.588132428167869e-05,
+      "loss": 0.567,
+      "step": 9572
+    },
+    {
+      "epoch": 1.7076977968067077,
+      "grad_norm": 0.4525754749774933,
+      "learning_rate": 2.5850288842724696e-05,
+      "loss": 0.3582,
+      "step": 9573
+    },
+    {
+      "epoch": 1.707876193024708,
+      "grad_norm": 0.4124913513660431,
+      "learning_rate": 2.5819271008392486e-05,
+      "loss": 0.3726,
+      "step": 9574
+    },
+    {
+      "epoch": 1.7080545892427081,
+      "grad_norm": 0.45310550928115845,
+      "learning_rate": 2.578827078111831e-05,
+      "loss": 0.4381,
+      "step": 9575
+    },
+    {
+      "epoch": 1.7082329854607083,
+      "grad_norm": 0.45527383685112,
+      "learning_rate": 2.5757288163336808e-05,
+      "loss": 0.4271,
+      "step": 9576
+    },
+    {
+      "epoch": 1.7084113816787085,
+      "grad_norm": 0.5173305869102478,
+      "learning_rate": 2.572632315748144e-05,
+      "loss": 0.5474,
+      "step": 9577
+    },
+    {
+      "epoch": 1.7085897778967087,
+      "grad_norm": 0.4721916913986206,
+      "learning_rate": 2.569537576598416e-05,
+      "loss": 0.37,
+      "step": 9578
+    },
+    {
+      "epoch": 1.7087681741147087,
+      "grad_norm": 0.503395676612854,
+      "learning_rate": 2.5664445991275486e-05,
+      "loss": 0.5476,
+      "step": 9579
+    },
+    {
+      "epoch": 1.708946570332709,
+      "grad_norm": 0.4126671254634857,
+      "learning_rate": 2.563353383578468e-05,
+      "loss": 0.3101,
+      "step": 9580
+    },
+    {
+      "epoch": 1.7091249665507091,
+      "grad_norm": 0.546122133731842,
+      "learning_rate": 2.560263930193957e-05,
+      "loss": 0.4748,
+      "step": 9581
+    },
+    {
+      "epoch": 1.7093033627687093,
+      "grad_norm": 0.4883154332637787,
+      "learning_rate": 2.557176239216666e-05,
+      "loss": 0.465,
+      "step": 9582
+    },
+    {
+      "epoch": 1.7094817589867095,
+      "grad_norm": 0.5338440537452698,
+      "learning_rate": 2.554090310889093e-05,
+      "loss": 0.4543,
+      "step": 9583
+    },
+    {
+      "epoch": 1.7096601552047097,
+      "grad_norm": 0.49619346857070923,
+      "learning_rate": 2.5510061454536048e-05,
+      "loss": 0.4506,
+      "step": 9584
+    },
+    {
+      "epoch": 1.7098385514227097,
+      "grad_norm": 0.5309063196182251,
+      "learning_rate": 2.5479237431524387e-05,
+      "loss": 0.6047,
+      "step": 9585
+    },
+    {
+      "epoch": 1.7100169476407099,
+      "grad_norm": 0.4647756516933441,
+      "learning_rate": 2.5448431042276732e-05,
+      "loss": 0.3435,
+      "step": 9586
+    },
+    {
+      "epoch": 1.71019534385871,
+      "grad_norm": 0.5021915435791016,
+      "learning_rate": 2.5417642289212756e-05,
+      "loss": 0.5529,
+      "step": 9587
+    },
+    {
+      "epoch": 1.7103737400767103,
+      "grad_norm": 0.5375409126281738,
+      "learning_rate": 2.5386871174750415e-05,
+      "loss": 0.5268,
+      "step": 9588
+    },
+    {
+      "epoch": 1.7105521362947105,
+      "grad_norm": 0.5530412793159485,
+      "learning_rate": 2.5356117701306637e-05,
+      "loss": 0.4953,
+      "step": 9589
+    },
+    {
+      "epoch": 1.7107305325127107,
+      "grad_norm": 0.5221925973892212,
+      "learning_rate": 2.532538187129668e-05,
+      "loss": 0.536,
+      "step": 9590
+    },
+    {
+      "epoch": 1.7109089287307109,
+      "grad_norm": 0.46529194712638855,
+      "learning_rate": 2.5294663687134512e-05,
+      "loss": 0.436,
+      "step": 9591
+    },
+    {
+      "epoch": 1.711087324948711,
+      "grad_norm": 0.5522665977478027,
+      "learning_rate": 2.5263963151232806e-05,
+      "loss": 0.4987,
+      "step": 9592
+    },
+    {
+      "epoch": 1.7112657211667113,
+      "grad_norm": 0.5286535024642944,
+      "learning_rate": 2.523328026600266e-05,
+      "loss": 0.5313,
+      "step": 9593
+    },
+    {
+      "epoch": 1.7114441173847115,
+      "grad_norm": 0.476859986782074,
+      "learning_rate": 2.5202615033854016e-05,
+      "loss": 0.4428,
+      "step": 9594
+    },
+    {
+      "epoch": 1.7116225136027117,
+      "grad_norm": 0.5317862629890442,
+      "learning_rate": 2.5171967457195217e-05,
+      "loss": 0.5632,
+      "step": 9595
+    },
+    {
+      "epoch": 1.7118009098207119,
+      "grad_norm": 0.5589151978492737,
+      "learning_rate": 2.5141337538433312e-05,
+      "loss": 0.6869,
+      "step": 9596
+    },
+    {
+      "epoch": 1.711979306038712,
+      "grad_norm": 0.4832451343536377,
+      "learning_rate": 2.5110725279974074e-05,
+      "loss": 0.4022,
+      "step": 9597
+    },
+    {
+      "epoch": 1.7121577022567123,
+      "grad_norm": 0.45813506841659546,
+      "learning_rate": 2.508013068422163e-05,
+      "loss": 0.3612,
+      "step": 9598
+    },
+    {
+      "epoch": 1.7123360984747125,
+      "grad_norm": 0.4622073471546173,
+      "learning_rate": 2.504955375357895e-05,
+      "loss": 0.3288,
+      "step": 9599
+    },
+    {
+      "epoch": 1.7125144946927127,
+      "grad_norm": 0.5384669899940491,
+      "learning_rate": 2.5018994490447505e-05,
+      "loss": 0.4338,
+      "step": 9600
+    },
+    {
+      "epoch": 1.7126928909107126,
+      "grad_norm": 0.5117572546005249,
+      "learning_rate": 2.4988452897227454e-05,
+      "loss": 0.4504,
+      "step": 9601
+    },
+    {
+      "epoch": 1.7128712871287128,
+      "grad_norm": 0.5058663487434387,
+      "learning_rate": 2.495792897631749e-05,
+      "loss": 0.41,
+      "step": 9602
+    },
+    {
+      "epoch": 1.713049683346713,
+      "grad_norm": 0.481413871049881,
+      "learning_rate": 2.492742273011489e-05,
+      "loss": 0.4433,
+      "step": 9603
+    },
+    {
+      "epoch": 1.7132280795647132,
+      "grad_norm": 0.5749334692955017,
+      "learning_rate": 2.4896934161015683e-05,
+      "loss": 0.542,
+      "step": 9604
+    },
+    {
+      "epoch": 1.7134064757827134,
+      "grad_norm": 0.4859156012535095,
+      "learning_rate": 2.486646327141437e-05,
+      "loss": 0.4481,
+      "step": 9605
+    },
+    {
+      "epoch": 1.7135848720007136,
+      "grad_norm": 0.5681634545326233,
+      "learning_rate": 2.4836010063704174e-05,
+      "loss": 0.5799,
+      "step": 9606
+    },
+    {
+      "epoch": 1.7137632682187136,
+      "grad_norm": 0.5673406720161438,
+      "learning_rate": 2.4805574540276822e-05,
+      "loss": 0.659,
+      "step": 9607
+    },
+    {
+      "epoch": 1.7139416644367138,
+      "grad_norm": 0.5422623753547668,
+      "learning_rate": 2.477515670352279e-05,
+      "loss": 0.4897,
+      "step": 9608
+    },
+    {
+      "epoch": 1.714120060654714,
+      "grad_norm": 0.4804922044277191,
+      "learning_rate": 2.4744756555830972e-05,
+      "loss": 0.4234,
+      "step": 9609
+    },
+    {
+      "epoch": 1.7142984568727142,
+      "grad_norm": 0.5365844368934631,
+      "learning_rate": 2.4714374099589042e-05,
+      "loss": 0.5374,
+      "step": 9610
+    },
+    {
+      "epoch": 1.7144768530907144,
+      "grad_norm": 0.5775295495986938,
+      "learning_rate": 2.468400933718326e-05,
+      "loss": 0.5003,
+      "step": 9611
+    },
+    {
+      "epoch": 1.7146552493087146,
+      "grad_norm": 0.5075268149375916,
+      "learning_rate": 2.4653662270998383e-05,
+      "loss": 0.4016,
+      "step": 9612
+    },
+    {
+      "epoch": 1.7148336455267148,
+      "grad_norm": 0.5659419298171997,
+      "learning_rate": 2.462333290341795e-05,
+      "loss": 0.6249,
+      "step": 9613
+    },
+    {
+      "epoch": 1.715012041744715,
+      "grad_norm": 0.5081014633178711,
+      "learning_rate": 2.4593021236823914e-05,
+      "loss": 0.394,
+      "step": 9614
+    },
+    {
+      "epoch": 1.7151904379627152,
+      "grad_norm": 0.5532371401786804,
+      "learning_rate": 2.456272727359704e-05,
+      "loss": 0.606,
+      "step": 9615
+    },
+    {
+      "epoch": 1.7153688341807154,
+      "grad_norm": 0.5273371934890747,
+      "learning_rate": 2.4532451016116535e-05,
+      "loss": 0.6084,
+      "step": 9616
+    },
+    {
+      "epoch": 1.7155472303987156,
+      "grad_norm": 0.5197193622589111,
+      "learning_rate": 2.4502192466760276e-05,
+      "loss": 0.5743,
+      "step": 9617
+    },
+    {
+      "epoch": 1.7157256266167158,
+      "grad_norm": 0.5154910683631897,
+      "learning_rate": 2.4471951627904804e-05,
+      "loss": 0.5972,
+      "step": 9618
+    },
+    {
+      "epoch": 1.715904022834716,
+      "grad_norm": 0.5434054732322693,
+      "learning_rate": 2.4441728501925165e-05,
+      "loss": 0.5537,
+      "step": 9619
+    },
+    {
+      "epoch": 1.7160824190527162,
+      "grad_norm": 0.5460578799247742,
+      "learning_rate": 2.4411523091195153e-05,
+      "loss": 0.6007,
+      "step": 9620
+    },
+    {
+      "epoch": 1.7162608152707164,
+      "grad_norm": 0.4608527421951294,
+      "learning_rate": 2.4381335398086985e-05,
+      "loss": 0.3695,
+      "step": 9621
+    },
+    {
+      "epoch": 1.7164392114887166,
+      "grad_norm": 0.6361474990844727,
+      "learning_rate": 2.4351165424971706e-05,
+      "loss": 0.7393,
+      "step": 9622
+    },
+    {
+      "epoch": 1.7166176077067166,
+      "grad_norm": 0.5027257204055786,
+      "learning_rate": 2.4321013174218782e-05,
+      "loss": 0.4491,
+      "step": 9623
+    },
+    {
+      "epoch": 1.7167960039247168,
+      "grad_norm": 0.6172974705696106,
+      "learning_rate": 2.4290878648196318e-05,
+      "loss": 0.6406,
+      "step": 9624
+    },
+    {
+      "epoch": 1.716974400142717,
+      "grad_norm": 0.48582372069358826,
+      "learning_rate": 2.4260761849271116e-05,
+      "loss": 0.5167,
+      "step": 9625
+    },
+    {
+      "epoch": 1.7171527963607172,
+      "grad_norm": 0.5107640624046326,
+      "learning_rate": 2.4230662779808538e-05,
+      "loss": 0.4779,
+      "step": 9626
+    },
+    {
+      "epoch": 1.7173311925787174,
+      "grad_norm": 0.591314971446991,
+      "learning_rate": 2.4200581442172576e-05,
+      "loss": 0.6189,
+      "step": 9627
+    },
+    {
+      "epoch": 1.7175095887967176,
+      "grad_norm": 0.5267913937568665,
+      "learning_rate": 2.4170517838725816e-05,
+      "loss": 0.4974,
+      "step": 9628
+    },
+    {
+      "epoch": 1.7176879850147175,
+      "grad_norm": 0.5801319479942322,
+      "learning_rate": 2.4140471971829338e-05,
+      "loss": 0.5212,
+      "step": 9629
+    },
+    {
+      "epoch": 1.7178663812327177,
+      "grad_norm": 0.5426841974258423,
+      "learning_rate": 2.4110443843843034e-05,
+      "loss": 0.4802,
+      "step": 9630
+    },
+    {
+      "epoch": 1.718044777450718,
+      "grad_norm": 0.5340462327003479,
+      "learning_rate": 2.408043345712521e-05,
+      "loss": 0.5782,
+      "step": 9631
+    },
+    {
+      "epoch": 1.7182231736687181,
+      "grad_norm": 0.5407845973968506,
+      "learning_rate": 2.4050440814032986e-05,
+      "loss": 0.5252,
+      "step": 9632
+    },
+    {
+      "epoch": 1.7184015698867183,
+      "grad_norm": 0.5170387625694275,
+      "learning_rate": 2.4020465916921862e-05,
+      "loss": 0.5455,
+      "step": 9633
+    },
+    {
+      "epoch": 1.7185799661047185,
+      "grad_norm": 0.45365792512893677,
+      "learning_rate": 2.3990508768146125e-05,
+      "loss": 0.3647,
+      "step": 9634
+    },
+    {
+      "epoch": 1.7187583623227187,
+      "grad_norm": 0.5068604946136475,
+      "learning_rate": 2.3960569370058528e-05,
+      "loss": 0.5216,
+      "step": 9635
+    },
+    {
+      "epoch": 1.718936758540719,
+      "grad_norm": 0.6078252792358398,
+      "learning_rate": 2.3930647725010607e-05,
+      "loss": 0.7029,
+      "step": 9636
+    },
+    {
+      "epoch": 1.7191151547587191,
+      "grad_norm": 0.4622660279273987,
+      "learning_rate": 2.3900743835352316e-05,
+      "loss": 0.4294,
+      "step": 9637
+    },
+    {
+      "epoch": 1.7192935509767193,
+      "grad_norm": 0.45116907358169556,
+      "learning_rate": 2.387085770343225e-05,
+      "loss": 0.3998,
+      "step": 9638
+    },
+    {
+      "epoch": 1.7194719471947195,
+      "grad_norm": 0.4865681231021881,
+      "learning_rate": 2.3840989331597757e-05,
+      "loss": 0.424,
+      "step": 9639
+    },
+    {
+      "epoch": 1.7196503434127197,
+      "grad_norm": 0.5520277619361877,
+      "learning_rate": 2.3811138722194593e-05,
+      "loss": 0.5527,
+      "step": 9640
+    },
+    {
+      "epoch": 1.71982873963072,
+      "grad_norm": 0.4993618130683899,
+      "learning_rate": 2.378130587756727e-05,
+      "loss": 0.4283,
+      "step": 9641
+    },
+    {
+      "epoch": 1.7200071358487201,
+      "grad_norm": 0.5111703276634216,
+      "learning_rate": 2.3751490800058865e-05,
+      "loss": 0.4524,
+      "step": 9642
+    },
+    {
+      "epoch": 1.7201855320667203,
+      "grad_norm": 0.554426372051239,
+      "learning_rate": 2.3721693492010977e-05,
+      "loss": 0.5797,
+      "step": 9643
+    },
+    {
+      "epoch": 1.7203639282847205,
+      "grad_norm": 0.4680947959423065,
+      "learning_rate": 2.369191395576395e-05,
+      "loss": 0.3494,
+      "step": 9644
+    },
+    {
+      "epoch": 1.7205423245027205,
+      "grad_norm": 0.5217585563659668,
+      "learning_rate": 2.3662152193656554e-05,
+      "loss": 0.6172,
+      "step": 9645
+    },
+    {
+      "epoch": 1.7207207207207207,
+      "grad_norm": 0.5365320444107056,
+      "learning_rate": 2.3632408208026395e-05,
+      "loss": 0.5242,
+      "step": 9646
+    },
+    {
+      "epoch": 1.7208991169387209,
+      "grad_norm": 0.6032186150550842,
+      "learning_rate": 2.3602682001209408e-05,
+      "loss": 0.6657,
+      "step": 9647
+    },
+    {
+      "epoch": 1.721077513156721,
+      "grad_norm": 0.5880352258682251,
+      "learning_rate": 2.3572973575540418e-05,
+      "loss": 0.665,
+      "step": 9648
+    },
+    {
+      "epoch": 1.7212559093747213,
+      "grad_norm": 0.5372354388237,
+      "learning_rate": 2.3543282933352645e-05,
+      "loss": 0.4893,
+      "step": 9649
+    },
+    {
+      "epoch": 1.7214343055927215,
+      "grad_norm": 0.44951343536376953,
+      "learning_rate": 2.3513610076977916e-05,
+      "loss": 0.3493,
+      "step": 9650
+    },
+    {
+      "epoch": 1.7216127018107215,
+      "grad_norm": 0.49766871333122253,
+      "learning_rate": 2.348395500874684e-05,
+      "loss": 0.4823,
+      "step": 9651
+    },
+    {
+      "epoch": 1.7217910980287217,
+      "grad_norm": 0.5321714282035828,
+      "learning_rate": 2.345431773098841e-05,
+      "loss": 0.5048,
+      "step": 9652
+    },
+    {
+      "epoch": 1.7219694942467219,
+      "grad_norm": 0.5095254182815552,
+      "learning_rate": 2.342469824603044e-05,
+      "loss": 0.4675,
+      "step": 9653
+    },
+    {
+      "epoch": 1.722147890464722,
+      "grad_norm": 0.48094990849494934,
+      "learning_rate": 2.339509655619909e-05,
+      "loss": 0.4484,
+      "step": 9654
+    },
+    {
+      "epoch": 1.7223262866827223,
+      "grad_norm": 0.5008235573768616,
+      "learning_rate": 2.336551266381942e-05,
+      "loss": 0.3872,
+      "step": 9655
+    },
+    {
+      "epoch": 1.7225046829007225,
+      "grad_norm": 0.552046537399292,
+      "learning_rate": 2.3335946571214795e-05,
+      "loss": 0.6718,
+      "step": 9656
+    },
+    {
+      "epoch": 1.7226830791187226,
+      "grad_norm": 0.5397679805755615,
+      "learning_rate": 2.330639828070738e-05,
+      "loss": 0.4931,
+      "step": 9657
+    },
+    {
+      "epoch": 1.7228614753367228,
+      "grad_norm": 0.5258644819259644,
+      "learning_rate": 2.3276867794617936e-05,
+      "loss": 0.3977,
+      "step": 9658
+    },
+    {
+      "epoch": 1.723039871554723,
+      "grad_norm": 0.5579936504364014,
+      "learning_rate": 2.3247355115265684e-05,
+      "loss": 0.5178,
+      "step": 9659
+    },
+    {
+      "epoch": 1.7232182677727232,
+      "grad_norm": 0.53934246301651,
+      "learning_rate": 2.3217860244968638e-05,
+      "loss": 0.5224,
+      "step": 9660
+    },
+    {
+      "epoch": 1.7233966639907234,
+      "grad_norm": 0.5465741157531738,
+      "learning_rate": 2.3188383186043187e-05,
+      "loss": 0.5329,
+      "step": 9661
+    },
+    {
+      "epoch": 1.7235750602087236,
+      "grad_norm": 0.6120946407318115,
+      "learning_rate": 2.3158923940804572e-05,
+      "loss": 0.7921,
+      "step": 9662
+    },
+    {
+      "epoch": 1.7237534564267238,
+      "grad_norm": 0.5330802798271179,
+      "learning_rate": 2.3129482511566463e-05,
+      "loss": 0.5436,
+      "step": 9663
+    },
+    {
+      "epoch": 1.723931852644724,
+      "grad_norm": 0.4863676428794861,
+      "learning_rate": 2.3100058900641125e-05,
+      "loss": 0.4046,
+      "step": 9664
+    },
+    {
+      "epoch": 1.7241102488627242,
+      "grad_norm": 0.43972915410995483,
+      "learning_rate": 2.3070653110339567e-05,
+      "loss": 0.4352,
+      "step": 9665
+    },
+    {
+      "epoch": 1.7242886450807244,
+      "grad_norm": 0.5199171304702759,
+      "learning_rate": 2.3041265142971196e-05,
+      "loss": 0.486,
+      "step": 9666
+    },
+    {
+      "epoch": 1.7244670412987244,
+      "grad_norm": 0.4947983920574188,
+      "learning_rate": 2.3011895000844247e-05,
+      "loss": 0.4544,
+      "step": 9667
+    },
+    {
+      "epoch": 1.7246454375167246,
+      "grad_norm": 0.565777599811554,
+      "learning_rate": 2.298254268626532e-05,
+      "loss": 0.5195,
+      "step": 9668
+    },
+    {
+      "epoch": 1.7248238337347248,
+      "grad_norm": 0.4736662805080414,
+      "learning_rate": 2.2953208201539873e-05,
+      "loss": 0.3791,
+      "step": 9669
+    },
+    {
+      "epoch": 1.725002229952725,
+      "grad_norm": 0.5035544633865356,
+      "learning_rate": 2.292389154897173e-05,
+      "loss": 0.488,
+      "step": 9670
+    },
+    {
+      "epoch": 1.7251806261707252,
+      "grad_norm": 0.6010121703147888,
+      "learning_rate": 2.2894592730863335e-05,
+      "loss": 0.6792,
+      "step": 9671
+    },
+    {
+      "epoch": 1.7253590223887254,
+      "grad_norm": 0.47702497243881226,
+      "learning_rate": 2.2865311749515978e-05,
+      "loss": 0.3775,
+      "step": 9672
+    },
+    {
+      "epoch": 1.7255374186067254,
+      "grad_norm": 0.5288291573524475,
+      "learning_rate": 2.2836048607229264e-05,
+      "loss": 0.4574,
+      "step": 9673
+    },
+    {
+      "epoch": 1.7257158148247256,
+      "grad_norm": 0.5164685845375061,
+      "learning_rate": 2.2806803306301583e-05,
+      "loss": 0.4143,
+      "step": 9674
+    },
+    {
+      "epoch": 1.7258942110427258,
+      "grad_norm": 0.5557328462600708,
+      "learning_rate": 2.2777575849029785e-05,
+      "loss": 0.5724,
+      "step": 9675
+    },
+    {
+      "epoch": 1.726072607260726,
+      "grad_norm": 0.5354199409484863,
+      "learning_rate": 2.2748366237709372e-05,
+      "loss": 0.499,
+      "step": 9676
+    },
+    {
+      "epoch": 1.7262510034787262,
+      "grad_norm": 0.5438181757926941,
+      "learning_rate": 2.271917447463451e-05,
+      "loss": 0.6275,
+      "step": 9677
+    },
+    {
+      "epoch": 1.7264293996967264,
+      "grad_norm": 0.5154425501823425,
+      "learning_rate": 2.2690000562097858e-05,
+      "loss": 0.4563,
+      "step": 9678
+    },
+    {
+      "epoch": 1.7266077959147266,
+      "grad_norm": 0.58868008852005,
+      "learning_rate": 2.2660844502390754e-05,
+      "loss": 0.632,
+      "step": 9679
+    },
+    {
+      "epoch": 1.7267861921327268,
+      "grad_norm": 0.541050136089325,
+      "learning_rate": 2.263170629780306e-05,
+      "loss": 0.59,
+      "step": 9680
+    },
+    {
+      "epoch": 1.726964588350727,
+      "grad_norm": 0.5016245245933533,
+      "learning_rate": 2.2602585950623367e-05,
+      "loss": 0.4542,
+      "step": 9681
+    },
+    {
+      "epoch": 1.7271429845687272,
+      "grad_norm": 0.5481529235839844,
+      "learning_rate": 2.25734834631387e-05,
+      "loss": 0.6274,
+      "step": 9682
+    },
+    {
+      "epoch": 1.7273213807867274,
+      "grad_norm": 0.5226790308952332,
+      "learning_rate": 2.2544398837634732e-05,
+      "loss": 0.4746,
+      "step": 9683
+    },
+    {
+      "epoch": 1.7274997770047276,
+      "grad_norm": 0.5335855484008789,
+      "learning_rate": 2.2515332076395862e-05,
+      "loss": 0.4868,
+      "step": 9684
+    },
+    {
+      "epoch": 1.7276781732227278,
+      "grad_norm": 0.5473877787590027,
+      "learning_rate": 2.2486283181704842e-05,
+      "loss": 0.4419,
+      "step": 9685
+    },
+    {
+      "epoch": 1.727856569440728,
+      "grad_norm": 0.5562229752540588,
+      "learning_rate": 2.2457252155843293e-05,
+      "loss": 0.6055,
+      "step": 9686
+    },
+    {
+      "epoch": 1.7280349656587282,
+      "grad_norm": 0.5625514984130859,
+      "learning_rate": 2.24282390010912e-05,
+      "loss": 0.5543,
+      "step": 9687
+    },
+    {
+      "epoch": 1.7282133618767284,
+      "grad_norm": 0.5255547165870667,
+      "learning_rate": 2.2399243719727265e-05,
+      "loss": 0.5099,
+      "step": 9688
+    },
+    {
+      "epoch": 1.7283917580947283,
+      "grad_norm": 0.6275233626365662,
+      "learning_rate": 2.237026631402883e-05,
+      "loss": 0.5615,
+      "step": 9689
+    },
+    {
+      "epoch": 1.7285701543127285,
+      "grad_norm": 0.5601621866226196,
+      "learning_rate": 2.234130678627169e-05,
+      "loss": 0.5324,
+      "step": 9690
+    },
+    {
+      "epoch": 1.7287485505307287,
+      "grad_norm": 0.5600411891937256,
+      "learning_rate": 2.231236513873039e-05,
+      "loss": 0.5877,
+      "step": 9691
+    },
+    {
+      "epoch": 1.728926946748729,
+      "grad_norm": 0.5259904861450195,
+      "learning_rate": 2.228344137367791e-05,
+      "loss": 0.5547,
+      "step": 9692
+    },
+    {
+      "epoch": 1.7291053429667291,
+      "grad_norm": 0.4963001012802124,
+      "learning_rate": 2.225453549338599e-05,
+      "loss": 0.4832,
+      "step": 9693
+    },
+    {
+      "epoch": 1.7292837391847293,
+      "grad_norm": 0.534983217716217,
+      "learning_rate": 2.222564750012479e-05,
+      "loss": 0.5112,
+      "step": 9694
+    },
+    {
+      "epoch": 1.7294621354027293,
+      "grad_norm": 0.47947463393211365,
+      "learning_rate": 2.219677739616327e-05,
+      "loss": 0.4251,
+      "step": 9695
+    },
+    {
+      "epoch": 1.7296405316207295,
+      "grad_norm": 0.5085344314575195,
+      "learning_rate": 2.216792518376884e-05,
+      "loss": 0.4215,
+      "step": 9696
+    },
+    {
+      "epoch": 1.7298189278387297,
+      "grad_norm": 0.48446446657180786,
+      "learning_rate": 2.213909086520746e-05,
+      "loss": 0.443,
+      "step": 9697
+    },
+    {
+      "epoch": 1.72999732405673,
+      "grad_norm": 0.4940386116504669,
+      "learning_rate": 2.2110274442743853e-05,
+      "loss": 0.4204,
+      "step": 9698
+    },
+    {
+      "epoch": 1.73017572027473,
+      "grad_norm": 0.5226745009422302,
+      "learning_rate": 2.2081475918641208e-05,
+      "loss": 0.5349,
+      "step": 9699
+    },
+    {
+      "epoch": 1.7303541164927303,
+      "grad_norm": 0.47551804780960083,
+      "learning_rate": 2.2052695295161407e-05,
+      "loss": 0.4946,
+      "step": 9700
+    },
+    {
+      "epoch": 1.7305325127107305,
+      "grad_norm": 0.5348621606826782,
+      "learning_rate": 2.2023932574564754e-05,
+      "loss": 0.5195,
+      "step": 9701
+    },
+    {
+      "epoch": 1.7307109089287307,
+      "grad_norm": 0.5494500398635864,
+      "learning_rate": 2.199518775911036e-05,
+      "loss": 0.5932,
+      "step": 9702
+    },
+    {
+      "epoch": 1.730889305146731,
+      "grad_norm": 0.5604888796806335,
+      "learning_rate": 2.196646085105583e-05,
+      "loss": 0.6247,
+      "step": 9703
+    },
+    {
+      "epoch": 1.731067701364731,
+      "grad_norm": 0.5602217316627502,
+      "learning_rate": 2.1937751852657285e-05,
+      "loss": 0.7553,
+      "step": 9704
+    },
+    {
+      "epoch": 1.7312460975827313,
+      "grad_norm": 0.5081519484519958,
+      "learning_rate": 2.190906076616961e-05,
+      "loss": 0.4189,
+      "step": 9705
+    },
+    {
+      "epoch": 1.7314244938007315,
+      "grad_norm": 0.5131904482841492,
+      "learning_rate": 2.188038759384611e-05,
+      "loss": 0.4722,
+      "step": 9706
+    },
+    {
+      "epoch": 1.7316028900187317,
+      "grad_norm": 0.47898441553115845,
+      "learning_rate": 2.1851732337938855e-05,
+      "loss": 0.393,
+      "step": 9707
+    },
+    {
+      "epoch": 1.7317812862367319,
+      "grad_norm": 0.4974140524864197,
+      "learning_rate": 2.1823095000698346e-05,
+      "loss": 0.5102,
+      "step": 9708
+    },
+    {
+      "epoch": 1.731959682454732,
+      "grad_norm": 0.5431427955627441,
+      "learning_rate": 2.1794475584373723e-05,
+      "loss": 0.4054,
+      "step": 9709
+    },
+    {
+      "epoch": 1.7321380786727323,
+      "grad_norm": 0.5118406414985657,
+      "learning_rate": 2.1765874091212834e-05,
+      "loss": 0.5844,
+      "step": 9710
+    },
+    {
+      "epoch": 1.7323164748907325,
+      "grad_norm": 0.5889133214950562,
+      "learning_rate": 2.1737290523461932e-05,
+      "loss": 0.662,
+      "step": 9711
+    },
+    {
+      "epoch": 1.7324948711087325,
+      "grad_norm": 0.5474311113357544,
+      "learning_rate": 2.170872488336606e-05,
+      "loss": 0.5537,
+      "step": 9712
+    },
+    {
+      "epoch": 1.7326732673267327,
+      "grad_norm": 0.5133411884307861,
+      "learning_rate": 2.1680177173168615e-05,
+      "loss": 0.5039,
+      "step": 9713
+    },
+    {
+      "epoch": 1.7328516635447329,
+      "grad_norm": 0.5092921257019043,
+      "learning_rate": 2.1651647395111884e-05,
+      "loss": 0.4727,
+      "step": 9714
+    },
+    {
+      "epoch": 1.733030059762733,
+      "grad_norm": 0.521981954574585,
+      "learning_rate": 2.1623135551436442e-05,
+      "loss": 0.5204,
+      "step": 9715
+    },
+    {
+      "epoch": 1.7332084559807333,
+      "grad_norm": 0.48075124621391296,
+      "learning_rate": 2.1594641644381684e-05,
+      "loss": 0.4316,
+      "step": 9716
+    },
+    {
+      "epoch": 1.7333868521987332,
+      "grad_norm": 0.5798913836479187,
+      "learning_rate": 2.156616567618544e-05,
+      "loss": 0.6439,
+      "step": 9717
+    },
+    {
+      "epoch": 1.7335652484167334,
+      "grad_norm": 0.42159613966941833,
+      "learning_rate": 2.153770764908425e-05,
+      "loss": 0.3031,
+      "step": 9718
+    },
+    {
+      "epoch": 1.7337436446347336,
+      "grad_norm": 0.7180142998695374,
+      "learning_rate": 2.150926756531324e-05,
+      "loss": 0.4069,
+      "step": 9719
+    },
+    {
+      "epoch": 1.7339220408527338,
+      "grad_norm": 0.4658437669277191,
+      "learning_rate": 2.148084542710596e-05,
+      "loss": 0.4876,
+      "step": 9720
+    },
+    {
+      "epoch": 1.734100437070734,
+      "grad_norm": 0.48459357023239136,
+      "learning_rate": 2.1452441236694792e-05,
+      "loss": 0.4056,
+      "step": 9721
+    },
+    {
+      "epoch": 1.7342788332887342,
+      "grad_norm": 0.5127313733100891,
+      "learning_rate": 2.142405499631056e-05,
+      "loss": 0.4503,
+      "step": 9722
+    },
+    {
+      "epoch": 1.7344572295067344,
+      "grad_norm": 0.47491639852523804,
+      "learning_rate": 2.139568670818262e-05,
+      "loss": 0.4279,
+      "step": 9723
+    },
+    {
+      "epoch": 1.7346356257247346,
+      "grad_norm": 0.5612772107124329,
+      "learning_rate": 2.136733637453911e-05,
+      "loss": 0.6575,
+      "step": 9724
+    },
+    {
+      "epoch": 1.7348140219427348,
+      "grad_norm": 0.566277027130127,
+      "learning_rate": 2.1339003997606577e-05,
+      "loss": 0.628,
+      "step": 9725
+    },
+    {
+      "epoch": 1.734992418160735,
+      "grad_norm": 0.5447231531143188,
+      "learning_rate": 2.1310689579610327e-05,
+      "loss": 0.5066,
+      "step": 9726
+    },
+    {
+      "epoch": 1.7351708143787352,
+      "grad_norm": 0.463008850812912,
+      "learning_rate": 2.1282393122774053e-05,
+      "loss": 0.3353,
+      "step": 9727
+    },
+    {
+      "epoch": 1.7353492105967354,
+      "grad_norm": 0.4631538689136505,
+      "learning_rate": 2.1254114629320227e-05,
+      "loss": 0.3234,
+      "step": 9728
+    },
+    {
+      "epoch": 1.7355276068147356,
+      "grad_norm": 0.5900436043739319,
+      "learning_rate": 2.1225854101469794e-05,
+      "loss": 0.5786,
+      "step": 9729
+    },
+    {
+      "epoch": 1.7357060030327358,
+      "grad_norm": 0.4738953113555908,
+      "learning_rate": 2.1197611541442313e-05,
+      "loss": 0.4288,
+      "step": 9730
+    },
+    {
+      "epoch": 1.735884399250736,
+      "grad_norm": 0.4944877028465271,
+      "learning_rate": 2.116938695145598e-05,
+      "loss": 0.3708,
+      "step": 9731
+    },
+    {
+      "epoch": 1.7360627954687362,
+      "grad_norm": 0.49892377853393555,
+      "learning_rate": 2.11411803337275e-05,
+      "loss": 0.4571,
+      "step": 9732
+    },
+    {
+      "epoch": 1.7362411916867364,
+      "grad_norm": 0.5799189209938049,
+      "learning_rate": 2.1112991690472234e-05,
+      "loss": 0.5227,
+      "step": 9733
+    },
+    {
+      "epoch": 1.7364195879047364,
+      "grad_norm": 0.5338386297225952,
+      "learning_rate": 2.1084821023904133e-05,
+      "loss": 0.4744,
+      "step": 9734
+    },
+    {
+      "epoch": 1.7365979841227366,
+      "grad_norm": 1.029298186302185,
+      "learning_rate": 2.1056668336235624e-05,
+      "loss": 0.5359,
+      "step": 9735
+    },
+    {
+      "epoch": 1.7367763803407368,
+      "grad_norm": 0.520614504814148,
+      "learning_rate": 2.1028533629677937e-05,
+      "loss": 0.5175,
+      "step": 9736
+    },
+    {
+      "epoch": 1.736954776558737,
+      "grad_norm": 0.5271267890930176,
+      "learning_rate": 2.1000416906440613e-05,
+      "loss": 0.5389,
+      "step": 9737
+    },
+    {
+      "epoch": 1.7371331727767372,
+      "grad_norm": 0.5082148313522339,
+      "learning_rate": 2.0972318168732048e-05,
+      "loss": 0.4963,
+      "step": 9738
+    },
+    {
+      "epoch": 1.7373115689947372,
+      "grad_norm": 0.42362144589424133,
+      "learning_rate": 2.094423741875903e-05,
+      "loss": 0.4004,
+      "step": 9739
+    },
+    {
+      "epoch": 1.7374899652127374,
+      "grad_norm": 0.5850268602371216,
+      "learning_rate": 2.0916174658727054e-05,
+      "loss": 0.5616,
+      "step": 9740
+    },
+    {
+      "epoch": 1.7376683614307376,
+      "grad_norm": 0.5065614581108093,
+      "learning_rate": 2.0888129890840102e-05,
+      "loss": 0.4137,
+      "step": 9741
+    },
+    {
+      "epoch": 1.7378467576487377,
+      "grad_norm": 0.535125732421875,
+      "learning_rate": 2.0860103117300882e-05,
+      "loss": 0.5246,
+      "step": 9742
+    },
+    {
+      "epoch": 1.738025153866738,
+      "grad_norm": 0.49829819798469543,
+      "learning_rate": 2.0832094340310554e-05,
+      "loss": 0.3795,
+      "step": 9743
+    },
+    {
+      "epoch": 1.7382035500847381,
+      "grad_norm": 0.5445170402526855,
+      "learning_rate": 2.0804103562068883e-05,
+      "loss": 0.5326,
+      "step": 9744
+    },
+    {
+      "epoch": 1.7383819463027383,
+      "grad_norm": 0.5501541495323181,
+      "learning_rate": 2.0776130784774333e-05,
+      "loss": 0.5399,
+      "step": 9745
+    },
+    {
+      "epoch": 1.7385603425207385,
+      "grad_norm": 0.48159241676330566,
+      "learning_rate": 2.0748176010623758e-05,
+      "loss": 0.4521,
+      "step": 9746
+    },
+    {
+      "epoch": 1.7387387387387387,
+      "grad_norm": 0.5003437399864197,
+      "learning_rate": 2.0720239241812848e-05,
+      "loss": 0.4789,
+      "step": 9747
+    },
+    {
+      "epoch": 1.738917134956739,
+      "grad_norm": 0.4868369400501251,
+      "learning_rate": 2.0692320480535624e-05,
+      "loss": 0.3965,
+      "step": 9748
+    },
+    {
+      "epoch": 1.7390955311747391,
+      "grad_norm": 0.5282782316207886,
+      "learning_rate": 2.066441972898489e-05,
+      "loss": 0.4808,
+      "step": 9749
+    },
+    {
+      "epoch": 1.7392739273927393,
+      "grad_norm": 0.4921809434890747,
+      "learning_rate": 2.0636536989351972e-05,
+      "loss": 0.458,
+      "step": 9750
+    },
+    {
+      "epoch": 1.7394523236107395,
+      "grad_norm": 0.5522050857543945,
+      "learning_rate": 2.0608672263826705e-05,
+      "loss": 0.5707,
+      "step": 9751
+    },
+    {
+      "epoch": 1.7396307198287397,
+      "grad_norm": 0.6109182834625244,
+      "learning_rate": 2.0580825554597644e-05,
+      "loss": 0.773,
+      "step": 9752
+    },
+    {
+      "epoch": 1.73980911604674,
+      "grad_norm": 0.5634390115737915,
+      "learning_rate": 2.055299686385176e-05,
+      "loss": 0.6855,
+      "step": 9753
+    },
+    {
+      "epoch": 1.7399875122647401,
+      "grad_norm": 0.47607871890068054,
+      "learning_rate": 2.0525186193774802e-05,
+      "loss": 0.4555,
+      "step": 9754
+    },
+    {
+      "epoch": 1.7401659084827403,
+      "grad_norm": 0.5108838677406311,
+      "learning_rate": 2.0497393546551003e-05,
+      "loss": 0.4981,
+      "step": 9755
+    },
+    {
+      "epoch": 1.7403443047007403,
+      "grad_norm": 0.522916316986084,
+      "learning_rate": 2.046961892436308e-05,
+      "loss": 0.5068,
+      "step": 9756
+    },
+    {
+      "epoch": 1.7405227009187405,
+      "grad_norm": 0.49157410860061646,
+      "learning_rate": 2.0441862329392547e-05,
+      "loss": 0.4636,
+      "step": 9757
+    },
+    {
+      "epoch": 1.7407010971367407,
+      "grad_norm": 0.5237782597541809,
+      "learning_rate": 2.0414123763819348e-05,
+      "loss": 0.4787,
+      "step": 9758
+    },
+    {
+      "epoch": 1.740879493354741,
+      "grad_norm": 0.521073043346405,
+      "learning_rate": 2.0386403229822102e-05,
+      "loss": 0.4384,
+      "step": 9759
+    },
+    {
+      "epoch": 1.741057889572741,
+      "grad_norm": 0.5777454376220703,
+      "learning_rate": 2.035870072957788e-05,
+      "loss": 0.6771,
+      "step": 9760
+    },
+    {
+      "epoch": 1.741236285790741,
+      "grad_norm": 0.5693924427032471,
+      "learning_rate": 2.0331016265262543e-05,
+      "loss": 0.6439,
+      "step": 9761
+    },
+    {
+      "epoch": 1.7414146820087413,
+      "grad_norm": 0.5041927099227905,
+      "learning_rate": 2.030334983905027e-05,
+      "loss": 0.4557,
+      "step": 9762
+    },
+    {
+      "epoch": 1.7415930782267415,
+      "grad_norm": 0.5025537014007568,
+      "learning_rate": 2.0275701453114108e-05,
+      "loss": 0.4614,
+      "step": 9763
+    },
+    {
+      "epoch": 1.7417714744447417,
+      "grad_norm": 0.5347465872764587,
+      "learning_rate": 2.024807110962551e-05,
+      "loss": 0.5502,
+      "step": 9764
+    },
+    {
+      "epoch": 1.7419498706627419,
+      "grad_norm": 0.4626553952693939,
+      "learning_rate": 2.0220458810754488e-05,
+      "loss": 0.3479,
+      "step": 9765
+    },
+    {
+      "epoch": 1.742128266880742,
+      "grad_norm": 0.4479665458202362,
+      "learning_rate": 2.019286455866981e-05,
+      "loss": 0.4113,
+      "step": 9766
+    },
+    {
+      "epoch": 1.7423066630987423,
+      "grad_norm": 0.5242836475372314,
+      "learning_rate": 2.0165288355538656e-05,
+      "loss": 0.522,
+      "step": 9767
+    },
+    {
+      "epoch": 1.7424850593167425,
+      "grad_norm": 0.4955342411994934,
+      "learning_rate": 2.01377302035268e-05,
+      "loss": 0.3913,
+      "step": 9768
+    },
+    {
+      "epoch": 1.7426634555347427,
+      "grad_norm": 0.5188047289848328,
+      "learning_rate": 2.0110190104798727e-05,
+      "loss": 0.4209,
+      "step": 9769
+    },
+    {
+      "epoch": 1.7428418517527429,
+      "grad_norm": 0.47911545634269714,
+      "learning_rate": 2.0082668061517373e-05,
+      "loss": 0.3811,
+      "step": 9770
+    },
+    {
+      "epoch": 1.743020247970743,
+      "grad_norm": 0.5450549125671387,
+      "learning_rate": 2.0055164075844345e-05,
+      "loss": 0.4722,
+      "step": 9771
+    },
+    {
+      "epoch": 1.7431986441887433,
+      "grad_norm": 0.5390350818634033,
+      "learning_rate": 2.0027678149939747e-05,
+      "loss": 0.558,
+      "step": 9772
+    },
+    {
+      "epoch": 1.7433770404067435,
+      "grad_norm": 0.5970989465713501,
+      "learning_rate": 2.0000210285962385e-05,
+      "loss": 0.5316,
+      "step": 9773
+    },
+    {
+      "epoch": 1.7435554366247437,
+      "grad_norm": 0.5830510258674622,
+      "learning_rate": 1.9972760486069498e-05,
+      "loss": 0.6375,
+      "step": 9774
+    },
+    {
+      "epoch": 1.7437338328427439,
+      "grad_norm": 0.5821729898452759,
+      "learning_rate": 1.9945328752417057e-05,
+      "loss": 0.5732,
+      "step": 9775
+    },
+    {
+      "epoch": 1.743912229060744,
+      "grad_norm": 0.5588756203651428,
+      "learning_rate": 1.9917915087159482e-05,
+      "loss": 0.5578,
+      "step": 9776
+    },
+    {
+      "epoch": 1.7440906252787443,
+      "grad_norm": 0.5491783618927002,
+      "learning_rate": 1.9890519492449798e-05,
+      "loss": 0.5814,
+      "step": 9777
+    },
+    {
+      "epoch": 1.7442690214967442,
+      "grad_norm": 0.5677632093429565,
+      "learning_rate": 1.9863141970439758e-05,
+      "loss": 0.5373,
+      "step": 9778
+    },
+    {
+      "epoch": 1.7444474177147444,
+      "grad_norm": 0.4712635576725006,
+      "learning_rate": 1.983578252327939e-05,
+      "loss": 0.3357,
+      "step": 9779
+    },
+    {
+      "epoch": 1.7446258139327446,
+      "grad_norm": 0.5437815189361572,
+      "learning_rate": 1.9808441153117723e-05,
+      "loss": 0.5909,
+      "step": 9780
+    },
+    {
+      "epoch": 1.7448042101507448,
+      "grad_norm": 0.49370652437210083,
+      "learning_rate": 1.9781117862102045e-05,
+      "loss": 0.5412,
+      "step": 9781
+    },
+    {
+      "epoch": 1.744982606368745,
+      "grad_norm": 0.5577530860900879,
+      "learning_rate": 1.9753812652378217e-05,
+      "loss": 0.672,
+      "step": 9782
+    },
+    {
+      "epoch": 1.745161002586745,
+      "grad_norm": 0.5543391108512878,
+      "learning_rate": 1.9726525526090917e-05,
+      "loss": 0.5612,
+      "step": 9783
+    },
+    {
+      "epoch": 1.7453393988047452,
+      "grad_norm": 0.4798511564731598,
+      "learning_rate": 1.9699256485383177e-05,
+      "loss": 0.4455,
+      "step": 9784
+    },
+    {
+      "epoch": 1.7455177950227454,
+      "grad_norm": 0.5388596057891846,
+      "learning_rate": 1.967200553239676e-05,
+      "loss": 0.4951,
+      "step": 9785
+    },
+    {
+      "epoch": 1.7456961912407456,
+      "grad_norm": 0.4370432496070862,
+      "learning_rate": 1.9644772669271894e-05,
+      "loss": 0.3437,
+      "step": 9786
+    },
+    {
+      "epoch": 1.7458745874587458,
+      "grad_norm": 0.5266339182853699,
+      "learning_rate": 1.9617557898147454e-05,
+      "loss": 0.5436,
+      "step": 9787
+    },
+    {
+      "epoch": 1.746052983676746,
+      "grad_norm": 0.4976480305194855,
+      "learning_rate": 1.9590361221160897e-05,
+      "loss": 0.396,
+      "step": 9788
+    },
+    {
+      "epoch": 1.7462313798947462,
+      "grad_norm": 0.5532608032226562,
+      "learning_rate": 1.956318264044818e-05,
+      "loss": 0.5835,
+      "step": 9789
+    },
+    {
+      "epoch": 1.7464097761127464,
+      "grad_norm": 0.45472416281700134,
+      "learning_rate": 1.9536022158143956e-05,
+      "loss": 0.3592,
+      "step": 9790
+    },
+    {
+      "epoch": 1.7465881723307466,
+      "grad_norm": 0.5193656086921692,
+      "learning_rate": 1.9508879776381355e-05,
+      "loss": 0.4665,
+      "step": 9791
+    },
+    {
+      "epoch": 1.7467665685487468,
+      "grad_norm": 0.5724708437919617,
+      "learning_rate": 1.948175549729217e-05,
+      "loss": 0.5152,
+      "step": 9792
+    },
+    {
+      "epoch": 1.746944964766747,
+      "grad_norm": 0.5208200216293335,
+      "learning_rate": 1.945464932300667e-05,
+      "loss": 0.6308,
+      "step": 9793
+    },
+    {
+      "epoch": 1.7471233609847472,
+      "grad_norm": 0.4895492494106293,
+      "learning_rate": 1.9427561255653815e-05,
+      "loss": 0.5369,
+      "step": 9794
+    },
+    {
+      "epoch": 1.7473017572027474,
+      "grad_norm": 0.46594980359077454,
+      "learning_rate": 1.94004912973611e-05,
+      "loss": 0.4012,
+      "step": 9795
+    },
+    {
+      "epoch": 1.7474801534207476,
+      "grad_norm": 0.5352393984794617,
+      "learning_rate": 1.937343945025455e-05,
+      "loss": 0.5697,
+      "step": 9796
+    },
+    {
+      "epoch": 1.7476585496387478,
+      "grad_norm": 0.5204436182975769,
+      "learning_rate": 1.934640571645882e-05,
+      "loss": 0.564,
+      "step": 9797
+    },
+    {
+      "epoch": 1.747836945856748,
+      "grad_norm": 0.5590188503265381,
+      "learning_rate": 1.9319390098097108e-05,
+      "loss": 0.5922,
+      "step": 9798
+    },
+    {
+      "epoch": 1.7480153420747482,
+      "grad_norm": 0.5261103510856628,
+      "learning_rate": 1.9292392597291293e-05,
+      "loss": 0.3612,
+      "step": 9799
+    },
+    {
+      "epoch": 1.7481937382927482,
+      "grad_norm": 0.4883411228656769,
+      "learning_rate": 1.9265413216161598e-05,
+      "loss": 0.4208,
+      "step": 9800
+    },
+    {
+      "epoch": 1.7483721345107484,
+      "grad_norm": 0.5407469272613525,
+      "learning_rate": 1.9238451956827135e-05,
+      "loss": 0.5281,
+      "step": 9801
+    },
+    {
+      "epoch": 1.7485505307287486,
+      "grad_norm": 0.5203365087509155,
+      "learning_rate": 1.921150882140532e-05,
+      "loss": 0.4411,
+      "step": 9802
+    },
+    {
+      "epoch": 1.7487289269467488,
+      "grad_norm": 0.5493040084838867,
+      "learning_rate": 1.9184583812012268e-05,
+      "loss": 0.5184,
+      "step": 9803
+    },
+    {
+      "epoch": 1.748907323164749,
+      "grad_norm": 0.41875502467155457,
+      "learning_rate": 1.9157676930762702e-05,
+      "loss": 0.3174,
+      "step": 9804
+    },
+    {
+      "epoch": 1.749085719382749,
+      "grad_norm": 0.5331047177314758,
+      "learning_rate": 1.913078817976982e-05,
+      "loss": 0.5325,
+      "step": 9805
+    },
+    {
+      "epoch": 1.7492641156007491,
+      "grad_norm": 0.5197538137435913,
+      "learning_rate": 1.9103917561145516e-05,
+      "loss": 0.4778,
+      "step": 9806
+    },
+    {
+      "epoch": 1.7494425118187493,
+      "grad_norm": 0.5139304995536804,
+      "learning_rate": 1.9077065077000104e-05,
+      "loss": 0.5117,
+      "step": 9807
+    },
+    {
+      "epoch": 1.7496209080367495,
+      "grad_norm": 0.540160596370697,
+      "learning_rate": 1.9050230729442702e-05,
+      "loss": 0.4415,
+      "step": 9808
+    },
+    {
+      "epoch": 1.7497993042547497,
+      "grad_norm": 0.5008589029312134,
+      "learning_rate": 1.9023414520580733e-05,
+      "loss": 0.4228,
+      "step": 9809
+    },
+    {
+      "epoch": 1.74997770047275,
+      "grad_norm": 0.5717454552650452,
+      "learning_rate": 1.8996616452520404e-05,
+      "loss": 0.5889,
+      "step": 9810
+    },
+    {
+      "epoch": 1.7501560966907501,
+      "grad_norm": 0.5014664530754089,
+      "learning_rate": 1.8969836527366448e-05,
+      "loss": 0.4555,
+      "step": 9811
+    },
+    {
+      "epoch": 1.7503344929087503,
+      "grad_norm": 0.5449042320251465,
+      "learning_rate": 1.8943074747222068e-05,
+      "loss": 0.5205,
+      "step": 9812
+    },
+    {
+      "epoch": 1.7505128891267505,
+      "grad_norm": 0.49682438373565674,
+      "learning_rate": 1.8916331114189195e-05,
+      "loss": 0.4651,
+      "step": 9813
+    },
+    {
+      "epoch": 1.7506912853447507,
+      "grad_norm": 0.5355203151702881,
+      "learning_rate": 1.8889605630368235e-05,
+      "loss": 0.5275,
+      "step": 9814
+    },
+    {
+      "epoch": 1.750869681562751,
+      "grad_norm": 0.5955283045768738,
+      "learning_rate": 1.8862898297858173e-05,
+      "loss": 0.6346,
+      "step": 9815
+    },
+    {
+      "epoch": 1.751048077780751,
+      "grad_norm": 0.5433852672576904,
+      "learning_rate": 1.8836209118756637e-05,
+      "loss": 0.5114,
+      "step": 9816
+    },
+    {
+      "epoch": 1.7512264739987513,
+      "grad_norm": 0.5149011611938477,
+      "learning_rate": 1.8809538095159727e-05,
+      "loss": 0.4793,
+      "step": 9817
+    },
+    {
+      "epoch": 1.7514048702167515,
+      "grad_norm": 0.4862879514694214,
+      "learning_rate": 1.8782885229162245e-05,
+      "loss": 0.3931,
+      "step": 9818
+    },
+    {
+      "epoch": 1.7515832664347517,
+      "grad_norm": 0.6665529012680054,
+      "learning_rate": 1.8756250522857397e-05,
+      "loss": 0.8526,
+      "step": 9819
+    },
+    {
+      "epoch": 1.751761662652752,
+      "grad_norm": 0.5120258927345276,
+      "learning_rate": 1.8729633978337184e-05,
+      "loss": 0.5787,
+      "step": 9820
+    },
+    {
+      "epoch": 1.751940058870752,
+      "grad_norm": 0.5091425776481628,
+      "learning_rate": 1.870303559769196e-05,
+      "loss": 0.4539,
+      "step": 9821
+    },
+    {
+      "epoch": 1.752118455088752,
+      "grad_norm": 0.5244202017784119,
+      "learning_rate": 1.86764553830108e-05,
+      "loss": 0.5742,
+      "step": 9822
+    },
+    {
+      "epoch": 1.7522968513067523,
+      "grad_norm": 0.4686015546321869,
+      "learning_rate": 1.8649893336381313e-05,
+      "loss": 0.4521,
+      "step": 9823
+    },
+    {
+      "epoch": 1.7524752475247525,
+      "grad_norm": 0.5084050893783569,
+      "learning_rate": 1.8623349459889582e-05,
+      "loss": 0.487,
+      "step": 9824
+    },
+    {
+      "epoch": 1.7526536437427527,
+      "grad_norm": 0.4738958477973938,
+      "learning_rate": 1.859682375562044e-05,
+      "loss": 0.3566,
+      "step": 9825
+    },
+    {
+      "epoch": 1.7528320399607529,
+      "grad_norm": 0.5699481964111328,
+      "learning_rate": 1.857031622565722e-05,
+      "loss": 0.5028,
+      "step": 9826
+    },
+    {
+      "epoch": 1.7530104361787529,
+      "grad_norm": 0.4751138985157013,
+      "learning_rate": 1.85438268720817e-05,
+      "loss": 0.3851,
+      "step": 9827
+    },
+    {
+      "epoch": 1.753188832396753,
+      "grad_norm": 0.5180573463439941,
+      "learning_rate": 1.851735569697449e-05,
+      "loss": 0.4933,
+      "step": 9828
+    },
+    {
+      "epoch": 1.7533672286147532,
+      "grad_norm": 0.5232348442077637,
+      "learning_rate": 1.849090270241449e-05,
+      "loss": 0.4913,
+      "step": 9829
+    },
+    {
+      "epoch": 1.7535456248327534,
+      "grad_norm": 0.43816691637039185,
+      "learning_rate": 1.8464467890479397e-05,
+      "loss": 0.3977,
+      "step": 9830
+    },
+    {
+      "epoch": 1.7537240210507536,
+      "grad_norm": 0.45382770895957947,
+      "learning_rate": 1.8438051263245326e-05,
+      "loss": 0.3,
+      "step": 9831
+    },
+    {
+      "epoch": 1.7539024172687538,
+      "grad_norm": 0.5431753396987915,
+      "learning_rate": 1.8411652822787118e-05,
+      "loss": 0.6459,
+      "step": 9832
+    },
+    {
+      "epoch": 1.754080813486754,
+      "grad_norm": 0.4964623749256134,
+      "learning_rate": 1.8385272571177974e-05,
+      "loss": 0.5321,
+      "step": 9833
+    },
+    {
+      "epoch": 1.7542592097047542,
+      "grad_norm": 0.5045402646064758,
+      "learning_rate": 1.8358910510489907e-05,
+      "loss": 0.4846,
+      "step": 9834
+    },
+    {
+      "epoch": 1.7544376059227544,
+      "grad_norm": 0.4275517761707306,
+      "learning_rate": 1.8332566642793312e-05,
+      "loss": 0.3162,
+      "step": 9835
+    },
+    {
+      "epoch": 1.7546160021407546,
+      "grad_norm": 0.5664304494857788,
+      "learning_rate": 1.8306240970157206e-05,
+      "loss": 0.5896,
+      "step": 9836
+    },
+    {
+      "epoch": 1.7547943983587548,
+      "grad_norm": 0.5230024456977844,
+      "learning_rate": 1.8279933494649265e-05,
+      "loss": 0.414,
+      "step": 9837
+    },
+    {
+      "epoch": 1.754972794576755,
+      "grad_norm": 0.5773944854736328,
+      "learning_rate": 1.8253644218335584e-05,
+      "loss": 0.557,
+      "step": 9838
+    },
+    {
+      "epoch": 1.7551511907947552,
+      "grad_norm": 0.5354689359664917,
+      "learning_rate": 1.8227373143281017e-05,
+      "loss": 0.516,
+      "step": 9839
+    },
+    {
+      "epoch": 1.7553295870127554,
+      "grad_norm": 0.47697752714157104,
+      "learning_rate": 1.820112027154877e-05,
+      "loss": 0.4222,
+      "step": 9840
+    },
+    {
+      "epoch": 1.7555079832307556,
+      "grad_norm": 0.5502009391784668,
+      "learning_rate": 1.8174885605200782e-05,
+      "loss": 0.4705,
+      "step": 9841
+    },
+    {
+      "epoch": 1.7556863794487558,
+      "grad_norm": 0.4850895404815674,
+      "learning_rate": 1.8148669146297565e-05,
+      "loss": 0.4318,
+      "step": 9842
+    },
+    {
+      "epoch": 1.755864775666756,
+      "grad_norm": 0.5271044969558716,
+      "learning_rate": 1.8122470896898057e-05,
+      "loss": 0.4357,
+      "step": 9843
+    },
+    {
+      "epoch": 1.756043171884756,
+      "grad_norm": 0.5360130071640015,
+      "learning_rate": 1.809629085905992e-05,
+      "loss": 0.4743,
+      "step": 9844
+    },
+    {
+      "epoch": 1.7562215681027562,
+      "grad_norm": 0.5907223224639893,
+      "learning_rate": 1.807012903483929e-05,
+      "loss": 0.4996,
+      "step": 9845
+    },
+    {
+      "epoch": 1.7563999643207564,
+      "grad_norm": 0.5198745131492615,
+      "learning_rate": 1.8043985426290958e-05,
+      "loss": 0.5376,
+      "step": 9846
+    },
+    {
+      "epoch": 1.7565783605387566,
+      "grad_norm": 0.7061692476272583,
+      "learning_rate": 1.801786003546818e-05,
+      "loss": 0.5927,
+      "step": 9847
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.4907400906085968,
+      "learning_rate": 1.799175286442281e-05,
+      "loss": 0.4041,
+      "step": 9848
+    },
+    {
+      "epoch": 1.7569351529747568,
+      "grad_norm": 0.4804750382900238,
+      "learning_rate": 1.7965663915205376e-05,
+      "loss": 0.4552,
+      "step": 9849
+    },
+    {
+      "epoch": 1.757113549192757,
+      "grad_norm": 0.5161197781562805,
+      "learning_rate": 1.7939593189864794e-05,
+      "loss": 0.4615,
+      "step": 9850
+    },
+    {
+      "epoch": 1.7572919454107572,
+      "grad_norm": 0.5448468327522278,
+      "learning_rate": 1.791354069044876e-05,
+      "loss": 0.4771,
+      "step": 9851
+    },
+    {
+      "epoch": 1.7574703416287574,
+      "grad_norm": 0.5089532136917114,
+      "learning_rate": 1.7887506419003303e-05,
+      "loss": 0.4607,
+      "step": 9852
+    },
+    {
+      "epoch": 1.7576487378467576,
+      "grad_norm": 0.49231788516044617,
+      "learning_rate": 1.786149037757326e-05,
+      "loss": 0.5082,
+      "step": 9853
+    },
+    {
+      "epoch": 1.7578271340647578,
+      "grad_norm": 0.5540265440940857,
+      "learning_rate": 1.78354925682018e-05,
+      "loss": 0.5348,
+      "step": 9854
+    },
+    {
+      "epoch": 1.758005530282758,
+      "grad_norm": 0.6183173060417175,
+      "learning_rate": 1.7809512992930875e-05,
+      "loss": 0.6396,
+      "step": 9855
+    },
+    {
+      "epoch": 1.7581839265007582,
+      "grad_norm": 0.4565494656562805,
+      "learning_rate": 1.778355165380091e-05,
+      "loss": 0.407,
+      "step": 9856
+    },
+    {
+      "epoch": 1.7583623227187584,
+      "grad_norm": 0.5190179347991943,
+      "learning_rate": 1.7757608552850828e-05,
+      "loss": 0.499,
+      "step": 9857
+    },
+    {
+      "epoch": 1.7585407189367586,
+      "grad_norm": 0.5426142811775208,
+      "learning_rate": 1.7731683692118277e-05,
+      "loss": 0.5268,
+      "step": 9858
+    },
+    {
+      "epoch": 1.7587191151547588,
+      "grad_norm": 0.5767265558242798,
+      "learning_rate": 1.770577707363927e-05,
+      "loss": 0.4803,
+      "step": 9859
+    },
+    {
+      "epoch": 1.758897511372759,
+      "grad_norm": 0.5167482495307922,
+      "learning_rate": 1.7679888699448644e-05,
+      "loss": 0.5015,
+      "step": 9860
+    },
+    {
+      "epoch": 1.7590759075907592,
+      "grad_norm": 0.4636717736721039,
+      "learning_rate": 1.7654018571579554e-05,
+      "loss": 0.3958,
+      "step": 9861
+    },
+    {
+      "epoch": 1.7592543038087594,
+      "grad_norm": 0.504102349281311,
+      "learning_rate": 1.7628166692063823e-05,
+      "loss": 0.3926,
+      "step": 9862
+    },
+    {
+      "epoch": 1.7594327000267596,
+      "grad_norm": 0.5384790301322937,
+      "learning_rate": 1.7602333062931935e-05,
+      "loss": 0.4652,
+      "step": 9863
+    },
+    {
+      "epoch": 1.7596110962447598,
+      "grad_norm": 0.48373332619667053,
+      "learning_rate": 1.757651768621274e-05,
+      "loss": 0.3906,
+      "step": 9864
+    },
+    {
+      "epoch": 1.75978949246276,
+      "grad_norm": 0.4555506110191345,
+      "learning_rate": 1.755072056393389e-05,
+      "loss": 0.3741,
+      "step": 9865
+    },
+    {
+      "epoch": 1.75996788868076,
+      "grad_norm": 0.6488031148910522,
+      "learning_rate": 1.7524941698121354e-05,
+      "loss": 0.7711,
+      "step": 9866
+    },
+    {
+      "epoch": 1.7601462848987601,
+      "grad_norm": 0.5170810222625732,
+      "learning_rate": 1.7499181090799928e-05,
+      "loss": 0.4957,
+      "step": 9867
+    },
+    {
+      "epoch": 1.7603246811167603,
+      "grad_norm": 0.5525889992713928,
+      "learning_rate": 1.7473438743992736e-05,
+      "loss": 0.6491,
+      "step": 9868
+    },
+    {
+      "epoch": 1.7605030773347605,
+      "grad_norm": 0.4711996018886566,
+      "learning_rate": 1.7447714659721586e-05,
+      "loss": 0.4518,
+      "step": 9869
+    },
+    {
+      "epoch": 1.7606814735527607,
+      "grad_norm": 0.43584319949150085,
+      "learning_rate": 1.742200884000686e-05,
+      "loss": 0.3699,
+      "step": 9870
+    },
+    {
+      "epoch": 1.7608598697707607,
+      "grad_norm": 0.49232321977615356,
+      "learning_rate": 1.7396321286867412e-05,
+      "loss": 0.4172,
+      "step": 9871
+    },
+    {
+      "epoch": 1.761038265988761,
+      "grad_norm": 0.6322119235992432,
+      "learning_rate": 1.737065200232088e-05,
+      "loss": 0.5614,
+      "step": 9872
+    },
+    {
+      "epoch": 1.761216662206761,
+      "grad_norm": 0.46853572130203247,
+      "learning_rate": 1.7345000988383208e-05,
+      "loss": 0.3119,
+      "step": 9873
+    },
+    {
+      "epoch": 1.7613950584247613,
+      "grad_norm": 0.5115934610366821,
+      "learning_rate": 1.7319368247069005e-05,
+      "loss": 0.5075,
+      "step": 9874
+    },
+    {
+      "epoch": 1.7615734546427615,
+      "grad_norm": 0.524381697177887,
+      "learning_rate": 1.729375378039155e-05,
+      "loss": 0.4831,
+      "step": 9875
+    },
+    {
+      "epoch": 1.7617518508607617,
+      "grad_norm": 0.5672167539596558,
+      "learning_rate": 1.7268157590362487e-05,
+      "loss": 0.5661,
+      "step": 9876
+    },
+    {
+      "epoch": 1.761930247078762,
+      "grad_norm": 0.5367761850357056,
+      "learning_rate": 1.7242579678992204e-05,
+      "loss": 0.5691,
+      "step": 9877
+    },
+    {
+      "epoch": 1.762108643296762,
+      "grad_norm": 0.5382869243621826,
+      "learning_rate": 1.721702004828951e-05,
+      "loss": 0.4331,
+      "step": 9878
+    },
+    {
+      "epoch": 1.7622870395147623,
+      "grad_norm": 0.4912665784358978,
+      "learning_rate": 1.719147870026194e-05,
+      "loss": 0.4589,
+      "step": 9879
+    },
+    {
+      "epoch": 1.7624654357327625,
+      "grad_norm": 0.45093002915382385,
+      "learning_rate": 1.716595563691539e-05,
+      "loss": 0.3703,
+      "step": 9880
+    },
+    {
+      "epoch": 1.7626438319507627,
+      "grad_norm": 0.614578127861023,
+      "learning_rate": 1.7140450860254535e-05,
+      "loss": 0.5906,
+      "step": 9881
+    },
+    {
+      "epoch": 1.7628222281687629,
+      "grad_norm": 0.45444604754447937,
+      "learning_rate": 1.7114964372282466e-05,
+      "loss": 0.4004,
+      "step": 9882
+    },
+    {
+      "epoch": 1.763000624386763,
+      "grad_norm": 0.5366213321685791,
+      "learning_rate": 1.708949617500083e-05,
+      "loss": 0.4764,
+      "step": 9883
+    },
+    {
+      "epoch": 1.7631790206047633,
+      "grad_norm": 0.48068365454673767,
+      "learning_rate": 1.7064046270409973e-05,
+      "loss": 0.447,
+      "step": 9884
+    },
+    {
+      "epoch": 1.7633574168227635,
+      "grad_norm": 0.5525119304656982,
+      "learning_rate": 1.7038614660508657e-05,
+      "loss": 0.6492,
+      "step": 9885
+    },
+    {
+      "epoch": 1.7635358130407637,
+      "grad_norm": 0.5292766094207764,
+      "learning_rate": 1.7013201347294284e-05,
+      "loss": 0.4177,
+      "step": 9886
+    },
+    {
+      "epoch": 1.7637142092587639,
+      "grad_norm": 0.48070111870765686,
+      "learning_rate": 1.698780633276284e-05,
+      "loss": 0.3513,
+      "step": 9887
+    },
+    {
+      "epoch": 1.7638926054767639,
+      "grad_norm": 0.4753524363040924,
+      "learning_rate": 1.6962429618908785e-05,
+      "loss": 0.4376,
+      "step": 9888
+    },
+    {
+      "epoch": 1.764071001694764,
+      "grad_norm": 0.5445675849914551,
+      "learning_rate": 1.693707120772528e-05,
+      "loss": 0.5026,
+      "step": 9889
+    },
+    {
+      "epoch": 1.7642493979127643,
+      "grad_norm": 0.5858935713768005,
+      "learning_rate": 1.6911731101203863e-05,
+      "loss": 0.7673,
+      "step": 9890
+    },
+    {
+      "epoch": 1.7644277941307644,
+      "grad_norm": 0.5799805521965027,
+      "learning_rate": 1.688640930133481e-05,
+      "loss": 0.6229,
+      "step": 9891
+    },
+    {
+      "epoch": 1.7646061903487646,
+      "grad_norm": 0.5168468952178955,
+      "learning_rate": 1.6861105810106804e-05,
+      "loss": 0.4652,
+      "step": 9892
+    },
+    {
+      "epoch": 1.7647845865667646,
+      "grad_norm": 0.5596296787261963,
+      "learning_rate": 1.683582062950728e-05,
+      "loss": 0.6326,
+      "step": 9893
+    },
+    {
+      "epoch": 1.7649629827847648,
+      "grad_norm": 0.5667998194694519,
+      "learning_rate": 1.6810553761522047e-05,
+      "loss": 0.5322,
+      "step": 9894
+    },
+    {
+      "epoch": 1.765141379002765,
+      "grad_norm": 0.5282207131385803,
+      "learning_rate": 1.6785305208135538e-05,
+      "loss": 0.4871,
+      "step": 9895
+    },
+    {
+      "epoch": 1.7653197752207652,
+      "grad_norm": 0.5011943578720093,
+      "learning_rate": 1.6760074971330863e-05,
+      "loss": 0.4561,
+      "step": 9896
+    },
+    {
+      "epoch": 1.7654981714387654,
+      "grad_norm": 0.49850714206695557,
+      "learning_rate": 1.6734863053089467e-05,
+      "loss": 0.4113,
+      "step": 9897
+    },
+    {
+      "epoch": 1.7656765676567656,
+      "grad_norm": 0.5289611220359802,
+      "learning_rate": 1.670966945539157e-05,
+      "loss": 0.5378,
+      "step": 9898
+    },
+    {
+      "epoch": 1.7658549638747658,
+      "grad_norm": 0.5059836506843567,
+      "learning_rate": 1.6684494180215837e-05,
+      "loss": 0.4601,
+      "step": 9899
+    },
+    {
+      "epoch": 1.766033360092766,
+      "grad_norm": 0.577877402305603,
+      "learning_rate": 1.6659337229539525e-05,
+      "loss": 0.6335,
+      "step": 9900
+    },
+    {
+      "epoch": 1.7662117563107662,
+      "grad_norm": 0.454922080039978,
+      "learning_rate": 1.6634198605338437e-05,
+      "loss": 0.3668,
+      "step": 9901
+    },
+    {
+      "epoch": 1.7663901525287664,
+      "grad_norm": 0.5340895652770996,
+      "learning_rate": 1.6609078309586967e-05,
+      "loss": 0.4897,
+      "step": 9902
+    },
+    {
+      "epoch": 1.7665685487467666,
+      "grad_norm": 0.5871651768684387,
+      "learning_rate": 1.6583976344258097e-05,
+      "loss": 0.7127,
+      "step": 9903
+    },
+    {
+      "epoch": 1.7667469449647668,
+      "grad_norm": 0.48758915066719055,
+      "learning_rate": 1.6558892711323215e-05,
+      "loss": 0.4231,
+      "step": 9904
+    },
+    {
+      "epoch": 1.766925341182767,
+      "grad_norm": 0.5873585343360901,
+      "learning_rate": 1.65338274127525e-05,
+      "loss": 0.6313,
+      "step": 9905
+    },
+    {
+      "epoch": 1.7671037374007672,
+      "grad_norm": 0.5112351775169373,
+      "learning_rate": 1.6508780450514516e-05,
+      "loss": 0.4986,
+      "step": 9906
+    },
+    {
+      "epoch": 1.7672821336187674,
+      "grad_norm": 0.5402355194091797,
+      "learning_rate": 1.6483751826576382e-05,
+      "loss": 0.6041,
+      "step": 9907
+    },
+    {
+      "epoch": 1.7674605298367676,
+      "grad_norm": 0.925947368144989,
+      "learning_rate": 1.6458741542903942e-05,
+      "loss": 0.4839,
+      "step": 9908
+    },
+    {
+      "epoch": 1.7676389260547678,
+      "grad_norm": 0.5026435256004333,
+      "learning_rate": 1.6433749601461378e-05,
+      "loss": 0.4032,
+      "step": 9909
+    },
+    {
+      "epoch": 1.7678173222727678,
+      "grad_norm": 0.5951597690582275,
+      "learning_rate": 1.6408776004211674e-05,
+      "loss": 0.5571,
+      "step": 9910
+    },
+    {
+      "epoch": 1.767995718490768,
+      "grad_norm": 0.4633539915084839,
+      "learning_rate": 1.6383820753116118e-05,
+      "loss": 0.4171,
+      "step": 9911
+    },
+    {
+      "epoch": 1.7681741147087682,
+      "grad_norm": 0.5063800811767578,
+      "learning_rate": 1.6358883850134816e-05,
+      "loss": 0.3949,
+      "step": 9912
+    },
+    {
+      "epoch": 1.7683525109267684,
+      "grad_norm": 0.5754553079605103,
+      "learning_rate": 1.6333965297226166e-05,
+      "loss": 0.6068,
+      "step": 9913
+    },
+    {
+      "epoch": 1.7685309071447686,
+      "grad_norm": 0.477913498878479,
+      "learning_rate": 1.6309065096347386e-05,
+      "loss": 0.4111,
+      "step": 9914
+    },
+    {
+      "epoch": 1.7687093033627685,
+      "grad_norm": 0.49996158480644226,
+      "learning_rate": 1.6284183249454048e-05,
+      "loss": 0.3918,
+      "step": 9915
+    },
+    {
+      "epoch": 1.7688876995807687,
+      "grad_norm": 0.6070882678031921,
+      "learning_rate": 1.6259319758500312e-05,
+      "loss": 0.6542,
+      "step": 9916
+    },
+    {
+      "epoch": 1.769066095798769,
+      "grad_norm": 0.5135297179222107,
+      "learning_rate": 1.6234474625439117e-05,
+      "loss": 0.4433,
+      "step": 9917
+    },
+    {
+      "epoch": 1.7692444920167691,
+      "grad_norm": 0.48473936319351196,
+      "learning_rate": 1.620964785222162e-05,
+      "loss": 0.5045,
+      "step": 9918
+    },
+    {
+      "epoch": 1.7694228882347693,
+      "grad_norm": 0.4841310679912567,
+      "learning_rate": 1.618483944079782e-05,
+      "loss": 0.3964,
+      "step": 9919
+    },
+    {
+      "epoch": 1.7696012844527695,
+      "grad_norm": 0.54714035987854,
+      "learning_rate": 1.6160049393116104e-05,
+      "loss": 0.5497,
+      "step": 9920
+    },
+    {
+      "epoch": 1.7697796806707697,
+      "grad_norm": 0.49543190002441406,
+      "learning_rate": 1.6135277711123443e-05,
+      "loss": 0.5597,
+      "step": 9921
+    },
+    {
+      "epoch": 1.76995807688877,
+      "grad_norm": 0.5189563035964966,
+      "learning_rate": 1.6110524396765496e-05,
+      "loss": 0.4857,
+      "step": 9922
+    },
+    {
+      "epoch": 1.7701364731067701,
+      "grad_norm": 0.5028755068778992,
+      "learning_rate": 1.6085789451986245e-05,
+      "loss": 0.4835,
+      "step": 9923
+    },
+    {
+      "epoch": 1.7703148693247703,
+      "grad_norm": 0.5071883201599121,
+      "learning_rate": 1.606107287872846e-05,
+      "loss": 0.5195,
+      "step": 9924
+    },
+    {
+      "epoch": 1.7704932655427705,
+      "grad_norm": 0.5900061130523682,
+      "learning_rate": 1.603637467893332e-05,
+      "loss": 0.6589,
+      "step": 9925
+    },
+    {
+      "epoch": 1.7706716617607707,
+      "grad_norm": 0.644641637802124,
+      "learning_rate": 1.6011694854540683e-05,
+      "loss": 0.4762,
+      "step": 9926
+    },
+    {
+      "epoch": 1.770850057978771,
+      "grad_norm": 0.543786883354187,
+      "learning_rate": 1.5987033407488806e-05,
+      "loss": 0.5394,
+      "step": 9927
+    },
+    {
+      "epoch": 1.7710284541967711,
+      "grad_norm": 0.5951870083808899,
+      "learning_rate": 1.5962390339714613e-05,
+      "loss": 0.6289,
+      "step": 9928
+    },
+    {
+      "epoch": 1.7712068504147713,
+      "grad_norm": 0.5071197748184204,
+      "learning_rate": 1.593776565315358e-05,
+      "loss": 0.4177,
+      "step": 9929
+    },
+    {
+      "epoch": 1.7713852466327715,
+      "grad_norm": 0.4966924786567688,
+      "learning_rate": 1.591315934973969e-05,
+      "loss": 0.4219,
+      "step": 9930
+    },
+    {
+      "epoch": 1.7715636428507717,
+      "grad_norm": 0.4919562339782715,
+      "learning_rate": 1.588857143140554e-05,
+      "loss": 0.4038,
+      "step": 9931
+    },
+    {
+      "epoch": 1.7717420390687717,
+      "grad_norm": 0.59256911277771,
+      "learning_rate": 1.5864001900082247e-05,
+      "loss": 0.5511,
+      "step": 9932
+    },
+    {
+      "epoch": 1.771920435286772,
+      "grad_norm": 0.5095023512840271,
+      "learning_rate": 1.5839450757699465e-05,
+      "loss": 0.5277,
+      "step": 9933
+    },
+    {
+      "epoch": 1.772098831504772,
+      "grad_norm": 0.5355534553527832,
+      "learning_rate": 1.581491800618548e-05,
+      "loss": 0.4272,
+      "step": 9934
+    },
+    {
+      "epoch": 1.7722772277227723,
+      "grad_norm": 0.5533571839332581,
+      "learning_rate": 1.5790403647467033e-05,
+      "loss": 0.5586,
+      "step": 9935
+    },
+    {
+      "epoch": 1.7724556239407725,
+      "grad_norm": 0.45834532380104065,
+      "learning_rate": 1.5765907683469527e-05,
+      "loss": 0.4001,
+      "step": 9936
+    },
+    {
+      "epoch": 1.7726340201587725,
+      "grad_norm": 0.5351871848106384,
+      "learning_rate": 1.5741430116116813e-05,
+      "loss": 0.5495,
+      "step": 9937
+    },
+    {
+      "epoch": 1.7728124163767727,
+      "grad_norm": 0.48931190371513367,
+      "learning_rate": 1.5716970947331376e-05,
+      "loss": 0.4277,
+      "step": 9938
+    },
+    {
+      "epoch": 1.7729908125947729,
+      "grad_norm": 0.48030897974967957,
+      "learning_rate": 1.5692530179034215e-05,
+      "loss": 0.4325,
+      "step": 9939
+    },
+    {
+      "epoch": 1.773169208812773,
+      "grad_norm": 0.5139449238777161,
+      "learning_rate": 1.5668107813144927e-05,
+      "loss": 0.5081,
+      "step": 9940
+    },
+    {
+      "epoch": 1.7733476050307733,
+      "grad_norm": 0.5078691840171814,
+      "learning_rate": 1.564370385158159e-05,
+      "loss": 0.3923,
+      "step": 9941
+    },
+    {
+      "epoch": 1.7735260012487735,
+      "grad_norm": 0.4742141664028168,
+      "learning_rate": 1.5619318296260897e-05,
+      "loss": 0.4197,
+      "step": 9942
+    },
+    {
+      "epoch": 1.7737043974667737,
+      "grad_norm": 0.5286515355110168,
+      "learning_rate": 1.5594951149098092e-05,
+      "loss": 0.4823,
+      "step": 9943
+    },
+    {
+      "epoch": 1.7738827936847739,
+      "grad_norm": 0.481815904378891,
+      "learning_rate": 1.5570602412006944e-05,
+      "loss": 0.4204,
+      "step": 9944
+    },
+    {
+      "epoch": 1.774061189902774,
+      "grad_norm": 0.6102718114852905,
+      "learning_rate": 1.554627208689982e-05,
+      "loss": 0.6736,
+      "step": 9945
+    },
+    {
+      "epoch": 1.7742395861207743,
+      "grad_norm": 0.534480094909668,
+      "learning_rate": 1.552196017568755e-05,
+      "loss": 0.4361,
+      "step": 9946
+    },
+    {
+      "epoch": 1.7744179823387745,
+      "grad_norm": 0.4984724521636963,
+      "learning_rate": 1.549766668027963e-05,
+      "loss": 0.4064,
+      "step": 9947
+    },
+    {
+      "epoch": 1.7745963785567747,
+      "grad_norm": 0.5009212493896484,
+      "learning_rate": 1.5473391602584096e-05,
+      "loss": 0.437,
+      "step": 9948
+    },
+    {
+      "epoch": 1.7747747747747749,
+      "grad_norm": 0.5126849412918091,
+      "learning_rate": 1.544913494450742e-05,
+      "loss": 0.4619,
+      "step": 9949
+    },
+    {
+      "epoch": 1.774953170992775,
+      "grad_norm": 0.5047542452812195,
+      "learning_rate": 1.5424896707954773e-05,
+      "loss": 0.451,
+      "step": 9950
+    },
+    {
+      "epoch": 1.7751315672107753,
+      "grad_norm": 0.5584752559661865,
+      "learning_rate": 1.5400676894829767e-05,
+      "loss": 0.5518,
+      "step": 9951
+    },
+    {
+      "epoch": 1.7753099634287755,
+      "grad_norm": 0.5159489512443542,
+      "learning_rate": 1.5376475507034694e-05,
+      "loss": 0.4518,
+      "step": 9952
+    },
+    {
+      "epoch": 1.7754883596467756,
+      "grad_norm": 0.589447021484375,
+      "learning_rate": 1.535229254647025e-05,
+      "loss": 0.5093,
+      "step": 9953
+    },
+    {
+      "epoch": 1.7756667558647756,
+      "grad_norm": 0.5014997720718384,
+      "learning_rate": 1.5328128015035746e-05,
+      "loss": 0.3893,
+      "step": 9954
+    },
+    {
+      "epoch": 1.7758451520827758,
+      "grad_norm": 0.5109811425209045,
+      "learning_rate": 1.5303981914629117e-05,
+      "loss": 0.4326,
+      "step": 9955
+    },
+    {
+      "epoch": 1.776023548300776,
+      "grad_norm": 0.5431187152862549,
+      "learning_rate": 1.52798542471467e-05,
+      "loss": 0.4706,
+      "step": 9956
+    },
+    {
+      "epoch": 1.7762019445187762,
+      "grad_norm": 0.6107049584388733,
+      "learning_rate": 1.5255745014483569e-05,
+      "loss": 0.4079,
+      "step": 9957
+    },
+    {
+      "epoch": 1.7763803407367764,
+      "grad_norm": 0.5987387299537659,
+      "learning_rate": 1.5231654218533175e-05,
+      "loss": 0.5319,
+      "step": 9958
+    },
+    {
+      "epoch": 1.7765587369547764,
+      "grad_norm": 0.5544418096542358,
+      "learning_rate": 1.5207581861187647e-05,
+      "loss": 0.4959,
+      "step": 9959
+    },
+    {
+      "epoch": 1.7767371331727766,
+      "grad_norm": 0.6012864112854004,
+      "learning_rate": 1.5183527944337583e-05,
+      "loss": 0.5722,
+      "step": 9960
+    },
+    {
+      "epoch": 1.7769155293907768,
+      "grad_norm": 0.5390208959579468,
+      "learning_rate": 1.5159492469872221e-05,
+      "loss": 0.6233,
+      "step": 9961
+    },
+    {
+      "epoch": 1.777093925608777,
+      "grad_norm": 0.5232547521591187,
+      "learning_rate": 1.513547543967922e-05,
+      "loss": 0.5397,
+      "step": 9962
+    },
+    {
+      "epoch": 1.7772723218267772,
+      "grad_norm": 0.508708655834198,
+      "learning_rate": 1.5111476855644901e-05,
+      "loss": 0.4217,
+      "step": 9963
+    },
+    {
+      "epoch": 1.7774507180447774,
+      "grad_norm": 0.49109214544296265,
+      "learning_rate": 1.5087496719654149e-05,
+      "loss": 0.4588,
+      "step": 9964
+    },
+    {
+      "epoch": 1.7776291142627776,
+      "grad_norm": 0.4775922894477844,
+      "learning_rate": 1.5063535033590287e-05,
+      "loss": 0.4091,
+      "step": 9965
+    },
+    {
+      "epoch": 1.7778075104807778,
+      "grad_norm": 0.4845198392868042,
+      "learning_rate": 1.5039591799335312e-05,
+      "loss": 0.374,
+      "step": 9966
+    },
+    {
+      "epoch": 1.777985906698778,
+      "grad_norm": 0.48541465401649475,
+      "learning_rate": 1.5015667018769692e-05,
+      "loss": 0.5168,
+      "step": 9967
+    },
+    {
+      "epoch": 1.7781643029167782,
+      "grad_norm": 0.5408685803413391,
+      "learning_rate": 1.4991760693772422e-05,
+      "loss": 0.617,
+      "step": 9968
+    },
+    {
+      "epoch": 1.7783426991347784,
+      "grad_norm": 0.48128771781921387,
+      "learning_rate": 1.4967872826221168e-05,
+      "loss": 0.4295,
+      "step": 9969
+    },
+    {
+      "epoch": 1.7785210953527786,
+      "grad_norm": 0.49828970432281494,
+      "learning_rate": 1.4944003417992014e-05,
+      "loss": 0.4724,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7786994915707788,
+      "grad_norm": 0.487578809261322,
+      "learning_rate": 1.4920152470959707e-05,
+      "loss": 0.4936,
+      "step": 9971
+    },
+    {
+      "epoch": 1.778877887788779,
+      "grad_norm": 0.47483158111572266,
+      "learning_rate": 1.489631998699742e-05,
+      "loss": 0.4905,
+      "step": 9972
+    },
+    {
+      "epoch": 1.7790562840067792,
+      "grad_norm": 0.4644508957862854,
+      "learning_rate": 1.487250596797704e-05,
+      "loss": 0.4285,
+      "step": 9973
+    },
+    {
+      "epoch": 1.7792346802247794,
+      "grad_norm": 0.5382997989654541,
+      "learning_rate": 1.4848710415768824e-05,
+      "loss": 0.5722,
+      "step": 9974
+    },
+    {
+      "epoch": 1.7794130764427796,
+      "grad_norm": 0.5405296087265015,
+      "learning_rate": 1.4824933332241692e-05,
+      "loss": 0.4465,
+      "step": 9975
+    },
+    {
+      "epoch": 1.7795914726607795,
+      "grad_norm": 0.526763916015625,
+      "learning_rate": 1.4801174719263122e-05,
+      "loss": 0.5126,
+      "step": 9976
+    },
+    {
+      "epoch": 1.7797698688787797,
+      "grad_norm": 0.4779605567455292,
+      "learning_rate": 1.477743457869904e-05,
+      "loss": 0.3936,
+      "step": 9977
+    },
+    {
+      "epoch": 1.77994826509678,
+      "grad_norm": 0.5108462572097778,
+      "learning_rate": 1.4753712912414035e-05,
+      "loss": 0.4744,
+      "step": 9978
+    },
+    {
+      "epoch": 1.7801266613147801,
+      "grad_norm": 0.5295320749282837,
+      "learning_rate": 1.4730009722271204e-05,
+      "loss": 0.5429,
+      "step": 9979
+    },
+    {
+      "epoch": 1.7803050575327803,
+      "grad_norm": 0.5670889019966125,
+      "learning_rate": 1.4706325010132137e-05,
+      "loss": 0.5606,
+      "step": 9980
+    },
+    {
+      "epoch": 1.7804834537507803,
+      "grad_norm": 0.5170961022377014,
+      "learning_rate": 1.4682658777857072e-05,
+      "loss": 0.5119,
+      "step": 9981
+    },
+    {
+      "epoch": 1.7806618499687805,
+      "grad_norm": 0.4892038106918335,
+      "learning_rate": 1.4659011027304686e-05,
+      "loss": 0.4014,
+      "step": 9982
+    },
+    {
+      "epoch": 1.7808402461867807,
+      "grad_norm": 0.46245473623275757,
+      "learning_rate": 1.4635381760332356e-05,
+      "loss": 0.3665,
+      "step": 9983
+    },
+    {
+      "epoch": 1.781018642404781,
+      "grad_norm": 0.5037046074867249,
+      "learning_rate": 1.461177097879579e-05,
+      "loss": 0.388,
+      "step": 9984
+    },
+    {
+      "epoch": 1.7811970386227811,
+      "grad_norm": 0.4744492471218109,
+      "learning_rate": 1.458817868454948e-05,
+      "loss": 0.3254,
+      "step": 9985
+    },
+    {
+      "epoch": 1.7813754348407813,
+      "grad_norm": 0.4990524649620056,
+      "learning_rate": 1.456460487944633e-05,
+      "loss": 0.4834,
+      "step": 9986
+    },
+    {
+      "epoch": 1.7815538310587815,
+      "grad_norm": 0.631146252155304,
+      "learning_rate": 1.4541049565337749e-05,
+      "loss": 0.5804,
+      "step": 9987
+    },
+    {
+      "epoch": 1.7817322272767817,
+      "grad_norm": 0.5024986863136292,
+      "learning_rate": 1.451751274407384e-05,
+      "loss": 0.5254,
+      "step": 9988
+    },
+    {
+      "epoch": 1.781910623494782,
+      "grad_norm": 0.5565928816795349,
+      "learning_rate": 1.4493994417503127e-05,
+      "loss": 0.4387,
+      "step": 9989
+    },
+    {
+      "epoch": 1.782089019712782,
+      "grad_norm": 0.5144158005714417,
+      "learning_rate": 1.4470494587472765e-05,
+      "loss": 0.4451,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7822674159307823,
+      "grad_norm": 0.5400242805480957,
+      "learning_rate": 1.4447013255828368e-05,
+      "loss": 0.5332,
+      "step": 9991
+    },
+    {
+      "epoch": 1.7824458121487825,
+      "grad_norm": 0.5203432440757751,
+      "learning_rate": 1.4423550424414234e-05,
+      "loss": 0.4738,
+      "step": 9992
+    },
+    {
+      "epoch": 1.7826242083667827,
+      "grad_norm": 0.5619713068008423,
+      "learning_rate": 1.4400106095073029e-05,
+      "loss": 0.5508,
+      "step": 9993
+    },
+    {
+      "epoch": 1.782802604584783,
+      "grad_norm": 0.5214311480522156,
+      "learning_rate": 1.4376680269646086e-05,
+      "loss": 0.4902,
+      "step": 9994
+    },
+    {
+      "epoch": 1.782981000802783,
+      "grad_norm": 0.4371115267276764,
+      "learning_rate": 1.4353272949973322e-05,
+      "loss": 0.3828,
+      "step": 9995
+    },
+    {
+      "epoch": 1.7831593970207833,
+      "grad_norm": 0.5262027382850647,
+      "learning_rate": 1.4329884137893074e-05,
+      "loss": 0.5594,
+      "step": 9996
+    },
+    {
+      "epoch": 1.7833377932387835,
+      "grad_norm": 0.5280055403709412,
+      "learning_rate": 1.430651383524234e-05,
+      "loss": 0.454,
+      "step": 9997
+    },
+    {
+      "epoch": 1.7835161894567835,
+      "grad_norm": 0.5150504112243652,
+      "learning_rate": 1.4283162043856546e-05,
+      "loss": 0.5101,
+      "step": 9998
+    },
+    {
+      "epoch": 1.7836945856747837,
+      "grad_norm": 0.5318194627761841,
+      "learning_rate": 1.4259828765569777e-05,
+      "loss": 0.4295,
+      "step": 9999
+    },
+    {
+      "epoch": 1.7838729818927839,
+      "grad_norm": 0.5350292921066284,
+      "learning_rate": 1.423651400221465e-05,
+      "loss": 0.4757,
+      "step": 10000
+    },
+    {
+      "epoch": 1.784051378110784,
+      "grad_norm": 0.5203887224197388,
+      "learning_rate": 1.4213217755622205e-05,
+      "loss": 0.5545,
+      "step": 10001
+    },
+    {
+      "epoch": 1.7842297743287843,
+      "grad_norm": 0.48459410667419434,
+      "learning_rate": 1.4189940027622194e-05,
+      "loss": 0.4402,
+      "step": 10002
+    },
+    {
+      "epoch": 1.7844081705467842,
+      "grad_norm": 0.5484018921852112,
+      "learning_rate": 1.41666808200428e-05,
+      "loss": 0.5066,
+      "step": 10003
+    },
+    {
+      "epoch": 1.7845865667647844,
+      "grad_norm": 0.6117169260978699,
+      "learning_rate": 1.4143440134710833e-05,
+      "loss": 0.7196,
+      "step": 10004
+    },
+    {
+      "epoch": 1.7847649629827846,
+      "grad_norm": 0.5282416939735413,
+      "learning_rate": 1.4120217973451533e-05,
+      "loss": 0.4336,
+      "step": 10005
+    },
+    {
+      "epoch": 1.7849433592007848,
+      "grad_norm": 0.5563318729400635,
+      "learning_rate": 1.4097014338088855e-05,
+      "loss": 0.5367,
+      "step": 10006
+    },
+    {
+      "epoch": 1.785121755418785,
+      "grad_norm": 0.6051769256591797,
+      "learning_rate": 1.4073829230445173e-05,
+      "loss": 0.6119,
+      "step": 10007
+    },
+    {
+      "epoch": 1.7853001516367852,
+      "grad_norm": 0.5828459858894348,
+      "learning_rate": 1.4050662652341312e-05,
+      "loss": 0.6749,
+      "step": 10008
+    },
+    {
+      "epoch": 1.7854785478547854,
+      "grad_norm": 0.5630053281784058,
+      "learning_rate": 1.4027514605596952e-05,
+      "loss": 0.5675,
+      "step": 10009
+    },
+    {
+      "epoch": 1.7856569440727856,
+      "grad_norm": 0.613758385181427,
+      "learning_rate": 1.4004385092030031e-05,
+      "loss": 0.5525,
+      "step": 10010
+    },
+    {
+      "epoch": 1.7858353402907858,
+      "grad_norm": 0.4936037063598633,
+      "learning_rate": 1.3981274113457148e-05,
+      "loss": 0.4094,
+      "step": 10011
+    },
+    {
+      "epoch": 1.786013736508786,
+      "grad_norm": 0.5002244114875793,
+      "learning_rate": 1.3958181671693466e-05,
+      "loss": 0.4153,
+      "step": 10012
+    },
+    {
+      "epoch": 1.7861921327267862,
+      "grad_norm": 0.5575555562973022,
+      "learning_rate": 1.3935107768552557e-05,
+      "loss": 0.6104,
+      "step": 10013
+    },
+    {
+      "epoch": 1.7863705289447864,
+      "grad_norm": 0.5280625224113464,
+      "learning_rate": 1.3912052405846754e-05,
+      "loss": 0.565,
+      "step": 10014
+    },
+    {
+      "epoch": 1.7865489251627866,
+      "grad_norm": 0.505584716796875,
+      "learning_rate": 1.3889015585386689e-05,
+      "loss": 0.4413,
+      "step": 10015
+    },
+    {
+      "epoch": 1.7867273213807868,
+      "grad_norm": 0.4564724862575531,
+      "learning_rate": 1.3865997308981804e-05,
+      "loss": 0.3544,
+      "step": 10016
+    },
+    {
+      "epoch": 1.786905717598787,
+      "grad_norm": 0.6348497271537781,
+      "learning_rate": 1.3842997578439819e-05,
+      "loss": 0.6875,
+      "step": 10017
+    },
+    {
+      "epoch": 1.7870841138167872,
+      "grad_norm": 0.5605414509773254,
+      "learning_rate": 1.3820016395567209e-05,
+      "loss": 0.5125,
+      "step": 10018
+    },
+    {
+      "epoch": 1.7872625100347874,
+      "grad_norm": 0.4736771285533905,
+      "learning_rate": 1.379705376216886e-05,
+      "loss": 0.4449,
+      "step": 10019
+    },
+    {
+      "epoch": 1.7874409062527874,
+      "grad_norm": 0.4950930178165436,
+      "learning_rate": 1.3774109680048274e-05,
+      "loss": 0.4283,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7876193024707876,
+      "grad_norm": 0.4793688654899597,
+      "learning_rate": 1.3751184151007485e-05,
+      "loss": 0.4695,
+      "step": 10021
+    },
+    {
+      "epoch": 1.7877976986887878,
+      "grad_norm": 0.4778786599636078,
+      "learning_rate": 1.3728277176846965e-05,
+      "loss": 0.4909,
+      "step": 10022
+    },
+    {
+      "epoch": 1.787976094906788,
+      "grad_norm": 0.5141474008560181,
+      "learning_rate": 1.3705388759365945e-05,
+      "loss": 0.4603,
+      "step": 10023
+    },
+    {
+      "epoch": 1.7881544911247882,
+      "grad_norm": 0.4962654411792755,
+      "learning_rate": 1.3682518900361902e-05,
+      "loss": 0.4038,
+      "step": 10024
+    },
+    {
+      "epoch": 1.7883328873427882,
+      "grad_norm": 0.44499000906944275,
+      "learning_rate": 1.3659667601631231e-05,
+      "loss": 0.3354,
+      "step": 10025
+    },
+    {
+      "epoch": 1.7885112835607884,
+      "grad_norm": 0.5469255447387695,
+      "learning_rate": 1.3636834864968556e-05,
+      "loss": 0.5772,
+      "step": 10026
+    },
+    {
+      "epoch": 1.7886896797787886,
+      "grad_norm": 0.5044131278991699,
+      "learning_rate": 1.3614020692167107e-05,
+      "loss": 0.4198,
+      "step": 10027
+    },
+    {
+      "epoch": 1.7888680759967888,
+      "grad_norm": 0.5094908475875854,
+      "learning_rate": 1.3591225085018782e-05,
+      "loss": 0.5186,
+      "step": 10028
+    },
+    {
+      "epoch": 1.789046472214789,
+      "grad_norm": 0.4827946424484253,
+      "learning_rate": 1.3568448045313874e-05,
+      "loss": 0.4773,
+      "step": 10029
+    },
+    {
+      "epoch": 1.7892248684327892,
+      "grad_norm": 0.5509223341941833,
+      "learning_rate": 1.354568957484134e-05,
+      "loss": 0.5378,
+      "step": 10030
+    },
+    {
+      "epoch": 1.7894032646507894,
+      "grad_norm": 0.49218085408210754,
+      "learning_rate": 1.3522949675388557e-05,
+      "loss": 0.4328,
+      "step": 10031
+    },
+    {
+      "epoch": 1.7895816608687896,
+      "grad_norm": 0.5489743947982788,
+      "learning_rate": 1.3500228348741594e-05,
+      "loss": 0.5715,
+      "step": 10032
+    },
+    {
+      "epoch": 1.7897600570867898,
+      "grad_norm": 0.5286942720413208,
+      "learning_rate": 1.3477525596684914e-05,
+      "loss": 0.4155,
+      "step": 10033
+    },
+    {
+      "epoch": 1.78993845330479,
+      "grad_norm": 0.5898469090461731,
+      "learning_rate": 1.3454841421001562e-05,
+      "loss": 0.6253,
+      "step": 10034
+    },
+    {
+      "epoch": 1.7901168495227902,
+      "grad_norm": 0.4971984326839447,
+      "learning_rate": 1.3432175823473197e-05,
+      "loss": 0.401,
+      "step": 10035
+    },
+    {
+      "epoch": 1.7902952457407904,
+      "grad_norm": 0.5394257307052612,
+      "learning_rate": 1.3409528805879895e-05,
+      "loss": 0.5684,
+      "step": 10036
+    },
+    {
+      "epoch": 1.7904736419587906,
+      "grad_norm": 0.47669845819473267,
+      "learning_rate": 1.3386900370000455e-05,
+      "loss": 0.4986,
+      "step": 10037
+    },
+    {
+      "epoch": 1.7906520381767908,
+      "grad_norm": 0.4922144114971161,
+      "learning_rate": 1.3364290517611982e-05,
+      "loss": 0.4396,
+      "step": 10038
+    },
+    {
+      "epoch": 1.790830434394791,
+      "grad_norm": 0.5299371480941772,
+      "learning_rate": 1.334169925049028e-05,
+      "loss": 0.4327,
+      "step": 10039
+    },
+    {
+      "epoch": 1.7910088306127911,
+      "grad_norm": 0.5726783871650696,
+      "learning_rate": 1.3319126570409734e-05,
+      "loss": 0.5291,
+      "step": 10040
+    },
+    {
+      "epoch": 1.7911872268307913,
+      "grad_norm": 0.5141592621803284,
+      "learning_rate": 1.3296572479143093e-05,
+      "loss": 0.4491,
+      "step": 10041
+    },
+    {
+      "epoch": 1.7913656230487913,
+      "grad_norm": 0.5207017660140991,
+      "learning_rate": 1.3274036978461829e-05,
+      "loss": 0.4182,
+      "step": 10042
+    },
+    {
+      "epoch": 1.7915440192667915,
+      "grad_norm": 0.5985919833183289,
+      "learning_rate": 1.3251520070135803e-05,
+      "loss": 0.6607,
+      "step": 10043
+    },
+    {
+      "epoch": 1.7917224154847917,
+      "grad_norm": 0.5005565285682678,
+      "learning_rate": 1.3229021755933546e-05,
+      "loss": 0.4086,
+      "step": 10044
+    },
+    {
+      "epoch": 1.791900811702792,
+      "grad_norm": 0.5269946455955505,
+      "learning_rate": 1.3206542037621978e-05,
+      "loss": 0.4658,
+      "step": 10045
+    },
+    {
+      "epoch": 1.7920792079207921,
+      "grad_norm": 0.5867735147476196,
+      "learning_rate": 1.318408091696674e-05,
+      "loss": 0.6541,
+      "step": 10046
+    },
+    {
+      "epoch": 1.792257604138792,
+      "grad_norm": 0.491451621055603,
+      "learning_rate": 1.3161638395731867e-05,
+      "loss": 0.3777,
+      "step": 10047
+    },
+    {
+      "epoch": 1.7924360003567923,
+      "grad_norm": 0.5263445377349854,
+      "learning_rate": 1.3139214475679977e-05,
+      "loss": 0.4627,
+      "step": 10048
+    },
+    {
+      "epoch": 1.7926143965747925,
+      "grad_norm": 0.4960615038871765,
+      "learning_rate": 1.3116809158572273e-05,
+      "loss": 0.4039,
+      "step": 10049
+    },
+    {
+      "epoch": 1.7927927927927927,
+      "grad_norm": 0.5893881916999817,
+      "learning_rate": 1.3094422446168403e-05,
+      "loss": 0.6089,
+      "step": 10050
+    },
+    {
+      "epoch": 1.7929711890107929,
+      "grad_norm": 0.636078953742981,
+      "learning_rate": 1.3072054340226708e-05,
+      "loss": 0.5522,
+      "step": 10051
+    },
+    {
+      "epoch": 1.793149585228793,
+      "grad_norm": 0.6112774610519409,
+      "learning_rate": 1.304970484250384e-05,
+      "loss": 0.6133,
+      "step": 10052
+    },
+    {
+      "epoch": 1.7933279814467933,
+      "grad_norm": 0.5139406323432922,
+      "learning_rate": 1.3027373954755229e-05,
+      "loss": 0.3866,
+      "step": 10053
+    },
+    {
+      "epoch": 1.7935063776647935,
+      "grad_norm": 0.5027743577957153,
+      "learning_rate": 1.3005061678734665e-05,
+      "loss": 0.4824,
+      "step": 10054
+    },
+    {
+      "epoch": 1.7936847738827937,
+      "grad_norm": 0.45598676800727844,
+      "learning_rate": 1.2982768016194551e-05,
+      "loss": 0.3925,
+      "step": 10055
+    },
+    {
+      "epoch": 1.7938631701007939,
+      "grad_norm": 0.5171281099319458,
+      "learning_rate": 1.2960492968885906e-05,
+      "loss": 0.499,
+      "step": 10056
+    },
+    {
+      "epoch": 1.794041566318794,
+      "grad_norm": 0.5457966923713684,
+      "learning_rate": 1.2938236538558079e-05,
+      "loss": 0.4644,
+      "step": 10057
+    },
+    {
+      "epoch": 1.7942199625367943,
+      "grad_norm": 0.5165960788726807,
+      "learning_rate": 1.2915998726959172e-05,
+      "loss": 0.5193,
+      "step": 10058
+    },
+    {
+      "epoch": 1.7943983587547945,
+      "grad_norm": 0.5514568090438843,
+      "learning_rate": 1.2893779535835703e-05,
+      "loss": 0.4803,
+      "step": 10059
+    },
+    {
+      "epoch": 1.7945767549727947,
+      "grad_norm": 0.47774699330329895,
+      "learning_rate": 1.2871578966932723e-05,
+      "loss": 0.3729,
+      "step": 10060
+    },
+    {
+      "epoch": 1.7947551511907949,
+      "grad_norm": 0.5756377577781677,
+      "learning_rate": 1.2849397021993947e-05,
+      "loss": 0.554,
+      "step": 10061
+    },
+    {
+      "epoch": 1.794933547408795,
+      "grad_norm": 0.5572470426559448,
+      "learning_rate": 1.2827233702761398e-05,
+      "loss": 0.4625,
+      "step": 10062
+    },
+    {
+      "epoch": 1.7951119436267953,
+      "grad_norm": 0.508305549621582,
+      "learning_rate": 1.2805089010975906e-05,
+      "loss": 0.4027,
+      "step": 10063
+    },
+    {
+      "epoch": 1.7952903398447952,
+      "grad_norm": 0.5296847224235535,
+      "learning_rate": 1.2782962948376608e-05,
+      "loss": 0.5182,
+      "step": 10064
+    },
+    {
+      "epoch": 1.7954687360627954,
+      "grad_norm": 0.5482270121574402,
+      "learning_rate": 1.2760855516701364e-05,
+      "loss": 0.6118,
+      "step": 10065
+    },
+    {
+      "epoch": 1.7956471322807956,
+      "grad_norm": 0.6912332773208618,
+      "learning_rate": 1.2738766717686396e-05,
+      "loss": 0.5096,
+      "step": 10066
+    },
+    {
+      "epoch": 1.7958255284987958,
+      "grad_norm": 0.5811026692390442,
+      "learning_rate": 1.271669655306662e-05,
+      "loss": 0.6013,
+      "step": 10067
+    },
+    {
+      "epoch": 1.796003924716796,
+      "grad_norm": 0.5055823922157288,
+      "learning_rate": 1.2694645024575374e-05,
+      "loss": 0.4401,
+      "step": 10068
+    },
+    {
+      "epoch": 1.796182320934796,
+      "grad_norm": 0.6301665306091309,
+      "learning_rate": 1.2672612133944578e-05,
+      "loss": 0.5362,
+      "step": 10069
+    },
+    {
+      "epoch": 1.7963607171527962,
+      "grad_norm": 0.4803159534931183,
+      "learning_rate": 1.265059788290468e-05,
+      "loss": 0.3668,
+      "step": 10070
+    },
+    {
+      "epoch": 1.7965391133707964,
+      "grad_norm": 0.4969823360443115,
+      "learning_rate": 1.2628602273184714e-05,
+      "loss": 0.4755,
+      "step": 10071
+    },
+    {
+      "epoch": 1.7967175095887966,
+      "grad_norm": 0.5457087755203247,
+      "learning_rate": 1.2606625306512159e-05,
+      "loss": 0.5974,
+      "step": 10072
+    },
+    {
+      "epoch": 1.7968959058067968,
+      "grad_norm": 0.6440889835357666,
+      "learning_rate": 1.2584666984613107e-05,
+      "loss": 0.6619,
+      "step": 10073
+    },
+    {
+      "epoch": 1.797074302024797,
+      "grad_norm": 0.5694631338119507,
+      "learning_rate": 1.2562727309212125e-05,
+      "loss": 0.5553,
+      "step": 10074
+    },
+    {
+      "epoch": 1.7972526982427972,
+      "grad_norm": 0.582349956035614,
+      "learning_rate": 1.2540806282032385e-05,
+      "loss": 0.5732,
+      "step": 10075
+    },
+    {
+      "epoch": 1.7974310944607974,
+      "grad_norm": 0.539436936378479,
+      "learning_rate": 1.2518903904795515e-05,
+      "loss": 0.4747,
+      "step": 10076
+    },
+    {
+      "epoch": 1.7976094906787976,
+      "grad_norm": 0.5120388865470886,
+      "learning_rate": 1.2497020179221747e-05,
+      "loss": 0.5279,
+      "step": 10077
+    },
+    {
+      "epoch": 1.7977878868967978,
+      "grad_norm": 0.5143983364105225,
+      "learning_rate": 1.247515510702979e-05,
+      "loss": 0.453,
+      "step": 10078
+    },
+    {
+      "epoch": 1.797966283114798,
+      "grad_norm": 0.46396833658218384,
+      "learning_rate": 1.2453308689936965e-05,
+      "loss": 0.3753,
+      "step": 10079
+    },
+    {
+      "epoch": 1.7981446793327982,
+      "grad_norm": 0.5515454411506653,
+      "learning_rate": 1.2431480929659066e-05,
+      "loss": 0.5563,
+      "step": 10080
+    },
+    {
+      "epoch": 1.7983230755507984,
+      "grad_norm": 0.5875598788261414,
+      "learning_rate": 1.2409671827910363e-05,
+      "loss": 0.5579,
+      "step": 10081
+    },
+    {
+      "epoch": 1.7985014717687986,
+      "grad_norm": 0.43758925795555115,
+      "learning_rate": 1.2387881386403844e-05,
+      "loss": 0.3783,
+      "step": 10082
+    },
+    {
+      "epoch": 1.7986798679867988,
+      "grad_norm": 0.4410098195075989,
+      "learning_rate": 1.2366109606850834e-05,
+      "loss": 0.336,
+      "step": 10083
+    },
+    {
+      "epoch": 1.798858264204799,
+      "grad_norm": 0.5217193365097046,
+      "learning_rate": 1.2344356490961328e-05,
+      "loss": 0.4554,
+      "step": 10084
+    },
+    {
+      "epoch": 1.7990366604227992,
+      "grad_norm": 0.5786857604980469,
+      "learning_rate": 1.2322622040443793e-05,
+      "loss": 0.5736,
+      "step": 10085
+    },
+    {
+      "epoch": 1.7992150566407992,
+      "grad_norm": 0.5340781807899475,
+      "learning_rate": 1.2300906257005196e-05,
+      "loss": 0.5068,
+      "step": 10086
+    },
+    {
+      "epoch": 1.7993934528587994,
+      "grad_norm": 0.5103334784507751,
+      "learning_rate": 1.22792091423512e-05,
+      "loss": 0.4845,
+      "step": 10087
+    },
+    {
+      "epoch": 1.7995718490767996,
+      "grad_norm": 0.5319827795028687,
+      "learning_rate": 1.2257530698185776e-05,
+      "loss": 0.5371,
+      "step": 10088
+    },
+    {
+      "epoch": 1.7997502452947998,
+      "grad_norm": 0.5461127161979675,
+      "learning_rate": 1.2235870926211617e-05,
+      "loss": 0.5464,
+      "step": 10089
+    },
+    {
+      "epoch": 1.7999286415128,
+      "grad_norm": 0.6167702078819275,
+      "learning_rate": 1.2214229828129808e-05,
+      "loss": 0.5783,
+      "step": 10090
+    },
+    {
+      "epoch": 1.8001070377308,
+      "grad_norm": 0.538575291633606,
+      "learning_rate": 1.2192607405640072e-05,
+      "loss": 0.5085,
+      "step": 10091
+    },
+    {
+      "epoch": 1.8002854339488001,
+      "grad_norm": 0.4790734648704529,
+      "learning_rate": 1.2171003660440633e-05,
+      "loss": 0.3925,
+      "step": 10092
+    },
+    {
+      "epoch": 1.8004638301668003,
+      "grad_norm": 0.5150331854820251,
+      "learning_rate": 1.214941859422819e-05,
+      "loss": 0.5111,
+      "step": 10093
+    },
+    {
+      "epoch": 1.8006422263848005,
+      "grad_norm": 0.5167433023452759,
+      "learning_rate": 1.2127852208698081e-05,
+      "loss": 0.4739,
+      "step": 10094
+    },
+    {
+      "epoch": 1.8008206226028007,
+      "grad_norm": 0.48076915740966797,
+      "learning_rate": 1.2106304505544063e-05,
+      "loss": 0.402,
+      "step": 10095
+    },
+    {
+      "epoch": 1.800999018820801,
+      "grad_norm": 0.5094106197357178,
+      "learning_rate": 1.208477548645856e-05,
+      "loss": 0.562,
+      "step": 10096
+    },
+    {
+      "epoch": 1.8011774150388011,
+      "grad_norm": 0.4915674030780792,
+      "learning_rate": 1.2063265153132359e-05,
+      "loss": 0.4165,
+      "step": 10097
+    },
+    {
+      "epoch": 1.8013558112568013,
+      "grad_norm": 0.4733083248138428,
+      "learning_rate": 1.2041773507254966e-05,
+      "loss": 0.3459,
+      "step": 10098
+    },
+    {
+      "epoch": 1.8015342074748015,
+      "grad_norm": 0.5540547966957092,
+      "learning_rate": 1.202030055051423e-05,
+      "loss": 0.4423,
+      "step": 10099
+    },
+    {
+      "epoch": 1.8017126036928017,
+      "grad_norm": 0.5337905883789062,
+      "learning_rate": 1.1998846284596687e-05,
+      "loss": 0.5369,
+      "step": 10100
+    },
+    {
+      "epoch": 1.801890999910802,
+      "grad_norm": 0.5766735672950745,
+      "learning_rate": 1.1977410711187381e-05,
+      "loss": 0.574,
+      "step": 10101
+    },
+    {
+      "epoch": 1.8020693961288021,
+      "grad_norm": 0.6233359575271606,
+      "learning_rate": 1.1955993831969769e-05,
+      "loss": 0.6326,
+      "step": 10102
+    },
+    {
+      "epoch": 1.8022477923468023,
+      "grad_norm": 0.538557231426239,
+      "learning_rate": 1.1934595648625978e-05,
+      "loss": 0.51,
+      "step": 10103
+    },
+    {
+      "epoch": 1.8024261885648025,
+      "grad_norm": 0.5933203101158142,
+      "learning_rate": 1.1913216162836582e-05,
+      "loss": 0.6019,
+      "step": 10104
+    },
+    {
+      "epoch": 1.8026045847828027,
+      "grad_norm": 0.5655239224433899,
+      "learning_rate": 1.1891855376280764e-05,
+      "loss": 0.5871,
+      "step": 10105
+    },
+    {
+      "epoch": 1.802782981000803,
+      "grad_norm": 0.4605049788951874,
+      "learning_rate": 1.1870513290636154e-05,
+      "loss": 0.3936,
+      "step": 10106
+    },
+    {
+      "epoch": 1.8029613772188031,
+      "grad_norm": 0.4446168541908264,
+      "learning_rate": 1.1849189907578938e-05,
+      "loss": 0.3525,
+      "step": 10107
+    },
+    {
+      "epoch": 1.803139773436803,
+      "grad_norm": 0.5252434611320496,
+      "learning_rate": 1.1827885228783863e-05,
+      "loss": 0.5285,
+      "step": 10108
+    },
+    {
+      "epoch": 1.8033181696548033,
+      "grad_norm": 0.49297425150871277,
+      "learning_rate": 1.1806599255924172e-05,
+      "loss": 0.4882,
+      "step": 10109
+    },
+    {
+      "epoch": 1.8034965658728035,
+      "grad_norm": 0.5996000170707703,
+      "learning_rate": 1.1785331990671722e-05,
+      "loss": 0.5251,
+      "step": 10110
+    },
+    {
+      "epoch": 1.8036749620908037,
+      "grad_norm": 0.5449775457382202,
+      "learning_rate": 1.1764083434696732e-05,
+      "loss": 0.5504,
+      "step": 10111
+    },
+    {
+      "epoch": 1.803853358308804,
+      "grad_norm": 0.5374523997306824,
+      "learning_rate": 1.1742853589668145e-05,
+      "loss": 0.5502,
+      "step": 10112
+    },
+    {
+      "epoch": 1.8040317545268039,
+      "grad_norm": 0.5069290399551392,
+      "learning_rate": 1.1721642457253323e-05,
+      "loss": 0.475,
+      "step": 10113
+    },
+    {
+      "epoch": 1.804210150744804,
+      "grad_norm": 0.5023181438446045,
+      "learning_rate": 1.1700450039118127e-05,
+      "loss": 0.4728,
+      "step": 10114
+    },
+    {
+      "epoch": 1.8043885469628043,
+      "grad_norm": 0.44995036721229553,
+      "learning_rate": 1.1679276336927058e-05,
+      "loss": 0.3337,
+      "step": 10115
+    },
+    {
+      "epoch": 1.8045669431808045,
+      "grad_norm": 0.5543752312660217,
+      "learning_rate": 1.1658121352342982e-05,
+      "loss": 0.5126,
+      "step": 10116
+    },
+    {
+      "epoch": 1.8047453393988047,
+      "grad_norm": 0.5311068296432495,
+      "learning_rate": 1.1636985087027597e-05,
+      "loss": 0.5781,
+      "step": 10117
+    },
+    {
+      "epoch": 1.8049237356168049,
+      "grad_norm": 0.5337813496589661,
+      "learning_rate": 1.1615867542640795e-05,
+      "loss": 0.4699,
+      "step": 10118
+    },
+    {
+      "epoch": 1.805102131834805,
+      "grad_norm": 0.5538474321365356,
+      "learning_rate": 1.1594768720841142e-05,
+      "loss": 0.616,
+      "step": 10119
+    },
+    {
+      "epoch": 1.8052805280528053,
+      "grad_norm": 0.5138535499572754,
+      "learning_rate": 1.157368862328581e-05,
+      "loss": 0.5966,
+      "step": 10120
+    },
+    {
+      "epoch": 1.8054589242708055,
+      "grad_norm": 0.5369413495063782,
+      "learning_rate": 1.1552627251630338e-05,
+      "loss": 0.6007,
+      "step": 10121
+    },
+    {
+      "epoch": 1.8056373204888057,
+      "grad_norm": 0.4994553029537201,
+      "learning_rate": 1.1531584607528928e-05,
+      "loss": 0.3921,
+      "step": 10122
+    },
+    {
+      "epoch": 1.8058157167068059,
+      "grad_norm": 0.6017369031906128,
+      "learning_rate": 1.1510560692634203e-05,
+      "loss": 0.553,
+      "step": 10123
+    },
+    {
+      "epoch": 1.805994112924806,
+      "grad_norm": 0.5081920027732849,
+      "learning_rate": 1.1489555508597455e-05,
+      "loss": 0.3915,
+      "step": 10124
+    },
+    {
+      "epoch": 1.8061725091428062,
+      "grad_norm": 0.5011181235313416,
+      "learning_rate": 1.1468569057068363e-05,
+      "loss": 0.4489,
+      "step": 10125
+    },
+    {
+      "epoch": 1.8063509053608064,
+      "grad_norm": 0.5123955607414246,
+      "learning_rate": 1.1447601339695218e-05,
+      "loss": 0.4911,
+      "step": 10126
+    },
+    {
+      "epoch": 1.8065293015788066,
+      "grad_norm": 0.6243327856063843,
+      "learning_rate": 1.1426652358124817e-05,
+      "loss": 0.7168,
+      "step": 10127
+    },
+    {
+      "epoch": 1.8067076977968068,
+      "grad_norm": 0.525888204574585,
+      "learning_rate": 1.1405722114002425e-05,
+      "loss": 0.512,
+      "step": 10128
+    },
+    {
+      "epoch": 1.806886094014807,
+      "grad_norm": 0.5606545209884644,
+      "learning_rate": 1.1384810608972007e-05,
+      "loss": 0.5632,
+      "step": 10129
+    },
+    {
+      "epoch": 1.807064490232807,
+      "grad_norm": 0.5493068695068359,
+      "learning_rate": 1.1363917844675803e-05,
+      "loss": 0.4691,
+      "step": 10130
+    },
+    {
+      "epoch": 1.8072428864508072,
+      "grad_norm": 0.6075304746627808,
+      "learning_rate": 1.1343043822754834e-05,
+      "loss": 0.7009,
+      "step": 10131
+    },
+    {
+      "epoch": 1.8074212826688074,
+      "grad_norm": 0.49980399012565613,
+      "learning_rate": 1.132218854484851e-05,
+      "loss": 0.4691,
+      "step": 10132
+    },
+    {
+      "epoch": 1.8075996788868076,
+      "grad_norm": 0.5099700093269348,
+      "learning_rate": 1.1301352012594774e-05,
+      "loss": 0.4577,
+      "step": 10133
+    },
+    {
+      "epoch": 1.8077780751048078,
+      "grad_norm": 0.5869147777557373,
+      "learning_rate": 1.1280534227630173e-05,
+      "loss": 0.6752,
+      "step": 10134
+    },
+    {
+      "epoch": 1.807956471322808,
+      "grad_norm": 0.5993606448173523,
+      "learning_rate": 1.1259735191589626e-05,
+      "loss": 0.6543,
+      "step": 10135
+    },
+    {
+      "epoch": 1.808134867540808,
+      "grad_norm": 0.5113434195518494,
+      "learning_rate": 1.123895490610677e-05,
+      "loss": 0.5406,
+      "step": 10136
+    },
+    {
+      "epoch": 1.8083132637588082,
+      "grad_norm": 0.5298689007759094,
+      "learning_rate": 1.1218193372813628e-05,
+      "loss": 0.4952,
+      "step": 10137
+    },
+    {
+      "epoch": 1.8084916599768084,
+      "grad_norm": 0.5689482092857361,
+      "learning_rate": 1.1197450593340875e-05,
+      "loss": 0.543,
+      "step": 10138
+    },
+    {
+      "epoch": 1.8086700561948086,
+      "grad_norm": 0.46555694937705994,
+      "learning_rate": 1.1176726569317563e-05,
+      "loss": 0.4123,
+      "step": 10139
+    },
+    {
+      "epoch": 1.8088484524128088,
+      "grad_norm": 0.4243280589580536,
+      "learning_rate": 1.1156021302371338e-05,
+      "loss": 0.4059,
+      "step": 10140
+    },
+    {
+      "epoch": 1.809026848630809,
+      "grad_norm": 0.6901927590370178,
+      "learning_rate": 1.1135334794128455e-05,
+      "loss": 0.6079,
+      "step": 10141
+    },
+    {
+      "epoch": 1.8092052448488092,
+      "grad_norm": 0.5594136714935303,
+      "learning_rate": 1.1114667046213555e-05,
+      "loss": 0.5835,
+      "step": 10142
+    },
+    {
+      "epoch": 1.8093836410668094,
+      "grad_norm": 0.4830605983734131,
+      "learning_rate": 1.1094018060249928e-05,
+      "loss": 0.3979,
+      "step": 10143
+    },
+    {
+      "epoch": 1.8095620372848096,
+      "grad_norm": 0.5415348410606384,
+      "learning_rate": 1.10733878378593e-05,
+      "loss": 0.5458,
+      "step": 10144
+    },
+    {
+      "epoch": 1.8097404335028098,
+      "grad_norm": 0.5037901401519775,
+      "learning_rate": 1.1052776380661988e-05,
+      "loss": 0.4389,
+      "step": 10145
+    },
+    {
+      "epoch": 1.80991882972081,
+      "grad_norm": 0.47683677077293396,
+      "learning_rate": 1.1032183690276754e-05,
+      "loss": 0.4552,
+      "step": 10146
+    },
+    {
+      "epoch": 1.8100972259388102,
+      "grad_norm": 0.5881403684616089,
+      "learning_rate": 1.1011609768320995e-05,
+      "loss": 0.617,
+      "step": 10147
+    },
+    {
+      "epoch": 1.8102756221568104,
+      "grad_norm": 0.5397078394889832,
+      "learning_rate": 1.099105461641059e-05,
+      "loss": 0.5371,
+      "step": 10148
+    },
+    {
+      "epoch": 1.8104540183748106,
+      "grad_norm": 0.5096381306648254,
+      "learning_rate": 1.0970518236159882e-05,
+      "loss": 0.4908,
+      "step": 10149
+    },
+    {
+      "epoch": 1.8106324145928108,
+      "grad_norm": 0.5621026754379272,
+      "learning_rate": 1.0950000629181806e-05,
+      "loss": 0.5941,
+      "step": 10150
+    },
+    {
+      "epoch": 1.810810810810811,
+      "grad_norm": 0.501204252243042,
+      "learning_rate": 1.0929501797087848e-05,
+      "loss": 0.4206,
+      "step": 10151
+    },
+    {
+      "epoch": 1.810989207028811,
+      "grad_norm": 0.48253610730171204,
+      "learning_rate": 1.0909021741487862e-05,
+      "loss": 0.4153,
+      "step": 10152
+    },
+    {
+      "epoch": 1.8111676032468111,
+      "grad_norm": 0.60355144739151,
+      "learning_rate": 1.0888560463990476e-05,
+      "loss": 0.518,
+      "step": 10153
+    },
+    {
+      "epoch": 1.8113459994648113,
+      "grad_norm": 0.4504523277282715,
+      "learning_rate": 1.086811796620263e-05,
+      "loss": 0.3618,
+      "step": 10154
+    },
+    {
+      "epoch": 1.8115243956828115,
+      "grad_norm": 0.5853896737098694,
+      "learning_rate": 1.0847694249729922e-05,
+      "loss": 0.6146,
+      "step": 10155
+    },
+    {
+      "epoch": 1.8117027919008117,
+      "grad_norm": 0.5694953203201294,
+      "learning_rate": 1.0827289316176353e-05,
+      "loss": 0.5908,
+      "step": 10156
+    },
+    {
+      "epoch": 1.811881188118812,
+      "grad_norm": 0.5596148371696472,
+      "learning_rate": 1.0806903167144583e-05,
+      "loss": 0.583,
+      "step": 10157
+    },
+    {
+      "epoch": 1.812059584336812,
+      "grad_norm": 0.4649914801120758,
+      "learning_rate": 1.0786535804235693e-05,
+      "loss": 0.4772,
+      "step": 10158
+    },
+    {
+      "epoch": 1.8122379805548121,
+      "grad_norm": 0.5577880144119263,
+      "learning_rate": 1.0766187229049345e-05,
+      "loss": 0.5872,
+      "step": 10159
+    },
+    {
+      "epoch": 1.8124163767728123,
+      "grad_norm": 0.6191621422767639,
+      "learning_rate": 1.0745857443183737e-05,
+      "loss": 0.6917,
+      "step": 10160
+    },
+    {
+      "epoch": 1.8125947729908125,
+      "grad_norm": 0.6169114112854004,
+      "learning_rate": 1.0725546448235424e-05,
+      "loss": 0.7141,
+      "step": 10161
+    },
+    {
+      "epoch": 1.8127731692088127,
+      "grad_norm": 0.5002092719078064,
+      "learning_rate": 1.0705254245799823e-05,
+      "loss": 0.435,
+      "step": 10162
+    },
+    {
+      "epoch": 1.812951565426813,
+      "grad_norm": 0.48867136240005493,
+      "learning_rate": 1.068498083747052e-05,
+      "loss": 0.4144,
+      "step": 10163
+    },
+    {
+      "epoch": 1.813129961644813,
+      "grad_norm": 0.5426896214485168,
+      "learning_rate": 1.0664726224839882e-05,
+      "loss": 0.4681,
+      "step": 10164
+    },
+    {
+      "epoch": 1.8133083578628133,
+      "grad_norm": 0.5177076458930969,
+      "learning_rate": 1.0644490409498636e-05,
+      "loss": 0.4917,
+      "step": 10165
+    },
+    {
+      "epoch": 1.8134867540808135,
+      "grad_norm": 0.5665025115013123,
+      "learning_rate": 1.0624273393036093e-05,
+      "loss": 0.5453,
+      "step": 10166
+    },
+    {
+      "epoch": 1.8136651502988137,
+      "grad_norm": 0.5852944254875183,
+      "learning_rate": 1.0604075177040151e-05,
+      "loss": 0.607,
+      "step": 10167
+    },
+    {
+      "epoch": 1.813843546516814,
+      "grad_norm": 0.5647523999214172,
+      "learning_rate": 1.0583895763097068e-05,
+      "loss": 0.6147,
+      "step": 10168
+    },
+    {
+      "epoch": 1.814021942734814,
+      "grad_norm": 0.4771578907966614,
+      "learning_rate": 1.0563735152791826e-05,
+      "loss": 0.3808,
+      "step": 10169
+    },
+    {
+      "epoch": 1.8142003389528143,
+      "grad_norm": 0.49404028058052063,
+      "learning_rate": 1.0543593347707742e-05,
+      "loss": 0.4938,
+      "step": 10170
+    },
+    {
+      "epoch": 1.8143787351708145,
+      "grad_norm": 0.5241032242774963,
+      "learning_rate": 1.0523470349426856e-05,
+      "loss": 0.4203,
+      "step": 10171
+    },
+    {
+      "epoch": 1.8145571313888147,
+      "grad_norm": 0.4687019884586334,
+      "learning_rate": 1.0503366159529515e-05,
+      "loss": 0.4542,
+      "step": 10172
+    },
+    {
+      "epoch": 1.814735527606815,
+      "grad_norm": 0.5116732120513916,
+      "learning_rate": 1.0483280779594707e-05,
+      "loss": 0.4266,
+      "step": 10173
+    },
+    {
+      "epoch": 1.8149139238248149,
+      "grad_norm": 0.593781054019928,
+      "learning_rate": 1.0463214211200001e-05,
+      "loss": 0.5538,
+      "step": 10174
+    },
+    {
+      "epoch": 1.815092320042815,
+      "grad_norm": 0.5247955918312073,
+      "learning_rate": 1.0443166455921332e-05,
+      "loss": 0.4487,
+      "step": 10175
+    },
+    {
+      "epoch": 1.8152707162608153,
+      "grad_norm": 0.5464117527008057,
+      "learning_rate": 1.042313751533333e-05,
+      "loss": 0.527,
+      "step": 10176
+    },
+    {
+      "epoch": 1.8154491124788155,
+      "grad_norm": 0.5567406415939331,
+      "learning_rate": 1.040312739100896e-05,
+      "loss": 0.6,
+      "step": 10177
+    },
+    {
+      "epoch": 1.8156275086968157,
+      "grad_norm": 0.5131311416625977,
+      "learning_rate": 1.038313608451985e-05,
+      "loss": 0.4397,
+      "step": 10178
+    },
+    {
+      "epoch": 1.8158059049148159,
+      "grad_norm": 0.45174816250801086,
+      "learning_rate": 1.0363163597436165e-05,
+      "loss": 0.389,
+      "step": 10179
+    },
+    {
+      "epoch": 1.8159843011328158,
+      "grad_norm": 0.47523677349090576,
+      "learning_rate": 1.0343209931326453e-05,
+      "loss": 0.3687,
+      "step": 10180
+    },
+    {
+      "epoch": 1.816162697350816,
+      "grad_norm": 0.49906423687934875,
+      "learning_rate": 1.0323275087757905e-05,
+      "loss": 0.4489,
+      "step": 10181
+    },
+    {
+      "epoch": 1.8163410935688162,
+      "grad_norm": 0.4328477382659912,
+      "learning_rate": 1.0303359068296187e-05,
+      "loss": 0.307,
+      "step": 10182
+    },
+    {
+      "epoch": 1.8165194897868164,
+      "grad_norm": 0.48864254355430603,
+      "learning_rate": 1.0283461874505545e-05,
+      "loss": 0.4578,
+      "step": 10183
+    },
+    {
+      "epoch": 1.8166978860048166,
+      "grad_norm": 0.5428241491317749,
+      "learning_rate": 1.0263583507948592e-05,
+      "loss": 0.5653,
+      "step": 10184
+    },
+    {
+      "epoch": 1.8168762822228168,
+      "grad_norm": 0.7537945508956909,
+      "learning_rate": 1.0243723970186663e-05,
+      "loss": 0.4563,
+      "step": 10185
+    },
+    {
+      "epoch": 1.817054678440817,
+      "grad_norm": 1.7523473501205444,
+      "learning_rate": 1.0223883262779455e-05,
+      "loss": 0.3763,
+      "step": 10186
+    },
+    {
+      "epoch": 1.8172330746588172,
+      "grad_norm": 0.5494785308837891,
+      "learning_rate": 1.0204061387285274e-05,
+      "loss": 0.5098,
+      "step": 10187
+    },
+    {
+      "epoch": 1.8174114708768174,
+      "grad_norm": 0.5195872187614441,
+      "learning_rate": 1.0184258345260933e-05,
+      "loss": 0.476,
+      "step": 10188
+    },
+    {
+      "epoch": 1.8175898670948176,
+      "grad_norm": 0.674954891204834,
+      "learning_rate": 1.0164474138261714e-05,
+      "loss": 0.447,
+      "step": 10189
+    },
+    {
+      "epoch": 1.8177682633128178,
+      "grad_norm": 0.5051981806755066,
+      "learning_rate": 1.0144708767841514e-05,
+      "loss": 0.4994,
+      "step": 10190
+    },
+    {
+      "epoch": 1.817946659530818,
+      "grad_norm": 0.5411693453788757,
+      "learning_rate": 1.0124962235552647e-05,
+      "loss": 0.504,
+      "step": 10191
+    },
+    {
+      "epoch": 1.8181250557488182,
+      "grad_norm": 0.5327993631362915,
+      "learning_rate": 1.0105234542946013e-05,
+      "loss": 0.5145,
+      "step": 10192
+    },
+    {
+      "epoch": 1.8183034519668184,
+      "grad_norm": 0.5137004256248474,
+      "learning_rate": 1.0085525691571063e-05,
+      "loss": 0.5566,
+      "step": 10193
+    },
+    {
+      "epoch": 1.8184818481848186,
+      "grad_norm": 0.4950316250324249,
+      "learning_rate": 1.0065835682975644e-05,
+      "loss": 0.4582,
+      "step": 10194
+    },
+    {
+      "epoch": 1.8186602444028188,
+      "grad_norm": 0.6043919324874878,
+      "learning_rate": 1.0046164518706269e-05,
+      "loss": 0.6262,
+      "step": 10195
+    },
+    {
+      "epoch": 1.8188386406208188,
+      "grad_norm": 0.48688915371894836,
+      "learning_rate": 1.0026512200307841e-05,
+      "loss": 0.4427,
+      "step": 10196
+    },
+    {
+      "epoch": 1.819017036838819,
+      "grad_norm": 0.5831645727157593,
+      "learning_rate": 1.0006878729323905e-05,
+      "loss": 0.6149,
+      "step": 10197
+    },
+    {
+      "epoch": 1.8191954330568192,
+      "grad_norm": 0.5436547994613647,
+      "learning_rate": 9.987264107296445e-06,
+      "loss": 0.541,
+      "step": 10198
+    },
+    {
+      "epoch": 1.8193738292748194,
+      "grad_norm": 0.4676104784011841,
+      "learning_rate": 9.967668335765927e-06,
+      "loss": 0.448,
+      "step": 10199
+    },
+    {
+      "epoch": 1.8195522254928196,
+      "grad_norm": 0.5124318599700928,
+      "learning_rate": 9.948091416271482e-06,
+      "loss": 0.4321,
+      "step": 10200
+    },
+    {
+      "epoch": 1.8197306217108198,
+      "grad_norm": 0.5362337231636047,
+      "learning_rate": 9.928533350350627e-06,
+      "loss": 0.4066,
+      "step": 10201
+    },
+    {
+      "epoch": 1.8199090179288198,
+      "grad_norm": 0.4841192066669464,
+      "learning_rate": 9.908994139539467e-06,
+      "loss": 0.4793,
+      "step": 10202
+    },
+    {
+      "epoch": 1.82008741414682,
+      "grad_norm": 0.577235996723175,
+      "learning_rate": 9.889473785372554e-06,
+      "loss": 0.5232,
+      "step": 10203
+    },
+    {
+      "epoch": 1.8202658103648202,
+      "grad_norm": 0.5391908288002014,
+      "learning_rate": 9.869972289383078e-06,
+      "loss": 0.4879,
+      "step": 10204
+    },
+    {
+      "epoch": 1.8204442065828204,
+      "grad_norm": 0.5783559679985046,
+      "learning_rate": 9.85048965310259e-06,
+      "loss": 0.5618,
+      "step": 10205
+    },
+    {
+      "epoch": 1.8206226028008206,
+      "grad_norm": 0.47223010659217834,
+      "learning_rate": 9.831025878061366e-06,
+      "loss": 0.4077,
+      "step": 10206
+    },
+    {
+      "epoch": 1.8208009990188208,
+      "grad_norm": 0.5232735872268677,
+      "learning_rate": 9.811580965787964e-06,
+      "loss": 0.5183,
+      "step": 10207
+    },
+    {
+      "epoch": 1.820979395236821,
+      "grad_norm": 0.5101804137229919,
+      "learning_rate": 9.792154917809631e-06,
+      "loss": 0.4681,
+      "step": 10208
+    },
+    {
+      "epoch": 1.8211577914548212,
+      "grad_norm": 0.5707786679267883,
+      "learning_rate": 9.772747735652122e-06,
+      "loss": 0.6249,
+      "step": 10209
+    },
+    {
+      "epoch": 1.8213361876728213,
+      "grad_norm": 0.5304009318351746,
+      "learning_rate": 9.753359420839631e-06,
+      "loss": 0.4669,
+      "step": 10210
+    },
+    {
+      "epoch": 1.8215145838908215,
+      "grad_norm": 0.555030107498169,
+      "learning_rate": 9.733989974894858e-06,
+      "loss": 0.5323,
+      "step": 10211
+    },
+    {
+      "epoch": 1.8216929801088217,
+      "grad_norm": 0.5161325931549072,
+      "learning_rate": 9.71463939933917e-06,
+      "loss": 0.4242,
+      "step": 10212
+    },
+    {
+      "epoch": 1.821871376326822,
+      "grad_norm": 0.4960889518260956,
+      "learning_rate": 9.69530769569224e-06,
+      "loss": 0.4857,
+      "step": 10213
+    },
+    {
+      "epoch": 1.8220497725448221,
+      "grad_norm": 0.5438405871391296,
+      "learning_rate": 9.675994865472492e-06,
+      "loss": 0.5828,
+      "step": 10214
+    },
+    {
+      "epoch": 1.8222281687628223,
+      "grad_norm": 0.5128703713417053,
+      "learning_rate": 9.656700910196631e-06,
+      "loss": 0.4462,
+      "step": 10215
+    },
+    {
+      "epoch": 1.8224065649808225,
+      "grad_norm": 0.506435751914978,
+      "learning_rate": 9.637425831380109e-06,
+      "loss": 0.4737,
+      "step": 10216
+    },
+    {
+      "epoch": 1.8225849611988227,
+      "grad_norm": 0.5082954168319702,
+      "learning_rate": 9.618169630536688e-06,
+      "loss": 0.5197,
+      "step": 10217
+    },
+    {
+      "epoch": 1.8227633574168227,
+      "grad_norm": 0.5157825946807861,
+      "learning_rate": 9.598932309178798e-06,
+      "loss": 0.4332,
+      "step": 10218
+    },
+    {
+      "epoch": 1.822941753634823,
+      "grad_norm": 0.5034956932067871,
+      "learning_rate": 9.579713868817313e-06,
+      "loss": 0.4036,
+      "step": 10219
+    },
+    {
+      "epoch": 1.8231201498528231,
+      "grad_norm": 0.566260576248169,
+      "learning_rate": 9.560514310961637e-06,
+      "loss": 0.6078,
+      "step": 10220
+    },
+    {
+      "epoch": 1.8232985460708233,
+      "grad_norm": 0.5446032881736755,
+      "learning_rate": 9.541333637119704e-06,
+      "loss": 0.449,
+      "step": 10221
+    },
+    {
+      "epoch": 1.8234769422888235,
+      "grad_norm": 0.5551034808158875,
+      "learning_rate": 9.522171848797917e-06,
+      "loss": 0.4385,
+      "step": 10222
+    },
+    {
+      "epoch": 1.8236553385068237,
+      "grad_norm": 0.5290305018424988,
+      "learning_rate": 9.503028947501269e-06,
+      "loss": 0.3873,
+      "step": 10223
+    },
+    {
+      "epoch": 1.8238337347248237,
+      "grad_norm": 0.5309109091758728,
+      "learning_rate": 9.483904934733278e-06,
+      "loss": 0.5638,
+      "step": 10224
+    },
+    {
+      "epoch": 1.8240121309428239,
+      "grad_norm": 0.5156473517417908,
+      "learning_rate": 9.464799811995855e-06,
+      "loss": 0.4597,
+      "step": 10225
+    },
+    {
+      "epoch": 1.824190527160824,
+      "grad_norm": 0.5108242630958557,
+      "learning_rate": 9.44571358078955e-06,
+      "loss": 0.4872,
+      "step": 10226
+    },
+    {
+      "epoch": 1.8243689233788243,
+      "grad_norm": 0.45744287967681885,
+      "learning_rate": 9.426646242613385e-06,
+      "loss": 0.3599,
+      "step": 10227
+    },
+    {
+      "epoch": 1.8245473195968245,
+      "grad_norm": 0.5380007028579712,
+      "learning_rate": 9.407597798964911e-06,
+      "loss": 0.4486,
+      "step": 10228
+    },
+    {
+      "epoch": 1.8247257158148247,
+      "grad_norm": 0.5029316544532776,
+      "learning_rate": 9.388568251340157e-06,
+      "loss": 0.4662,
+      "step": 10229
+    },
+    {
+      "epoch": 1.8249041120328249,
+      "grad_norm": 0.5693724751472473,
+      "learning_rate": 9.369557601233703e-06,
+      "loss": 0.4787,
+      "step": 10230
+    },
+    {
+      "epoch": 1.825082508250825,
+      "grad_norm": 0.49826717376708984,
+      "learning_rate": 9.350565850138688e-06,
+      "loss": 0.4471,
+      "step": 10231
+    },
+    {
+      "epoch": 1.8252609044688253,
+      "grad_norm": 0.6438023447990417,
+      "learning_rate": 9.331592999546612e-06,
+      "loss": 0.6015,
+      "step": 10232
+    },
+    {
+      "epoch": 1.8254393006868255,
+      "grad_norm": 0.6153572797775269,
+      "learning_rate": 9.312639050947702e-06,
+      "loss": 0.7253,
+      "step": 10233
+    },
+    {
+      "epoch": 1.8256176969048257,
+      "grad_norm": 0.508865237236023,
+      "learning_rate": 9.293704005830488e-06,
+      "loss": 0.4195,
+      "step": 10234
+    },
+    {
+      "epoch": 1.8257960931228259,
+      "grad_norm": 0.533497154712677,
+      "learning_rate": 9.274787865682227e-06,
+      "loss": 0.4233,
+      "step": 10235
+    },
+    {
+      "epoch": 1.825974489340826,
+      "grad_norm": 0.5289534330368042,
+      "learning_rate": 9.255890631988505e-06,
+      "loss": 0.5847,
+      "step": 10236
+    },
+    {
+      "epoch": 1.8261528855588263,
+      "grad_norm": 0.5981556177139282,
+      "learning_rate": 9.237012306233555e-06,
+      "loss": 0.7135,
+      "step": 10237
+    },
+    {
+      "epoch": 1.8263312817768265,
+      "grad_norm": 0.5175567269325256,
+      "learning_rate": 9.21815288990005e-06,
+      "loss": 0.4407,
+      "step": 10238
+    },
+    {
+      "epoch": 1.8265096779948267,
+      "grad_norm": 0.4687216877937317,
+      "learning_rate": 9.199312384469166e-06,
+      "loss": 0.408,
+      "step": 10239
+    },
+    {
+      "epoch": 1.8266880742128266,
+      "grad_norm": 0.5143327713012695,
+      "learning_rate": 9.180490791420693e-06,
+      "loss": 0.5345,
+      "step": 10240
+    },
+    {
+      "epoch": 1.8268664704308268,
+      "grad_norm": 0.5955470204353333,
+      "learning_rate": 9.161688112232836e-06,
+      "loss": 0.6327,
+      "step": 10241
+    },
+    {
+      "epoch": 1.827044866648827,
+      "grad_norm": 0.482693612575531,
+      "learning_rate": 9.142904348382359e-06,
+      "loss": 0.39,
+      "step": 10242
+    },
+    {
+      "epoch": 1.8272232628668272,
+      "grad_norm": 0.6863899230957031,
+      "learning_rate": 9.124139501344496e-06,
+      "loss": 0.6892,
+      "step": 10243
+    },
+    {
+      "epoch": 1.8274016590848274,
+      "grad_norm": 0.49588871002197266,
+      "learning_rate": 9.105393572593102e-06,
+      "loss": 0.3929,
+      "step": 10244
+    },
+    {
+      "epoch": 1.8275800553028276,
+      "grad_norm": 0.5905642509460449,
+      "learning_rate": 9.086666563600437e-06,
+      "loss": 0.7023,
+      "step": 10245
+    },
+    {
+      "epoch": 1.8277584515208276,
+      "grad_norm": 0.5300746560096741,
+      "learning_rate": 9.067958475837274e-06,
+      "loss": 0.5067,
+      "step": 10246
+    },
+    {
+      "epoch": 1.8279368477388278,
+      "grad_norm": 0.5370733141899109,
+      "learning_rate": 9.04926931077299e-06,
+      "loss": 0.5826,
+      "step": 10247
+    },
+    {
+      "epoch": 1.828115243956828,
+      "grad_norm": 0.5125030279159546,
+      "learning_rate": 9.030599069875383e-06,
+      "loss": 0.4693,
+      "step": 10248
+    },
+    {
+      "epoch": 1.8282936401748282,
+      "grad_norm": 0.48797720670700073,
+      "learning_rate": 9.011947754610839e-06,
+      "loss": 0.4299,
+      "step": 10249
+    },
+    {
+      "epoch": 1.8284720363928284,
+      "grad_norm": 0.5317216515541077,
+      "learning_rate": 8.99331536644421e-06,
+      "loss": 0.4746,
+      "step": 10250
+    },
+    {
+      "epoch": 1.8286504326108286,
+      "grad_norm": 0.5748763680458069,
+      "learning_rate": 8.974701906838884e-06,
+      "loss": 0.51,
+      "step": 10251
+    },
+    {
+      "epoch": 1.8288288288288288,
+      "grad_norm": 0.573381781578064,
+      "learning_rate": 8.956107377256772e-06,
+      "loss": 0.6049,
+      "step": 10252
+    },
+    {
+      "epoch": 1.829007225046829,
+      "grad_norm": 0.4922345280647278,
+      "learning_rate": 8.937531779158181e-06,
+      "loss": 0.371,
+      "step": 10253
+    },
+    {
+      "epoch": 1.8291856212648292,
+      "grad_norm": 0.494431734085083,
+      "learning_rate": 8.918975114002192e-06,
+      "loss": 0.4184,
+      "step": 10254
+    },
+    {
+      "epoch": 1.8293640174828294,
+      "grad_norm": 0.5176249146461487,
+      "learning_rate": 8.900437383246084e-06,
+      "loss": 0.45,
+      "step": 10255
+    },
+    {
+      "epoch": 1.8295424137008296,
+      "grad_norm": 0.4918590188026428,
+      "learning_rate": 8.881918588345917e-06,
+      "loss": 0.5091,
+      "step": 10256
+    },
+    {
+      "epoch": 1.8297208099188298,
+      "grad_norm": 0.5216599702835083,
+      "learning_rate": 8.863418730756106e-06,
+      "loss": 0.4921,
+      "step": 10257
+    },
+    {
+      "epoch": 1.82989920613683,
+      "grad_norm": 0.45257261395454407,
+      "learning_rate": 8.844937811929605e-06,
+      "loss": 0.3359,
+      "step": 10258
+    },
+    {
+      "epoch": 1.8300776023548302,
+      "grad_norm": 0.5172678828239441,
+      "learning_rate": 8.826475833317914e-06,
+      "loss": 0.5071,
+      "step": 10259
+    },
+    {
+      "epoch": 1.8302559985728304,
+      "grad_norm": 0.5505502223968506,
+      "learning_rate": 8.808032796371018e-06,
+      "loss": 0.6637,
+      "step": 10260
+    },
+    {
+      "epoch": 1.8304343947908306,
+      "grad_norm": 0.48044058680534363,
+      "learning_rate": 8.78960870253745e-06,
+      "loss": 0.4304,
+      "step": 10261
+    },
+    {
+      "epoch": 1.8306127910088306,
+      "grad_norm": 0.4884861707687378,
+      "learning_rate": 8.77120355326419e-06,
+      "loss": 0.3665,
+      "step": 10262
+    },
+    {
+      "epoch": 1.8307911872268308,
+      "grad_norm": 0.4926454424858093,
+      "learning_rate": 8.752817349996806e-06,
+      "loss": 0.3983,
+      "step": 10263
+    },
+    {
+      "epoch": 1.830969583444831,
+      "grad_norm": 0.5212196111679077,
+      "learning_rate": 8.734450094179309e-06,
+      "loss": 0.5036,
+      "step": 10264
+    },
+    {
+      "epoch": 1.8311479796628312,
+      "grad_norm": 0.4858270287513733,
+      "learning_rate": 8.716101787254321e-06,
+      "loss": 0.467,
+      "step": 10265
+    },
+    {
+      "epoch": 1.8313263758808314,
+      "grad_norm": 0.5602033138275146,
+      "learning_rate": 8.697772430662858e-06,
+      "loss": 0.5441,
+      "step": 10266
+    },
+    {
+      "epoch": 1.8315047720988316,
+      "grad_norm": 0.5834726095199585,
+      "learning_rate": 8.679462025844464e-06,
+      "loss": 0.6227,
+      "step": 10267
+    },
+    {
+      "epoch": 1.8316831683168315,
+      "grad_norm": 0.5684033632278442,
+      "learning_rate": 8.66117057423732e-06,
+      "loss": 0.57,
+      "step": 10268
+    },
+    {
+      "epoch": 1.8318615645348317,
+      "grad_norm": 0.4957300126552582,
+      "learning_rate": 8.642898077277944e-06,
+      "loss": 0.4303,
+      "step": 10269
+    },
+    {
+      "epoch": 1.832039960752832,
+      "grad_norm": 0.5206865668296814,
+      "learning_rate": 8.624644536401521e-06,
+      "loss": 0.5213,
+      "step": 10270
+    },
+    {
+      "epoch": 1.8322183569708321,
+      "grad_norm": 0.5535826683044434,
+      "learning_rate": 8.606409953041627e-06,
+      "loss": 0.5366,
+      "step": 10271
+    },
+    {
+      "epoch": 1.8323967531888323,
+      "grad_norm": 0.5583673715591431,
+      "learning_rate": 8.588194328630422e-06,
+      "loss": 0.5131,
+      "step": 10272
+    },
+    {
+      "epoch": 1.8325751494068325,
+      "grad_norm": 0.45444154739379883,
+      "learning_rate": 8.569997664598567e-06,
+      "loss": 0.2948,
+      "step": 10273
+    },
+    {
+      "epoch": 1.8327535456248327,
+      "grad_norm": 0.5591797232627869,
+      "learning_rate": 8.55181996237514e-06,
+      "loss": 0.546,
+      "step": 10274
+    },
+    {
+      "epoch": 1.832931941842833,
+      "grad_norm": 0.6100978255271912,
+      "learning_rate": 8.533661223387946e-06,
+      "loss": 0.4099,
+      "step": 10275
+    },
+    {
+      "epoch": 1.8331103380608331,
+      "grad_norm": 0.5809532403945923,
+      "learning_rate": 8.515521449063036e-06,
+      "loss": 0.5211,
+      "step": 10276
+    },
+    {
+      "epoch": 1.8332887342788333,
+      "grad_norm": 0.5336730480194092,
+      "learning_rate": 8.497400640825186e-06,
+      "loss": 0.5875,
+      "step": 10277
+    },
+    {
+      "epoch": 1.8334671304968335,
+      "grad_norm": 0.5325056910514832,
+      "learning_rate": 8.47929880009754e-06,
+      "loss": 0.4842,
+      "step": 10278
+    },
+    {
+      "epoch": 1.8336455267148337,
+      "grad_norm": 0.528160810470581,
+      "learning_rate": 8.461215928301819e-06,
+      "loss": 0.5506,
+      "step": 10279
+    },
+    {
+      "epoch": 1.833823922932834,
+      "grad_norm": 0.5754840970039368,
+      "learning_rate": 8.443152026858303e-06,
+      "loss": 0.4679,
+      "step": 10280
+    },
+    {
+      "epoch": 1.8340023191508341,
+      "grad_norm": 0.5175067782402039,
+      "learning_rate": 8.425107097185636e-06,
+      "loss": 0.5501,
+      "step": 10281
+    },
+    {
+      "epoch": 1.8341807153688343,
+      "grad_norm": 0.47843798995018005,
+      "learning_rate": 8.407081140701128e-06,
+      "loss": 0.3725,
+      "step": 10282
+    },
+    {
+      "epoch": 1.8343591115868345,
+      "grad_norm": 0.5563110709190369,
+      "learning_rate": 8.389074158820453e-06,
+      "loss": 0.5422,
+      "step": 10283
+    },
+    {
+      "epoch": 1.8345375078048345,
+      "grad_norm": 0.5997392535209656,
+      "learning_rate": 8.371086152957952e-06,
+      "loss": 0.6135,
+      "step": 10284
+    },
+    {
+      "epoch": 1.8347159040228347,
+      "grad_norm": 0.5224518775939941,
+      "learning_rate": 8.353117124526382e-06,
+      "loss": 0.4452,
+      "step": 10285
+    },
+    {
+      "epoch": 1.8348943002408349,
+      "grad_norm": 0.5603434443473816,
+      "learning_rate": 8.33516707493695e-06,
+      "loss": 0.5088,
+      "step": 10286
+    },
+    {
+      "epoch": 1.835072696458835,
+      "grad_norm": 0.5151491761207581,
+      "learning_rate": 8.317236005599554e-06,
+      "loss": 0.5069,
+      "step": 10287
+    },
+    {
+      "epoch": 1.8352510926768353,
+      "grad_norm": 0.46216633915901184,
+      "learning_rate": 8.299323917922402e-06,
+      "loss": 0.3475,
+      "step": 10288
+    },
+    {
+      "epoch": 1.8354294888948355,
+      "grad_norm": 0.5850088000297546,
+      "learning_rate": 8.281430813312368e-06,
+      "loss": 0.6099,
+      "step": 10289
+    },
+    {
+      "epoch": 1.8356078851128355,
+      "grad_norm": 0.510653018951416,
+      "learning_rate": 8.263556693174745e-06,
+      "loss": 0.4093,
+      "step": 10290
+    },
+    {
+      "epoch": 1.8357862813308357,
+      "grad_norm": 0.6367735862731934,
+      "learning_rate": 8.245701558913327e-06,
+      "loss": 0.5877,
+      "step": 10291
+    },
+    {
+      "epoch": 1.8359646775488359,
+      "grad_norm": 0.5245457887649536,
+      "learning_rate": 8.227865411930492e-06,
+      "loss": 0.4942,
+      "step": 10292
+    },
+    {
+      "epoch": 1.836143073766836,
+      "grad_norm": 0.5388858318328857,
+      "learning_rate": 8.210048253627034e-06,
+      "loss": 0.5077,
+      "step": 10293
+    },
+    {
+      "epoch": 1.8363214699848363,
+      "grad_norm": 0.4671160876750946,
+      "learning_rate": 8.192250085402364e-06,
+      "loss": 0.4215,
+      "step": 10294
+    },
+    {
+      "epoch": 1.8364998662028365,
+      "grad_norm": 0.5302553772926331,
+      "learning_rate": 8.174470908654309e-06,
+      "loss": 0.4563,
+      "step": 10295
+    },
+    {
+      "epoch": 1.8366782624208366,
+      "grad_norm": 0.5391623377799988,
+      "learning_rate": 8.156710724779249e-06,
+      "loss": 0.5735,
+      "step": 10296
+    },
+    {
+      "epoch": 1.8368566586388368,
+      "grad_norm": 0.5140558481216431,
+      "learning_rate": 8.138969535172014e-06,
+      "loss": 0.4466,
+      "step": 10297
+    },
+    {
+      "epoch": 1.837035054856837,
+      "grad_norm": 0.497837096452713,
+      "learning_rate": 8.121247341226074e-06,
+      "loss": 0.4269,
+      "step": 10298
+    },
+    {
+      "epoch": 1.8372134510748372,
+      "grad_norm": 0.49367204308509827,
+      "learning_rate": 8.103544144333259e-06,
+      "loss": 0.4433,
+      "step": 10299
+    },
+    {
+      "epoch": 1.8373918472928374,
+      "grad_norm": 0.5372571349143982,
+      "learning_rate": 8.085859945883984e-06,
+      "loss": 0.5503,
+      "step": 10300
+    },
+    {
+      "epoch": 1.8375702435108376,
+      "grad_norm": 0.5628436207771301,
+      "learning_rate": 8.068194747267193e-06,
+      "loss": 0.5568,
+      "step": 10301
+    },
+    {
+      "epoch": 1.8377486397288378,
+      "grad_norm": 0.5379055142402649,
+      "learning_rate": 8.050548549870252e-06,
+      "loss": 0.5478,
+      "step": 10302
+    },
+    {
+      "epoch": 1.837927035946838,
+      "grad_norm": 1.1105915307998657,
+      "learning_rate": 8.032921355079132e-06,
+      "loss": 0.4982,
+      "step": 10303
+    },
+    {
+      "epoch": 1.8381054321648382,
+      "grad_norm": 0.4970918893814087,
+      "learning_rate": 8.015313164278227e-06,
+      "loss": 0.5111,
+      "step": 10304
+    },
+    {
+      "epoch": 1.8382838283828384,
+      "grad_norm": 0.4896716773509979,
+      "learning_rate": 7.997723978850486e-06,
+      "loss": 0.3981,
+      "step": 10305
+    },
+    {
+      "epoch": 1.8384622246008384,
+      "grad_norm": 0.5600361824035645,
+      "learning_rate": 7.980153800177387e-06,
+      "loss": 0.4834,
+      "step": 10306
+    },
+    {
+      "epoch": 1.8386406208188386,
+      "grad_norm": 0.560856819152832,
+      "learning_rate": 7.962602629638827e-06,
+      "loss": 0.6468,
+      "step": 10307
+    },
+    {
+      "epoch": 1.8388190170368388,
+      "grad_norm": 0.46853670477867126,
+      "learning_rate": 7.945070468613313e-06,
+      "loss": 0.4358,
+      "step": 10308
+    },
+    {
+      "epoch": 1.838997413254839,
+      "grad_norm": 0.4796409606933594,
+      "learning_rate": 7.9275573184778e-06,
+      "loss": 0.4112,
+      "step": 10309
+    },
+    {
+      "epoch": 1.8391758094728392,
+      "grad_norm": 0.5527827143669128,
+      "learning_rate": 7.910063180607775e-06,
+      "loss": 0.63,
+      "step": 10310
+    },
+    {
+      "epoch": 1.8393542056908394,
+      "grad_norm": 0.5861801505088806,
+      "learning_rate": 7.892588056377214e-06,
+      "loss": 0.6435,
+      "step": 10311
+    },
+    {
+      "epoch": 1.8395326019088394,
+      "grad_norm": 0.5089249610900879,
+      "learning_rate": 7.875131947158554e-06,
+      "loss": 0.4791,
+      "step": 10312
+    },
+    {
+      "epoch": 1.8397109981268396,
+      "grad_norm": 0.5814892053604126,
+      "learning_rate": 7.857694854322888e-06,
+      "loss": 0.5377,
+      "step": 10313
+    },
+    {
+      "epoch": 1.8398893943448398,
+      "grad_norm": 0.516287088394165,
+      "learning_rate": 7.840276779239625e-06,
+      "loss": 0.3958,
+      "step": 10314
+    },
+    {
+      "epoch": 1.84006779056284,
+      "grad_norm": 0.5036550164222717,
+      "learning_rate": 7.822877723276834e-06,
+      "loss": 0.518,
+      "step": 10315
+    },
+    {
+      "epoch": 1.8402461867808402,
+      "grad_norm": 0.4933060109615326,
+      "learning_rate": 7.805497687801006e-06,
+      "loss": 0.4199,
+      "step": 10316
+    },
+    {
+      "epoch": 1.8404245829988404,
+      "grad_norm": 0.4610743522644043,
+      "learning_rate": 7.78813667417716e-06,
+      "loss": 0.3651,
+      "step": 10317
+    },
+    {
+      "epoch": 1.8406029792168406,
+      "grad_norm": 0.5514711737632751,
+      "learning_rate": 7.770794683768845e-06,
+      "loss": 0.4906,
+      "step": 10318
+    },
+    {
+      "epoch": 1.8407813754348408,
+      "grad_norm": 0.6051323413848877,
+      "learning_rate": 7.753471717938054e-06,
+      "loss": 0.6172,
+      "step": 10319
+    },
+    {
+      "epoch": 1.840959771652841,
+      "grad_norm": 0.4640166461467743,
+      "learning_rate": 7.736167778045367e-06,
+      "loss": 0.4343,
+      "step": 10320
+    },
+    {
+      "epoch": 1.8411381678708412,
+      "grad_norm": 0.4809337556362152,
+      "learning_rate": 7.718882865449806e-06,
+      "loss": 0.3689,
+      "step": 10321
+    },
+    {
+      "epoch": 1.8413165640888414,
+      "grad_norm": 0.5500307083129883,
+      "learning_rate": 7.701616981508924e-06,
+      "loss": 0.5312,
+      "step": 10322
+    },
+    {
+      "epoch": 1.8414949603068416,
+      "grad_norm": 0.5196850299835205,
+      "learning_rate": 7.684370127578749e-06,
+      "loss": 0.4428,
+      "step": 10323
+    },
+    {
+      "epoch": 1.8416733565248418,
+      "grad_norm": 0.5277876853942871,
+      "learning_rate": 7.66714230501392e-06,
+      "loss": 0.5147,
+      "step": 10324
+    },
+    {
+      "epoch": 1.841851752742842,
+      "grad_norm": 0.4931204915046692,
+      "learning_rate": 7.649933515167407e-06,
+      "loss": 0.4189,
+      "step": 10325
+    },
+    {
+      "epoch": 1.8420301489608422,
+      "grad_norm": 0.5108685493469238,
+      "learning_rate": 7.632743759390826e-06,
+      "loss": 0.5112,
+      "step": 10326
+    },
+    {
+      "epoch": 1.8422085451788424,
+      "grad_norm": 0.5715261697769165,
+      "learning_rate": 7.61557303903429e-06,
+      "loss": 0.5237,
+      "step": 10327
+    },
+    {
+      "epoch": 1.8423869413968423,
+      "grad_norm": 0.5731346607208252,
+      "learning_rate": 7.5984213554462775e-06,
+      "loss": 0.6463,
+      "step": 10328
+    },
+    {
+      "epoch": 1.8425653376148425,
+      "grad_norm": 0.4316110610961914,
+      "learning_rate": 7.581288709973988e-06,
+      "loss": 0.3569,
+      "step": 10329
+    },
+    {
+      "epoch": 1.8427437338328427,
+      "grad_norm": 0.512586772441864,
+      "learning_rate": 7.564175103962956e-06,
+      "loss": 0.4411,
+      "step": 10330
+    },
+    {
+      "epoch": 1.842922130050843,
+      "grad_norm": 0.5421055555343628,
+      "learning_rate": 7.547080538757245e-06,
+      "loss": 0.4933,
+      "step": 10331
+    },
+    {
+      "epoch": 1.8431005262688431,
+      "grad_norm": 0.5586483478546143,
+      "learning_rate": 7.5300050156995315e-06,
+      "loss": 0.5065,
+      "step": 10332
+    },
+    {
+      "epoch": 1.8432789224868433,
+      "grad_norm": 0.6223009824752808,
+      "learning_rate": 7.5129485361308534e-06,
+      "loss": 0.4358,
+      "step": 10333
+    },
+    {
+      "epoch": 1.8434573187048433,
+      "grad_norm": 0.5178797841072083,
+      "learning_rate": 7.49591110139089e-06,
+      "loss": 0.4661,
+      "step": 10334
+    },
+    {
+      "epoch": 1.8436357149228435,
+      "grad_norm": 0.5489305257797241,
+      "learning_rate": 7.478892712817681e-06,
+      "loss": 0.5131,
+      "step": 10335
+    },
+    {
+      "epoch": 1.8438141111408437,
+      "grad_norm": 0.5495460033416748,
+      "learning_rate": 7.4618933717478796e-06,
+      "loss": 0.6006,
+      "step": 10336
+    },
+    {
+      "epoch": 1.843992507358844,
+      "grad_norm": 0.4898929297924042,
+      "learning_rate": 7.444913079516613e-06,
+      "loss": 0.3961,
+      "step": 10337
+    },
+    {
+      "epoch": 1.844170903576844,
+      "grad_norm": 0.48516249656677246,
+      "learning_rate": 7.42795183745748e-06,
+      "loss": 0.3852,
+      "step": 10338
+    },
+    {
+      "epoch": 1.8443492997948443,
+      "grad_norm": 0.6237598061561584,
+      "learning_rate": 7.411009646902639e-06,
+      "loss": 0.6107,
+      "step": 10339
+    },
+    {
+      "epoch": 1.8445276960128445,
+      "grad_norm": 0.49778953194618225,
+      "learning_rate": 7.394086509182663e-06,
+      "loss": 0.3913,
+      "step": 10340
+    },
+    {
+      "epoch": 1.8447060922308447,
+      "grad_norm": 0.5452266931533813,
+      "learning_rate": 7.377182425626766e-06,
+      "loss": 0.5521,
+      "step": 10341
+    },
+    {
+      "epoch": 1.844884488448845,
+      "grad_norm": 0.6213571429252625,
+      "learning_rate": 7.360297397562527e-06,
+      "loss": 0.6084,
+      "step": 10342
+    },
+    {
+      "epoch": 1.845062884666845,
+      "grad_norm": 0.5283211469650269,
+      "learning_rate": 7.34343142631616e-06,
+      "loss": 0.5813,
+      "step": 10343
+    },
+    {
+      "epoch": 1.8452412808848453,
+      "grad_norm": 0.45486849546432495,
+      "learning_rate": 7.32658451321222e-06,
+      "loss": 0.2873,
+      "step": 10344
+    },
+    {
+      "epoch": 1.8454196771028455,
+      "grad_norm": 0.544691264629364,
+      "learning_rate": 7.3097566595738965e-06,
+      "loss": 0.5815,
+      "step": 10345
+    },
+    {
+      "epoch": 1.8455980733208457,
+      "grad_norm": 0.6090754866600037,
+      "learning_rate": 7.292947866722882e-06,
+      "loss": 0.567,
+      "step": 10346
+    },
+    {
+      "epoch": 1.8457764695388459,
+      "grad_norm": 0.5459246635437012,
+      "learning_rate": 7.276158135979288e-06,
+      "loss": 0.6544,
+      "step": 10347
+    },
+    {
+      "epoch": 1.845954865756846,
+      "grad_norm": 0.54819256067276,
+      "learning_rate": 7.259387468661782e-06,
+      "loss": 0.5531,
+      "step": 10348
+    },
+    {
+      "epoch": 1.8461332619748463,
+      "grad_norm": 0.5134831070899963,
+      "learning_rate": 7.242635866087505e-06,
+      "loss": 0.4531,
+      "step": 10349
+    },
+    {
+      "epoch": 1.8463116581928463,
+      "grad_norm": 0.5044652223587036,
+      "learning_rate": 7.225903329572181e-06,
+      "loss": 0.4553,
+      "step": 10350
+    },
+    {
+      "epoch": 1.8464900544108465,
+      "grad_norm": 0.4595658779144287,
+      "learning_rate": 7.209189860429899e-06,
+      "loss": 0.34,
+      "step": 10351
+    },
+    {
+      "epoch": 1.8466684506288467,
+      "grad_norm": 0.5554966330528259,
+      "learning_rate": 7.1924954599733864e-06,
+      "loss": 0.4627,
+      "step": 10352
+    },
+    {
+      "epoch": 1.8468468468468469,
+      "grad_norm": 0.49979686737060547,
+      "learning_rate": 7.175820129513788e-06,
+      "loss": 0.4567,
+      "step": 10353
+    },
+    {
+      "epoch": 1.847025243064847,
+      "grad_norm": 0.527464747428894,
+      "learning_rate": 7.159163870360752e-06,
+      "loss": 0.4319,
+      "step": 10354
+    },
+    {
+      "epoch": 1.8472036392828473,
+      "grad_norm": 0.5300561785697937,
+      "learning_rate": 7.142526683822537e-06,
+      "loss": 0.4145,
+      "step": 10355
+    },
+    {
+      "epoch": 1.8473820355008472,
+      "grad_norm": 0.5047762393951416,
+      "learning_rate": 7.125908571205708e-06,
+      "loss": 0.411,
+      "step": 10356
+    },
+    {
+      "epoch": 1.8475604317188474,
+      "grad_norm": 0.5059321522712708,
+      "learning_rate": 7.109309533815556e-06,
+      "loss": 0.4051,
+      "step": 10357
+    },
+    {
+      "epoch": 1.8477388279368476,
+      "grad_norm": 0.5237871408462524,
+      "learning_rate": 7.092729572955675e-06,
+      "loss": 0.4884,
+      "step": 10358
+    },
+    {
+      "epoch": 1.8479172241548478,
+      "grad_norm": 0.4973069131374359,
+      "learning_rate": 7.076168689928275e-06,
+      "loss": 0.3889,
+      "step": 10359
+    },
+    {
+      "epoch": 1.848095620372848,
+      "grad_norm": 0.5443868637084961,
+      "learning_rate": 7.059626886034093e-06,
+      "loss": 0.4669,
+      "step": 10360
+    },
+    {
+      "epoch": 1.8482740165908482,
+      "grad_norm": 0.5395627021789551,
+      "learning_rate": 7.0431041625722e-06,
+      "loss": 0.4496,
+      "step": 10361
+    },
+    {
+      "epoch": 1.8484524128088484,
+      "grad_norm": 0.5488491654396057,
+      "learning_rate": 7.026600520840393e-06,
+      "loss": 0.5293,
+      "step": 10362
+    },
+    {
+      "epoch": 1.8486308090268486,
+      "grad_norm": 0.5110378265380859,
+      "learning_rate": 7.010115962134855e-06,
+      "loss": 0.4904,
+      "step": 10363
+    },
+    {
+      "epoch": 1.8488092052448488,
+      "grad_norm": 0.5103933215141296,
+      "learning_rate": 6.993650487750192e-06,
+      "loss": 0.5069,
+      "step": 10364
+    },
+    {
+      "epoch": 1.848987601462849,
+      "grad_norm": 0.5104700326919556,
+      "learning_rate": 6.9772040989796725e-06,
+      "loss": 0.468,
+      "step": 10365
+    },
+    {
+      "epoch": 1.8491659976808492,
+      "grad_norm": 0.4703051745891571,
+      "learning_rate": 6.960776797114931e-06,
+      "loss": 0.4197,
+      "step": 10366
+    },
+    {
+      "epoch": 1.8493443938988494,
+      "grad_norm": 0.5158678889274597,
+      "learning_rate": 6.944368583446242e-06,
+      "loss": 0.4147,
+      "step": 10367
+    },
+    {
+      "epoch": 1.8495227901168496,
+      "grad_norm": 0.5142411589622498,
+      "learning_rate": 6.927979459262212e-06,
+      "loss": 0.4338,
+      "step": 10368
+    },
+    {
+      "epoch": 1.8497011863348498,
+      "grad_norm": 0.4859728515148163,
+      "learning_rate": 6.9116094258500905e-06,
+      "loss": 0.38,
+      "step": 10369
+    },
+    {
+      "epoch": 1.84987958255285,
+      "grad_norm": 0.5308949947357178,
+      "learning_rate": 6.895258484495515e-06,
+      "loss": 0.5024,
+      "step": 10370
+    },
+    {
+      "epoch": 1.8500579787708502,
+      "grad_norm": 0.46590912342071533,
+      "learning_rate": 6.878926636482791e-06,
+      "loss": 0.3446,
+      "step": 10371
+    },
+    {
+      "epoch": 1.8502363749888502,
+      "grad_norm": 0.5768349170684814,
+      "learning_rate": 6.862613883094504e-06,
+      "loss": 0.593,
+      "step": 10372
+    },
+    {
+      "epoch": 1.8504147712068504,
+      "grad_norm": 0.6099775433540344,
+      "learning_rate": 6.84632022561188e-06,
+      "loss": 0.5706,
+      "step": 10373
+    },
+    {
+      "epoch": 1.8505931674248506,
+      "grad_norm": 0.6040098071098328,
+      "learning_rate": 6.830045665314672e-06,
+      "loss": 0.6825,
+      "step": 10374
+    },
+    {
+      "epoch": 1.8507715636428508,
+      "grad_norm": 0.4853060841560364,
+      "learning_rate": 6.813790203480996e-06,
+      "loss": 0.4622,
+      "step": 10375
+    },
+    {
+      "epoch": 1.850949959860851,
+      "grad_norm": 0.5379960536956787,
+      "learning_rate": 6.7975538413875825e-06,
+      "loss": 0.5277,
+      "step": 10376
+    },
+    {
+      "epoch": 1.8511283560788512,
+      "grad_norm": 0.6292576193809509,
+      "learning_rate": 6.781336580309661e-06,
+      "loss": 0.6256,
+      "step": 10377
+    },
+    {
+      "epoch": 1.8513067522968512,
+      "grad_norm": 0.5415869355201721,
+      "learning_rate": 6.765138421520878e-06,
+      "loss": 0.3965,
+      "step": 10378
+    },
+    {
+      "epoch": 1.8514851485148514,
+      "grad_norm": 0.5575872659683228,
+      "learning_rate": 6.748959366293467e-06,
+      "loss": 0.5549,
+      "step": 10379
+    },
+    {
+      "epoch": 1.8516635447328516,
+      "grad_norm": 0.5051519870758057,
+      "learning_rate": 6.732799415898078e-06,
+      "loss": 0.4057,
+      "step": 10380
+    },
+    {
+      "epoch": 1.8518419409508518,
+      "grad_norm": 0.48078250885009766,
+      "learning_rate": 6.716658571603973e-06,
+      "loss": 0.3774,
+      "step": 10381
+    },
+    {
+      "epoch": 1.852020337168852,
+      "grad_norm": 0.6021243333816528,
+      "learning_rate": 6.7005368346787775e-06,
+      "loss": 0.5807,
+      "step": 10382
+    },
+    {
+      "epoch": 1.8521987333868521,
+      "grad_norm": 0.5509424209594727,
+      "learning_rate": 6.6844342063887565e-06,
+      "loss": 0.5441,
+      "step": 10383
+    },
+    {
+      "epoch": 1.8523771296048523,
+      "grad_norm": 0.5380014181137085,
+      "learning_rate": 6.6683506879985645e-06,
+      "loss": 0.583,
+      "step": 10384
+    },
+    {
+      "epoch": 1.8525555258228525,
+      "grad_norm": 0.5455968976020813,
+      "learning_rate": 6.652286280771358e-06,
+      "loss": 0.4776,
+      "step": 10385
+    },
+    {
+      "epoch": 1.8527339220408527,
+      "grad_norm": 0.5762505531311035,
+      "learning_rate": 6.636240985968906e-06,
+      "loss": 0.5963,
+      "step": 10386
+    },
+    {
+      "epoch": 1.852912318258853,
+      "grad_norm": 0.5499910712242126,
+      "learning_rate": 6.620214804851338e-06,
+      "loss": 0.5036,
+      "step": 10387
+    },
+    {
+      "epoch": 1.8530907144768531,
+      "grad_norm": 0.46007874608039856,
+      "learning_rate": 6.60420773867737e-06,
+      "loss": 0.3352,
+      "step": 10388
+    },
+    {
+      "epoch": 1.8532691106948533,
+      "grad_norm": 0.5644509792327881,
+      "learning_rate": 6.588219788704164e-06,
+      "loss": 0.5467,
+      "step": 10389
+    },
+    {
+      "epoch": 1.8534475069128535,
+      "grad_norm": 0.5105949640274048,
+      "learning_rate": 6.572250956187465e-06,
+      "loss": 0.4846,
+      "step": 10390
+    },
+    {
+      "epoch": 1.8536259031308537,
+      "grad_norm": 0.505388617515564,
+      "learning_rate": 6.556301242381379e-06,
+      "loss": 0.4123,
+      "step": 10391
+    },
+    {
+      "epoch": 1.853804299348854,
+      "grad_norm": 0.49062561988830566,
+      "learning_rate": 6.540370648538657e-06,
+      "loss": 0.3651,
+      "step": 10392
+    },
+    {
+      "epoch": 1.8539826955668541,
+      "grad_norm": 0.5342191457748413,
+      "learning_rate": 6.524459175910464e-06,
+      "loss": 0.5833,
+      "step": 10393
+    },
+    {
+      "epoch": 1.854161091784854,
+      "grad_norm": 0.5109807848930359,
+      "learning_rate": 6.508566825746437e-06,
+      "loss": 0.402,
+      "step": 10394
+    },
+    {
+      "epoch": 1.8543394880028543,
+      "grad_norm": 0.5528099536895752,
+      "learning_rate": 6.49269359929483e-06,
+      "loss": 0.5504,
+      "step": 10395
+    },
+    {
+      "epoch": 1.8545178842208545,
+      "grad_norm": 0.479790061712265,
+      "learning_rate": 6.476839497802256e-06,
+      "loss": 0.3927,
+      "step": 10396
+    },
+    {
+      "epoch": 1.8546962804388547,
+      "grad_norm": 0.5817665457725525,
+      "learning_rate": 6.461004522513913e-06,
+      "loss": 0.5102,
+      "step": 10397
+    },
+    {
+      "epoch": 1.854874676656855,
+      "grad_norm": 0.5208722352981567,
+      "learning_rate": 6.445188674673474e-06,
+      "loss": 0.489,
+      "step": 10398
+    },
+    {
+      "epoch": 1.855053072874855,
+      "grad_norm": 0.4927591383457184,
+      "learning_rate": 6.429391955523112e-06,
+      "loss": 0.469,
+      "step": 10399
+    },
+    {
+      "epoch": 1.855231469092855,
+      "grad_norm": 0.5992792844772339,
+      "learning_rate": 6.413614366303472e-06,
+      "loss": 0.5953,
+      "step": 10400
+    },
+    {
+      "epoch": 1.8554098653108553,
+      "grad_norm": 0.5083011984825134,
+      "learning_rate": 6.397855908253758e-06,
+      "loss": 0.5346,
+      "step": 10401
+    },
+    {
+      "epoch": 1.8555882615288555,
+      "grad_norm": 0.557491660118103,
+      "learning_rate": 6.382116582611591e-06,
+      "loss": 0.6326,
+      "step": 10402
+    },
+    {
+      "epoch": 1.8557666577468557,
+      "grad_norm": 0.5369499325752258,
+      "learning_rate": 6.36639639061315e-06,
+      "loss": 0.5539,
+      "step": 10403
+    },
+    {
+      "epoch": 1.8559450539648559,
+      "grad_norm": 0.5842743515968323,
+      "learning_rate": 6.350695333493112e-06,
+      "loss": 0.5387,
+      "step": 10404
+    },
+    {
+      "epoch": 1.856123450182856,
+      "grad_norm": 0.5030590295791626,
+      "learning_rate": 6.33501341248463e-06,
+      "loss": 0.4925,
+      "step": 10405
+    },
+    {
+      "epoch": 1.8563018464008563,
+      "grad_norm": 0.5017548203468323,
+      "learning_rate": 6.319350628819304e-06,
+      "loss": 0.4771,
+      "step": 10406
+    },
+    {
+      "epoch": 1.8564802426188565,
+      "grad_norm": 0.5642099380493164,
+      "learning_rate": 6.303706983727286e-06,
+      "loss": 0.5488,
+      "step": 10407
+    },
+    {
+      "epoch": 1.8566586388368567,
+      "grad_norm": 0.5220343470573425,
+      "learning_rate": 6.28808247843729e-06,
+      "loss": 0.5252,
+      "step": 10408
+    },
+    {
+      "epoch": 1.8568370350548569,
+      "grad_norm": 0.5058274865150452,
+      "learning_rate": 6.272477114176417e-06,
+      "loss": 0.4143,
+      "step": 10409
+    },
+    {
+      "epoch": 1.857015431272857,
+      "grad_norm": 0.5488356351852417,
+      "learning_rate": 6.2568908921703245e-06,
+      "loss": 0.5446,
+      "step": 10410
+    },
+    {
+      "epoch": 1.8571938274908573,
+      "grad_norm": 0.5424718856811523,
+      "learning_rate": 6.241323813643091e-06,
+      "loss": 0.4847,
+      "step": 10411
+    },
+    {
+      "epoch": 1.8573722237088575,
+      "grad_norm": 0.5151358246803284,
+      "learning_rate": 6.22577587981743e-06,
+      "loss": 0.4357,
+      "step": 10412
+    },
+    {
+      "epoch": 1.8575506199268577,
+      "grad_norm": 0.5469593405723572,
+      "learning_rate": 6.210247091914395e-06,
+      "loss": 0.4538,
+      "step": 10413
+    },
+    {
+      "epoch": 1.8577290161448579,
+      "grad_norm": 0.5348469614982605,
+      "learning_rate": 6.194737451153648e-06,
+      "loss": 0.5542,
+      "step": 10414
+    },
+    {
+      "epoch": 1.857907412362858,
+      "grad_norm": 0.6335954070091248,
+      "learning_rate": 6.179246958753298e-06,
+      "loss": 0.57,
+      "step": 10415
+    },
+    {
+      "epoch": 1.858085808580858,
+      "grad_norm": 0.5610979795455933,
+      "learning_rate": 6.163775615929984e-06,
+      "loss": 0.585,
+      "step": 10416
+    },
+    {
+      "epoch": 1.8582642047988582,
+      "grad_norm": 0.4940567910671234,
+      "learning_rate": 6.148323423898816e-06,
+      "loss": 0.4222,
+      "step": 10417
+    },
+    {
+      "epoch": 1.8584426010168584,
+      "grad_norm": 0.5531600117683411,
+      "learning_rate": 6.132890383873352e-06,
+      "loss": 0.4795,
+      "step": 10418
+    },
+    {
+      "epoch": 1.8586209972348586,
+      "grad_norm": 0.5417333841323853,
+      "learning_rate": 6.11747649706576e-06,
+      "loss": 0.5194,
+      "step": 10419
+    },
+    {
+      "epoch": 1.8587993934528588,
+      "grad_norm": 0.5340380668640137,
+      "learning_rate": 6.1020817646866014e-06,
+      "loss": 0.534,
+      "step": 10420
+    },
+    {
+      "epoch": 1.858977789670859,
+      "grad_norm": 0.5391452312469482,
+      "learning_rate": 6.08670618794499e-06,
+      "loss": 0.4972,
+      "step": 10421
+    },
+    {
+      "epoch": 1.859156185888859,
+      "grad_norm": 0.4881535470485687,
+      "learning_rate": 6.07134976804849e-06,
+      "loss": 0.3307,
+      "step": 10422
+    },
+    {
+      "epoch": 1.8593345821068592,
+      "grad_norm": 0.4824979305267334,
+      "learning_rate": 6.056012506203218e-06,
+      "loss": 0.4743,
+      "step": 10423
+    },
+    {
+      "epoch": 1.8595129783248594,
+      "grad_norm": 0.5851719975471497,
+      "learning_rate": 6.040694403613767e-06,
+      "loss": 0.5364,
+      "step": 10424
+    },
+    {
+      "epoch": 1.8596913745428596,
+      "grad_norm": 0.5611156821250916,
+      "learning_rate": 6.025395461483174e-06,
+      "loss": 0.5917,
+      "step": 10425
+    },
+    {
+      "epoch": 1.8598697707608598,
+      "grad_norm": 0.5232062935829163,
+      "learning_rate": 6.010115681013034e-06,
+      "loss": 0.381,
+      "step": 10426
+    },
+    {
+      "epoch": 1.86004816697886,
+      "grad_norm": 0.5068747401237488,
+      "learning_rate": 5.994855063403415e-06,
+      "loss": 0.453,
+      "step": 10427
+    },
+    {
+      "epoch": 1.8602265631968602,
+      "grad_norm": 0.5189545154571533,
+      "learning_rate": 5.979613609852885e-06,
+      "loss": 0.3813,
+      "step": 10428
+    },
+    {
+      "epoch": 1.8604049594148604,
+      "grad_norm": 0.5572920441627502,
+      "learning_rate": 5.96439132155846e-06,
+      "loss": 0.597,
+      "step": 10429
+    },
+    {
+      "epoch": 1.8605833556328606,
+      "grad_norm": 0.6158145666122437,
+      "learning_rate": 5.949188199715766e-06,
+      "loss": 0.6569,
+      "step": 10430
+    },
+    {
+      "epoch": 1.8607617518508608,
+      "grad_norm": 0.5191422700881958,
+      "learning_rate": 5.934004245518793e-06,
+      "loss": 0.401,
+      "step": 10431
+    },
+    {
+      "epoch": 1.860940148068861,
+      "grad_norm": 0.5226636528968811,
+      "learning_rate": 5.918839460160086e-06,
+      "loss": 0.4098,
+      "step": 10432
+    },
+    {
+      "epoch": 1.8611185442868612,
+      "grad_norm": 0.5896599888801575,
+      "learning_rate": 5.903693844830693e-06,
+      "loss": 0.634,
+      "step": 10433
+    },
+    {
+      "epoch": 1.8612969405048614,
+      "grad_norm": 0.5435173511505127,
+      "learning_rate": 5.888567400720135e-06,
+      "loss": 0.4192,
+      "step": 10434
+    },
+    {
+      "epoch": 1.8614753367228616,
+      "grad_norm": 0.5945151448249817,
+      "learning_rate": 5.8734601290164615e-06,
+      "loss": 0.5485,
+      "step": 10435
+    },
+    {
+      "epoch": 1.8616537329408618,
+      "grad_norm": 0.5320826172828674,
+      "learning_rate": 5.858372030906167e-06,
+      "loss": 0.5008,
+      "step": 10436
+    },
+    {
+      "epoch": 1.861832129158862,
+      "grad_norm": 0.569065272808075,
+      "learning_rate": 5.843303107574249e-06,
+      "loss": 0.5214,
+      "step": 10437
+    },
+    {
+      "epoch": 1.862010525376862,
+      "grad_norm": 0.5648977756500244,
+      "learning_rate": 5.828253360204261e-06,
+      "loss": 0.6008,
+      "step": 10438
+    },
+    {
+      "epoch": 1.8621889215948622,
+      "grad_norm": 0.5260228514671326,
+      "learning_rate": 5.813222789978173e-06,
+      "loss": 0.4568,
+      "step": 10439
+    },
+    {
+      "epoch": 1.8623673178128624,
+      "grad_norm": 0.4943072497844696,
+      "learning_rate": 5.798211398076486e-06,
+      "loss": 0.411,
+      "step": 10440
+    },
+    {
+      "epoch": 1.8625457140308626,
+      "grad_norm": 0.5590642094612122,
+      "learning_rate": 5.783219185678173e-06,
+      "loss": 0.4807,
+      "step": 10441
+    },
+    {
+      "epoch": 1.8627241102488628,
+      "grad_norm": 0.490158349275589,
+      "learning_rate": 5.768246153960766e-06,
+      "loss": 0.4316,
+      "step": 10442
+    },
+    {
+      "epoch": 1.862902506466863,
+      "grad_norm": 0.5113316774368286,
+      "learning_rate": 5.7532923041001825e-06,
+      "loss": 0.3966,
+      "step": 10443
+    },
+    {
+      "epoch": 1.863080902684863,
+      "grad_norm": 0.5623192191123962,
+      "learning_rate": 5.73835763727093e-06,
+      "loss": 0.5611,
+      "step": 10444
+    },
+    {
+      "epoch": 1.8632592989028631,
+      "grad_norm": 0.5246427059173584,
+      "learning_rate": 5.723442154645931e-06,
+      "loss": 0.4583,
+      "step": 10445
+    },
+    {
+      "epoch": 1.8634376951208633,
+      "grad_norm": 0.5438576936721802,
+      "learning_rate": 5.708545857396663e-06,
+      "loss": 0.4384,
+      "step": 10446
+    },
+    {
+      "epoch": 1.8636160913388635,
+      "grad_norm": 0.4891131818294525,
+      "learning_rate": 5.693668746693109e-06,
+      "loss": 0.4494,
+      "step": 10447
+    },
+    {
+      "epoch": 1.8637944875568637,
+      "grad_norm": 0.504887044429779,
+      "learning_rate": 5.678810823703639e-06,
+      "loss": 0.414,
+      "step": 10448
+    },
+    {
+      "epoch": 1.863972883774864,
+      "grad_norm": 0.5959895253181458,
+      "learning_rate": 5.663972089595265e-06,
+      "loss": 0.6671,
+      "step": 10449
+    },
+    {
+      "epoch": 1.8641512799928641,
+      "grad_norm": 0.45866143703460693,
+      "learning_rate": 5.649152545533331e-06,
+      "loss": 0.347,
+      "step": 10450
+    },
+    {
+      "epoch": 1.8643296762108643,
+      "grad_norm": 0.5574930310249329,
+      "learning_rate": 5.634352192681852e-06,
+      "loss": 0.6087,
+      "step": 10451
+    },
+    {
+      "epoch": 1.8645080724288645,
+      "grad_norm": 0.5660715103149414,
+      "learning_rate": 5.619571032203147e-06,
+      "loss": 0.5658,
+      "step": 10452
+    },
+    {
+      "epoch": 1.8646864686468647,
+      "grad_norm": 0.5469009280204773,
+      "learning_rate": 5.604809065258176e-06,
+      "loss": 0.5725,
+      "step": 10453
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 0.5283670425415039,
+      "learning_rate": 5.590066293006374e-06,
+      "loss": 0.4507,
+      "step": 10454
+    },
+    {
+      "epoch": 1.8650432610828651,
+      "grad_norm": 0.5461406111717224,
+      "learning_rate": 5.5753427166055635e-06,
+      "loss": 0.5125,
+      "step": 10455
+    },
+    {
+      "epoch": 1.8652216573008653,
+      "grad_norm": 0.5304574370384216,
+      "learning_rate": 5.560638337212126e-06,
+      "loss": 0.504,
+      "step": 10456
+    },
+    {
+      "epoch": 1.8654000535188655,
+      "grad_norm": 0.5879049301147461,
+      "learning_rate": 5.545953155980998e-06,
+      "loss": 0.5871,
+      "step": 10457
+    },
+    {
+      "epoch": 1.8655784497368657,
+      "grad_norm": 0.4767381250858307,
+      "learning_rate": 5.531287174065508e-06,
+      "loss": 0.4185,
+      "step": 10458
+    },
+    {
+      "epoch": 1.865756845954866,
+      "grad_norm": 0.5429925918579102,
+      "learning_rate": 5.516640392617511e-06,
+      "loss": 0.5435,
+      "step": 10459
+    },
+    {
+      "epoch": 1.8659352421728659,
+      "grad_norm": 0.5666899085044861,
+      "learning_rate": 5.502012812787366e-06,
+      "loss": 0.4597,
+      "step": 10460
+    },
+    {
+      "epoch": 1.866113638390866,
+      "grad_norm": 0.4842609167098999,
+      "learning_rate": 5.4874044357239305e-06,
+      "loss": 0.4877,
+      "step": 10461
+    },
+    {
+      "epoch": 1.8662920346088663,
+      "grad_norm": 0.5365511178970337,
+      "learning_rate": 5.4728152625745094e-06,
+      "loss": 0.5841,
+      "step": 10462
+    },
+    {
+      "epoch": 1.8664704308268665,
+      "grad_norm": 0.4799635112285614,
+      "learning_rate": 5.458245294484964e-06,
+      "loss": 0.4223,
+      "step": 10463
+    },
+    {
+      "epoch": 1.8666488270448667,
+      "grad_norm": 0.5694946050643921,
+      "learning_rate": 5.443694532599602e-06,
+      "loss": 0.6436,
+      "step": 10464
+    },
+    {
+      "epoch": 1.8668272232628669,
+      "grad_norm": 0.5620405673980713,
+      "learning_rate": 5.429162978061203e-06,
+      "loss": 0.5037,
+      "step": 10465
+    },
+    {
+      "epoch": 1.8670056194808669,
+      "grad_norm": 0.5125569105148315,
+      "learning_rate": 5.414650632011131e-06,
+      "loss": 0.5161,
+      "step": 10466
+    },
+    {
+      "epoch": 1.867184015698867,
+      "grad_norm": 0.5853519439697266,
+      "learning_rate": 5.400157495589114e-06,
+      "loss": 0.6433,
+      "step": 10467
+    },
+    {
+      "epoch": 1.8673624119168672,
+      "grad_norm": 0.46554625034332275,
+      "learning_rate": 5.385683569933464e-06,
+      "loss": 0.3893,
+      "step": 10468
+    },
+    {
+      "epoch": 1.8675408081348674,
+      "grad_norm": 0.547414243221283,
+      "learning_rate": 5.371228856180993e-06,
+      "loss": 0.4324,
+      "step": 10469
+    },
+    {
+      "epoch": 1.8677192043528676,
+      "grad_norm": 0.6247468590736389,
+      "learning_rate": 5.356793355466933e-06,
+      "loss": 0.729,
+      "step": 10470
+    },
+    {
+      "epoch": 1.8678976005708678,
+      "grad_norm": 0.5195764303207397,
+      "learning_rate": 5.342377068925041e-06,
+      "loss": 0.4048,
+      "step": 10471
+    },
+    {
+      "epoch": 1.868075996788868,
+      "grad_norm": 0.4851647615432739,
+      "learning_rate": 5.327979997687554e-06,
+      "loss": 0.342,
+      "step": 10472
+    },
+    {
+      "epoch": 1.8682543930068682,
+      "grad_norm": 0.5746437907218933,
+      "learning_rate": 5.313602142885232e-06,
+      "loss": 0.4786,
+      "step": 10473
+    },
+    {
+      "epoch": 1.8684327892248684,
+      "grad_norm": 0.6188861131668091,
+      "learning_rate": 5.299243505647283e-06,
+      "loss": 0.6486,
+      "step": 10474
+    },
+    {
+      "epoch": 1.8686111854428686,
+      "grad_norm": 0.48796316981315613,
+      "learning_rate": 5.2849040871015e-06,
+      "loss": 0.3936,
+      "step": 10475
+    },
+    {
+      "epoch": 1.8687895816608688,
+      "grad_norm": 0.5425949692726135,
+      "learning_rate": 5.270583888374009e-06,
+      "loss": 0.5362,
+      "step": 10476
+    },
+    {
+      "epoch": 1.868967977878869,
+      "grad_norm": 0.5946719646453857,
+      "learning_rate": 5.256282910589521e-06,
+      "loss": 0.5137,
+      "step": 10477
+    },
+    {
+      "epoch": 1.8691463740968692,
+      "grad_norm": 0.5937526822090149,
+      "learning_rate": 5.242001154871306e-06,
+      "loss": 0.5385,
+      "step": 10478
+    },
+    {
+      "epoch": 1.8693247703148694,
+      "grad_norm": 0.5041108727455139,
+      "learning_rate": 5.227738622340938e-06,
+      "loss": 0.4807,
+      "step": 10479
+    },
+    {
+      "epoch": 1.8695031665328696,
+      "grad_norm": 0.45467957854270935,
+      "learning_rate": 5.213495314118688e-06,
+      "loss": 0.3457,
+      "step": 10480
+    },
+    {
+      "epoch": 1.8696815627508698,
+      "grad_norm": 0.5164430141448975,
+      "learning_rate": 5.199271231323133e-06,
+      "loss": 0.4477,
+      "step": 10481
+    },
+    {
+      "epoch": 1.8698599589688698,
+      "grad_norm": 0.5357136726379395,
+      "learning_rate": 5.185066375071518e-06,
+      "loss": 0.4991,
+      "step": 10482
+    },
+    {
+      "epoch": 1.87003835518687,
+      "grad_norm": 0.4412253797054291,
+      "learning_rate": 5.170880746479395e-06,
+      "loss": 0.3304,
+      "step": 10483
+    },
+    {
+      "epoch": 1.8702167514048702,
+      "grad_norm": 0.4293851852416992,
+      "learning_rate": 5.156714346660957e-06,
+      "loss": 0.3241,
+      "step": 10484
+    },
+    {
+      "epoch": 1.8703951476228704,
+      "grad_norm": 0.5267345309257507,
+      "learning_rate": 5.142567176728813e-06,
+      "loss": 0.5351,
+      "step": 10485
+    },
+    {
+      "epoch": 1.8705735438408706,
+      "grad_norm": 0.5791170001029968,
+      "learning_rate": 5.128439237794047e-06,
+      "loss": 0.4566,
+      "step": 10486
+    },
+    {
+      "epoch": 1.8707519400588708,
+      "grad_norm": 0.4116140305995941,
+      "learning_rate": 5.114330530966326e-06,
+      "loss": 0.2923,
+      "step": 10487
+    },
+    {
+      "epoch": 1.8709303362768708,
+      "grad_norm": 0.4969407618045807,
+      "learning_rate": 5.100241057353683e-06,
+      "loss": 0.4083,
+      "step": 10488
+    },
+    {
+      "epoch": 1.871108732494871,
+      "grad_norm": 0.6181281208992004,
+      "learning_rate": 5.08617081806273e-06,
+      "loss": 0.6443,
+      "step": 10489
+    },
+    {
+      "epoch": 1.8712871287128712,
+      "grad_norm": 0.5408806204795837,
+      "learning_rate": 5.072119814198528e-06,
+      "loss": 0.4914,
+      "step": 10490
+    },
+    {
+      "epoch": 1.8714655249308714,
+      "grad_norm": 0.4671943783760071,
+      "learning_rate": 5.058088046864611e-06,
+      "loss": 0.3844,
+      "step": 10491
+    },
+    {
+      "epoch": 1.8716439211488716,
+      "grad_norm": 0.5404434204101562,
+      "learning_rate": 5.044075517163071e-06,
+      "loss": 0.4863,
+      "step": 10492
+    },
+    {
+      "epoch": 1.8718223173668718,
+      "grad_norm": 0.5988584756851196,
+      "learning_rate": 5.030082226194415e-06,
+      "loss": 0.5877,
+      "step": 10493
+    },
+    {
+      "epoch": 1.872000713584872,
+      "grad_norm": 0.5672785639762878,
+      "learning_rate": 5.01610817505771e-06,
+      "loss": 0.4992,
+      "step": 10494
+    },
+    {
+      "epoch": 1.8721791098028722,
+      "grad_norm": 0.5256212949752808,
+      "learning_rate": 5.002153364850409e-06,
+      "loss": 0.4614,
+      "step": 10495
+    },
+    {
+      "epoch": 1.8723575060208724,
+      "grad_norm": 0.47382408380508423,
+      "learning_rate": 4.9882177966685814e-06,
+      "loss": 0.3875,
+      "step": 10496
+    },
+    {
+      "epoch": 1.8725359022388726,
+      "grad_norm": 0.5471598505973816,
+      "learning_rate": 4.974301471606685e-06,
+      "loss": 0.4792,
+      "step": 10497
+    },
+    {
+      "epoch": 1.8727142984568728,
+      "grad_norm": 0.514196515083313,
+      "learning_rate": 4.96040439075765e-06,
+      "loss": 0.5545,
+      "step": 10498
+    },
+    {
+      "epoch": 1.872892694674873,
+      "grad_norm": 0.5053452849388123,
+      "learning_rate": 4.946526555213077e-06,
+      "loss": 0.4108,
+      "step": 10499
+    },
+    {
+      "epoch": 1.8730710908928732,
+      "grad_norm": 0.6003931164741516,
+      "learning_rate": 4.9326679660628145e-06,
+      "loss": 0.5834,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8732494871108734,
+      "grad_norm": 0.559822678565979,
+      "learning_rate": 4.918828624395383e-06,
+      "loss": 0.517,
+      "step": 10501
+    },
+    {
+      "epoch": 1.8734278833288736,
+      "grad_norm": 0.5788659453392029,
+      "learning_rate": 4.905008531297661e-06,
+      "loss": 0.5528,
+      "step": 10502
+    },
+    {
+      "epoch": 1.8736062795468738,
+      "grad_norm": 0.4903022050857544,
+      "learning_rate": 4.891207687855115e-06,
+      "loss": 0.4657,
+      "step": 10503
+    },
+    {
+      "epoch": 1.8737846757648737,
+      "grad_norm": 0.529029369354248,
+      "learning_rate": 4.877426095151627e-06,
+      "loss": 0.4519,
+      "step": 10504
+    },
+    {
+      "epoch": 1.873963071982874,
+      "grad_norm": 0.48290225863456726,
+      "learning_rate": 4.863663754269609e-06,
+      "loss": 0.4803,
+      "step": 10505
+    },
+    {
+      "epoch": 1.8741414682008741,
+      "grad_norm": 0.5195373296737671,
+      "learning_rate": 4.849920666289947e-06,
+      "loss": 0.3987,
+      "step": 10506
+    },
+    {
+      "epoch": 1.8743198644188743,
+      "grad_norm": 0.537219226360321,
+      "learning_rate": 4.836196832292e-06,
+      "loss": 0.5085,
+      "step": 10507
+    },
+    {
+      "epoch": 1.8744982606368745,
+      "grad_norm": 0.4911697208881378,
+      "learning_rate": 4.8224922533536834e-06,
+      "loss": 0.4654,
+      "step": 10508
+    },
+    {
+      "epoch": 1.8746766568548747,
+      "grad_norm": 0.5637452006340027,
+      "learning_rate": 4.8088069305513015e-06,
+      "loss": 0.431,
+      "step": 10509
+    },
+    {
+      "epoch": 1.8748550530728747,
+      "grad_norm": 0.4910629391670227,
+      "learning_rate": 4.795140864959718e-06,
+      "loss": 0.3722,
+      "step": 10510
+    },
+    {
+      "epoch": 1.875033449290875,
+      "grad_norm": 0.5422062873840332,
+      "learning_rate": 4.781494057652269e-06,
+      "loss": 0.5239,
+      "step": 10511
+    },
+    {
+      "epoch": 1.875211845508875,
+      "grad_norm": 0.5688182711601257,
+      "learning_rate": 4.767866509700708e-06,
+      "loss": 0.4991,
+      "step": 10512
+    },
+    {
+      "epoch": 1.8753902417268753,
+      "grad_norm": 0.5716597437858582,
+      "learning_rate": 4.754258222175428e-06,
+      "loss": 0.5913,
+      "step": 10513
+    },
+    {
+      "epoch": 1.8755686379448755,
+      "grad_norm": 0.5582311749458313,
+      "learning_rate": 4.740669196145131e-06,
+      "loss": 0.5269,
+      "step": 10514
+    },
+    {
+      "epoch": 1.8757470341628757,
+      "grad_norm": 0.5111207365989685,
+      "learning_rate": 4.727099432677129e-06,
+      "loss": 0.4636,
+      "step": 10515
+    },
+    {
+      "epoch": 1.875925430380876,
+      "grad_norm": 0.5934797525405884,
+      "learning_rate": 4.713548932837208e-06,
+      "loss": 0.5653,
+      "step": 10516
+    },
+    {
+      "epoch": 1.876103826598876,
+      "grad_norm": 0.602021336555481,
+      "learning_rate": 4.700017697689574e-06,
+      "loss": 0.6561,
+      "step": 10517
+    },
+    {
+      "epoch": 1.8762822228168763,
+      "grad_norm": 0.5633701086044312,
+      "learning_rate": 4.686505728297013e-06,
+      "loss": 0.5126,
+      "step": 10518
+    },
+    {
+      "epoch": 1.8764606190348765,
+      "grad_norm": 0.5579110383987427,
+      "learning_rate": 4.6730130257207345e-06,
+      "loss": 0.5513,
+      "step": 10519
+    },
+    {
+      "epoch": 1.8766390152528767,
+      "grad_norm": 0.5543150901794434,
+      "learning_rate": 4.659539591020417e-06,
+      "loss": 0.5097,
+      "step": 10520
+    },
+    {
+      "epoch": 1.8768174114708769,
+      "grad_norm": 0.4759593904018402,
+      "learning_rate": 4.646085425254298e-06,
+      "loss": 0.4002,
+      "step": 10521
+    },
+    {
+      "epoch": 1.876995807688877,
+      "grad_norm": 0.6420212388038635,
+      "learning_rate": 4.632650529479032e-06,
+      "loss": 0.6743,
+      "step": 10522
+    },
+    {
+      "epoch": 1.8771742039068773,
+      "grad_norm": 0.483822762966156,
+      "learning_rate": 4.61923490474983e-06,
+      "loss": 0.4431,
+      "step": 10523
+    },
+    {
+      "epoch": 1.8773526001248775,
+      "grad_norm": 0.5354387760162354,
+      "learning_rate": 4.605838552120295e-06,
+      "loss": 0.4565,
+      "step": 10524
+    },
+    {
+      "epoch": 1.8775309963428777,
+      "grad_norm": 0.5774738192558289,
+      "learning_rate": 4.592461472642611e-06,
+      "loss": 0.635,
+      "step": 10525
+    },
+    {
+      "epoch": 1.8777093925608777,
+      "grad_norm": 0.4764508605003357,
+      "learning_rate": 4.579103667367385e-06,
+      "loss": 0.357,
+      "step": 10526
+    },
+    {
+      "epoch": 1.8778877887788779,
+      "grad_norm": 0.5291125774383545,
+      "learning_rate": 4.565765137343775e-06,
+      "loss": 0.4858,
+      "step": 10527
+    },
+    {
+      "epoch": 1.878066184996878,
+      "grad_norm": 0.5577571392059326,
+      "learning_rate": 4.552445883619305e-06,
+      "loss": 0.4481,
+      "step": 10528
+    },
+    {
+      "epoch": 1.8782445812148783,
+      "grad_norm": 0.489429771900177,
+      "learning_rate": 4.539145907240139e-06,
+      "loss": 0.4101,
+      "step": 10529
+    },
+    {
+      "epoch": 1.8784229774328784,
+      "grad_norm": 0.49116069078445435,
+      "learning_rate": 4.525865209250829e-06,
+      "loss": 0.3959,
+      "step": 10530
+    },
+    {
+      "epoch": 1.8786013736508786,
+      "grad_norm": 0.5530174970626831,
+      "learning_rate": 4.512603790694403e-06,
+      "loss": 0.6012,
+      "step": 10531
+    },
+    {
+      "epoch": 1.8787797698688786,
+      "grad_norm": 0.5700563788414001,
+      "learning_rate": 4.499361652612444e-06,
+      "loss": 0.5586,
+      "step": 10532
+    },
+    {
+      "epoch": 1.8789581660868788,
+      "grad_norm": 0.562254011631012,
+      "learning_rate": 4.486138796044981e-06,
+      "loss": 0.4436,
+      "step": 10533
+    },
+    {
+      "epoch": 1.879136562304879,
+      "grad_norm": 0.4984350800514221,
+      "learning_rate": 4.472935222030544e-06,
+      "loss": 0.4187,
+      "step": 10534
+    },
+    {
+      "epoch": 1.8793149585228792,
+      "grad_norm": 0.6067535281181335,
+      "learning_rate": 4.459750931606083e-06,
+      "loss": 0.6892,
+      "step": 10535
+    },
+    {
+      "epoch": 1.8794933547408794,
+      "grad_norm": 0.5302973389625549,
+      "learning_rate": 4.446585925807129e-06,
+      "loss": 0.4837,
+      "step": 10536
+    },
+    {
+      "epoch": 1.8796717509588796,
+      "grad_norm": 0.466941773891449,
+      "learning_rate": 4.43344020566766e-06,
+      "loss": 0.393,
+      "step": 10537
+    },
+    {
+      "epoch": 1.8798501471768798,
+      "grad_norm": 0.4711865186691284,
+      "learning_rate": 4.420313772220103e-06,
+      "loss": 0.4012,
+      "step": 10538
+    },
+    {
+      "epoch": 1.88002854339488,
+      "grad_norm": 0.574239194393158,
+      "learning_rate": 4.4072066264954355e-06,
+      "loss": 0.5279,
+      "step": 10539
+    },
+    {
+      "epoch": 1.8802069396128802,
+      "grad_norm": 0.5025227069854736,
+      "learning_rate": 4.394118769523059e-06,
+      "loss": 0.4958,
+      "step": 10540
+    },
+    {
+      "epoch": 1.8803853358308804,
+      "grad_norm": 0.47873345017433167,
+      "learning_rate": 4.381050202330927e-06,
+      "loss": 0.3807,
+      "step": 10541
+    },
+    {
+      "epoch": 1.8805637320488806,
+      "grad_norm": 0.5627789497375488,
+      "learning_rate": 4.368000925945386e-06,
+      "loss": 0.5274,
+      "step": 10542
+    },
+    {
+      "epoch": 1.8807421282668808,
+      "grad_norm": 0.5269836187362671,
+      "learning_rate": 4.3549709413913675e-06,
+      "loss": 0.4638,
+      "step": 10543
+    },
+    {
+      "epoch": 1.880920524484881,
+      "grad_norm": 0.5568525791168213,
+      "learning_rate": 4.34196024969219e-06,
+      "loss": 0.4697,
+      "step": 10544
+    },
+    {
+      "epoch": 1.8810989207028812,
+      "grad_norm": 0.4760674834251404,
+      "learning_rate": 4.328968851869758e-06,
+      "loss": 0.4065,
+      "step": 10545
+    },
+    {
+      "epoch": 1.8812773169208814,
+      "grad_norm": 0.5974029898643494,
+      "learning_rate": 4.3159967489443955e-06,
+      "loss": 0.6253,
+      "step": 10546
+    },
+    {
+      "epoch": 1.8814557131388816,
+      "grad_norm": 0.49886420369148254,
+      "learning_rate": 4.3030439419349255e-06,
+      "loss": 0.4455,
+      "step": 10547
+    },
+    {
+      "epoch": 1.8816341093568816,
+      "grad_norm": 0.5672736167907715,
+      "learning_rate": 4.290110431858646e-06,
+      "loss": 0.4771,
+      "step": 10548
+    },
+    {
+      "epoch": 1.8818125055748818,
+      "grad_norm": 0.5267834067344666,
+      "learning_rate": 4.277196219731383e-06,
+      "loss": 0.5392,
+      "step": 10549
+    },
+    {
+      "epoch": 1.881990901792882,
+      "grad_norm": 0.5531571507453918,
+      "learning_rate": 4.264301306567353e-06,
+      "loss": 0.6175,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8821692980108822,
+      "grad_norm": 0.5343933701515198,
+      "learning_rate": 4.251425693379357e-06,
+      "loss": 0.5271,
+      "step": 10551
+    },
+    {
+      "epoch": 1.8823476942288824,
+      "grad_norm": 0.5798811316490173,
+      "learning_rate": 4.238569381178642e-06,
+      "loss": 0.6559,
+      "step": 10552
+    },
+    {
+      "epoch": 1.8825260904468826,
+      "grad_norm": 0.5240869522094727,
+      "learning_rate": 4.225732370974928e-06,
+      "loss": 0.4699,
+      "step": 10553
+    },
+    {
+      "epoch": 1.8827044866648825,
+      "grad_norm": 0.5387951731681824,
+      "learning_rate": 4.212914663776407e-06,
+      "loss": 0.4743,
+      "step": 10554
+    },
+    {
+      "epoch": 1.8828828828828827,
+      "grad_norm": 0.5581191778182983,
+      "learning_rate": 4.200116260589831e-06,
+      "loss": 0.4566,
+      "step": 10555
+    },
+    {
+      "epoch": 1.883061279100883,
+      "grad_norm": 0.5226394534111023,
+      "learning_rate": 4.1873371624203406e-06,
+      "loss": 0.4776,
+      "step": 10556
+    },
+    {
+      "epoch": 1.8832396753188831,
+      "grad_norm": 0.5839105844497681,
+      "learning_rate": 4.174577370271576e-06,
+      "loss": 0.5059,
+      "step": 10557
+    },
+    {
+      "epoch": 1.8834180715368833,
+      "grad_norm": 0.5159413814544678,
+      "learning_rate": 4.161836885145765e-06,
+      "loss": 0.4771,
+      "step": 10558
+    },
+    {
+      "epoch": 1.8835964677548835,
+      "grad_norm": 0.5594849586486816,
+      "learning_rate": 4.149115708043438e-06,
+      "loss": 0.5485,
+      "step": 10559
+    },
+    {
+      "epoch": 1.8837748639728837,
+      "grad_norm": 0.5460865497589111,
+      "learning_rate": 4.136413839963799e-06,
+      "loss": 0.4878,
+      "step": 10560
+    },
+    {
+      "epoch": 1.883953260190884,
+      "grad_norm": 0.5533373355865479,
+      "learning_rate": 4.123731281904408e-06,
+      "loss": 0.4856,
+      "step": 10561
+    },
+    {
+      "epoch": 1.8841316564088841,
+      "grad_norm": 0.6217470169067383,
+      "learning_rate": 4.111068034861359e-06,
+      "loss": 0.6534,
+      "step": 10562
+    },
+    {
+      "epoch": 1.8843100526268843,
+      "grad_norm": 0.5605418682098389,
+      "learning_rate": 4.098424099829218e-06,
+      "loss": 0.5772,
+      "step": 10563
+    },
+    {
+      "epoch": 1.8844884488448845,
+      "grad_norm": 0.6147893667221069,
+      "learning_rate": 4.085799477800995e-06,
+      "loss": 0.739,
+      "step": 10564
+    },
+    {
+      "epoch": 1.8846668450628847,
+      "grad_norm": 0.5503641963005066,
+      "learning_rate": 4.073194169768285e-06,
+      "loss": 0.5422,
+      "step": 10565
+    },
+    {
+      "epoch": 1.884845241280885,
+      "grad_norm": 0.5137640833854675,
+      "learning_rate": 4.0606081767210755e-06,
+      "loss": 0.417,
+      "step": 10566
+    },
+    {
+      "epoch": 1.8850236374988851,
+      "grad_norm": 0.46685469150543213,
+      "learning_rate": 4.048041499647853e-06,
+      "loss": 0.3761,
+      "step": 10567
+    },
+    {
+      "epoch": 1.8852020337168853,
+      "grad_norm": 0.4922786056995392,
+      "learning_rate": 4.035494139535606e-06,
+      "loss": 0.4437,
+      "step": 10568
+    },
+    {
+      "epoch": 1.8853804299348855,
+      "grad_norm": 0.5689695477485657,
+      "learning_rate": 4.0229660973698235e-06,
+      "loss": 0.5262,
+      "step": 10569
+    },
+    {
+      "epoch": 1.8855588261528857,
+      "grad_norm": 0.5004351139068604,
+      "learning_rate": 4.010457374134441e-06,
+      "loss": 0.389,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8857372223708857,
+      "grad_norm": 0.5173558592796326,
+      "learning_rate": 3.997967970811839e-06,
+      "loss": 0.5562,
+      "step": 10571
+    },
+    {
+      "epoch": 1.885915618588886,
+      "grad_norm": 0.48343634605407715,
+      "learning_rate": 3.985497888382983e-06,
+      "loss": 0.4431,
+      "step": 10572
+    },
+    {
+      "epoch": 1.886094014806886,
+      "grad_norm": 0.5336623787879944,
+      "learning_rate": 3.973047127827256e-06,
+      "loss": 0.4717,
+      "step": 10573
+    },
+    {
+      "epoch": 1.8862724110248863,
+      "grad_norm": 0.5859199166297913,
+      "learning_rate": 3.960615690122543e-06,
+      "loss": 0.5706,
+      "step": 10574
+    },
+    {
+      "epoch": 1.8864508072428865,
+      "grad_norm": 0.5274314880371094,
+      "learning_rate": 3.948203576245174e-06,
+      "loss": 0.5479,
+      "step": 10575
+    },
+    {
+      "epoch": 1.8866292034608865,
+      "grad_norm": 0.5471269488334656,
+      "learning_rate": 3.935810787170036e-06,
+      "loss": 0.5647,
+      "step": 10576
+    },
+    {
+      "epoch": 1.8868075996788867,
+      "grad_norm": 0.48193615674972534,
+      "learning_rate": 3.923437323870405e-06,
+      "loss": 0.4555,
+      "step": 10577
+    },
+    {
+      "epoch": 1.8869859958968869,
+      "grad_norm": 0.574404776096344,
+      "learning_rate": 3.911083187318115e-06,
+      "loss": 0.6331,
+      "step": 10578
+    },
+    {
+      "epoch": 1.887164392114887,
+      "grad_norm": 0.5169478058815002,
+      "learning_rate": 3.898748378483474e-06,
+      "loss": 0.5231,
+      "step": 10579
+    },
+    {
+      "epoch": 1.8873427883328873,
+      "grad_norm": 0.56248939037323,
+      "learning_rate": 3.886432898335207e-06,
+      "loss": 0.6228,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8875211845508875,
+      "grad_norm": 0.516521692276001,
+      "learning_rate": 3.874136747840623e-06,
+      "loss": 0.4946,
+      "step": 10581
+    },
+    {
+      "epoch": 1.8876995807688877,
+      "grad_norm": 0.5308520793914795,
+      "learning_rate": 3.861859927965394e-06,
+      "loss": 0.5038,
+      "step": 10582
+    },
+    {
+      "epoch": 1.8878779769868879,
+      "grad_norm": 0.5017712712287903,
+      "learning_rate": 3.849602439673749e-06,
+      "loss": 0.4414,
+      "step": 10583
+    },
+    {
+      "epoch": 1.888056373204888,
+      "grad_norm": 0.5355345010757446,
+      "learning_rate": 3.837364283928446e-06,
+      "loss": 0.4509,
+      "step": 10584
+    },
+    {
+      "epoch": 1.8882347694228883,
+      "grad_norm": 0.5393890142440796,
+      "learning_rate": 3.825145461690577e-06,
+      "loss": 0.6249,
+      "step": 10585
+    },
+    {
+      "epoch": 1.8884131656408885,
+      "grad_norm": 0.5255075693130493,
+      "learning_rate": 3.8129459739198737e-06,
+      "loss": 0.5318,
+      "step": 10586
+    },
+    {
+      "epoch": 1.8885915618588887,
+      "grad_norm": 0.5181835293769836,
+      "learning_rate": 3.800765821574431e-06,
+      "loss": 0.5787,
+      "step": 10587
+    },
+    {
+      "epoch": 1.8887699580768889,
+      "grad_norm": 0.5990747213363647,
+      "learning_rate": 3.7886050056109287e-06,
+      "loss": 0.5349,
+      "step": 10588
+    },
+    {
+      "epoch": 1.888948354294889,
+      "grad_norm": 0.49007025361061096,
+      "learning_rate": 3.7764635269843804e-06,
+      "loss": 0.3994,
+      "step": 10589
+    },
+    {
+      "epoch": 1.8891267505128893,
+      "grad_norm": 0.45775508880615234,
+      "learning_rate": 3.7643413866484678e-06,
+      "loss": 0.325,
+      "step": 10590
+    },
+    {
+      "epoch": 1.8893051467308895,
+      "grad_norm": 0.5210441946983337,
+      "learning_rate": 3.7522385855552067e-06,
+      "loss": 0.5073,
+      "step": 10591
+    },
+    {
+      "epoch": 1.8894835429488896,
+      "grad_norm": 0.5008862018585205,
+      "learning_rate": 3.7401551246551703e-06,
+      "loss": 0.4124,
+      "step": 10592
+    },
+    {
+      "epoch": 1.8896619391668896,
+      "grad_norm": 0.5803366899490356,
+      "learning_rate": 3.728091004897377e-06,
+      "loss": 0.5425,
+      "step": 10593
+    },
+    {
+      "epoch": 1.8898403353848898,
+      "grad_norm": 0.48922333121299744,
+      "learning_rate": 3.7160462272293195e-06,
+      "loss": 0.461,
+      "step": 10594
+    },
+    {
+      "epoch": 1.89001873160289,
+      "grad_norm": 0.6023316979408264,
+      "learning_rate": 3.704020792597018e-06,
+      "loss": 0.7902,
+      "step": 10595
+    },
+    {
+      "epoch": 1.8901971278208902,
+      "grad_norm": 0.5757283568382263,
+      "learning_rate": 3.69201470194494e-06,
+      "loss": 0.5757,
+      "step": 10596
+    },
+    {
+      "epoch": 1.8903755240388904,
+      "grad_norm": 0.5134090185165405,
+      "learning_rate": 3.6800279562160257e-06,
+      "loss": 0.5602,
+      "step": 10597
+    },
+    {
+      "epoch": 1.8905539202568904,
+      "grad_norm": 0.49674350023269653,
+      "learning_rate": 3.6680605563517153e-06,
+      "loss": 0.4284,
+      "step": 10598
+    },
+    {
+      "epoch": 1.8907323164748906,
+      "grad_norm": 0.5035196542739868,
+      "learning_rate": 3.6561125032918975e-06,
+      "loss": 0.4645,
+      "step": 10599
+    },
+    {
+      "epoch": 1.8909107126928908,
+      "grad_norm": 0.5700138807296753,
+      "learning_rate": 3.6441837979750427e-06,
+      "loss": 0.6076,
+      "step": 10600
+    },
+    {
+      "epoch": 1.891089108910891,
+      "grad_norm": 0.5219821333885193,
+      "learning_rate": 3.632274441337957e-06,
+      "loss": 0.4652,
+      "step": 10601
+    },
+    {
+      "epoch": 1.8912675051288912,
+      "grad_norm": 0.5027645826339722,
+      "learning_rate": 3.620384434316004e-06,
+      "loss": 0.4392,
+      "step": 10602
+    },
+    {
+      "epoch": 1.8914459013468914,
+      "grad_norm": 0.4853323996067047,
+      "learning_rate": 3.608513777843048e-06,
+      "loss": 0.4215,
+      "step": 10603
+    },
+    {
+      "epoch": 1.8916242975648916,
+      "grad_norm": 0.5268037915229797,
+      "learning_rate": 3.59666247285137e-06,
+      "loss": 0.5088,
+      "step": 10604
+    },
+    {
+      "epoch": 1.8918026937828918,
+      "grad_norm": 0.5432520508766174,
+      "learning_rate": 3.584830520271809e-06,
+      "loss": 0.4998,
+      "step": 10605
+    },
+    {
+      "epoch": 1.891981090000892,
+      "grad_norm": 0.6279560327529907,
+      "learning_rate": 3.5730179210335946e-06,
+      "loss": 0.5981,
+      "step": 10606
+    },
+    {
+      "epoch": 1.8921594862188922,
+      "grad_norm": 0.5216326713562012,
+      "learning_rate": 3.5612246760645118e-06,
+      "loss": 0.4698,
+      "step": 10607
+    },
+    {
+      "epoch": 1.8923378824368924,
+      "grad_norm": 0.5420396327972412,
+      "learning_rate": 3.54945078629082e-06,
+      "loss": 0.5827,
+      "step": 10608
+    },
+    {
+      "epoch": 1.8925162786548926,
+      "grad_norm": 0.5069664716720581,
+      "learning_rate": 3.5376962526371682e-06,
+      "loss": 0.4561,
+      "step": 10609
+    },
+    {
+      "epoch": 1.8926946748728928,
+      "grad_norm": 0.4788225293159485,
+      "learning_rate": 3.525961076026818e-06,
+      "loss": 0.4292,
+      "step": 10610
+    },
+    {
+      "epoch": 1.892873071090893,
+      "grad_norm": 0.5387194156646729,
+      "learning_rate": 3.514245257381421e-06,
+      "loss": 0.5415,
+      "step": 10611
+    },
+    {
+      "epoch": 1.8930514673088932,
+      "grad_norm": 0.5902103185653687,
+      "learning_rate": 3.50254879762113e-06,
+      "loss": 0.6707,
+      "step": 10612
+    },
+    {
+      "epoch": 1.8932298635268934,
+      "grad_norm": 0.4795367121696472,
+      "learning_rate": 3.490871697664544e-06,
+      "loss": 0.3785,
+      "step": 10613
+    },
+    {
+      "epoch": 1.8934082597448936,
+      "grad_norm": 0.5698316097259521,
+      "learning_rate": 3.4792139584288728e-06,
+      "loss": 0.5774,
+      "step": 10614
+    },
+    {
+      "epoch": 1.8935866559628936,
+      "grad_norm": 0.530458927154541,
+      "learning_rate": 3.4675755808296073e-06,
+      "loss": 0.5076,
+      "step": 10615
+    },
+    {
+      "epoch": 1.8937650521808937,
+      "grad_norm": 0.5088088512420654,
+      "learning_rate": 3.4559565657808766e-06,
+      "loss": 0.4608,
+      "step": 10616
+    },
+    {
+      "epoch": 1.893943448398894,
+      "grad_norm": 0.46197885274887085,
+      "learning_rate": 3.444356914195229e-06,
+      "loss": 0.355,
+      "step": 10617
+    },
+    {
+      "epoch": 1.8941218446168941,
+      "grad_norm": 0.5059583783149719,
+      "learning_rate": 3.432776626983658e-06,
+      "loss": 0.4677,
+      "step": 10618
+    },
+    {
+      "epoch": 1.8943002408348943,
+      "grad_norm": 0.519899308681488,
+      "learning_rate": 3.421215705055741e-06,
+      "loss": 0.485,
+      "step": 10619
+    },
+    {
+      "epoch": 1.8944786370528943,
+      "grad_norm": 0.544313907623291,
+      "learning_rate": 3.4096741493194194e-06,
+      "loss": 0.5031,
+      "step": 10620
+    },
+    {
+      "epoch": 1.8946570332708945,
+      "grad_norm": 0.5371688008308411,
+      "learning_rate": 3.3981519606811616e-06,
+      "loss": 0.5519,
+      "step": 10621
+    },
+    {
+      "epoch": 1.8948354294888947,
+      "grad_norm": 0.47906023263931274,
+      "learning_rate": 3.3866491400459387e-06,
+      "loss": 0.4269,
+      "step": 10622
+    },
+    {
+      "epoch": 1.895013825706895,
+      "grad_norm": 0.5047534704208374,
+      "learning_rate": 3.3751656883171668e-06,
+      "loss": 0.4814,
+      "step": 10623
+    },
+    {
+      "epoch": 1.8951922219248951,
+      "grad_norm": 0.5101923942565918,
+      "learning_rate": 3.363701606396735e-06,
+      "loss": 0.4778,
+      "step": 10624
+    },
+    {
+      "epoch": 1.8953706181428953,
+      "grad_norm": 0.46724262833595276,
+      "learning_rate": 3.352256895185063e-06,
+      "loss": 0.3659,
+      "step": 10625
+    },
+    {
+      "epoch": 1.8955490143608955,
+      "grad_norm": 0.5907710194587708,
+      "learning_rate": 3.3408315555809863e-06,
+      "loss": 0.4965,
+      "step": 10626
+    },
+    {
+      "epoch": 1.8957274105788957,
+      "grad_norm": 0.5587756037712097,
+      "learning_rate": 3.3294255884818435e-06,
+      "loss": 0.5713,
+      "step": 10627
+    },
+    {
+      "epoch": 1.895905806796896,
+      "grad_norm": 0.5773804783821106,
+      "learning_rate": 3.318038994783501e-06,
+      "loss": 0.5113,
+      "step": 10628
+    },
+    {
+      "epoch": 1.896084203014896,
+      "grad_norm": 0.5329833626747131,
+      "learning_rate": 3.306671775380188e-06,
+      "loss": 0.5319,
+      "step": 10629
+    },
+    {
+      "epoch": 1.8962625992328963,
+      "grad_norm": 0.541481077671051,
+      "learning_rate": 3.2953239311647175e-06,
+      "loss": 0.3925,
+      "step": 10630
+    },
+    {
+      "epoch": 1.8964409954508965,
+      "grad_norm": 0.5258289575576782,
+      "learning_rate": 3.2839954630283497e-06,
+      "loss": 0.4991,
+      "step": 10631
+    },
+    {
+      "epoch": 1.8966193916688967,
+      "grad_norm": 0.6551702618598938,
+      "learning_rate": 3.272686371860789e-06,
+      "loss": 0.6916,
+      "step": 10632
+    },
+    {
+      "epoch": 1.896797787886897,
+      "grad_norm": 0.5645005702972412,
+      "learning_rate": 3.2613966585502976e-06,
+      "loss": 0.5999,
+      "step": 10633
+    },
+    {
+      "epoch": 1.896976184104897,
+      "grad_norm": 0.7276445627212524,
+      "learning_rate": 3.2501263239834987e-06,
+      "loss": 0.4629,
+      "step": 10634
+    },
+    {
+      "epoch": 1.8971545803228973,
+      "grad_norm": 0.5128322839736938,
+      "learning_rate": 3.2388753690456296e-06,
+      "loss": 0.5149,
+      "step": 10635
+    },
+    {
+      "epoch": 1.8973329765408975,
+      "grad_norm": 0.4655255377292633,
+      "learning_rate": 3.2276437946202607e-06,
+      "loss": 0.3665,
+      "step": 10636
+    },
+    {
+      "epoch": 1.8975113727588975,
+      "grad_norm": 0.5978597402572632,
+      "learning_rate": 3.2164316015895476e-06,
+      "loss": 0.6666,
+      "step": 10637
+    },
+    {
+      "epoch": 1.8976897689768977,
+      "grad_norm": 0.5160049796104431,
+      "learning_rate": 3.205238790834147e-06,
+      "loss": 0.4178,
+      "step": 10638
+    },
+    {
+      "epoch": 1.8978681651948979,
+      "grad_norm": 0.5093833804130554,
+      "learning_rate": 3.194065363233051e-06,
+      "loss": 0.5388,
+      "step": 10639
+    },
+    {
+      "epoch": 1.898046561412898,
+      "grad_norm": 0.5614250898361206,
+      "learning_rate": 3.1829113196638614e-06,
+      "loss": 0.4887,
+      "step": 10640
+    },
+    {
+      "epoch": 1.8982249576308983,
+      "grad_norm": 0.7175207734107971,
+      "learning_rate": 3.171776661002601e-06,
+      "loss": 0.4711,
+      "step": 10641
+    },
+    {
+      "epoch": 1.8984033538488982,
+      "grad_norm": 0.5213533043861389,
+      "learning_rate": 3.1606613881237924e-06,
+      "loss": 0.4632,
+      "step": 10642
+    },
+    {
+      "epoch": 1.8985817500668984,
+      "grad_norm": 0.5014880895614624,
+      "learning_rate": 3.1495655019004032e-06,
+      "loss": 0.4144,
+      "step": 10643
+    },
+    {
+      "epoch": 1.8987601462848986,
+      "grad_norm": 0.5747901201248169,
+      "learning_rate": 3.138489003203904e-06,
+      "loss": 0.6008,
+      "step": 10644
+    },
+    {
+      "epoch": 1.8989385425028988,
+      "grad_norm": 0.5374071598052979,
+      "learning_rate": 3.1274318929042644e-06,
+      "loss": 0.435,
+      "step": 10645
+    },
+    {
+      "epoch": 1.899116938720899,
+      "grad_norm": 0.457996129989624,
+      "learning_rate": 3.116394171869874e-06,
+      "loss": 0.4792,
+      "step": 10646
+    },
+    {
+      "epoch": 1.8992953349388992,
+      "grad_norm": 0.4922611713409424,
+      "learning_rate": 3.105375840967678e-06,
+      "loss": 0.3421,
+      "step": 10647
+    },
+    {
+      "epoch": 1.8994737311568994,
+      "grad_norm": 0.5241499543190002,
+      "learning_rate": 3.0943769010629565e-06,
+      "loss": 0.5663,
+      "step": 10648
+    },
+    {
+      "epoch": 1.8996521273748996,
+      "grad_norm": 0.4872463345527649,
+      "learning_rate": 3.0833973530196846e-06,
+      "loss": 0.3615,
+      "step": 10649
+    },
+    {
+      "epoch": 1.8998305235928998,
+      "grad_norm": 0.5093347430229187,
+      "learning_rate": 3.07243719770009e-06,
+      "loss": 0.477,
+      "step": 10650
+    },
+    {
+      "epoch": 1.9000089198109,
+      "grad_norm": 0.48740455508232117,
+      "learning_rate": 3.0614964359650112e-06,
+      "loss": 0.4094,
+      "step": 10651
+    },
+    {
+      "epoch": 1.9001873160289002,
+      "grad_norm": 0.46468567848205566,
+      "learning_rate": 3.0505750686737332e-06,
+      "loss": 0.3435,
+      "step": 10652
+    },
+    {
+      "epoch": 1.9003657122469004,
+      "grad_norm": 0.49886494874954224,
+      "learning_rate": 3.0396730966840423e-06,
+      "loss": 0.4067,
+      "step": 10653
+    },
+    {
+      "epoch": 1.9005441084649006,
+      "grad_norm": 0.5205758213996887,
+      "learning_rate": 3.0287905208521427e-06,
+      "loss": 0.3478,
+      "step": 10654
+    },
+    {
+      "epoch": 1.9007225046829008,
+      "grad_norm": 0.5665127635002136,
+      "learning_rate": 3.017927342032767e-06,
+      "loss": 0.5246,
+      "step": 10655
+    },
+    {
+      "epoch": 1.900900900900901,
+      "grad_norm": 0.5088105797767639,
+      "learning_rate": 3.007083561079066e-06,
+      "loss": 0.3591,
+      "step": 10656
+    },
+    {
+      "epoch": 1.9010792971189012,
+      "grad_norm": 0.49926015734672546,
+      "learning_rate": 2.9962591788427473e-06,
+      "loss": 0.4791,
+      "step": 10657
+    },
+    {
+      "epoch": 1.9012576933369014,
+      "grad_norm": 0.5580708384513855,
+      "learning_rate": 2.985454196173937e-06,
+      "loss": 0.4928,
+      "step": 10658
+    },
+    {
+      "epoch": 1.9014360895549014,
+      "grad_norm": 0.5072367787361145,
+      "learning_rate": 2.974668613921261e-06,
+      "loss": 0.4386,
+      "step": 10659
+    },
+    {
+      "epoch": 1.9016144857729016,
+      "grad_norm": 0.5622830986976624,
+      "learning_rate": 2.963902432931792e-06,
+      "loss": 0.4739,
+      "step": 10660
+    },
+    {
+      "epoch": 1.9017928819909018,
+      "grad_norm": 0.5466598272323608,
+      "learning_rate": 2.9531556540511593e-06,
+      "loss": 0.5038,
+      "step": 10661
+    },
+    {
+      "epoch": 1.901971278208902,
+      "grad_norm": 0.5235751271247864,
+      "learning_rate": 2.942428278123327e-06,
+      "loss": 0.5349,
+      "step": 10662
+    },
+    {
+      "epoch": 1.9021496744269022,
+      "grad_norm": 0.5213757753372192,
+      "learning_rate": 2.9317203059908703e-06,
+      "loss": 0.4482,
+      "step": 10663
+    },
+    {
+      "epoch": 1.9023280706449022,
+      "grad_norm": 0.537200391292572,
+      "learning_rate": 2.921031738494784e-06,
+      "loss": 0.473,
+      "step": 10664
+    },
+    {
+      "epoch": 1.9025064668629024,
+      "grad_norm": 0.5032278895378113,
+      "learning_rate": 2.9103625764745344e-06,
+      "loss": 0.4604,
+      "step": 10665
+    },
+    {
+      "epoch": 1.9026848630809026,
+      "grad_norm": 0.5438551902770996,
+      "learning_rate": 2.899712820768091e-06,
+      "loss": 0.5475,
+      "step": 10666
+    },
+    {
+      "epoch": 1.9028632592989028,
+      "grad_norm": 0.5733838677406311,
+      "learning_rate": 2.8890824722118403e-06,
+      "loss": 0.4181,
+      "step": 10667
+    },
+    {
+      "epoch": 1.903041655516903,
+      "grad_norm": 0.5377230644226074,
+      "learning_rate": 2.878471531640725e-06,
+      "loss": 0.4601,
+      "step": 10668
+    },
+    {
+      "epoch": 1.9032200517349032,
+      "grad_norm": 0.5529724955558777,
+      "learning_rate": 2.8678799998881345e-06,
+      "loss": 0.4844,
+      "step": 10669
+    },
+    {
+      "epoch": 1.9033984479529034,
+      "grad_norm": 0.5532283186912537,
+      "learning_rate": 2.8573078777858753e-06,
+      "loss": 0.5625,
+      "step": 10670
+    },
+    {
+      "epoch": 1.9035768441709036,
+      "grad_norm": 0.5340506434440613,
+      "learning_rate": 2.846755166164311e-06,
+      "loss": 0.4384,
+      "step": 10671
+    },
+    {
+      "epoch": 1.9037552403889038,
+      "grad_norm": 0.520977258682251,
+      "learning_rate": 2.836221865852223e-06,
+      "loss": 0.4364,
+      "step": 10672
+    },
+    {
+      "epoch": 1.903933636606904,
+      "grad_norm": 0.47905829548835754,
+      "learning_rate": 2.8257079776769224e-06,
+      "loss": 0.415,
+      "step": 10673
+    },
+    {
+      "epoch": 1.9041120328249042,
+      "grad_norm": 0.540252149105072,
+      "learning_rate": 2.8152135024641366e-06,
+      "loss": 0.5036,
+      "step": 10674
+    },
+    {
+      "epoch": 1.9042904290429044,
+      "grad_norm": 0.4911253750324249,
+      "learning_rate": 2.8047384410381237e-06,
+      "loss": 0.4359,
+      "step": 10675
+    },
+    {
+      "epoch": 1.9044688252609046,
+      "grad_norm": 0.5184650421142578,
+      "learning_rate": 2.7942827942215585e-06,
+      "loss": 0.4715,
+      "step": 10676
+    },
+    {
+      "epoch": 1.9046472214789048,
+      "grad_norm": 0.4889717102050781,
+      "learning_rate": 2.783846562835646e-06,
+      "loss": 0.3955,
+      "step": 10677
+    },
+    {
+      "epoch": 1.904825617696905,
+      "grad_norm": 0.5702592134475708,
+      "learning_rate": 2.7734297477000626e-06,
+      "loss": 0.5709,
+      "step": 10678
+    },
+    {
+      "epoch": 1.9050040139149051,
+      "grad_norm": 0.5835392475128174,
+      "learning_rate": 2.7630323496328767e-06,
+      "loss": 0.4535,
+      "step": 10679
+    },
+    {
+      "epoch": 1.9051824101329053,
+      "grad_norm": 0.5429273843765259,
+      "learning_rate": 2.752654369450769e-06,
+      "loss": 0.5521,
+      "step": 10680
+    },
+    {
+      "epoch": 1.9053608063509053,
+      "grad_norm": 0.5422603487968445,
+      "learning_rate": 2.742295807968781e-06,
+      "loss": 0.5759,
+      "step": 10681
+    },
+    {
+      "epoch": 1.9055392025689055,
+      "grad_norm": 0.5456951856613159,
+      "learning_rate": 2.731956666000457e-06,
+      "loss": 0.4278,
+      "step": 10682
+    },
+    {
+      "epoch": 1.9057175987869057,
+      "grad_norm": 0.5333420038223267,
+      "learning_rate": 2.7216369443578694e-06,
+      "loss": 0.5113,
+      "step": 10683
+    },
+    {
+      "epoch": 1.905895995004906,
+      "grad_norm": 0.6626531481742859,
+      "learning_rate": 2.7113366438515085e-06,
+      "loss": 0.6753,
+      "step": 10684
+    },
+    {
+      "epoch": 1.9060743912229061,
+      "grad_norm": 0.5112046003341675,
+      "learning_rate": 2.701055765290339e-06,
+      "loss": 0.4424,
+      "step": 10685
+    },
+    {
+      "epoch": 1.906252787440906,
+      "grad_norm": 0.48551252484321594,
+      "learning_rate": 2.6907943094818255e-06,
+      "loss": 0.4418,
+      "step": 10686
+    },
+    {
+      "epoch": 1.9064311836589063,
+      "grad_norm": 0.5245196223258972,
+      "learning_rate": 2.680552277231907e-06,
+      "loss": 0.4654,
+      "step": 10687
+    },
+    {
+      "epoch": 1.9066095798769065,
+      "grad_norm": 0.5276603102684021,
+      "learning_rate": 2.6703296693449953e-06,
+      "loss": 0.4482,
+      "step": 10688
+    },
+    {
+      "epoch": 1.9067879760949067,
+      "grad_norm": 0.5473247766494751,
+      "learning_rate": 2.6601264866239485e-06,
+      "loss": 0.5329,
+      "step": 10689
+    },
+    {
+      "epoch": 1.9069663723129069,
+      "grad_norm": 0.6117315292358398,
+      "learning_rate": 2.649942729870125e-06,
+      "loss": 0.5775,
+      "step": 10690
+    },
+    {
+      "epoch": 1.907144768530907,
+      "grad_norm": 0.5208797454833984,
+      "learning_rate": 2.639778399883358e-06,
+      "loss": 0.5727,
+      "step": 10691
+    },
+    {
+      "epoch": 1.9073231647489073,
+      "grad_norm": 0.5800744295120239,
+      "learning_rate": 2.6296334974619528e-06,
+      "loss": 0.6938,
+      "step": 10692
+    },
+    {
+      "epoch": 1.9075015609669075,
+      "grad_norm": 0.49815651774406433,
+      "learning_rate": 2.6195080234026613e-06,
+      "loss": 0.4241,
+      "step": 10693
+    },
+    {
+      "epoch": 1.9076799571849077,
+      "grad_norm": 0.515434980392456,
+      "learning_rate": 2.6094019785007914e-06,
+      "loss": 0.4804,
+      "step": 10694
+    },
+    {
+      "epoch": 1.9078583534029079,
+      "grad_norm": 0.5178319215774536,
+      "learning_rate": 2.5993153635500143e-06,
+      "loss": 0.3885,
+      "step": 10695
+    },
+    {
+      "epoch": 1.908036749620908,
+      "grad_norm": 0.48950934410095215,
+      "learning_rate": 2.589248179342529e-06,
+      "loss": 0.395,
+      "step": 10696
+    },
+    {
+      "epoch": 1.9082151458389083,
+      "grad_norm": 0.5800197124481201,
+      "learning_rate": 2.5792004266690094e-06,
+      "loss": 0.6178,
+      "step": 10697
+    },
+    {
+      "epoch": 1.9083935420569085,
+      "grad_norm": 0.6045670509338379,
+      "learning_rate": 2.569172106318629e-06,
+      "loss": 0.511,
+      "step": 10698
+    },
+    {
+      "epoch": 1.9085719382749087,
+      "grad_norm": 0.4140417277812958,
+      "learning_rate": 2.5591632190790084e-06,
+      "loss": 0.2933,
+      "step": 10699
+    },
+    {
+      "epoch": 1.9087503344929089,
+      "grad_norm": 0.5823186635971069,
+      "learning_rate": 2.5491737657362123e-06,
+      "loss": 0.4683,
+      "step": 10700
+    },
+    {
+      "epoch": 1.908928730710909,
+      "grad_norm": 0.5681852102279663,
+      "learning_rate": 2.5392037470748365e-06,
+      "loss": 0.5934,
+      "step": 10701
+    },
+    {
+      "epoch": 1.9091071269289093,
+      "grad_norm": 0.5418094992637634,
+      "learning_rate": 2.5292531638778926e-06,
+      "loss": 0.5774,
+      "step": 10702
+    },
+    {
+      "epoch": 1.9092855231469092,
+      "grad_norm": 0.47225964069366455,
+      "learning_rate": 2.519322016926895e-06,
+      "loss": 0.3773,
+      "step": 10703
+    },
+    {
+      "epoch": 1.9094639193649094,
+      "grad_norm": 0.5559024810791016,
+      "learning_rate": 2.5094103070018858e-06,
+      "loss": 0.3676,
+      "step": 10704
+    },
+    {
+      "epoch": 1.9096423155829096,
+      "grad_norm": 0.5408554077148438,
+      "learning_rate": 2.499518034881271e-06,
+      "loss": 0.3963,
+      "step": 10705
+    },
+    {
+      "epoch": 1.9098207118009098,
+      "grad_norm": 0.4998513162136078,
+      "learning_rate": 2.4896452013420114e-06,
+      "loss": 0.3837,
+      "step": 10706
+    },
+    {
+      "epoch": 1.90999910801891,
+      "grad_norm": 0.5326787233352661,
+      "learning_rate": 2.4797918071594594e-06,
+      "loss": 0.5206,
+      "step": 10707
+    },
+    {
+      "epoch": 1.91017750423691,
+      "grad_norm": 0.5712464451789856,
+      "learning_rate": 2.469957853107607e-06,
+      "loss": 0.5556,
+      "step": 10708
+    },
+    {
+      "epoch": 1.9103559004549102,
+      "grad_norm": 0.5410659909248352,
+      "learning_rate": 2.460143339958726e-06,
+      "loss": 0.4366,
+      "step": 10709
+    },
+    {
+      "epoch": 1.9105342966729104,
+      "grad_norm": 0.5795556902885437,
+      "learning_rate": 2.4503482684836154e-06,
+      "loss": 0.5793,
+      "step": 10710
+    },
+    {
+      "epoch": 1.9107126928909106,
+      "grad_norm": 0.6724327802658081,
+      "learning_rate": 2.440572639451688e-06,
+      "loss": 0.6335,
+      "step": 10711
+    },
+    {
+      "epoch": 1.9108910891089108,
+      "grad_norm": 0.44440463185310364,
+      "learning_rate": 2.4308164536306075e-06,
+      "loss": 0.3103,
+      "step": 10712
+    },
+    {
+      "epoch": 1.911069485326911,
+      "grad_norm": 0.49097147583961487,
+      "learning_rate": 2.421079711786678e-06,
+      "loss": 0.3918,
+      "step": 10713
+    },
+    {
+      "epoch": 1.9112478815449112,
+      "grad_norm": 0.5756399035453796,
+      "learning_rate": 2.4113624146846205e-06,
+      "loss": 0.669,
+      "step": 10714
+    },
+    {
+      "epoch": 1.9114262777629114,
+      "grad_norm": 0.5272496938705444,
+      "learning_rate": 2.4016645630875744e-06,
+      "loss": 0.4483,
+      "step": 10715
+    },
+    {
+      "epoch": 1.9116046739809116,
+      "grad_norm": 0.561396062374115,
+      "learning_rate": 2.391986157757292e-06,
+      "loss": 0.6147,
+      "step": 10716
+    },
+    {
+      "epoch": 1.9117830701989118,
+      "grad_norm": 0.6622545123100281,
+      "learning_rate": 2.382327199453832e-06,
+      "loss": 0.4827,
+      "step": 10717
+    },
+    {
+      "epoch": 1.911961466416912,
+      "grad_norm": 0.5634920597076416,
+      "learning_rate": 2.372687688935837e-06,
+      "loss": 0.5104,
+      "step": 10718
+    },
+    {
+      "epoch": 1.9121398626349122,
+      "grad_norm": 0.5393690466880798,
+      "learning_rate": 2.363067626960397e-06,
+      "loss": 0.4723,
+      "step": 10719
+    },
+    {
+      "epoch": 1.9123182588529124,
+      "grad_norm": 0.501996636390686,
+      "learning_rate": 2.353467014283073e-06,
+      "loss": 0.4674,
+      "step": 10720
+    },
+    {
+      "epoch": 1.9124966550709126,
+      "grad_norm": 0.5242270827293396,
+      "learning_rate": 2.3438858516578733e-06,
+      "loss": 0.4293,
+      "step": 10721
+    },
+    {
+      "epoch": 1.9126750512889128,
+      "grad_norm": 0.5785874128341675,
+      "learning_rate": 2.3343241398372796e-06,
+      "loss": 0.5556,
+      "step": 10722
+    },
+    {
+      "epoch": 1.912853447506913,
+      "grad_norm": 0.49174314737319946,
+      "learning_rate": 2.324781879572302e-06,
+      "loss": 0.3852,
+      "step": 10723
+    },
+    {
+      "epoch": 1.9130318437249132,
+      "grad_norm": 0.56098473072052,
+      "learning_rate": 2.3152590716123688e-06,
+      "loss": 0.5528,
+      "step": 10724
+    },
+    {
+      "epoch": 1.9132102399429132,
+      "grad_norm": 0.5013071894645691,
+      "learning_rate": 2.3057557167054377e-06,
+      "loss": 0.4798,
+      "step": 10725
+    },
+    {
+      "epoch": 1.9133886361609134,
+      "grad_norm": 0.5687433481216431,
+      "learning_rate": 2.2962718155978e-06,
+      "loss": 0.5398,
+      "step": 10726
+    },
+    {
+      "epoch": 1.9135670323789136,
+      "grad_norm": 0.5144028067588806,
+      "learning_rate": 2.286807369034416e-06,
+      "loss": 0.3893,
+      "step": 10727
+    },
+    {
+      "epoch": 1.9137454285969138,
+      "grad_norm": 0.48423799872398376,
+      "learning_rate": 2.2773623777585796e-06,
+      "loss": 0.5236,
+      "step": 10728
+    },
+    {
+      "epoch": 1.913923824814914,
+      "grad_norm": 0.5465452671051025,
+      "learning_rate": 2.267936842512086e-06,
+      "loss": 0.5458,
+      "step": 10729
+    },
+    {
+      "epoch": 1.914102221032914,
+      "grad_norm": 0.4663107991218567,
+      "learning_rate": 2.2585307640352326e-06,
+      "loss": 0.4132,
+      "step": 10730
+    },
+    {
+      "epoch": 1.9142806172509141,
+      "grad_norm": 0.5530097484588623,
+      "learning_rate": 2.2491441430667614e-06,
+      "loss": 0.6104,
+      "step": 10731
+    },
+    {
+      "epoch": 1.9144590134689143,
+      "grad_norm": 0.5279607176780701,
+      "learning_rate": 2.239776980343916e-06,
+      "loss": 0.4833,
+      "step": 10732
+    },
+    {
+      "epoch": 1.9146374096869145,
+      "grad_norm": 0.554326057434082,
+      "learning_rate": 2.2304292766023304e-06,
+      "loss": 0.5667,
+      "step": 10733
+    },
+    {
+      "epoch": 1.9148158059049147,
+      "grad_norm": 0.5720328688621521,
+      "learning_rate": 2.2211010325762227e-06,
+      "loss": 0.5472,
+      "step": 10734
+    },
+    {
+      "epoch": 1.914994202122915,
+      "grad_norm": 0.6049355864524841,
+      "learning_rate": 2.211792248998229e-06,
+      "loss": 0.7378,
+      "step": 10735
+    },
+    {
+      "epoch": 1.9151725983409151,
+      "grad_norm": 0.5384811758995056,
+      "learning_rate": 2.2025029265994036e-06,
+      "loss": 0.588,
+      "step": 10736
+    },
+    {
+      "epoch": 1.9153509945589153,
+      "grad_norm": 0.5090281963348389,
+      "learning_rate": 2.1932330661093845e-06,
+      "loss": 0.5302,
+      "step": 10737
+    },
+    {
+      "epoch": 1.9155293907769155,
+      "grad_norm": 0.5153109431266785,
+      "learning_rate": 2.183982668256201e-06,
+      "loss": 0.5014,
+      "step": 10738
+    },
+    {
+      "epoch": 1.9157077869949157,
+      "grad_norm": 0.5636624693870544,
+      "learning_rate": 2.1747517337663836e-06,
+      "loss": 0.6512,
+      "step": 10739
+    },
+    {
+      "epoch": 1.915886183212916,
+      "grad_norm": 0.5594334006309509,
+      "learning_rate": 2.1655402633648792e-06,
+      "loss": 0.5636,
+      "step": 10740
+    },
+    {
+      "epoch": 1.9160645794309161,
+      "grad_norm": 0.5760837197303772,
+      "learning_rate": 2.1563482577752482e-06,
+      "loss": 0.5173,
+      "step": 10741
+    },
+    {
+      "epoch": 1.9162429756489163,
+      "grad_norm": 0.5558215379714966,
+      "learning_rate": 2.1471757177193295e-06,
+      "loss": 0.486,
+      "step": 10742
+    },
+    {
+      "epoch": 1.9164213718669165,
+      "grad_norm": 0.4847973883152008,
+      "learning_rate": 2.1380226439175478e-06,
+      "loss": 0.3862,
+      "step": 10743
+    },
+    {
+      "epoch": 1.9165997680849167,
+      "grad_norm": 0.548984944820404,
+      "learning_rate": 2.1288890370888546e-06,
+      "loss": 0.546,
+      "step": 10744
+    },
+    {
+      "epoch": 1.916778164302917,
+      "grad_norm": 0.5168630480766296,
+      "learning_rate": 2.119774897950538e-06,
+      "loss": 0.5107,
+      "step": 10745
+    },
+    {
+      "epoch": 1.9169565605209171,
+      "grad_norm": 0.5281267762184143,
+      "learning_rate": 2.1106802272184146e-06,
+      "loss": 0.4927,
+      "step": 10746
+    },
+    {
+      "epoch": 1.917134956738917,
+      "grad_norm": 0.5126060843467712,
+      "learning_rate": 2.1016050256068007e-06,
+      "loss": 0.3642,
+      "step": 10747
+    },
+    {
+      "epoch": 1.9173133529569173,
+      "grad_norm": 0.5802842378616333,
+      "learning_rate": 2.092549293828433e-06,
+      "loss": 0.5518,
+      "step": 10748
+    },
+    {
+      "epoch": 1.9174917491749175,
+      "grad_norm": 0.5918566584587097,
+      "learning_rate": 2.0835130325946026e-06,
+      "loss": 0.5455,
+      "step": 10749
+    },
+    {
+      "epoch": 1.9176701453929177,
+      "grad_norm": 0.5723239779472351,
+      "learning_rate": 2.0744962426149374e-06,
+      "loss": 0.5907,
+      "step": 10750
+    },
+    {
+      "epoch": 1.917848541610918,
+      "grad_norm": 0.4773622751235962,
+      "learning_rate": 2.0654989245976473e-06,
+      "loss": 0.4396,
+      "step": 10751
+    },
+    {
+      "epoch": 1.9180269378289179,
+      "grad_norm": 0.5784128308296204,
+      "learning_rate": 2.0565210792493903e-06,
+      "loss": 0.6173,
+      "step": 10752
+    },
+    {
+      "epoch": 1.918205334046918,
+      "grad_norm": 0.5458529591560364,
+      "learning_rate": 2.047562707275269e-06,
+      "loss": 0.5314,
+      "step": 10753
+    },
+    {
+      "epoch": 1.9183837302649183,
+      "grad_norm": 0.4943179488182068,
+      "learning_rate": 2.038623809378859e-06,
+      "loss": 0.3984,
+      "step": 10754
+    },
+    {
+      "epoch": 1.9185621264829185,
+      "grad_norm": 0.6198917031288147,
+      "learning_rate": 2.029704386262238e-06,
+      "loss": 0.7195,
+      "step": 10755
+    },
+    {
+      "epoch": 1.9187405227009187,
+      "grad_norm": 0.5691092610359192,
+      "learning_rate": 2.020804438625928e-06,
+      "loss": 0.6608,
+      "step": 10756
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 0.6087114214897156,
+      "learning_rate": 2.011923967168899e-06,
+      "loss": 0.7876,
+      "step": 10757
+    },
+    {
+      "epoch": 1.919097315136919,
+      "grad_norm": 0.5224511027336121,
+      "learning_rate": 2.0030629725886763e-06,
+      "loss": 0.418,
+      "step": 10758
+    },
+    {
+      "epoch": 1.9192757113549193,
+      "grad_norm": 0.5266249775886536,
+      "learning_rate": 1.9942214555811187e-06,
+      "loss": 0.4646,
+      "step": 10759
+    },
+    {
+      "epoch": 1.9194541075729195,
+      "grad_norm": 0.5271503925323486,
+      "learning_rate": 1.9853994168407273e-06,
+      "loss": 0.4795,
+      "step": 10760
+    },
+    {
+      "epoch": 1.9196325037909197,
+      "grad_norm": 0.5206458568572998,
+      "learning_rate": 1.9765968570603365e-06,
+      "loss": 0.449,
+      "step": 10761
+    },
+    {
+      "epoch": 1.9198109000089199,
+      "grad_norm": 0.5074752569198608,
+      "learning_rate": 1.9678137769312543e-06,
+      "loss": 0.4446,
+      "step": 10762
+    },
+    {
+      "epoch": 1.91998929622692,
+      "grad_norm": 0.4914679229259491,
+      "learning_rate": 1.9590501771433735e-06,
+      "loss": 0.4738,
+      "step": 10763
+    },
+    {
+      "epoch": 1.9201676924449202,
+      "grad_norm": 0.5547556281089783,
+      "learning_rate": 1.9503060583849485e-06,
+      "loss": 0.5181,
+      "step": 10764
+    },
+    {
+      "epoch": 1.9203460886629204,
+      "grad_norm": 0.45473459362983704,
+      "learning_rate": 1.941581421342764e-06,
+      "loss": 0.3413,
+      "step": 10765
+    },
+    {
+      "epoch": 1.9205244848809206,
+      "grad_norm": 0.5488914251327515,
+      "learning_rate": 1.932876266701994e-06,
+      "loss": 0.5253,
+      "step": 10766
+    },
+    {
+      "epoch": 1.9207028810989208,
+      "grad_norm": 0.5307526588439941,
+      "learning_rate": 1.924190595146369e-06,
+      "loss": 0.5097,
+      "step": 10767
+    },
+    {
+      "epoch": 1.920881277316921,
+      "grad_norm": 0.530610203742981,
+      "learning_rate": 1.9155244073580936e-06,
+      "loss": 0.4198,
+      "step": 10768
+    },
+    {
+      "epoch": 1.921059673534921,
+      "grad_norm": 0.48655280470848083,
+      "learning_rate": 1.9068777040177342e-06,
+      "loss": 0.4815,
+      "step": 10769
+    },
+    {
+      "epoch": 1.9212380697529212,
+      "grad_norm": 0.5858240127563477,
+      "learning_rate": 1.8982504858044703e-06,
+      "loss": 0.5908,
+      "step": 10770
+    },
+    {
+      "epoch": 1.9214164659709214,
+      "grad_norm": 0.5390949249267578,
+      "learning_rate": 1.889642753395815e-06,
+      "loss": 0.5308,
+      "step": 10771
+    },
+    {
+      "epoch": 1.9215948621889216,
+      "grad_norm": 0.4654117524623871,
+      "learning_rate": 1.8810545074678664e-06,
+      "loss": 0.3985,
+      "step": 10772
+    },
+    {
+      "epoch": 1.9217732584069218,
+      "grad_norm": 0.570939838886261,
+      "learning_rate": 1.872485748695113e-06,
+      "loss": 0.5217,
+      "step": 10773
+    },
+    {
+      "epoch": 1.9219516546249218,
+      "grad_norm": 0.5508378148078918,
+      "learning_rate": 1.8639364777505442e-06,
+      "loss": 0.5657,
+      "step": 10774
+    },
+    {
+      "epoch": 1.922130050842922,
+      "grad_norm": 0.49652281403541565,
+      "learning_rate": 1.8554066953056502e-06,
+      "loss": 0.4462,
+      "step": 10775
+    },
+    {
+      "epoch": 1.9223084470609222,
+      "grad_norm": 0.5410851836204529,
+      "learning_rate": 1.8468964020302847e-06,
+      "loss": 0.5473,
+      "step": 10776
+    },
+    {
+      "epoch": 1.9224868432789224,
+      "grad_norm": 0.5557335615158081,
+      "learning_rate": 1.8384055985929405e-06,
+      "loss": 0.6211,
+      "step": 10777
+    },
+    {
+      "epoch": 1.9226652394969226,
+      "grad_norm": 0.49317964911460876,
+      "learning_rate": 1.8299342856603895e-06,
+      "loss": 0.4219,
+      "step": 10778
+    },
+    {
+      "epoch": 1.9228436357149228,
+      "grad_norm": 0.5330275297164917,
+      "learning_rate": 1.8214824638980166e-06,
+      "loss": 0.4602,
+      "step": 10779
+    },
+    {
+      "epoch": 1.923022031932923,
+      "grad_norm": 0.5371294617652893,
+      "learning_rate": 1.8130501339696237e-06,
+      "loss": 0.4941,
+      "step": 10780
+    },
+    {
+      "epoch": 1.9232004281509232,
+      "grad_norm": 0.5647792220115662,
+      "learning_rate": 1.8046372965374592e-06,
+      "loss": 0.5879,
+      "step": 10781
+    },
+    {
+      "epoch": 1.9233788243689234,
+      "grad_norm": 0.5888603329658508,
+      "learning_rate": 1.796243952262272e-06,
+      "loss": 0.6028,
+      "step": 10782
+    },
+    {
+      "epoch": 1.9235572205869236,
+      "grad_norm": 0.5580531358718872,
+      "learning_rate": 1.787870101803285e-06,
+      "loss": 0.5612,
+      "step": 10783
+    },
+    {
+      "epoch": 1.9237356168049238,
+      "grad_norm": 0.5175449848175049,
+      "learning_rate": 1.7795157458181389e-06,
+      "loss": 0.4524,
+      "step": 10784
+    },
+    {
+      "epoch": 1.923914013022924,
+      "grad_norm": 0.5534098148345947,
+      "learning_rate": 1.7711808849630307e-06,
+      "loss": 0.5132,
+      "step": 10785
+    },
+    {
+      "epoch": 1.9240924092409242,
+      "grad_norm": 0.5773023962974548,
+      "learning_rate": 1.7628655198925481e-06,
+      "loss": 0.5995,
+      "step": 10786
+    },
+    {
+      "epoch": 1.9242708054589244,
+      "grad_norm": 0.4896835386753082,
+      "learning_rate": 1.7545696512597797e-06,
+      "loss": 0.4349,
+      "step": 10787
+    },
+    {
+      "epoch": 1.9244492016769246,
+      "grad_norm": 0.4547155797481537,
+      "learning_rate": 1.7462932797163156e-06,
+      "loss": 0.3644,
+      "step": 10788
+    },
+    {
+      "epoch": 1.9246275978949248,
+      "grad_norm": 0.5199649333953857,
+      "learning_rate": 1.738036405912108e-06,
+      "loss": 0.5188,
+      "step": 10789
+    },
+    {
+      "epoch": 1.924805994112925,
+      "grad_norm": 0.5165232419967651,
+      "learning_rate": 1.7297990304956934e-06,
+      "loss": 0.4517,
+      "step": 10790
+    },
+    {
+      "epoch": 1.924984390330925,
+      "grad_norm": 0.5304819345474243,
+      "learning_rate": 1.721581154114027e-06,
+      "loss": 0.4473,
+      "step": 10791
+    },
+    {
+      "epoch": 1.9251627865489251,
+      "grad_norm": 0.582391619682312,
+      "learning_rate": 1.7133827774125365e-06,
+      "loss": 0.5413,
+      "step": 10792
+    },
+    {
+      "epoch": 1.9253411827669253,
+      "grad_norm": 0.5489014983177185,
+      "learning_rate": 1.7052039010350962e-06,
+      "loss": 0.5362,
+      "step": 10793
+    },
+    {
+      "epoch": 1.9255195789849255,
+      "grad_norm": 0.560737669467926,
+      "learning_rate": 1.6970445256241363e-06,
+      "loss": 0.5321,
+      "step": 10794
+    },
+    {
+      "epoch": 1.9256979752029257,
+      "grad_norm": 0.5768745541572571,
+      "learning_rate": 1.6889046518203943e-06,
+      "loss": 0.5531,
+      "step": 10795
+    },
+    {
+      "epoch": 1.9258763714209257,
+      "grad_norm": 0.4823215901851654,
+      "learning_rate": 1.6807842802632756e-06,
+      "loss": 0.4796,
+      "step": 10796
+    },
+    {
+      "epoch": 1.926054767638926,
+      "grad_norm": 0.5584346055984497,
+      "learning_rate": 1.6726834115904643e-06,
+      "loss": 0.5695,
+      "step": 10797
+    },
+    {
+      "epoch": 1.9262331638569261,
+      "grad_norm": 0.6224119067192078,
+      "learning_rate": 1.6646020464382294e-06,
+      "loss": 0.59,
+      "step": 10798
+    },
+    {
+      "epoch": 1.9264115600749263,
+      "grad_norm": 0.5279660820960999,
+      "learning_rate": 1.6565401854413132e-06,
+      "loss": 0.5258,
+      "step": 10799
+    },
+    {
+      "epoch": 1.9265899562929265,
+      "grad_norm": 0.529108464717865,
+      "learning_rate": 1.648497829232848e-06,
+      "loss": 0.4812,
+      "step": 10800
+    },
+    {
+      "epoch": 1.9267683525109267,
+      "grad_norm": 0.6457714438438416,
+      "learning_rate": 1.6404749784444673e-06,
+      "loss": 0.5662,
+      "step": 10801
+    },
+    {
+      "epoch": 1.926946748728927,
+      "grad_norm": 0.5989267230033875,
+      "learning_rate": 1.632471633706334e-06,
+      "loss": 0.5611,
+      "step": 10802
+    },
+    {
+      "epoch": 1.927125144946927,
+      "grad_norm": 0.4550947844982147,
+      "learning_rate": 1.6244877956469728e-06,
+      "loss": 0.3234,
+      "step": 10803
+    },
+    {
+      "epoch": 1.9273035411649273,
+      "grad_norm": 0.49629008769989014,
+      "learning_rate": 1.6165234648934657e-06,
+      "loss": 0.4037,
+      "step": 10804
+    },
+    {
+      "epoch": 1.9274819373829275,
+      "grad_norm": 0.44870060682296753,
+      "learning_rate": 1.6085786420713123e-06,
+      "loss": 0.2973,
+      "step": 10805
+    },
+    {
+      "epoch": 1.9276603336009277,
+      "grad_norm": 0.5849354267120361,
+      "learning_rate": 1.6006533278045131e-06,
+      "loss": 0.6132,
+      "step": 10806
+    },
+    {
+      "epoch": 1.927838729818928,
+      "grad_norm": 0.4785587787628174,
+      "learning_rate": 1.5927475227155152e-06,
+      "loss": 0.3464,
+      "step": 10807
+    },
+    {
+      "epoch": 1.928017126036928,
+      "grad_norm": 0.4874267876148224,
+      "learning_rate": 1.5848612274252105e-06,
+      "loss": 0.4023,
+      "step": 10808
+    },
+    {
+      "epoch": 1.9281955222549283,
+      "grad_norm": 0.48776915669441223,
+      "learning_rate": 1.5769944425530202e-06,
+      "loss": 0.3752,
+      "step": 10809
+    },
+    {
+      "epoch": 1.9283739184729285,
+      "grad_norm": 0.612381100654602,
+      "learning_rate": 1.5691471687167558e-06,
+      "loss": 0.5973,
+      "step": 10810
+    },
+    {
+      "epoch": 1.9285523146909287,
+      "grad_norm": 0.5721031427383423,
+      "learning_rate": 1.561319406532785e-06,
+      "loss": 0.5869,
+      "step": 10811
+    },
+    {
+      "epoch": 1.928730710908929,
+      "grad_norm": 0.524873673915863,
+      "learning_rate": 1.5535111566158667e-06,
+      "loss": 0.4592,
+      "step": 10812
+    },
+    {
+      "epoch": 1.9289091071269289,
+      "grad_norm": 0.5512157678604126,
+      "learning_rate": 1.5457224195792873e-06,
+      "loss": 0.5546,
+      "step": 10813
+    },
+    {
+      "epoch": 1.929087503344929,
+      "grad_norm": 0.562592625617981,
+      "learning_rate": 1.5379531960347247e-06,
+      "loss": 0.4041,
+      "step": 10814
+    },
+    {
+      "epoch": 1.9292658995629293,
+      "grad_norm": 0.5566669702529907,
+      "learning_rate": 1.530203486592413e-06,
+      "loss": 0.4858,
+      "step": 10815
+    },
+    {
+      "epoch": 1.9294442957809295,
+      "grad_norm": 0.5376592874526978,
+      "learning_rate": 1.5224732918609762e-06,
+      "loss": 0.5448,
+      "step": 10816
+    },
+    {
+      "epoch": 1.9296226919989297,
+      "grad_norm": 0.46925434470176697,
+      "learning_rate": 1.5147626124475955e-06,
+      "loss": 0.4111,
+      "step": 10817
+    },
+    {
+      "epoch": 1.9298010882169296,
+      "grad_norm": 0.5781516432762146,
+      "learning_rate": 1.5070714489577864e-06,
+      "loss": 0.471,
+      "step": 10818
+    },
+    {
+      "epoch": 1.9299794844349298,
+      "grad_norm": 0.7423487305641174,
+      "learning_rate": 1.4993998019956767e-06,
+      "loss": 0.4972,
+      "step": 10819
+    },
+    {
+      "epoch": 1.93015788065293,
+      "grad_norm": 0.487798273563385,
+      "learning_rate": 1.4917476721637569e-06,
+      "loss": 0.4267,
+      "step": 10820
+    },
+    {
+      "epoch": 1.9303362768709302,
+      "grad_norm": 0.47426673769950867,
+      "learning_rate": 1.4841150600630183e-06,
+      "loss": 0.3958,
+      "step": 10821
+    },
+    {
+      "epoch": 1.9305146730889304,
+      "grad_norm": 0.4360140562057495,
+      "learning_rate": 1.4765019662929812e-06,
+      "loss": 0.3328,
+      "step": 10822
+    },
+    {
+      "epoch": 1.9306930693069306,
+      "grad_norm": 0.49829891324043274,
+      "learning_rate": 1.468908391451501e-06,
+      "loss": 0.3546,
+      "step": 10823
+    },
+    {
+      "epoch": 1.9308714655249308,
+      "grad_norm": 0.5402979254722595,
+      "learning_rate": 1.4613343361349897e-06,
+      "loss": 0.5048,
+      "step": 10824
+    },
+    {
+      "epoch": 1.931049861742931,
+      "grad_norm": 0.49174410104751587,
+      "learning_rate": 1.4537798009383596e-06,
+      "loss": 0.4204,
+      "step": 10825
+    },
+    {
+      "epoch": 1.9312282579609312,
+      "grad_norm": 0.5174482464790344,
+      "learning_rate": 1.4462447864548866e-06,
+      "loss": 0.4889,
+      "step": 10826
+    },
+    {
+      "epoch": 1.9314066541789314,
+      "grad_norm": 0.5101271271705627,
+      "learning_rate": 1.4387292932764029e-06,
+      "loss": 0.447,
+      "step": 10827
+    },
+    {
+      "epoch": 1.9315850503969316,
+      "grad_norm": 0.5471371412277222,
+      "learning_rate": 1.4312333219931307e-06,
+      "loss": 0.4726,
+      "step": 10828
+    },
+    {
+      "epoch": 1.9317634466149318,
+      "grad_norm": 0.5995847582817078,
+      "learning_rate": 1.4237568731938488e-06,
+      "loss": 0.5217,
+      "step": 10829
+    },
+    {
+      "epoch": 1.931941842832932,
+      "grad_norm": 0.5153140425682068,
+      "learning_rate": 1.4162999474657267e-06,
+      "loss": 0.4917,
+      "step": 10830
+    },
+    {
+      "epoch": 1.9321202390509322,
+      "grad_norm": 0.4703426957130432,
+      "learning_rate": 1.4088625453944348e-06,
+      "loss": 0.4425,
+      "step": 10831
+    },
+    {
+      "epoch": 1.9322986352689324,
+      "grad_norm": 0.5764087438583374,
+      "learning_rate": 1.401444667564089e-06,
+      "loss": 0.553,
+      "step": 10832
+    },
+    {
+      "epoch": 1.9324770314869326,
+      "grad_norm": 0.7190420627593994,
+      "learning_rate": 1.3940463145573068e-06,
+      "loss": 0.5883,
+      "step": 10833
+    },
+    {
+      "epoch": 1.9326554277049328,
+      "grad_norm": 0.5590774416923523,
+      "learning_rate": 1.3866674869551232e-06,
+      "loss": 0.5026,
+      "step": 10834
+    },
+    {
+      "epoch": 1.9328338239229328,
+      "grad_norm": 0.5439345240592957,
+      "learning_rate": 1.3793081853371026e-06,
+      "loss": 0.5682,
+      "step": 10835
+    },
+    {
+      "epoch": 1.933012220140933,
+      "grad_norm": 0.52024245262146,
+      "learning_rate": 1.3719684102812547e-06,
+      "loss": 0.428,
+      "step": 10836
+    },
+    {
+      "epoch": 1.9331906163589332,
+      "grad_norm": 0.5542047023773193,
+      "learning_rate": 1.3646481623639794e-06,
+      "loss": 0.5323,
+      "step": 10837
+    },
+    {
+      "epoch": 1.9333690125769334,
+      "grad_norm": 0.5421310663223267,
+      "learning_rate": 1.3573474421602617e-06,
+      "loss": 0.4396,
+      "step": 10838
+    },
+    {
+      "epoch": 1.9335474087949336,
+      "grad_norm": 0.4774705767631531,
+      "learning_rate": 1.3500662502434758e-06,
+      "loss": 0.4457,
+      "step": 10839
+    },
+    {
+      "epoch": 1.9337258050129336,
+      "grad_norm": 0.5488468408584595,
+      "learning_rate": 1.34280458718547e-06,
+      "loss": 0.5401,
+      "step": 10840
+    },
+    {
+      "epoch": 1.9339042012309338,
+      "grad_norm": 0.4551815986633301,
+      "learning_rate": 1.3355624535565936e-06,
+      "loss": 0.3325,
+      "step": 10841
+    },
+    {
+      "epoch": 1.934082597448934,
+      "grad_norm": 0.5352870225906372,
+      "learning_rate": 1.3283398499256138e-06,
+      "loss": 0.5896,
+      "step": 10842
+    },
+    {
+      "epoch": 1.9342609936669342,
+      "grad_norm": 0.5666049718856812,
+      "learning_rate": 1.3211367768598548e-06,
+      "loss": 0.5675,
+      "step": 10843
+    },
+    {
+      "epoch": 1.9344393898849344,
+      "grad_norm": 0.5233065485954285,
+      "learning_rate": 1.3139532349249473e-06,
+      "loss": 0.5271,
+      "step": 10844
+    },
+    {
+      "epoch": 1.9346177861029346,
+      "grad_norm": 0.47093701362609863,
+      "learning_rate": 1.3067892246851897e-06,
+      "loss": 0.3446,
+      "step": 10845
+    },
+    {
+      "epoch": 1.9347961823209348,
+      "grad_norm": 0.5583943724632263,
+      "learning_rate": 1.2996447467031326e-06,
+      "loss": 0.5945,
+      "step": 10846
+    },
+    {
+      "epoch": 1.934974578538935,
+      "grad_norm": 0.4825393259525299,
+      "learning_rate": 1.292519801539993e-06,
+      "loss": 0.369,
+      "step": 10847
+    },
+    {
+      "epoch": 1.9351529747569352,
+      "grad_norm": 0.6124469637870789,
+      "learning_rate": 1.2854143897552961e-06,
+      "loss": 0.5456,
+      "step": 10848
+    },
+    {
+      "epoch": 1.9353313709749353,
+      "grad_norm": 0.4597680866718292,
+      "learning_rate": 1.278328511907123e-06,
+      "loss": 0.3464,
+      "step": 10849
+    },
+    {
+      "epoch": 1.9355097671929355,
+      "grad_norm": 0.529517650604248,
+      "learning_rate": 1.2712621685520287e-06,
+      "loss": 0.4432,
+      "step": 10850
+    },
+    {
+      "epoch": 1.9356881634109357,
+      "grad_norm": 0.52036052942276,
+      "learning_rate": 1.2642153602449303e-06,
+      "loss": 0.488,
+      "step": 10851
+    },
+    {
+      "epoch": 1.935866559628936,
+      "grad_norm": 0.5084417462348938,
+      "learning_rate": 1.2571880875393293e-06,
+      "loss": 0.4408,
+      "step": 10852
+    },
+    {
+      "epoch": 1.9360449558469361,
+      "grad_norm": 0.5051507353782654,
+      "learning_rate": 1.2501803509871457e-06,
+      "loss": 0.4353,
+      "step": 10853
+    },
+    {
+      "epoch": 1.9362233520649363,
+      "grad_norm": 0.4483596980571747,
+      "learning_rate": 1.2431921511387167e-06,
+      "loss": 0.3902,
+      "step": 10854
+    },
+    {
+      "epoch": 1.9364017482829365,
+      "grad_norm": 0.5327636003494263,
+      "learning_rate": 1.2362234885429646e-06,
+      "loss": 0.5355,
+      "step": 10855
+    },
+    {
+      "epoch": 1.9365801445009367,
+      "grad_norm": 0.5151218175888062,
+      "learning_rate": 1.229274363747146e-06,
+      "loss": 0.5091,
+      "step": 10856
+    },
+    {
+      "epoch": 1.9367585407189367,
+      "grad_norm": 0.4956924319267273,
+      "learning_rate": 1.2223447772970742e-06,
+      "loss": 0.4205,
+      "step": 10857
+    },
+    {
+      "epoch": 1.936936936936937,
+      "grad_norm": 0.4500170052051544,
+      "learning_rate": 1.2154347297369806e-06,
+      "loss": 0.3676,
+      "step": 10858
+    },
+    {
+      "epoch": 1.9371153331549371,
+      "grad_norm": 0.44707950949668884,
+      "learning_rate": 1.2085442216095977e-06,
+      "loss": 0.3637,
+      "step": 10859
+    },
+    {
+      "epoch": 1.9372937293729373,
+      "grad_norm": 0.5490984916687012,
+      "learning_rate": 1.201673253456076e-06,
+      "loss": 0.498,
+      "step": 10860
+    },
+    {
+      "epoch": 1.9374721255909375,
+      "grad_norm": 0.5158944725990295,
+      "learning_rate": 1.194821825816067e-06,
+      "loss": 0.4,
+      "step": 10861
+    },
+    {
+      "epoch": 1.9376505218089375,
+      "grad_norm": 0.519051194190979,
+      "learning_rate": 1.1879899392276961e-06,
+      "loss": 0.4514,
+      "step": 10862
+    },
+    {
+      "epoch": 1.9378289180269377,
+      "grad_norm": 0.5193673968315125,
+      "learning_rate": 1.1811775942275061e-06,
+      "loss": 0.5199,
+      "step": 10863
+    },
+    {
+      "epoch": 1.9380073142449379,
+      "grad_norm": 0.600447952747345,
+      "learning_rate": 1.174384791350569e-06,
+      "loss": 0.6023,
+      "step": 10864
+    },
+    {
+      "epoch": 1.938185710462938,
+      "grad_norm": 0.5528981685638428,
+      "learning_rate": 1.1676115311303747e-06,
+      "loss": 0.4367,
+      "step": 10865
+    },
+    {
+      "epoch": 1.9383641066809383,
+      "grad_norm": 0.5125691890716553,
+      "learning_rate": 1.160857814098859e-06,
+      "loss": 0.4637,
+      "step": 10866
+    },
+    {
+      "epoch": 1.9385425028989385,
+      "grad_norm": 0.5695079565048218,
+      "learning_rate": 1.154123640786514e-06,
+      "loss": 0.573,
+      "step": 10867
+    },
+    {
+      "epoch": 1.9387208991169387,
+      "grad_norm": 0.5509238243103027,
+      "learning_rate": 1.1474090117221947e-06,
+      "loss": 0.5835,
+      "step": 10868
+    },
+    {
+      "epoch": 1.9388992953349389,
+      "grad_norm": 0.5293681025505066,
+      "learning_rate": 1.1407139274333124e-06,
+      "loss": 0.5052,
+      "step": 10869
+    },
+    {
+      "epoch": 1.939077691552939,
+      "grad_norm": 0.5039731860160828,
+      "learning_rate": 1.1340383884456407e-06,
+      "loss": 0.4013,
+      "step": 10870
+    },
+    {
+      "epoch": 1.9392560877709393,
+      "grad_norm": 0.5404682159423828,
+      "learning_rate": 1.1273823952835106e-06,
+      "loss": 0.5339,
+      "step": 10871
+    },
+    {
+      "epoch": 1.9394344839889395,
+      "grad_norm": 0.49914366006851196,
+      "learning_rate": 1.1207459484696424e-06,
+      "loss": 0.5104,
+      "step": 10872
+    },
+    {
+      "epoch": 1.9396128802069397,
+      "grad_norm": 0.5963449478149414,
+      "learning_rate": 1.1141290485253141e-06,
+      "loss": 0.5313,
+      "step": 10873
+    },
+    {
+      "epoch": 1.9397912764249399,
+      "grad_norm": 0.5142577886581421,
+      "learning_rate": 1.107531695970193e-06,
+      "loss": 0.4771,
+      "step": 10874
+    },
+    {
+      "epoch": 1.93996967264294,
+      "grad_norm": 0.49809107184410095,
+      "learning_rate": 1.1009538913223927e-06,
+      "loss": 0.4144,
+      "step": 10875
+    },
+    {
+      "epoch": 1.9401480688609403,
+      "grad_norm": 0.5837324261665344,
+      "learning_rate": 1.0943956350985828e-06,
+      "loss": 0.6171,
+      "step": 10876
+    },
+    {
+      "epoch": 1.9403264650789405,
+      "grad_norm": 0.48720839619636536,
+      "learning_rate": 1.0878569278138239e-06,
+      "loss": 0.331,
+      "step": 10877
+    },
+    {
+      "epoch": 1.9405048612969407,
+      "grad_norm": 0.5243699550628662,
+      "learning_rate": 1.0813377699816773e-06,
+      "loss": 0.4904,
+      "step": 10878
+    },
+    {
+      "epoch": 1.9406832575149406,
+      "grad_norm": 0.564947783946991,
+      "learning_rate": 1.07483816211415e-06,
+      "loss": 0.6612,
+      "step": 10879
+    },
+    {
+      "epoch": 1.9408616537329408,
+      "grad_norm": 0.5145286321640015,
+      "learning_rate": 1.068358104721695e-06,
+      "loss": 0.5127,
+      "step": 10880
+    },
+    {
+      "epoch": 1.941040049950941,
+      "grad_norm": 0.5018771290779114,
+      "learning_rate": 1.0618975983132662e-06,
+      "loss": 0.4624,
+      "step": 10881
+    },
+    {
+      "epoch": 1.9412184461689412,
+      "grad_norm": 0.490432471036911,
+      "learning_rate": 1.055456643396291e-06,
+      "loss": 0.4464,
+      "step": 10882
+    },
+    {
+      "epoch": 1.9413968423869414,
+      "grad_norm": 0.45111989974975586,
+      "learning_rate": 1.0490352404766146e-06,
+      "loss": 0.3621,
+      "step": 10883
+    },
+    {
+      "epoch": 1.9415752386049414,
+      "grad_norm": 0.5172296762466431,
+      "learning_rate": 1.042633390058556e-06,
+      "loss": 0.4011,
+      "step": 10884
+    },
+    {
+      "epoch": 1.9417536348229416,
+      "grad_norm": 0.5376213192939758,
+      "learning_rate": 1.036251092644963e-06,
+      "loss": 0.5768,
+      "step": 10885
+    },
+    {
+      "epoch": 1.9419320310409418,
+      "grad_norm": 0.5566000938415527,
+      "learning_rate": 1.0298883487370736e-06,
+      "loss": 0.5342,
+      "step": 10886
+    },
+    {
+      "epoch": 1.942110427258942,
+      "grad_norm": 0.5264571309089661,
+      "learning_rate": 1.023545158834599e-06,
+      "loss": 0.4918,
+      "step": 10887
+    },
+    {
+      "epoch": 1.9422888234769422,
+      "grad_norm": 0.4640972316265106,
+      "learning_rate": 1.0172215234357519e-06,
+      "loss": 0.3828,
+      "step": 10888
+    },
+    {
+      "epoch": 1.9424672196949424,
+      "grad_norm": 0.5521480441093445,
+      "learning_rate": 1.0109174430371905e-06,
+      "loss": 0.4931,
+      "step": 10889
+    },
+    {
+      "epoch": 1.9426456159129426,
+      "grad_norm": 0.48122310638427734,
+      "learning_rate": 1.0046329181340187e-06,
+      "loss": 0.3782,
+      "step": 10890
+    },
+    {
+      "epoch": 1.9428240121309428,
+      "grad_norm": 0.5191644430160522,
+      "learning_rate": 9.983679492198139e-07,
+      "loss": 0.4283,
+      "step": 10891
+    },
+    {
+      "epoch": 1.943002408348943,
+      "grad_norm": 0.48529642820358276,
+      "learning_rate": 9.92122536786627e-07,
+      "loss": 0.3826,
+      "step": 10892
+    },
+    {
+      "epoch": 1.9431808045669432,
+      "grad_norm": 0.5107811689376831,
+      "learning_rate": 9.858966813250102e-07,
+      "loss": 0.497,
+      "step": 10893
+    },
+    {
+      "epoch": 1.9433592007849434,
+      "grad_norm": 0.4362351596355438,
+      "learning_rate": 9.796903833239057e-07,
+      "loss": 0.3437,
+      "step": 10894
+    },
+    {
+      "epoch": 1.9435375970029436,
+      "grad_norm": 0.5843741297721863,
+      "learning_rate": 9.73503643270729e-07,
+      "loss": 0.6999,
+      "step": 10895
+    },
+    {
+      "epoch": 1.9437159932209438,
+      "grad_norm": 0.49152183532714844,
+      "learning_rate": 9.673364616514247e-07,
+      "loss": 0.4538,
+      "step": 10896
+    },
+    {
+      "epoch": 1.943894389438944,
+      "grad_norm": 0.6454296112060547,
+      "learning_rate": 9.611888389503553e-07,
+      "loss": 0.5617,
+      "step": 10897
+    },
+    {
+      "epoch": 1.9440727856569442,
+      "grad_norm": 0.5631362795829773,
+      "learning_rate": 9.55060775650357e-07,
+      "loss": 0.502,
+      "step": 10898
+    },
+    {
+      "epoch": 1.9442511818749444,
+      "grad_norm": 0.5519803762435913,
+      "learning_rate": 9.489522722326838e-07,
+      "loss": 0.6406,
+      "step": 10899
+    },
+    {
+      "epoch": 1.9444295780929446,
+      "grad_norm": 0.5168196558952332,
+      "learning_rate": 9.42863329177146e-07,
+      "loss": 0.3918,
+      "step": 10900
+    },
+    {
+      "epoch": 1.9446079743109446,
+      "grad_norm": 0.5088583827018738,
+      "learning_rate": 9.367939469619169e-07,
+      "loss": 0.3789,
+      "step": 10901
+    },
+    {
+      "epoch": 1.9447863705289448,
+      "grad_norm": 0.5261327624320984,
+      "learning_rate": 9.307441260637261e-07,
+      "loss": 0.5678,
+      "step": 10902
+    },
+    {
+      "epoch": 1.944964766746945,
+      "grad_norm": 0.5280705094337463,
+      "learning_rate": 9.247138669577215e-07,
+      "loss": 0.4663,
+      "step": 10903
+    },
+    {
+      "epoch": 1.9451431629649452,
+      "grad_norm": 0.5464169383049011,
+      "learning_rate": 9.187031701174963e-07,
+      "loss": 0.5453,
+      "step": 10904
+    },
+    {
+      "epoch": 1.9453215591829454,
+      "grad_norm": 0.587669312953949,
+      "learning_rate": 9.127120360151175e-07,
+      "loss": 0.5612,
+      "step": 10905
+    },
+    {
+      "epoch": 1.9454999554009453,
+      "grad_norm": 0.512403130531311,
+      "learning_rate": 9.067404651211808e-07,
+      "loss": 0.4129,
+      "step": 10906
+    },
+    {
+      "epoch": 1.9456783516189455,
+      "grad_norm": 0.6009140014648438,
+      "learning_rate": 9.007884579046444e-07,
+      "loss": 0.5581,
+      "step": 10907
+    },
+    {
+      "epoch": 1.9458567478369457,
+      "grad_norm": 0.5417647957801819,
+      "learning_rate": 8.948560148329676e-07,
+      "loss": 0.5139,
+      "step": 10908
+    },
+    {
+      "epoch": 1.946035144054946,
+      "grad_norm": 0.5130847692489624,
+      "learning_rate": 8.889431363721112e-07,
+      "loss": 0.4498,
+      "step": 10909
+    },
+    {
+      "epoch": 1.9462135402729461,
+      "grad_norm": 0.5233637094497681,
+      "learning_rate": 8.830498229864537e-07,
+      "loss": 0.4592,
+      "step": 10910
+    },
+    {
+      "epoch": 1.9463919364909463,
+      "grad_norm": 0.5232832431793213,
+      "learning_rate": 8.771760751388746e-07,
+      "loss": 0.44,
+      "step": 10911
+    },
+    {
+      "epoch": 1.9465703327089465,
+      "grad_norm": 0.47777479887008667,
+      "learning_rate": 8.713218932906719e-07,
+      "loss": 0.4183,
+      "step": 10912
+    },
+    {
+      "epoch": 1.9467487289269467,
+      "grad_norm": 0.5267250537872314,
+      "learning_rate": 8.654872779016443e-07,
+      "loss": 0.4499,
+      "step": 10913
+    },
+    {
+      "epoch": 1.946927125144947,
+      "grad_norm": 0.5199578404426575,
+      "learning_rate": 8.596722294300364e-07,
+      "loss": 0.4039,
+      "step": 10914
+    },
+    {
+      "epoch": 1.9471055213629471,
+      "grad_norm": 0.4668145477771759,
+      "learning_rate": 8.538767483325383e-07,
+      "loss": 0.3787,
+      "step": 10915
+    },
+    {
+      "epoch": 1.9472839175809473,
+      "grad_norm": 0.5497430562973022,
+      "learning_rate": 8.481008350643693e-07,
+      "loss": 0.5505,
+      "step": 10916
+    },
+    {
+      "epoch": 1.9474623137989475,
+      "grad_norm": 0.5314509272575378,
+      "learning_rate": 8.423444900791111e-07,
+      "loss": 0.4944,
+      "step": 10917
+    },
+    {
+      "epoch": 1.9476407100169477,
+      "grad_norm": 0.46878379583358765,
+      "learning_rate": 8.366077138289296e-07,
+      "loss": 0.3517,
+      "step": 10918
+    },
+    {
+      "epoch": 1.947819106234948,
+      "grad_norm": 0.4314865469932556,
+      "learning_rate": 8.308905067643536e-07,
+      "loss": 0.3136,
+      "step": 10919
+    },
+    {
+      "epoch": 1.9479975024529481,
+      "grad_norm": 0.5057112574577332,
+      "learning_rate": 8.251928693343846e-07,
+      "loss": 0.5543,
+      "step": 10920
+    },
+    {
+      "epoch": 1.9481758986709483,
+      "grad_norm": 0.5008751153945923,
+      "learning_rate": 8.19514801986554e-07,
+      "loss": 0.4436,
+      "step": 10921
+    },
+    {
+      "epoch": 1.9483542948889485,
+      "grad_norm": 0.49664172530174255,
+      "learning_rate": 8.138563051667824e-07,
+      "loss": 0.3823,
+      "step": 10922
+    },
+    {
+      "epoch": 1.9485326911069485,
+      "grad_norm": 0.9543033838272095,
+      "learning_rate": 8.082173793195479e-07,
+      "loss": 0.448,
+      "step": 10923
+    },
+    {
+      "epoch": 1.9487110873249487,
+      "grad_norm": 0.49950093030929565,
+      "learning_rate": 8.025980248876352e-07,
+      "loss": 0.3344,
+      "step": 10924
+    },
+    {
+      "epoch": 1.9488894835429489,
+      "grad_norm": 0.5399156808853149,
+      "learning_rate": 7.969982423124689e-07,
+      "loss": 0.5778,
+      "step": 10925
+    },
+    {
+      "epoch": 1.949067879760949,
+      "grad_norm": 0.5570793151855469,
+      "learning_rate": 7.914180320338082e-07,
+      "loss": 0.511,
+      "step": 10926
+    },
+    {
+      "epoch": 1.9492462759789493,
+      "grad_norm": 0.5615115761756897,
+      "learning_rate": 7.858573944899139e-07,
+      "loss": 0.5364,
+      "step": 10927
+    },
+    {
+      "epoch": 1.9494246721969493,
+      "grad_norm": 0.5255836844444275,
+      "learning_rate": 7.803163301175753e-07,
+      "loss": 0.5906,
+      "step": 10928
+    },
+    {
+      "epoch": 1.9496030684149495,
+      "grad_norm": 0.5021457672119141,
+      "learning_rate": 7.747948393519167e-07,
+      "loss": 0.4357,
+      "step": 10929
+    },
+    {
+      "epoch": 1.9497814646329497,
+      "grad_norm": 0.5015770196914673,
+      "learning_rate": 7.692929226266188e-07,
+      "loss": 0.4276,
+      "step": 10930
+    },
+    {
+      "epoch": 1.9499598608509499,
+      "grad_norm": 0.5243127346038818,
+      "learning_rate": 7.638105803738083e-07,
+      "loss": 0.5425,
+      "step": 10931
+    },
+    {
+      "epoch": 1.95013825706895,
+      "grad_norm": 0.47330746054649353,
+      "learning_rate": 7.583478130240851e-07,
+      "loss": 0.3797,
+      "step": 10932
+    },
+    {
+      "epoch": 1.9503166532869503,
+      "grad_norm": 0.49246537685394287,
+      "learning_rate": 7.529046210064394e-07,
+      "loss": 0.435,
+      "step": 10933
+    },
+    {
+      "epoch": 1.9504950495049505,
+      "grad_norm": 0.5126248598098755,
+      "learning_rate": 7.474810047484182e-07,
+      "loss": 0.4749,
+      "step": 10934
+    },
+    {
+      "epoch": 1.9506734457229506,
+      "grad_norm": 0.5344600081443787,
+      "learning_rate": 7.420769646759584e-07,
+      "loss": 0.4556,
+      "step": 10935
+    },
+    {
+      "epoch": 1.9508518419409508,
+      "grad_norm": 0.5256651043891907,
+      "learning_rate": 7.366925012135262e-07,
+      "loss": 0.5272,
+      "step": 10936
+    },
+    {
+      "epoch": 1.951030238158951,
+      "grad_norm": 0.49008166790008545,
+      "learning_rate": 7.313276147840053e-07,
+      "loss": 0.4495,
+      "step": 10937
+    },
+    {
+      "epoch": 1.9512086343769512,
+      "grad_norm": 0.4859989285469055,
+      "learning_rate": 7.259823058087256e-07,
+      "loss": 0.3527,
+      "step": 10938
+    },
+    {
+      "epoch": 1.9513870305949514,
+      "grad_norm": 0.4771299958229065,
+      "learning_rate": 7.206565747075178e-07,
+      "loss": 0.4565,
+      "step": 10939
+    },
+    {
+      "epoch": 1.9515654268129516,
+      "grad_norm": 0.49878621101379395,
+      "learning_rate": 7.153504218986862e-07,
+      "loss": 0.39,
+      "step": 10940
+    },
+    {
+      "epoch": 1.9517438230309518,
+      "grad_norm": 0.6032863855361938,
+      "learning_rate": 7.10063847798953e-07,
+      "loss": 0.4377,
+      "step": 10941
+    },
+    {
+      "epoch": 1.951922219248952,
+      "grad_norm": 0.5695745348930359,
+      "learning_rate": 7.047968528235416e-07,
+      "loss": 0.5126,
+      "step": 10942
+    },
+    {
+      "epoch": 1.9521006154669522,
+      "grad_norm": 0.567709743976593,
+      "learning_rate": 6.995494373860656e-07,
+      "loss": 0.5469,
+      "step": 10943
+    },
+    {
+      "epoch": 1.9522790116849524,
+      "grad_norm": 0.516247034072876,
+      "learning_rate": 6.943216018987508e-07,
+      "loss": 0.437,
+      "step": 10944
+    },
+    {
+      "epoch": 1.9524574079029524,
+      "grad_norm": 0.5342845916748047,
+      "learning_rate": 6.891133467721022e-07,
+      "loss": 0.4906,
+      "step": 10945
+    },
+    {
+      "epoch": 1.9526358041209526,
+      "grad_norm": 0.5445780158042908,
+      "learning_rate": 6.839246724151815e-07,
+      "loss": 0.4014,
+      "step": 10946
+    },
+    {
+      "epoch": 1.9528142003389528,
+      "grad_norm": 0.5402600169181824,
+      "learning_rate": 6.787555792355793e-07,
+      "loss": 0.5584,
+      "step": 10947
+    },
+    {
+      "epoch": 1.952992596556953,
+      "grad_norm": 0.5336621403694153,
+      "learning_rate": 6.736060676391653e-07,
+      "loss": 0.435,
+      "step": 10948
+    },
+    {
+      "epoch": 1.9531709927749532,
+      "grad_norm": 0.6429981589317322,
+      "learning_rate": 6.684761380304772e-07,
+      "loss": 0.4705,
+      "step": 10949
+    },
+    {
+      "epoch": 1.9533493889929532,
+      "grad_norm": 0.4981113374233246,
+      "learning_rate": 6.633657908123592e-07,
+      "loss": 0.484,
+      "step": 10950
+    },
+    {
+      "epoch": 1.9535277852109534,
+      "grad_norm": 0.5144602656364441,
+      "learning_rate": 6.582750263862125e-07,
+      "loss": 0.5402,
+      "step": 10951
+    },
+    {
+      "epoch": 1.9537061814289536,
+      "grad_norm": 0.5602140426635742,
+      "learning_rate": 6.532038451518286e-07,
+      "loss": 0.498,
+      "step": 10952
+    },
+    {
+      "epoch": 1.9538845776469538,
+      "grad_norm": 0.5775460600852966,
+      "learning_rate": 6.481522475075274e-07,
+      "loss": 0.5634,
+      "step": 10953
+    },
+    {
+      "epoch": 1.954062973864954,
+      "grad_norm": 0.4967274069786072,
+      "learning_rate": 6.431202338500475e-07,
+      "loss": 0.3654,
+      "step": 10954
+    },
+    {
+      "epoch": 1.9542413700829542,
+      "grad_norm": 0.5827997326850891,
+      "learning_rate": 6.381078045745725e-07,
+      "loss": 0.6665,
+      "step": 10955
+    },
+    {
+      "epoch": 1.9544197663009544,
+      "grad_norm": 0.5111951231956482,
+      "learning_rate": 6.331149600748154e-07,
+      "loss": 0.4386,
+      "step": 10956
+    },
+    {
+      "epoch": 1.9545981625189546,
+      "grad_norm": 0.4768127202987671,
+      "learning_rate": 6.281417007429069e-07,
+      "loss": 0.4033,
+      "step": 10957
+    },
+    {
+      "epoch": 1.9547765587369548,
+      "grad_norm": 0.5813576579093933,
+      "learning_rate": 6.231880269694235e-07,
+      "loss": 0.7205,
+      "step": 10958
+    },
+    {
+      "epoch": 1.954954954954955,
+      "grad_norm": 0.4979632496833801,
+      "learning_rate": 6.182539391434428e-07,
+      "loss": 0.3926,
+      "step": 10959
+    },
+    {
+      "epoch": 1.9551333511729552,
+      "grad_norm": 0.4942147433757782,
+      "learning_rate": 6.133394376524604e-07,
+      "loss": 0.4401,
+      "step": 10960
+    },
+    {
+      "epoch": 1.9553117473909554,
+      "grad_norm": 0.5690099000930786,
+      "learning_rate": 6.084445228825009e-07,
+      "loss": 0.5989,
+      "step": 10961
+    },
+    {
+      "epoch": 1.9554901436089556,
+      "grad_norm": 0.5280899405479431,
+      "learning_rate": 6.035691952179789e-07,
+      "loss": 0.4661,
+      "step": 10962
+    },
+    {
+      "epoch": 1.9556685398269558,
+      "grad_norm": 0.49601706862449646,
+      "learning_rate": 5.987134550418105e-07,
+      "loss": 0.4591,
+      "step": 10963
+    },
+    {
+      "epoch": 1.955846936044956,
+      "grad_norm": 0.5549279451370239,
+      "learning_rate": 5.938773027353572e-07,
+      "loss": 0.4277,
+      "step": 10964
+    },
+    {
+      "epoch": 1.9560253322629562,
+      "grad_norm": 0.6416307687759399,
+      "learning_rate": 5.890607386784818e-07,
+      "loss": 0.6958,
+      "step": 10965
+    },
+    {
+      "epoch": 1.9562037284809564,
+      "grad_norm": 0.5070635080337524,
+      "learning_rate": 5.842637632494097e-07,
+      "loss": 0.4569,
+      "step": 10966
+    },
+    {
+      "epoch": 1.9563821246989563,
+      "grad_norm": 0.5254069566726685,
+      "learning_rate": 5.794863768249503e-07,
+      "loss": 0.5185,
+      "step": 10967
+    },
+    {
+      "epoch": 1.9565605209169565,
+      "grad_norm": 0.48455336689949036,
+      "learning_rate": 5.747285797802759e-07,
+      "loss": 0.403,
+      "step": 10968
+    },
+    {
+      "epoch": 1.9567389171349567,
+      "grad_norm": 0.574587345123291,
+      "learning_rate": 5.699903724891153e-07,
+      "loss": 0.7152,
+      "step": 10969
+    },
+    {
+      "epoch": 1.956917313352957,
+      "grad_norm": 0.5534584522247314,
+      "learning_rate": 5.652717553235597e-07,
+      "loss": 0.5475,
+      "step": 10970
+    },
+    {
+      "epoch": 1.9570957095709571,
+      "grad_norm": 0.5564422011375427,
+      "learning_rate": 5.605727286542017e-07,
+      "loss": 0.544,
+      "step": 10971
+    },
+    {
+      "epoch": 1.957274105788957,
+      "grad_norm": 0.5110240578651428,
+      "learning_rate": 5.558932928501347e-07,
+      "loss": 0.4253,
+      "step": 10972
+    },
+    {
+      "epoch": 1.9574525020069573,
+      "grad_norm": 0.5717257857322693,
+      "learning_rate": 5.512334482788428e-07,
+      "loss": 0.685,
+      "step": 10973
+    },
+    {
+      "epoch": 1.9576308982249575,
+      "grad_norm": 0.5493577718734741,
+      "learning_rate": 5.465931953063663e-07,
+      "loss": 0.459,
+      "step": 10974
+    },
+    {
+      "epoch": 1.9578092944429577,
+      "grad_norm": 0.5325579643249512,
+      "learning_rate": 5.419725342970805e-07,
+      "loss": 0.4834,
+      "step": 10975
+    },
+    {
+      "epoch": 1.957987690660958,
+      "grad_norm": 0.5637785792350769,
+      "learning_rate": 5.37371465613945e-07,
+      "loss": 0.5612,
+      "step": 10976
+    },
+    {
+      "epoch": 1.958166086878958,
+      "grad_norm": 0.5901339054107666,
+      "learning_rate": 5.32789989618282e-07,
+      "loss": 0.6958,
+      "step": 10977
+    },
+    {
+      "epoch": 1.9583444830969583,
+      "grad_norm": 0.49308058619499207,
+      "learning_rate": 5.282281066699701e-07,
+      "loss": 0.4053,
+      "step": 10978
+    },
+    {
+      "epoch": 1.9585228793149585,
+      "grad_norm": 0.5078552961349487,
+      "learning_rate": 5.236858171272229e-07,
+      "loss": 0.433,
+      "step": 10979
+    },
+    {
+      "epoch": 1.9587012755329587,
+      "grad_norm": 0.5317704677581787,
+      "learning_rate": 5.191631213468661e-07,
+      "loss": 0.4935,
+      "step": 10980
+    },
+    {
+      "epoch": 1.958879671750959,
+      "grad_norm": 0.606121301651001,
+      "learning_rate": 5.1466001968406e-07,
+      "loss": 0.6044,
+      "step": 10981
+    },
+    {
+      "epoch": 1.959058067968959,
+      "grad_norm": 0.49278080463409424,
+      "learning_rate": 5.101765124925217e-07,
+      "loss": 0.3807,
+      "step": 10982
+    },
+    {
+      "epoch": 1.9592364641869593,
+      "grad_norm": 0.5795280933380127,
+      "learning_rate": 5.057126001243306e-07,
+      "loss": 0.5853,
+      "step": 10983
+    },
+    {
+      "epoch": 1.9594148604049595,
+      "grad_norm": 0.5039269924163818,
+      "learning_rate": 5.01268282930123e-07,
+      "loss": 0.4857,
+      "step": 10984
+    },
+    {
+      "epoch": 1.9595932566229597,
+      "grad_norm": 0.5751108527183533,
+      "learning_rate": 4.968435612588973e-07,
+      "loss": 0.5788,
+      "step": 10985
+    },
+    {
+      "epoch": 1.9597716528409599,
+      "grad_norm": 0.5224045515060425,
+      "learning_rate": 4.924384354582645e-07,
+      "loss": 0.505,
+      "step": 10986
+    },
+    {
+      "epoch": 1.95995004905896,
+      "grad_norm": 0.6626994609832764,
+      "learning_rate": 4.880529058741146e-07,
+      "loss": 0.586,
+      "step": 10987
+    },
+    {
+      "epoch": 1.9601284452769603,
+      "grad_norm": 0.571682333946228,
+      "learning_rate": 4.836869728508941e-07,
+      "loss": 0.4711,
+      "step": 10988
+    },
+    {
+      "epoch": 1.9603068414949603,
+      "grad_norm": 0.5564180016517639,
+      "learning_rate": 4.793406367315512e-07,
+      "loss": 0.5446,
+      "step": 10989
+    },
+    {
+      "epoch": 1.9604852377129605,
+      "grad_norm": 0.5148413777351379,
+      "learning_rate": 4.750138978574237e-07,
+      "loss": 0.5736,
+      "step": 10990
+    },
+    {
+      "epoch": 1.9606636339309607,
+      "grad_norm": 0.5848867297172546,
+      "learning_rate": 4.7070675656832316e-07,
+      "loss": 0.5588,
+      "step": 10991
+    },
+    {
+      "epoch": 1.9608420301489609,
+      "grad_norm": 0.5551292896270752,
+      "learning_rate": 4.6641921320253465e-07,
+      "loss": 0.5043,
+      "step": 10992
+    },
+    {
+      "epoch": 1.961020426366961,
+      "grad_norm": 0.543438732624054,
+      "learning_rate": 4.6215126809678875e-07,
+      "loss": 0.4619,
+      "step": 10993
+    },
+    {
+      "epoch": 1.961198822584961,
+      "grad_norm": 0.4963773190975189,
+      "learning_rate": 4.579029215862895e-07,
+      "loss": 0.4392,
+      "step": 10994
+    },
+    {
+      "epoch": 1.9613772188029612,
+      "grad_norm": 0.5400619506835938,
+      "learning_rate": 4.5367417400471454e-07,
+      "loss": 0.6237,
+      "step": 10995
+    },
+    {
+      "epoch": 1.9615556150209614,
+      "grad_norm": 0.5189566612243652,
+      "learning_rate": 4.49465025684187e-07,
+      "loss": 0.416,
+      "step": 10996
+    },
+    {
+      "epoch": 1.9617340112389616,
+      "grad_norm": 0.5119401216506958,
+      "learning_rate": 4.452754769553036e-07,
+      "loss": 0.4979,
+      "step": 10997
+    },
+    {
+      "epoch": 1.9619124074569618,
+      "grad_norm": 0.4972620904445648,
+      "learning_rate": 4.4110552814707884e-07,
+      "loss": 0.4132,
+      "step": 10998
+    },
+    {
+      "epoch": 1.962090803674962,
+      "grad_norm": 0.6130009889602661,
+      "learning_rate": 4.3695517958702856e-07,
+      "loss": 0.6817,
+      "step": 10999
+    },
+    {
+      "epoch": 1.9622691998929622,
+      "grad_norm": 0.5110456347465515,
+      "learning_rate": 4.328244316011143e-07,
+      "loss": 0.3981,
+      "step": 11000
+    },
+    {
+      "epoch": 1.9624475961109624,
+      "grad_norm": 0.5126025676727295,
+      "learning_rate": 4.287132845137709e-07,
+      "loss": 0.4441,
+      "step": 11001
+    },
+    {
+      "epoch": 1.9626259923289626,
+      "grad_norm": 0.5288227796554565,
+      "learning_rate": 4.246217386479068e-07,
+      "loss": 0.4437,
+      "step": 11002
+    },
+    {
+      "epoch": 1.9628043885469628,
+      "grad_norm": 0.5382393002510071,
+      "learning_rate": 4.2054979432482044e-07,
+      "loss": 0.571,
+      "step": 11003
+    },
+    {
+      "epoch": 1.962982784764963,
+      "grad_norm": 0.4760775566101074,
+      "learning_rate": 4.164974518643672e-07,
+      "loss": 0.4336,
+      "step": 11004
+    },
+    {
+      "epoch": 1.9631611809829632,
+      "grad_norm": 0.5154813528060913,
+      "learning_rate": 4.1246471158482015e-07,
+      "loss": 0.4456,
+      "step": 11005
+    },
+    {
+      "epoch": 1.9633395772009634,
+      "grad_norm": 0.5530751943588257,
+      "learning_rate": 4.0845157380287047e-07,
+      "loss": 0.5216,
+      "step": 11006
+    },
+    {
+      "epoch": 1.9635179734189636,
+      "grad_norm": 0.5821408033370972,
+      "learning_rate": 4.044580388337105e-07,
+      "loss": 0.6407,
+      "step": 11007
+    },
+    {
+      "epoch": 1.9636963696369638,
+      "grad_norm": 0.5298261642456055,
+      "learning_rate": 4.0048410699103365e-07,
+      "loss": 0.5755,
+      "step": 11008
+    },
+    {
+      "epoch": 1.963874765854964,
+      "grad_norm": 0.5816518068313599,
+      "learning_rate": 3.9652977858692375e-07,
+      "loss": 0.6441,
+      "step": 11009
+    },
+    {
+      "epoch": 1.9640531620729642,
+      "grad_norm": 0.5004158020019531,
+      "learning_rate": 3.9259505393193785e-07,
+      "loss": 0.4389,
+      "step": 11010
+    },
+    {
+      "epoch": 1.9642315582909642,
+      "grad_norm": 0.4865829348564148,
+      "learning_rate": 3.886799333351343e-07,
+      "loss": 0.4019,
+      "step": 11011
+    },
+    {
+      "epoch": 1.9644099545089644,
+      "grad_norm": 0.48600196838378906,
+      "learning_rate": 3.847844171039616e-07,
+      "loss": 0.3581,
+      "step": 11012
+    },
+    {
+      "epoch": 1.9645883507269646,
+      "grad_norm": 0.4642302393913269,
+      "learning_rate": 3.809085055444528e-07,
+      "loss": 0.3521,
+      "step": 11013
+    },
+    {
+      "epoch": 1.9647667469449648,
+      "grad_norm": 0.45133063197135925,
+      "learning_rate": 3.770521989609199e-07,
+      "loss": 0.3936,
+      "step": 11014
+    },
+    {
+      "epoch": 1.964945143162965,
+      "grad_norm": 0.4991499185562134,
+      "learning_rate": 3.7321549765631514e-07,
+      "loss": 0.3983,
+      "step": 11015
+    },
+    {
+      "epoch": 1.9651235393809652,
+      "grad_norm": 0.526542067527771,
+      "learning_rate": 3.6939840193195296e-07,
+      "loss": 0.4563,
+      "step": 11016
+    },
+    {
+      "epoch": 1.9653019355989652,
+      "grad_norm": 0.5464054346084595,
+      "learning_rate": 3.656009120875936e-07,
+      "loss": 0.4337,
+      "step": 11017
+    },
+    {
+      "epoch": 1.9654803318169654,
+      "grad_norm": 0.43672722578048706,
+      "learning_rate": 3.618230284215263e-07,
+      "loss": 0.2778,
+      "step": 11018
+    },
+    {
+      "epoch": 1.9656587280349656,
+      "grad_norm": 0.6000391244888306,
+      "learning_rate": 3.580647512304303e-07,
+      "loss": 0.5887,
+      "step": 11019
+    },
+    {
+      "epoch": 1.9658371242529658,
+      "grad_norm": 0.5648049712181091,
+      "learning_rate": 3.543260808095139e-07,
+      "loss": 0.623,
+      "step": 11020
+    },
+    {
+      "epoch": 1.966015520470966,
+      "grad_norm": 0.4601522386074066,
+      "learning_rate": 3.5060701745240344e-07,
+      "loss": 0.3363,
+      "step": 11021
+    },
+    {
+      "epoch": 1.9661939166889661,
+      "grad_norm": 0.48688170313835144,
+      "learning_rate": 3.4690756145117074e-07,
+      "loss": 0.431,
+      "step": 11022
+    },
+    {
+      "epoch": 1.9663723129069663,
+      "grad_norm": 0.5293704271316528,
+      "learning_rate": 3.43227713096389e-07,
+      "loss": 0.5696,
+      "step": 11023
+    },
+    {
+      "epoch": 1.9665507091249665,
+      "grad_norm": 0.511298418045044,
+      "learning_rate": 3.39567472677077e-07,
+      "loss": 0.4419,
+      "step": 11024
+    },
+    {
+      "epoch": 1.9667291053429667,
+      "grad_norm": 0.580195963382721,
+      "learning_rate": 3.3592684048067144e-07,
+      "loss": 0.6206,
+      "step": 11025
+    },
+    {
+      "epoch": 1.966907501560967,
+      "grad_norm": 0.5301962494850159,
+      "learning_rate": 3.3230581679316584e-07,
+      "loss": 0.4893,
+      "step": 11026
+    },
+    {
+      "epoch": 1.9670858977789671,
+      "grad_norm": 0.48305174708366394,
+      "learning_rate": 3.287044018988883e-07,
+      "loss": 0.4104,
+      "step": 11027
+    },
+    {
+      "epoch": 1.9672642939969673,
+      "grad_norm": 0.5521623492240906,
+      "learning_rate": 3.2512259608075134e-07,
+      "loss": 0.5381,
+      "step": 11028
+    },
+    {
+      "epoch": 1.9674426902149675,
+      "grad_norm": 0.5551265478134155,
+      "learning_rate": 3.2156039962003e-07,
+      "loss": 0.5689,
+      "step": 11029
+    },
+    {
+      "epoch": 1.9676210864329677,
+      "grad_norm": 0.5386922359466553,
+      "learning_rate": 3.180178127965283e-07,
+      "loss": 0.5749,
+      "step": 11030
+    },
+    {
+      "epoch": 1.967799482650968,
+      "grad_norm": 0.5408132672309875,
+      "learning_rate": 3.144948358884403e-07,
+      "loss": 0.4842,
+      "step": 11031
+    },
+    {
+      "epoch": 1.9679778788689681,
+      "grad_norm": 0.5893284678459167,
+      "learning_rate": 3.109914691724891e-07,
+      "loss": 0.6408,
+      "step": 11032
+    },
+    {
+      "epoch": 1.968156275086968,
+      "grad_norm": 0.5585829019546509,
+      "learning_rate": 3.0750771292381573e-07,
+      "loss": 0.5564,
+      "step": 11033
+    },
+    {
+      "epoch": 1.9683346713049683,
+      "grad_norm": 0.544394850730896,
+      "learning_rate": 3.0404356741603467e-07,
+      "loss": 0.4695,
+      "step": 11034
+    },
+    {
+      "epoch": 1.9685130675229685,
+      "grad_norm": 0.4653702676296234,
+      "learning_rate": 3.0059903292120605e-07,
+      "loss": 0.3943,
+      "step": 11035
+    },
+    {
+      "epoch": 1.9686914637409687,
+      "grad_norm": 0.5404059886932373,
+      "learning_rate": 2.971741097098912e-07,
+      "loss": 0.5084,
+      "step": 11036
+    },
+    {
+      "epoch": 1.968869859958969,
+      "grad_norm": 0.5264713168144226,
+      "learning_rate": 2.9376879805106947e-07,
+      "loss": 0.4465,
+      "step": 11037
+    },
+    {
+      "epoch": 1.969048256176969,
+      "grad_norm": 0.48295858502388,
+      "learning_rate": 2.903830982121658e-07,
+      "loss": 0.3769,
+      "step": 11038
+    },
+    {
+      "epoch": 1.969226652394969,
+      "grad_norm": 0.5343292355537415,
+      "learning_rate": 2.870170104591341e-07,
+      "loss": 0.533,
+      "step": 11039
+    },
+    {
+      "epoch": 1.9694050486129693,
+      "grad_norm": 0.5259873270988464,
+      "learning_rate": 2.8367053505631846e-07,
+      "loss": 0.4229,
+      "step": 11040
+    },
+    {
+      "epoch": 1.9695834448309695,
+      "grad_norm": 0.46989312767982483,
+      "learning_rate": 2.803436722665642e-07,
+      "loss": 0.4079,
+      "step": 11041
+    },
+    {
+      "epoch": 1.9697618410489697,
+      "grad_norm": 0.50926274061203,
+      "learning_rate": 2.770364223511623e-07,
+      "loss": 0.494,
+      "step": 11042
+    },
+    {
+      "epoch": 1.9699402372669699,
+      "grad_norm": 0.756600022315979,
+      "learning_rate": 2.737487855698495e-07,
+      "loss": 0.5463,
+      "step": 11043
+    },
+    {
+      "epoch": 1.97011863348497,
+      "grad_norm": 0.47964367270469666,
+      "learning_rate": 2.7048076218083585e-07,
+      "loss": 0.3886,
+      "step": 11044
+    },
+    {
+      "epoch": 1.9702970297029703,
+      "grad_norm": 0.5303892493247986,
+      "learning_rate": 2.672323524408049e-07,
+      "loss": 0.5722,
+      "step": 11045
+    },
+    {
+      "epoch": 1.9704754259209705,
+      "grad_norm": 0.4234614670276642,
+      "learning_rate": 2.6400355660488594e-07,
+      "loss": 0.2525,
+      "step": 11046
+    },
+    {
+      "epoch": 1.9706538221389707,
+      "grad_norm": 0.5104755163192749,
+      "learning_rate": 2.607943749266262e-07,
+      "loss": 0.4572,
+      "step": 11047
+    },
+    {
+      "epoch": 1.9708322183569709,
+      "grad_norm": 0.5379624366760254,
+      "learning_rate": 2.5760480765812946e-07,
+      "loss": 0.5121,
+      "step": 11048
+    },
+    {
+      "epoch": 1.971010614574971,
+      "grad_norm": 0.5435539484024048,
+      "learning_rate": 2.544348550498621e-07,
+      "loss": 0.5189,
+      "step": 11049
+    },
+    {
+      "epoch": 1.9711890107929713,
+      "grad_norm": 0.6087327599525452,
+      "learning_rate": 2.512845173508194e-07,
+      "loss": 0.5903,
+      "step": 11050
+    },
+    {
+      "epoch": 1.9713674070109715,
+      "grad_norm": 0.5233466029167175,
+      "learning_rate": 2.481537948084145e-07,
+      "loss": 0.5574,
+      "step": 11051
+    },
+    {
+      "epoch": 1.9715458032289717,
+      "grad_norm": 0.4774353802204132,
+      "learning_rate": 2.4504268766853413e-07,
+      "loss": 0.4658,
+      "step": 11052
+    },
+    {
+      "epoch": 1.9717241994469719,
+      "grad_norm": 0.5861150622367859,
+      "learning_rate": 2.4195119617551054e-07,
+      "loss": 0.4947,
+      "step": 11053
+    },
+    {
+      "epoch": 1.971902595664972,
+      "grad_norm": 0.4522158205509186,
+      "learning_rate": 2.388793205721773e-07,
+      "loss": 0.2857,
+      "step": 11054
+    },
+    {
+      "epoch": 1.972080991882972,
+      "grad_norm": 0.4695200026035309,
+      "learning_rate": 2.358270610997859e-07,
+      "loss": 0.4349,
+      "step": 11055
+    },
+    {
+      "epoch": 1.9722593881009722,
+      "grad_norm": 0.5388724207878113,
+      "learning_rate": 2.3279441799803347e-07,
+      "loss": 0.5289,
+      "step": 11056
+    },
+    {
+      "epoch": 1.9724377843189724,
+      "grad_norm": 0.5425492525100708,
+      "learning_rate": 2.297813915051461e-07,
+      "loss": 0.4849,
+      "step": 11057
+    },
+    {
+      "epoch": 1.9726161805369726,
+      "grad_norm": 0.5144857168197632,
+      "learning_rate": 2.2678798185771233e-07,
+      "loss": 0.4728,
+      "step": 11058
+    },
+    {
+      "epoch": 1.9727945767549728,
+      "grad_norm": 0.5156108736991882,
+      "learning_rate": 2.2381418929090515e-07,
+      "loss": 0.4917,
+      "step": 11059
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 0.4618896543979645,
+      "learning_rate": 2.208600140382322e-07,
+      "loss": 0.3896,
+      "step": 11060
+    },
+    {
+      "epoch": 1.973151369190973,
+      "grad_norm": 0.484020471572876,
+      "learning_rate": 2.1792545633170236e-07,
+      "loss": 0.4211,
+      "step": 11061
+    },
+    {
+      "epoch": 1.9733297654089732,
+      "grad_norm": 0.5049816370010376,
+      "learning_rate": 2.1501051640182566e-07,
+      "loss": 0.4561,
+      "step": 11062
+    },
+    {
+      "epoch": 1.9735081616269734,
+      "grad_norm": 0.5276315808296204,
+      "learning_rate": 2.121151944775579e-07,
+      "loss": 0.4653,
+      "step": 11063
+    },
+    {
+      "epoch": 1.9736865578449736,
+      "grad_norm": 0.45448240637779236,
+      "learning_rate": 2.0923949078624493e-07,
+      "loss": 0.3645,
+      "step": 11064
+    },
+    {
+      "epoch": 1.9738649540629738,
+      "grad_norm": 0.5729450583457947,
+      "learning_rate": 2.0638340555376168e-07,
+      "loss": 0.5265,
+      "step": 11065
+    },
+    {
+      "epoch": 1.974043350280974,
+      "grad_norm": 0.47857052087783813,
+      "learning_rate": 2.0354693900445643e-07,
+      "loss": 0.3919,
+      "step": 11066
+    },
+    {
+      "epoch": 1.9742217464989742,
+      "grad_norm": 0.5773352980613708,
+      "learning_rate": 2.0073009136106768e-07,
+      "loss": 0.4904,
+      "step": 11067
+    },
+    {
+      "epoch": 1.9744001427169744,
+      "grad_norm": 0.47996997833251953,
+      "learning_rate": 1.9793286284483514e-07,
+      "loss": 0.4176,
+      "step": 11068
+    },
+    {
+      "epoch": 1.9745785389349746,
+      "grad_norm": 0.45277294516563416,
+      "learning_rate": 1.9515525367547192e-07,
+      "loss": 0.3085,
+      "step": 11069
+    },
+    {
+      "epoch": 1.9747569351529748,
+      "grad_norm": 0.49644288420677185,
+      "learning_rate": 1.9239726407110913e-07,
+      "loss": 0.4873,
+      "step": 11070
+    },
+    {
+      "epoch": 1.974935331370975,
+      "grad_norm": 0.5908966064453125,
+      "learning_rate": 1.8965889424835126e-07,
+      "loss": 0.6073,
+      "step": 11071
+    },
+    {
+      "epoch": 1.9751137275889752,
+      "grad_norm": 0.487192839384079,
+      "learning_rate": 1.86940144422304e-07,
+      "loss": 0.4247,
+      "step": 11072
+    },
+    {
+      "epoch": 1.9752921238069754,
+      "grad_norm": 0.5491554141044617,
+      "learning_rate": 1.8424101480646326e-07,
+      "loss": 0.4814,
+      "step": 11073
+    },
+    {
+      "epoch": 1.9754705200249756,
+      "grad_norm": 0.5391243696212769,
+      "learning_rate": 1.8156150561282614e-07,
+      "loss": 0.4927,
+      "step": 11074
+    },
+    {
+      "epoch": 1.9756489162429758,
+      "grad_norm": 0.5700384974479675,
+      "learning_rate": 1.7890161705183538e-07,
+      "loss": 0.6383,
+      "step": 11075
+    },
+    {
+      "epoch": 1.975827312460976,
+      "grad_norm": 0.5709267258644104,
+      "learning_rate": 1.7626134933243498e-07,
+      "loss": 0.5058,
+      "step": 11076
+    },
+    {
+      "epoch": 1.976005708678976,
+      "grad_norm": 0.5328453779220581,
+      "learning_rate": 1.7364070266193133e-07,
+      "loss": 0.4764,
+      "step": 11077
+    },
+    {
+      "epoch": 1.9761841048969762,
+      "grad_norm": 0.5492943525314331,
+      "learning_rate": 1.7103967724618753e-07,
+      "loss": 0.5105,
+      "step": 11078
+    },
+    {
+      "epoch": 1.9763625011149764,
+      "grad_norm": 0.6113517880439758,
+      "learning_rate": 1.684582732894846e-07,
+      "loss": 0.53,
+      "step": 11079
+    },
+    {
+      "epoch": 1.9765408973329766,
+      "grad_norm": 0.6604119539260864,
+      "learning_rate": 1.658964909945493e-07,
+      "loss": 0.5217,
+      "step": 11080
+    },
+    {
+      "epoch": 1.9767192935509768,
+      "grad_norm": 0.5008206367492676,
+      "learning_rate": 1.6335433056258176e-07,
+      "loss": 0.4225,
+      "step": 11081
+    },
+    {
+      "epoch": 1.976897689768977,
+      "grad_norm": 0.5335331559181213,
+      "learning_rate": 1.608317921932556e-07,
+      "loss": 0.4722,
+      "step": 11082
+    },
+    {
+      "epoch": 1.977076085986977,
+      "grad_norm": 0.5418149828910828,
+      "learning_rate": 1.5832887608471792e-07,
+      "loss": 0.4254,
+      "step": 11083
+    },
+    {
+      "epoch": 1.9772544822049771,
+      "grad_norm": 0.5159653425216675,
+      "learning_rate": 1.5584558243347813e-07,
+      "loss": 0.4687,
+      "step": 11084
+    },
+    {
+      "epoch": 1.9774328784229773,
+      "grad_norm": 0.5415779948234558,
+      "learning_rate": 1.5338191143463022e-07,
+      "loss": 0.4405,
+      "step": 11085
+    },
+    {
+      "epoch": 1.9776112746409775,
+      "grad_norm": 0.6081345677375793,
+      "learning_rate": 1.5093786328163052e-07,
+      "loss": 0.4643,
+      "step": 11086
+    },
+    {
+      "epoch": 1.9777896708589777,
+      "grad_norm": 0.5560696721076965,
+      "learning_rate": 1.4851343816646434e-07,
+      "loss": 0.6136,
+      "step": 11087
+    },
+    {
+      "epoch": 1.977968067076978,
+      "grad_norm": 0.521943986415863,
+      "learning_rate": 1.4610863627953496e-07,
+      "loss": 0.5582,
+      "step": 11088
+    },
+    {
+      "epoch": 1.9781464632949781,
+      "grad_norm": 0.5156327486038208,
+      "learning_rate": 1.4372345780971907e-07,
+      "loss": 0.4443,
+      "step": 11089
+    },
+    {
+      "epoch": 1.9783248595129783,
+      "grad_norm": 0.5138960480690002,
+      "learning_rate": 1.4135790294433904e-07,
+      "loss": 0.4079,
+      "step": 11090
+    },
+    {
+      "epoch": 1.9785032557309785,
+      "grad_norm": 0.503109335899353,
+      "learning_rate": 1.3901197186919067e-07,
+      "loss": 0.4661,
+      "step": 11091
+    },
+    {
+      "epoch": 1.9786816519489787,
+      "grad_norm": 0.601678192615509,
+      "learning_rate": 1.3668566476848775e-07,
+      "loss": 0.6345,
+      "step": 11092
+    },
+    {
+      "epoch": 1.978860048166979,
+      "grad_norm": 0.5237042903900146,
+      "learning_rate": 1.3437898182500075e-07,
+      "loss": 0.504,
+      "step": 11093
+    },
+    {
+      "epoch": 1.9790384443849791,
+      "grad_norm": 0.4751412868499756,
+      "learning_rate": 1.3209192321986252e-07,
+      "loss": 0.3427,
+      "step": 11094
+    },
+    {
+      "epoch": 1.9792168406029793,
+      "grad_norm": 0.4530685245990753,
+      "learning_rate": 1.298244891326794e-07,
+      "loss": 0.3887,
+      "step": 11095
+    },
+    {
+      "epoch": 1.9793952368209795,
+      "grad_norm": 0.5451707243919373,
+      "learning_rate": 1.2757667974155895e-07,
+      "loss": 0.627,
+      "step": 11096
+    },
+    {
+      "epoch": 1.9795736330389797,
+      "grad_norm": 0.5313323140144348,
+      "learning_rate": 1.253484952230266e-07,
+      "loss": 0.4812,
+      "step": 11097
+    },
+    {
+      "epoch": 1.97975202925698,
+      "grad_norm": 0.5562189221382141,
+      "learning_rate": 1.2313993575210901e-07,
+      "loss": 0.5047,
+      "step": 11098
+    },
+    {
+      "epoch": 1.9799304254749799,
+      "grad_norm": 0.622951865196228,
+      "learning_rate": 1.209510015022508e-07,
+      "loss": 0.8287,
+      "step": 11099
+    },
+    {
+      "epoch": 1.98010882169298,
+      "grad_norm": 0.5174034833908081,
+      "learning_rate": 1.1878169264536997e-07,
+      "loss": 0.4743,
+      "step": 11100
+    },
+    {
+      "epoch": 1.9802872179109803,
+      "grad_norm": 0.5302353501319885,
+      "learning_rate": 1.1663200935183028e-07,
+      "loss": 0.5506,
+      "step": 11101
+    },
+    {
+      "epoch": 1.9804656141289805,
+      "grad_norm": 0.6083629727363586,
+      "learning_rate": 1.145019517904966e-07,
+      "loss": 0.6897,
+      "step": 11102
+    },
+    {
+      "epoch": 1.9806440103469807,
+      "grad_norm": 0.591913640499115,
+      "learning_rate": 1.1239152012865183e-07,
+      "loss": 0.6142,
+      "step": 11103
+    },
+    {
+      "epoch": 1.9808224065649809,
+      "grad_norm": 0.49568477272987366,
+      "learning_rate": 1.103007145320245e-07,
+      "loss": 0.4795,
+      "step": 11104
+    },
+    {
+      "epoch": 1.9810008027829809,
+      "grad_norm": 0.6029313802719116,
+      "learning_rate": 1.0822953516484436e-07,
+      "loss": 0.627,
+      "step": 11105
+    },
+    {
+      "epoch": 1.981179199000981,
+      "grad_norm": 0.5010586977005005,
+      "learning_rate": 1.061779821897868e-07,
+      "loss": 0.4972,
+      "step": 11106
+    },
+    {
+      "epoch": 1.9813575952189812,
+      "grad_norm": 0.4987967312335968,
+      "learning_rate": 1.0414605576797298e-07,
+      "loss": 0.4968,
+      "step": 11107
+    },
+    {
+      "epoch": 1.9815359914369814,
+      "grad_norm": 0.6345359683036804,
+      "learning_rate": 1.0213375605896969e-07,
+      "loss": 0.6357,
+      "step": 11108
+    },
+    {
+      "epoch": 1.9817143876549816,
+      "grad_norm": 0.5932630896568298,
+      "learning_rate": 1.0014108322084492e-07,
+      "loss": 0.6181,
+      "step": 11109
+    },
+    {
+      "epoch": 1.9818927838729818,
+      "grad_norm": 0.48465976119041443,
+      "learning_rate": 9.816803741011237e-08,
+      "loss": 0.4411,
+      "step": 11110
+    },
+    {
+      "epoch": 1.982071180090982,
+      "grad_norm": 0.5137790441513062,
+      "learning_rate": 9.621461878173143e-08,
+      "loss": 0.4103,
+      "step": 11111
+    },
+    {
+      "epoch": 1.9822495763089822,
+      "grad_norm": 0.5033047199249268,
+      "learning_rate": 9.428082748910716e-08,
+      "loss": 0.4768,
+      "step": 11112
+    },
+    {
+      "epoch": 1.9824279725269824,
+      "grad_norm": 0.713068425655365,
+      "learning_rate": 9.236666368411806e-08,
+      "loss": 0.4846,
+      "step": 11113
+    },
+    {
+      "epoch": 1.9826063687449826,
+      "grad_norm": 0.5444625616073608,
+      "learning_rate": 9.047212751708834e-08,
+      "loss": 0.5477,
+      "step": 11114
+    },
+    {
+      "epoch": 1.9827847649629828,
+      "grad_norm": 0.49971529841423035,
+      "learning_rate": 8.859721913684337e-08,
+      "loss": 0.5182,
+      "step": 11115
+    },
+    {
+      "epoch": 1.982963161180983,
+      "grad_norm": 0.5162731409072876,
+      "learning_rate": 8.674193869065428e-08,
+      "loss": 0.4779,
+      "step": 11116
+    },
+    {
+      "epoch": 1.9831415573989832,
+      "grad_norm": 0.5822873115539551,
+      "learning_rate": 8.49062863241823e-08,
+      "loss": 0.6038,
+      "step": 11117
+    },
+    {
+      "epoch": 1.9833199536169834,
+      "grad_norm": 0.5523039698600769,
+      "learning_rate": 8.309026218161764e-08,
+      "loss": 0.5589,
+      "step": 11118
+    },
+    {
+      "epoch": 1.9834983498349836,
+      "grad_norm": 0.49146732687950134,
+      "learning_rate": 8.129386640562398e-08,
+      "loss": 0.4082,
+      "step": 11119
+    },
+    {
+      "epoch": 1.9836767460529838,
+      "grad_norm": 0.5370675325393677,
+      "learning_rate": 7.951709913722738e-08,
+      "loss": 0.5361,
+      "step": 11120
+    },
+    {
+      "epoch": 1.9838551422709838,
+      "grad_norm": 0.5307615399360657,
+      "learning_rate": 7.77599605160384e-08,
+      "loss": 0.4153,
+      "step": 11121
+    },
+    {
+      "epoch": 1.984033538488984,
+      "grad_norm": 0.49418723583221436,
+      "learning_rate": 7.602245068003e-08,
+      "loss": 0.4469,
+      "step": 11122
+    },
+    {
+      "epoch": 1.9842119347069842,
+      "grad_norm": 0.4878632724285126,
+      "learning_rate": 7.430456976564859e-08,
+      "loss": 0.4843,
+      "step": 11123
+    },
+    {
+      "epoch": 1.9843903309249844,
+      "grad_norm": 0.45123180747032166,
+      "learning_rate": 7.260631790784178e-08,
+      "loss": 0.4106,
+      "step": 11124
+    },
+    {
+      "epoch": 1.9845687271429846,
+      "grad_norm": 0.4762474000453949,
+      "learning_rate": 7.09276952399751e-08,
+      "loss": 0.349,
+      "step": 11125
+    },
+    {
+      "epoch": 1.9847471233609848,
+      "grad_norm": 0.547519862651825,
+      "learning_rate": 6.926870189391532e-08,
+      "loss": 0.4515,
+      "step": 11126
+    },
+    {
+      "epoch": 1.9849255195789848,
+      "grad_norm": 0.6240761876106262,
+      "learning_rate": 6.762933799991933e-08,
+      "loss": 0.6499,
+      "step": 11127
+    },
+    {
+      "epoch": 1.985103915796985,
+      "grad_norm": 0.5219815969467163,
+      "learning_rate": 6.6009603686773e-08,
+      "loss": 0.5382,
+      "step": 11128
+    },
+    {
+      "epoch": 1.9852823120149852,
+      "grad_norm": 0.5049837231636047,
+      "learning_rate": 6.440949908168014e-08,
+      "loss": 0.4419,
+      "step": 11129
+    },
+    {
+      "epoch": 1.9854607082329854,
+      "grad_norm": 0.5972961783409119,
+      "learning_rate": 6.282902431029025e-08,
+      "loss": 0.6477,
+      "step": 11130
+    },
+    {
+      "epoch": 1.9856391044509856,
+      "grad_norm": 0.5097864270210266,
+      "learning_rate": 6.126817949678176e-08,
+      "loss": 0.4747,
+      "step": 11131
+    },
+    {
+      "epoch": 1.9858175006689858,
+      "grad_norm": 0.5179688930511475,
+      "learning_rate": 5.972696476369555e-08,
+      "loss": 0.4006,
+      "step": 11132
+    },
+    {
+      "epoch": 1.985995896886986,
+      "grad_norm": 0.5777585506439209,
+      "learning_rate": 5.8205380232073666e-08,
+      "loss": 0.6213,
+      "step": 11133
+    },
+    {
+      "epoch": 1.9861742931049862,
+      "grad_norm": 0.5305327773094177,
+      "learning_rate": 5.6703426021487146e-08,
+      "loss": 0.4677,
+      "step": 11134
+    },
+    {
+      "epoch": 1.9863526893229864,
+      "grad_norm": 0.5374904870986938,
+      "learning_rate": 5.522110224981391e-08,
+      "loss": 0.492,
+      "step": 11135
+    },
+    {
+      "epoch": 1.9865310855409866,
+      "grad_norm": 0.5351753830909729,
+      "learning_rate": 5.375840903354412e-08,
+      "loss": 0.5088,
+      "step": 11136
+    },
+    {
+      "epoch": 1.9867094817589868,
+      "grad_norm": 0.513270378112793,
+      "learning_rate": 5.2315346487530334e-08,
+      "loss": 0.4432,
+      "step": 11137
+    },
+    {
+      "epoch": 1.986887877976987,
+      "grad_norm": 0.4735656678676605,
+      "learning_rate": 5.089191472507082e-08,
+      "loss": 0.4919,
+      "step": 11138
+    },
+    {
+      "epoch": 1.9870662741949872,
+      "grad_norm": 0.5290320515632629,
+      "learning_rate": 4.94881138580483e-08,
+      "loss": 0.4908,
+      "step": 11139
+    },
+    {
+      "epoch": 1.9872446704129874,
+      "grad_norm": 0.5589138865470886,
+      "learning_rate": 4.8103943996624654e-08,
+      "loss": 0.5107,
+      "step": 11140
+    },
+    {
+      "epoch": 1.9874230666309876,
+      "grad_norm": 0.5663526058197021,
+      "learning_rate": 4.673940524957398e-08,
+      "loss": 0.5412,
+      "step": 11141
+    },
+    {
+      "epoch": 1.9876014628489878,
+      "grad_norm": 0.5641487240791321,
+      "learning_rate": 4.539449772406057e-08,
+      "loss": 0.5338,
+      "step": 11142
+    },
+    {
+      "epoch": 1.9877798590669877,
+      "grad_norm": 0.5815507769584656,
+      "learning_rate": 4.406922152566661e-08,
+      "loss": 0.6252,
+      "step": 11143
+    },
+    {
+      "epoch": 1.987958255284988,
+      "grad_norm": 0.49604350328445435,
+      "learning_rate": 4.276357675853104e-08,
+      "loss": 0.4181,
+      "step": 11144
+    },
+    {
+      "epoch": 1.9881366515029881,
+      "grad_norm": 0.5443685054779053,
+      "learning_rate": 4.1477563525182945e-08,
+      "loss": 0.4321,
+      "step": 11145
+    },
+    {
+      "epoch": 1.9883150477209883,
+      "grad_norm": 0.532589316368103,
+      "learning_rate": 4.021118192662487e-08,
+      "loss": 0.5099,
+      "step": 11146
+    },
+    {
+      "epoch": 1.9884934439389885,
+      "grad_norm": 0.5429589152336121,
+      "learning_rate": 3.8964432062305046e-08,
+      "loss": 0.5191,
+      "step": 11147
+    },
+    {
+      "epoch": 1.9886718401569887,
+      "grad_norm": 0.6189258098602295,
+      "learning_rate": 3.773731403014513e-08,
+      "loss": 0.4195,
+      "step": 11148
+    },
+    {
+      "epoch": 1.9888502363749887,
+      "grad_norm": 0.5345398187637329,
+      "learning_rate": 3.652982792654025e-08,
+      "loss": 0.4932,
+      "step": 11149
+    },
+    {
+      "epoch": 1.989028632592989,
+      "grad_norm": 0.5087316036224365,
+      "learning_rate": 3.534197384630344e-08,
+      "loss": 0.4493,
+      "step": 11150
+    },
+    {
+      "epoch": 1.989207028810989,
+      "grad_norm": 0.7846596240997314,
+      "learning_rate": 3.417375188274896e-08,
+      "loss": 0.4459,
+      "step": 11151
+    },
+    {
+      "epoch": 1.9893854250289893,
+      "grad_norm": 0.5246078968048096,
+      "learning_rate": 3.302516212763673e-08,
+      "loss": 0.4156,
+      "step": 11152
+    },
+    {
+      "epoch": 1.9895638212469895,
+      "grad_norm": 0.5263677835464478,
+      "learning_rate": 3.1896204671144625e-08,
+      "loss": 0.5612,
+      "step": 11153
+    },
+    {
+      "epoch": 1.9897422174649897,
+      "grad_norm": 0.5278795957565308,
+      "learning_rate": 3.07868796019517e-08,
+      "loss": 0.474,
+      "step": 11154
+    },
+    {
+      "epoch": 1.98992061368299,
+      "grad_norm": 0.5510560274124146,
+      "learning_rate": 2.9697187007182715e-08,
+      "loss": 0.5469,
+      "step": 11155
+    },
+    {
+      "epoch": 1.99009900990099,
+      "grad_norm": 0.47594988346099854,
+      "learning_rate": 2.8627126972435857e-08,
+      "loss": 0.4399,
+      "step": 11156
+    },
+    {
+      "epoch": 1.9902774061189903,
+      "grad_norm": 0.5883272290229797,
+      "learning_rate": 2.757669958172726e-08,
+      "loss": 0.6997,
+      "step": 11157
+    },
+    {
+      "epoch": 1.9904558023369905,
+      "grad_norm": 0.5753530859947205,
+      "learning_rate": 2.654590491757425e-08,
+      "loss": 0.5808,
+      "step": 11158
+    },
+    {
+      "epoch": 1.9906341985549907,
+      "grad_norm": 0.45445308089256287,
+      "learning_rate": 2.5534743060939836e-08,
+      "loss": 0.3831,
+      "step": 11159
+    },
+    {
+      "epoch": 1.9908125947729909,
+      "grad_norm": 0.5948703289031982,
+      "learning_rate": 2.4543214091232723e-08,
+      "loss": 0.5709,
+      "step": 11160
+    },
+    {
+      "epoch": 1.990990990990991,
+      "grad_norm": 0.4452953338623047,
+      "learning_rate": 2.357131808633506e-08,
+      "loss": 0.3258,
+      "step": 11161
+    },
+    {
+      "epoch": 1.9911693872089913,
+      "grad_norm": 0.485113263130188,
+      "learning_rate": 2.2619055122574674e-08,
+      "loss": 0.3959,
+      "step": 11162
+    },
+    {
+      "epoch": 1.9913477834269915,
+      "grad_norm": 0.48504024744033813,
+      "learning_rate": 2.168642527475284e-08,
+      "loss": 0.4657,
+      "step": 11163
+    },
+    {
+      "epoch": 1.9915261796449917,
+      "grad_norm": 0.557052731513977,
+      "learning_rate": 2.0773428616088773e-08,
+      "loss": 0.5872,
+      "step": 11164
+    },
+    {
+      "epoch": 1.9917045758629917,
+      "grad_norm": 0.5224217176437378,
+      "learning_rate": 1.9880065218302877e-08,
+      "loss": 0.5756,
+      "step": 11165
+    },
+    {
+      "epoch": 1.9918829720809919,
+      "grad_norm": 0.5543699264526367,
+      "learning_rate": 1.900633515156125e-08,
+      "loss": 0.5897,
+      "step": 11166
+    },
+    {
+      "epoch": 1.992061368298992,
+      "grad_norm": 0.5769318342208862,
+      "learning_rate": 1.815223848447567e-08,
+      "loss": 0.5921,
+      "step": 11167
+    },
+    {
+      "epoch": 1.9922397645169923,
+      "grad_norm": 0.519289493560791,
+      "learning_rate": 1.731777528415912e-08,
+      "loss": 0.5332,
+      "step": 11168
+    },
+    {
+      "epoch": 1.9924181607349924,
+      "grad_norm": 0.4625517725944519,
+      "learning_rate": 1.650294561611476e-08,
+      "loss": 0.3986,
+      "step": 11169
+    },
+    {
+      "epoch": 1.9925965569529926,
+      "grad_norm": 0.7756746411323547,
+      "learning_rate": 1.5707749544374705e-08,
+      "loss": 0.4403,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9927749531709926,
+      "grad_norm": 0.5089825987815857,
+      "learning_rate": 1.4932187131333485e-08,
+      "loss": 0.4378,
+      "step": 11171
+    },
+    {
+      "epoch": 1.9929533493889928,
+      "grad_norm": 0.5764268040657043,
+      "learning_rate": 1.4176258437970102e-08,
+      "loss": 0.5419,
+      "step": 11172
+    },
+    {
+      "epoch": 1.993131745606993,
+      "grad_norm": 0.5734716653823853,
+      "learning_rate": 1.3439963523625976e-08,
+      "loss": 0.5719,
+      "step": 11173
+    },
+    {
+      "epoch": 1.9933101418249932,
+      "grad_norm": 0.5470873117446899,
+      "learning_rate": 1.2723302446115969e-08,
+      "loss": 0.4679,
+      "step": 11174
+    },
+    {
+      "epoch": 1.9934885380429934,
+      "grad_norm": 0.5397417545318604,
+      "learning_rate": 1.2026275261756148e-08,
+      "loss": 0.6085,
+      "step": 11175
+    },
+    {
+      "epoch": 1.9936669342609936,
+      "grad_norm": 0.5251044034957886,
+      "learning_rate": 1.1348882025252749e-08,
+      "loss": 0.4125,
+      "step": 11176
+    },
+    {
+      "epoch": 1.9938453304789938,
+      "grad_norm": 0.5922536253929138,
+      "learning_rate": 1.0691122789840969e-08,
+      "loss": 0.5207,
+      "step": 11177
+    },
+    {
+      "epoch": 1.994023726696994,
+      "grad_norm": 0.5680334568023682,
+      "learning_rate": 1.005299760717393e-08,
+      "loss": 0.5015,
+      "step": 11178
+    },
+    {
+      "epoch": 1.9942021229149942,
+      "grad_norm": 0.4970249831676483,
+      "learning_rate": 9.434506527378206e-09,
+      "loss": 0.3887,
+      "step": 11179
+    },
+    {
+      "epoch": 1.9943805191329944,
+      "grad_norm": 0.46873077750205994,
+      "learning_rate": 8.835649598998297e-09,
+      "loss": 0.4312,
+      "step": 11180
+    },
+    {
+      "epoch": 1.9945589153509946,
+      "grad_norm": 0.5385526418685913,
+      "learning_rate": 8.256426869079903e-09,
+      "loss": 0.4843,
+      "step": 11181
+    },
+    {
+      "epoch": 1.9947373115689948,
+      "grad_norm": 0.520590603351593,
+      "learning_rate": 7.696838383114412e-09,
+      "loss": 0.5141,
+      "step": 11182
+    },
+    {
+      "epoch": 1.994915707786995,
+      "grad_norm": 0.568450927734375,
+      "learning_rate": 7.156884185094414e-09,
+      "loss": 0.5097,
+      "step": 11183
+    },
+    {
+      "epoch": 1.9950941040049952,
+      "grad_norm": 0.5293620228767395,
+      "learning_rate": 6.636564317374916e-09,
+      "loss": 0.4605,
+      "step": 11184
+    },
+    {
+      "epoch": 1.9952725002229954,
+      "grad_norm": 0.5584262013435364,
+      "learning_rate": 6.13587882083988e-09,
+      "loss": 0.5948,
+      "step": 11185
+    },
+    {
+      "epoch": 1.9954508964409956,
+      "grad_norm": 0.5276932120323181,
+      "learning_rate": 5.654827734791201e-09,
+      "loss": 0.4695,
+      "step": 11186
+    },
+    {
+      "epoch": 1.9956292926589956,
+      "grad_norm": 0.5315849781036377,
+      "learning_rate": 5.19341109705973e-09,
+      "loss": 0.4228,
+      "step": 11187
+    },
+    {
+      "epoch": 1.9958076888769958,
+      "grad_norm": 0.5869420170783997,
+      "learning_rate": 4.751628943838737e-09,
+      "loss": 0.5691,
+      "step": 11188
+    },
+    {
+      "epoch": 1.995986085094996,
+      "grad_norm": 0.5161817669868469,
+      "learning_rate": 4.329481309850447e-09,
+      "loss": 0.4109,
+      "step": 11189
+    },
+    {
+      "epoch": 1.9961644813129962,
+      "grad_norm": 0.5489330887794495,
+      "learning_rate": 3.926968228262773e-09,
+      "loss": 0.5284,
+      "step": 11190
+    },
+    {
+      "epoch": 1.9963428775309964,
+      "grad_norm": 0.49122124910354614,
+      "learning_rate": 3.5440897306338037e-09,
+      "loss": 0.386,
+      "step": 11191
+    },
+    {
+      "epoch": 1.9965212737489966,
+      "grad_norm": 0.4992600679397583,
+      "learning_rate": 3.1808458470783395e-09,
+      "loss": 0.5033,
+      "step": 11192
+    },
+    {
+      "epoch": 1.9966996699669965,
+      "grad_norm": 0.4739280045032501,
+      "learning_rate": 2.837236606129112e-09,
+      "loss": 0.3841,
+      "step": 11193
+    },
+    {
+      "epoch": 1.9968780661849967,
+      "grad_norm": 0.5642136931419373,
+      "learning_rate": 2.5132620347645407e-09,
+      "loss": 0.5123,
+      "step": 11194
+    },
+    {
+      "epoch": 1.997056462402997,
+      "grad_norm": 0.4581596553325653,
+      "learning_rate": 2.2089221584087328e-09,
+      "loss": 0.3491,
+      "step": 11195
+    },
+    {
+      "epoch": 1.9972348586209971,
+      "grad_norm": 0.5008403062820435,
+      "learning_rate": 1.9242170010147497e-09,
+      "loss": 0.4724,
+      "step": 11196
+    },
+    {
+      "epoch": 1.9974132548389973,
+      "grad_norm": 0.5869413614273071,
+      "learning_rate": 1.6591465848703192e-09,
+      "loss": 0.5462,
+      "step": 11197
+    },
+    {
+      "epoch": 1.9975916510569975,
+      "grad_norm": 0.48155996203422546,
+      "learning_rate": 1.4137109308476338e-09,
+      "loss": 0.3609,
+      "step": 11198
+    },
+    {
+      "epoch": 1.9977700472749977,
+      "grad_norm": 0.5589842796325684,
+      "learning_rate": 1.187910058209063e-09,
+      "loss": 0.5697,
+      "step": 11199
+    },
+    {
+      "epoch": 1.997948443492998,
+      "grad_norm": 0.5098150968551636,
+      "learning_rate": 9.817439847181753e-10,
+      "loss": 0.3743,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9981268397109981,
+      "grad_norm": 0.47274699807167053,
+      "learning_rate": 7.952127265009601e-10,
+      "loss": 0.2905,
+      "step": 11201
+    },
+    {
+      "epoch": 1.9983052359289983,
+      "grad_norm": 0.48699966073036194,
+      "learning_rate": 6.283162982678725e-10,
+      "loss": 0.5163,
+      "step": 11202
+    },
+    {
+      "epoch": 1.9984836321469985,
+      "grad_norm": 0.5384899973869324,
+      "learning_rate": 4.810547130917886e-10,
+      "loss": 0.5131,
+      "step": 11203
+    },
+    {
+      "epoch": 1.9986620283649987,
+      "grad_norm": 0.5598852038383484,
+      "learning_rate": 3.534279825467834e-10,
+      "loss": 0.6017,
+      "step": 11204
+    },
+    {
+      "epoch": 1.998840424582999,
+      "grad_norm": 0.513744592666626,
+      "learning_rate": 2.454361166526198e-10,
+      "loss": 0.4185,
+      "step": 11205
+    },
+    {
+      "epoch": 1.9990188208009991,
+      "grad_norm": 0.5312053561210632,
+      "learning_rate": 1.570791239025038e-10,
+      "loss": 0.4817,
+      "step": 11206
+    },
+    {
+      "epoch": 1.9991972170189993,
+      "grad_norm": 0.48965707421302795,
+      "learning_rate": 8.835701123532935e-11,
+      "loss": 0.4257,
+      "step": 11207
+    },
+    {
+      "epoch": 1.9993756132369995,
+      "grad_norm": 0.6015014052391052,
+      "learning_rate": 3.9269784063433736e-11,
+      "loss": 0.7296,
+      "step": 11208
+    },
+    {
+      "epoch": 1.9995540094549995,
+      "grad_norm": 0.4663368761539459,
+      "learning_rate": 9.817446217086356e-12,
+      "loss": 0.3784,
+      "step": 11209
+    },
+    {
+      "epoch": 1.9997324056729997,
+      "grad_norm": 0.471098393201828,
+      "learning_rate": 0.0,
+      "loss": 0.4405,
+      "step": 11210
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 11210,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.680882782008115e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}