diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8200 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.0,
+  "global_step": 676224,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019998669080068145,
+      "loss": 10.1742,
+      "step": 500
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00019997338160136289,
+      "loss": 5.5137,
+      "step": 1000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0001999600724020443,
+      "loss": 4.8097,
+      "step": 1500
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019994676320272574,
+      "loss": 4.638,
+      "step": 2000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00019993345400340717,
+      "loss": 4.2904,
+      "step": 2500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0001999201448040886,
+      "loss": 4.1718,
+      "step": 3000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019990683560477005,
+      "loss": 4.1966,
+      "step": 3500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00019989352640545146,
+      "loss": 4.0892,
+      "step": 4000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00019988021720613287,
+      "loss": 4.0691,
+      "step": 4500
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0001998669080068143,
+      "loss": 3.9735,
+      "step": 5000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00019985359880749575,
+      "loss": 3.8196,
+      "step": 5500
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00019984028960817719,
+      "loss": 3.8869,
+      "step": 6000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0001998269804088586,
+      "loss": 3.6496,
+      "step": 6500
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00019981367120954004,
+      "loss": 3.787,
+      "step": 7000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00019980036201022147,
+      "loss": 3.679,
+      "step": 7500
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0001997870528109029,
+      "loss": 3.5967,
+      "step": 8000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00019977374361158435,
+      "loss": 3.459,
+      "step": 8500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00019976043441226576,
+      "loss": 3.563,
+      "step": 9000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0001997471252129472,
+      "loss": 3.4225,
+      "step": 9500
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00019973381601362864,
+      "loss": 3.3701,
+      "step": 10000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00019972050681431008,
+      "loss": 3.4006,
+      "step": 10500
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019970719761499149,
+      "loss": 3.4251,
+      "step": 11000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00019969388841567292,
+      "loss": 3.4543,
+      "step": 11500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00019968057921635434,
+      "loss": 3.2496,
+      "step": 12000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00019966727001703577,
+      "loss": 3.2102,
+      "step": 12500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0001996539608177172,
+      "loss": 3.3174,
+      "step": 13000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00019964065161839865,
+      "loss": 3.2625,
+      "step": 13500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001996273424190801,
+      "loss": 3.1891,
+      "step": 14000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0001996140332197615,
+      "loss": 3.1652,
+      "step": 14500
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00019960072402044294,
+      "loss": 3.1784,
+      "step": 15000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019958741482112438,
+      "loss": 3.1258,
+      "step": 15500
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00019957410562180581,
+      "loss": 3.0577,
+      "step": 16000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00019956079642248725,
+      "loss": 3.121,
+      "step": 16500
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00019954748722316866,
+      "loss": 3.0403,
+      "step": 17000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0001995341780238501,
+      "loss": 3.0235,
+      "step": 17500
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0001995208688245315,
+      "loss": 2.9968,
+      "step": 18000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00019950755962521295,
+      "loss": 3.0215,
+      "step": 18500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0001994942504258944,
+      "loss": 2.9497,
+      "step": 19000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0001994809412265758,
+      "loss": 2.9783,
+      "step": 19500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019946763202725724,
+      "loss": 2.8336,
+      "step": 20000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00019945432282793868,
+      "loss": 2.8082,
+      "step": 20500
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00019944101362862011,
+      "loss": 2.9229,
+      "step": 21000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019942770442930155,
+      "loss": 2.8394,
+      "step": 21500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00019941439522998296,
+      "loss": 2.8064,
+      "step": 22000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0001994010860306644,
+      "loss": 2.8957,
+      "step": 22500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019938777683134584,
+      "loss": 2.8115,
+      "step": 23000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00019937446763202728,
+      "loss": 2.8356,
+      "step": 23500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00019936115843270872,
+      "loss": 2.7868,
+      "step": 24000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019934784923339013,
+      "loss": 2.8692,
+      "step": 24500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00019933454003407156,
+      "loss": 2.7465,
+      "step": 25000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00019932123083475298,
+      "loss": 2.6548,
+      "step": 25500
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019930792163543441,
+      "loss": 2.7644,
+      "step": 26000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00019929461243611585,
+      "loss": 2.7362,
+      "step": 26500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0001992813032367973,
+      "loss": 2.773,
+      "step": 27000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0001992679940374787,
+      "loss": 2.6761,
+      "step": 27500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00019925468483816014,
+      "loss": 2.6885,
+      "step": 28000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00019924137563884158,
+      "loss": 2.79,
+      "step": 28500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019922806643952302,
+      "loss": 2.5552,
+      "step": 29000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00019921475724020445,
+      "loss": 2.6489,
+      "step": 29500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00019920144804088586,
+      "loss": 2.6368,
+      "step": 30000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0001991881388415673,
+      "loss": 2.6442,
+      "step": 30500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00019917482964224874,
+      "loss": 2.594,
+      "step": 31000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00019916152044293018,
+      "loss": 2.6498,
+      "step": 31500
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00019914821124361162,
+      "loss": 2.5306,
+      "step": 32000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.000199134902044293,
+      "loss": 2.6558,
+      "step": 32500
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00019912159284497444,
+      "loss": 2.5459,
+      "step": 33000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019910828364565588,
+      "loss": 2.6651,
+      "step": 33500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00019909497444633732,
+      "loss": 2.559,
+      "step": 34000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00019908166524701875,
+      "loss": 2.6479,
+      "step": 34500
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00019906835604770017,
+      "loss": 2.6017,
+      "step": 35000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0001990550468483816,
+      "loss": 2.4582,
+      "step": 35500
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00019904173764906304,
+      "loss": 2.4567,
+      "step": 36000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019902842844974448,
+      "loss": 2.4983,
+      "step": 36500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00019901511925042592,
+      "loss": 2.5343,
+      "step": 37000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00019900181005110733,
+      "loss": 2.4445,
+      "step": 37500
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00019898850085178877,
+      "loss": 2.5262,
+      "step": 38000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0001989751916524702,
+      "loss": 2.4907,
+      "step": 38500
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019896188245315164,
+      "loss": 2.6002,
+      "step": 39000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019894857325383305,
+      "loss": 2.4953,
+      "step": 39500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001989352640545145,
+      "loss": 2.4148,
+      "step": 40000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001989219548551959,
+      "loss": 2.4096,
+      "step": 40500
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019890864565587734,
+      "loss": 2.3756,
+      "step": 41000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00019889533645655878,
+      "loss": 2.4706,
+      "step": 41500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00019888202725724022,
+      "loss": 2.4323,
+      "step": 42000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019886871805792166,
+      "loss": 2.39,
+      "step": 42500
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00019885540885860307,
+      "loss": 2.3738,
+      "step": 43000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001988420996592845,
+      "loss": 2.3418,
+      "step": 43500
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019882879045996594,
+      "loss": 2.451,
+      "step": 44000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00019881548126064738,
+      "loss": 2.4559,
+      "step": 44500
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00019880217206132882,
+      "loss": 2.3337,
+      "step": 45000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00019878886286201023,
+      "loss": 2.3844,
+      "step": 45500
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00019877555366269167,
+      "loss": 2.365,
+      "step": 46000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001987622444633731,
+      "loss": 2.4264,
+      "step": 46500
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019874893526405452,
+      "loss": 2.4404,
+      "step": 47000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00019873562606473596,
+      "loss": 2.3269,
+      "step": 47500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00019872231686541737,
+      "loss": 2.2934,
+      "step": 48000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001987090076660988,
+      "loss": 2.2699,
+      "step": 48500
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00019869569846678024,
+      "loss": 2.3035,
+      "step": 49000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00019868238926746168,
+      "loss": 2.3646,
+      "step": 49500
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00019866908006814312,
+      "loss": 2.3598,
+      "step": 50000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00019865577086882453,
+      "loss": 2.2457,
+      "step": 50500
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00019864246166950597,
+      "loss": 2.3047,
+      "step": 51000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.0001986291524701874,
+      "loss": 2.3037,
+      "step": 51500
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00019861584327086885,
+      "loss": 2.1934,
+      "step": 52000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 0.00019860253407155028,
+      "loss": 2.2807,
+      "step": 52500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.0001985892248722317,
+      "loss": 2.4267,
+      "step": 53000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 0.00019857591567291313,
+      "loss": 2.2867,
+      "step": 53500
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 0.00019856260647359454,
+      "loss": 2.2614,
+      "step": 54000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00019854929727427598,
+      "loss": 2.2015,
+      "step": 54500
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 0.00019853598807495742,
+      "loss": 2.1686,
+      "step": 55000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.00019852267887563886,
+      "loss": 2.2909,
+      "step": 55500
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.00019850936967632027,
+      "loss": 2.2395,
+      "step": 56000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 0.0001984960604770017,
+      "loss": 2.2176,
+      "step": 56500
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 0.00019848275127768315,
+      "loss": 2.228,
+      "step": 57000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00019846944207836458,
+      "loss": 2.2459,
+      "step": 57500
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 0.00019845613287904602,
+      "loss": 2.277,
+      "step": 58000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 0.00019844282367972743,
+      "loss": 2.3463,
+      "step": 58500
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.00019842951448040887,
+      "loss": 2.2644,
+      "step": 59000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0001984162052810903,
+      "loss": 2.2869,
+      "step": 59500
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 0.00019840289608177175,
+      "loss": 2.1746,
+      "step": 60000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00019838958688245316,
+      "loss": 2.2261,
+      "step": 60500
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 0.00019837627768313457,
+      "loss": 2.2695,
+      "step": 61000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 0.000198362968483816,
+      "loss": 2.2722,
+      "step": 61500
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00019834965928449745,
+      "loss": 2.082,
+      "step": 62000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.00019833635008517888,
+      "loss": 2.2108,
+      "step": 62500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 0.00019832304088586032,
+      "loss": 2.2651,
+      "step": 63000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00019830973168654173,
+      "loss": 2.1761,
+      "step": 63500
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 0.00019829642248722317,
+      "loss": 2.2045,
+      "step": 64000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 0.0001982831132879046,
+      "loss": 2.2311,
+      "step": 64500
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00019826980408858605,
+      "loss": 2.2373,
+      "step": 65000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 0.00019825649488926749,
+      "loss": 2.165,
+      "step": 65500
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.0001982431856899489,
+      "loss": 2.2365,
+      "step": 66000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00019822987649063033,
+      "loss": 2.3064,
+      "step": 66500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 0.00019821656729131177,
+      "loss": 2.0983,
+      "step": 67000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 0.0001982032580919932,
+      "loss": 2.0742,
+      "step": 67500
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00019818994889267465,
+      "loss": 2.1454,
+      "step": 68000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 0.00019817663969335606,
+      "loss": 2.2284,
+      "step": 68500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.00019816333049403747,
+      "loss": 2.1694,
+      "step": 69000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 0.0001981500212947189,
+      "loss": 2.1209,
+      "step": 69500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00019813671209540035,
+      "loss": 2.0298,
+      "step": 70000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.00019812340289608179,
+      "loss": 2.1896,
+      "step": 70500
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 0.0001981100936967632,
+      "loss": 2.1888,
+      "step": 71000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 0.00019809678449744464,
+      "loss": 2.0871,
+      "step": 71500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.00019808347529812607,
+      "loss": 2.0753,
+      "step": 72000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 0.0001980701660988075,
+      "loss": 2.1047,
+      "step": 72500
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00019805685689948895,
+      "loss": 2.08,
+      "step": 73000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.00019804354770017036,
+      "loss": 2.1196,
+      "step": 73500
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 0.0001980302385008518,
+      "loss": 2.0891,
+      "step": 74000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 0.00019801692930153324,
+      "loss": 2.0471,
+      "step": 74500
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00019800362010221468,
+      "loss": 2.1587,
+      "step": 75000
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.852659821510315,
+      "eval_runtime": 152.8059,
+      "eval_samples_per_second": 61.47,
+      "eval_steps_per_second": 61.47,
+      "step": 75136
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.00019799031090289609,
+      "loss": 2.0473,
+      "step": 75500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 0.00019797700170357752,
+      "loss": 2.03,
+      "step": 76000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00019796369250425894,
+      "loss": 2.01,
+      "step": 76500
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.00019795038330494037,
+      "loss": 2.1045,
+      "step": 77000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 0.0001979370741056218,
+      "loss": 1.9704,
+      "step": 77500
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.00019792376490630325,
+      "loss": 1.9728,
+      "step": 78000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 0.0001979104557069847,
+      "loss": 2.0347,
+      "step": 78500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 0.0001978971465076661,
+      "loss": 2.029,
+      "step": 79000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00019788383730834754,
+      "loss": 2.0278,
+      "step": 79500
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 0.00019787052810902898,
+      "loss": 1.9847,
+      "step": 80000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0001978572189097104,
+      "loss": 2.0862,
+      "step": 80500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00019784390971039185,
+      "loss": 2.0602,
+      "step": 81000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 0.00019783060051107326,
+      "loss": 1.9392,
+      "step": 81500
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 0.0001978172913117547,
+      "loss": 2.0476,
+      "step": 82000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.0001978039821124361,
+      "loss": 2.022,
+      "step": 82500
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 0.00019779067291311755,
+      "loss": 1.9251,
+      "step": 83000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.000197777363713799,
+      "loss": 1.9334,
+      "step": 83500
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.0001977640545144804,
+      "loss": 1.9369,
+      "step": 84000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 0.00019775074531516184,
+      "loss": 2.0145,
+      "step": 84500
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 0.00019773743611584328,
+      "loss": 2.0389,
+      "step": 85000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00019772412691652471,
+      "loss": 2.0435,
+      "step": 85500
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 0.00019771081771720615,
+      "loss": 1.9536,
+      "step": 86000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 0.00019769750851788756,
+      "loss": 2.0962,
+      "step": 86500
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.000197684199318569,
+      "loss": 1.9748,
+      "step": 87000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00019767089011925044,
+      "loss": 1.9873,
+      "step": 87500
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 0.00019765758091993188,
+      "loss": 1.9304,
+      "step": 88000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019764427172061332,
+      "loss": 1.9514,
+      "step": 88500
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 0.00019763096252129473,
+      "loss": 1.9115,
+      "step": 89000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 0.00019761765332197614,
+      "loss": 2.015,
+      "step": 89500
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019760434412265758,
+      "loss": 1.8767,
+      "step": 90000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00019759103492333901,
+      "loss": 1.9433,
+      "step": 90500
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 0.00019757772572402045,
+      "loss": 1.8935,
+      "step": 91000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001975644165247019,
+      "loss": 1.9653,
+      "step": 91500
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 0.0001975511073253833,
+      "loss": 1.927,
+      "step": 92000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 0.00019753779812606474,
+      "loss": 1.9577,
+      "step": 92500
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019752448892674618,
+      "loss": 1.9963,
+      "step": 93000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 0.00019751117972742762,
+      "loss": 1.8967,
+      "step": 93500
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.00019749787052810905,
+      "loss": 1.9731,
+      "step": 94000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.00019748456132879046,
+      "loss": 1.9941,
+      "step": 94500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 0.0001974712521294719,
+      "loss": 1.9225,
+      "step": 95000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 0.00019745794293015334,
+      "loss": 1.9181,
+      "step": 95500
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00019744463373083478,
+      "loss": 1.9383,
+      "step": 96000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 0.00019743132453151622,
+      "loss": 1.981,
+      "step": 96500
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 0.0001974180153321976,
+      "loss": 1.9374,
+      "step": 97000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00019740470613287904,
+      "loss": 1.9689,
+      "step": 97500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.00019739139693356048,
+      "loss": 1.9001,
+      "step": 98000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 0.00019737808773424192,
+      "loss": 1.8901,
+      "step": 98500
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00019736477853492335,
+      "loss": 1.9768,
+      "step": 99000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 0.00019735146933560476,
+      "loss": 1.8889,
+      "step": 99500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 0.0001973381601362862,
+      "loss": 1.7857,
+      "step": 100000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00019732485093696764,
+      "loss": 1.89,
+      "step": 100500
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.00019731154173764908,
+      "loss": 1.9971,
+      "step": 101000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 0.00019729823253833052,
+      "loss": 1.9359,
+      "step": 101500
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00019728492333901193,
+      "loss": 1.787,
+      "step": 102000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 0.00019727161413969337,
+      "loss": 2.013,
+      "step": 102500
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 0.0001972583049403748,
+      "loss": 1.9847,
+      "step": 103000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00019724499574105624,
+      "loss": 1.9503,
+      "step": 103500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 0.00019723168654173765,
+      "loss": 1.9554,
+      "step": 104000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.0001972183773424191,
+      "loss": 1.9334,
+      "step": 104500
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.0001972050681431005,
+      "loss": 1.8504,
+      "step": 105000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 0.00019719175894378194,
+      "loss": 1.9733,
+      "step": 105500
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 0.00019717844974446338,
+      "loss": 1.8735,
+      "step": 106000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00019716514054514482,
+      "loss": 1.9075,
+      "step": 106500
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 0.00019715183134582626,
+      "loss": 1.8733,
+      "step": 107000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 0.00019713852214650767,
+      "loss": 2.0068,
+      "step": 107500
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.0001971252129471891,
+      "loss": 1.8536,
+      "step": 108000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00019711190374787054,
+      "loss": 1.8932,
+      "step": 108500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 0.00019709859454855198,
+      "loss": 1.9241,
+      "step": 109000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00019708528534923342,
+      "loss": 1.9382,
+      "step": 109500
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 0.00019707197614991483,
+      "loss": 1.8566,
+      "step": 110000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 0.00019705866695059627,
+      "loss": 1.9342,
+      "step": 110500
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00019704535775127768,
+      "loss": 1.8722,
+      "step": 111000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.00019703204855195912,
+      "loss": 1.8558,
+      "step": 111500
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 0.00019701873935264056,
+      "loss": 1.8737,
+      "step": 112000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.00019700543015332197,
+      "loss": 1.9147,
+      "step": 112500
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 0.0001969921209540034,
+      "loss": 1.958,
+      "step": 113000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 0.00019697881175468484,
+      "loss": 1.805,
+      "step": 113500
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00019696550255536628,
+      "loss": 1.7978,
+      "step": 114000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 0.00019695219335604772,
+      "loss": 1.8977,
+      "step": 114500
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00019693888415672913,
+      "loss": 1.9328,
+      "step": 115000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.00019692557495741057,
+      "loss": 1.8249,
+      "step": 115500
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 0.000196912265758092,
+      "loss": 1.9425,
+      "step": 116000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 0.00019689895655877345,
+      "loss": 1.7741,
+      "step": 116500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.00019688564735945488,
+      "loss": 1.925,
+      "step": 117000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 0.0001968723381601363,
+      "loss": 1.8892,
+      "step": 117500
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.00019685902896081773,
+      "loss": 1.832,
+      "step": 118000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00019684571976149914,
+      "loss": 1.774,
+      "step": 118500
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 0.00019683241056218058,
+      "loss": 1.8796,
+      "step": 119000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 0.00019681910136286202,
+      "loss": 1.8493,
+      "step": 119500
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00019680579216354346,
+      "loss": 1.8939,
+      "step": 120000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 0.00019679248296422487,
+      "loss": 1.8701,
+      "step": 120500
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 0.0001967791737649063,
+      "loss": 1.8433,
+      "step": 121000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00019676586456558775,
+      "loss": 1.9049,
+      "step": 121500
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.00019675255536626918,
+      "loss": 1.7999,
+      "step": 122000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 0.0001967392461669506,
+      "loss": 1.918,
+      "step": 122500
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00019672593696763203,
+      "loss": 1.8511,
+      "step": 123000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 0.00019671262776831347,
+      "loss": 1.8173,
+      "step": 123500
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 0.0001966993185689949,
+      "loss": 1.8211,
+      "step": 124000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00019668600936967635,
+      "loss": 1.7833,
+      "step": 124500
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 0.00019667270017035776,
+      "loss": 1.7822,
+      "step": 125000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00019665939097103917,
+      "loss": 2.0203,
+      "step": 125500
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.0001966460817717206,
+      "loss": 1.8281,
+      "step": 126000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 0.00019663277257240205,
+      "loss": 1.8007,
+      "step": 126500
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 0.00019661946337308348,
+      "loss": 1.8349,
+      "step": 127000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00019660615417376492,
+      "loss": 1.8092,
+      "step": 127500
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 0.00019659284497444633,
+      "loss": 1.8903,
+      "step": 128000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.00019657953577512777,
+      "loss": 1.9048,
+      "step": 128500
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.0001965662265758092,
+      "loss": 1.796,
+      "step": 129000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 0.00019655291737649065,
+      "loss": 1.8246,
+      "step": 129500
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 0.00019653960817717209,
+      "loss": 1.8512,
+      "step": 130000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.0001965262989778535,
+      "loss": 1.7868,
+      "step": 130500
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 0.00019651298977853493,
+      "loss": 1.8051,
+      "step": 131000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 0.00019649968057921637,
+      "loss": 1.8618,
+      "step": 131500
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0001964863713798978,
+      "loss": 1.8571,
+      "step": 132000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.00019647306218057922,
+      "loss": 1.8886,
+      "step": 132500
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 0.00019645975298126066,
+      "loss": 1.9144,
+      "step": 133000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.00019644644378194207,
+      "loss": 1.804,
+      "step": 133500
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 0.0001964331345826235,
+      "loss": 1.7999,
+      "step": 134000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 0.00019641982538330495,
+      "loss": 1.8522,
+      "step": 134500
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.00019640651618398639,
+      "loss": 1.6701,
+      "step": 135000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 0.0001963932069846678,
+      "loss": 1.8355,
+      "step": 135500
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00019637989778534923,
+      "loss": 1.7975,
+      "step": 136000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.00019636658858603067,
+      "loss": 1.8187,
+      "step": 136500
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 0.0001963532793867121,
+      "loss": 1.8684,
+      "step": 137000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 0.00019633997018739355,
+      "loss": 1.829,
+      "step": 137500
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.00019632666098807496,
+      "loss": 1.8728,
+      "step": 138000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 0.0001963133517887564,
+      "loss": 1.7919,
+      "step": 138500
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00019630004258943784,
+      "loss": 1.8923,
+      "step": 139000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00019628673339011927,
+      "loss": 1.8129,
+      "step": 139500
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 0.00019627342419080069,
+      "loss": 1.7684,
+      "step": 140000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 0.00019626011499148212,
+      "loss": 1.7987,
+      "step": 140500
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00019624680579216354,
+      "loss": 1.806,
+      "step": 141000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 0.00019623349659284497,
+      "loss": 1.7626,
+      "step": 141500
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 0.0001962201873935264,
+      "loss": 1.7647,
+      "step": 142000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.00019620687819420785,
+      "loss": 1.7437,
+      "step": 142500
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0001961935689948893,
+      "loss": 1.7604,
+      "step": 143000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 0.0001961802597955707,
+      "loss": 1.7737,
+      "step": 143500
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00019616695059625214,
+      "loss": 1.8022,
+      "step": 144000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 0.00019615364139693358,
+      "loss": 1.7409,
+      "step": 144500
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 0.000196140332197615,
+      "loss": 1.7658,
+      "step": 145000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00019612702299829645,
+      "loss": 1.68,
+      "step": 145500
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 0.00019611371379897786,
+      "loss": 1.8822,
+      "step": 146000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.0001961004045996593,
+      "loss": 1.8034,
+      "step": 146500
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.0001960870954003407,
+      "loss": 1.7649,
+      "step": 147000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 0.00019607378620102215,
+      "loss": 1.7675,
+      "step": 147500
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 0.0001960604770017036,
+      "loss": 1.7329,
+      "step": 148000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.000196047167802385,
+      "loss": 1.7883,
+      "step": 148500
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 0.00019603385860306644,
+      "loss": 1.7564,
+      "step": 149000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00019602054940374788,
+      "loss": 1.7273,
+      "step": 149500
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.0001960072402044293,
+      "loss": 1.8357,
+      "step": 150000
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.5852015018463135,
+      "eval_runtime": 154.5122,
+      "eval_samples_per_second": 60.791,
+      "eval_steps_per_second": 60.791,
+      "step": 150272
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 0.00019599393100511075,
+      "loss": 1.752,
+      "step": 150500
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 0.00019598062180579216,
+      "loss": 1.7735,
+      "step": 151000
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.0001959673126064736,
+      "loss": 1.7071,
+      "step": 151500
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 0.00019595400340715504,
+      "loss": 1.6139,
+      "step": 152000
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.00019594069420783648,
+      "loss": 1.7057,
+      "step": 152500
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00019592738500851792,
+      "loss": 1.645,
+      "step": 153000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.00019591407580919933,
+      "loss": 1.5747,
+      "step": 153500
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 0.00019590076660988074,
+      "loss": 1.6816,
+      "step": 154000
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00019588745741056218,
+      "loss": 1.6257,
+      "step": 154500
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 0.00019587414821124361,
+      "loss": 1.694,
+      "step": 155000
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 0.00019586083901192505,
+      "loss": 1.543,
+      "step": 155500
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0001958475298126065,
+      "loss": 1.6464,
+      "step": 156000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0001958342206132879,
+      "loss": 1.6821,
+      "step": 156500
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 0.00019582091141396934,
+      "loss": 1.6968,
+      "step": 157000
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00019580760221465078,
+      "loss": 1.6546,
+      "step": 157500
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 0.00019579429301533222,
+      "loss": 1.6658,
+      "step": 158000
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 0.00019578098381601365,
+      "loss": 1.6306,
+      "step": 158500
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.00019576767461669506,
+      "loss": 1.6541,
+      "step": 159000
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 0.0001957543654173765,
+      "loss": 1.6013,
+      "step": 159500
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00019574105621805794,
+      "loss": 1.7092,
+      "step": 160000
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00019572774701873938,
+      "loss": 1.6469,
+      "step": 160500
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 0.00019571443781942082,
+      "loss": 1.7672,
+      "step": 161000
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 0.0001957011286201022,
+      "loss": 1.5536,
+      "step": 161500
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00019568781942078364,
+      "loss": 1.6306,
+      "step": 162000
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 0.00019567451022146508,
+      "loss": 1.6919,
+      "step": 162500
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 0.00019566120102214652,
+      "loss": 1.5898,
+      "step": 163000
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00019564789182282795,
+      "loss": 1.6933,
+      "step": 163500
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.00019563458262350936,
+      "loss": 1.6183,
+      "step": 164000
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 0.0001956212734241908,
+      "loss": 1.6215,
+      "step": 164500
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00019560796422487224,
+      "loss": 1.6982,
+      "step": 165000
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 0.00019559465502555368,
+      "loss": 1.6352,
+      "step": 165500
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 0.00019558134582623512,
+      "loss": 1.659,
+      "step": 166000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00019556803662691653,
+      "loss": 1.7314,
+      "step": 166500
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.00019555472742759797,
+      "loss": 1.679,
+      "step": 167000
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 0.0001955414182282794,
+      "loss": 1.6426,
+      "step": 167500
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00019552810902896084,
+      "loss": 1.7539,
+      "step": 168000
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 0.00019551479982964225,
+      "loss": 1.6371,
+      "step": 168500
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 0.0001955014906303237,
+      "loss": 1.5544,
+      "step": 169000
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.0001954881814310051,
+      "loss": 1.7934,
+      "step": 169500
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 0.00019547487223168654,
+      "loss": 1.6451,
+      "step": 170000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00019546156303236798,
+      "loss": 1.6743,
+      "step": 170500
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00019544825383304942,
+      "loss": 1.7578,
+      "step": 171000
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 0.00019543494463373086,
+      "loss": 1.6256,
+      "step": 171500
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 0.00019542163543441227,
+      "loss": 1.6422,
+      "step": 172000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.0001954083262350937,
+      "loss": 1.6602,
+      "step": 172500
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 0.00019539501703577514,
+      "loss": 1.6107,
+      "step": 173000
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 0.00019538170783645658,
+      "loss": 1.6988,
+      "step": 173500
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00019536839863713802,
+      "loss": 1.6518,
+      "step": 174000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.00019535508943781943,
+      "loss": 1.6225,
+      "step": 174500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 0.00019534178023850087,
+      "loss": 1.6268,
+      "step": 175000
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00019532847103918228,
+      "loss": 1.6202,
+      "step": 175500
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 0.00019531516183986372,
+      "loss": 1.5834,
+      "step": 176000
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 0.00019530185264054516,
+      "loss": 1.6375,
+      "step": 176500
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.00019528854344122657,
+      "loss": 1.6556,
+      "step": 177000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.000195275234241908,
+      "loss": 1.634,
+      "step": 177500
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 0.00019526192504258944,
+      "loss": 1.5574,
+      "step": 178000
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00019524861584327088,
+      "loss": 1.6763,
+      "step": 178500
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 0.00019523530664395232,
+      "loss": 1.6812,
+      "step": 179000
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 0.00019522199744463373,
+      "loss": 1.7062,
+      "step": 179500
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.00019520868824531517,
+      "loss": 1.6767,
+      "step": 180000
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 0.0001951953790459966,
+      "loss": 1.7315,
+      "step": 180500
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.00019518206984667805,
+      "loss": 1.6142,
+      "step": 181000
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.00019516876064735948,
+      "loss": 1.6534,
+      "step": 181500
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 0.0001951554514480409,
+      "loss": 1.7295,
+      "step": 182000
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 0.00019514214224872233,
+      "loss": 1.5371,
+      "step": 182500
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00019512883304940374,
+      "loss": 1.7184,
+      "step": 183000
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 0.00019511552385008518,
+      "loss": 1.6156,
+      "step": 183500
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00019510221465076662,
+      "loss": 1.6871,
+      "step": 184000
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00019508890545144806,
+      "loss": 1.6362,
+      "step": 184500
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 0.00019507559625212947,
+      "loss": 1.5297,
+      "step": 185000
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 0.0001950622870528109,
+      "loss": 1.6025,
+      "step": 185500
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00019504897785349235,
+      "loss": 1.659,
+      "step": 186000
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 0.00019503566865417378,
+      "loss": 1.6596,
+      "step": 186500
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 0.0001950223594548552,
+      "loss": 1.6745,
+      "step": 187000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00019500905025553663,
+      "loss": 1.5484,
+      "step": 187500
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00019499574105621807,
+      "loss": 1.6971,
+      "step": 188000
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 0.0001949824318568995,
+      "loss": 1.5821,
+      "step": 188500
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00019496912265758095,
+      "loss": 1.602,
+      "step": 189000
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 0.00019495581345826236,
+      "loss": 1.6921,
+      "step": 189500
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 0.00019494250425894377,
+      "loss": 1.6905,
+      "step": 190000
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.0001949291950596252,
+      "loss": 1.7078,
+      "step": 190500
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 0.00019491588586030665,
+      "loss": 1.6151,
+      "step": 191000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00019490257666098808,
+      "loss": 1.6262,
+      "step": 191500
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00019488926746166952,
+      "loss": 1.7011,
+      "step": 192000
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 0.00019487595826235093,
+      "loss": 1.6587,
+      "step": 192500
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 0.00019486264906303237,
+      "loss": 1.6809,
+      "step": 193000
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.0001948493398637138,
+      "loss": 1.5958,
+      "step": 193500
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 0.00019483603066439525,
+      "loss": 1.6319,
+      "step": 194000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.00019482272146507669,
+      "loss": 1.5954,
+      "step": 194500
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.0001948094122657581,
+      "loss": 1.6271,
+      "step": 195000
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 0.00019479610306643953,
+      "loss": 1.7188,
+      "step": 195500
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 0.00019478279386712097,
+      "loss": 1.5758,
+      "step": 196000
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.0001947694846678024,
+      "loss": 1.6076,
+      "step": 196500
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 0.00019475617546848382,
+      "loss": 1.6151,
+      "step": 197000
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 0.00019474286626916526,
+      "loss": 1.6977,
+      "step": 197500
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.00019472955706984667,
+      "loss": 1.5838,
+      "step": 198000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0001947162478705281,
+      "loss": 1.708,
+      "step": 198500
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 0.00019470293867120955,
+      "loss": 1.6652,
+      "step": 199000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.00019468962947189099,
+      "loss": 1.5048,
+      "step": 199500
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 0.0001946763202725724,
+      "loss": 1.5494,
+      "step": 200000
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 0.00019466301107325383,
+      "loss": 1.6802,
+      "step": 200500
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.00019464970187393527,
+      "loss": 1.5599,
+      "step": 201000
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 0.0001946363926746167,
+      "loss": 1.6305,
+      "step": 201500
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.00019462308347529815,
+      "loss": 1.6988,
+      "step": 202000
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.00019460977427597956,
+      "loss": 1.5617,
+      "step": 202500
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 0.000194596465076661,
+      "loss": 1.6741,
+      "step": 203000
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 0.00019458315587734244,
+      "loss": 1.7877,
+      "step": 203500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00019456984667802387,
+      "loss": 1.698,
+      "step": 204000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 0.00019455653747870529,
+      "loss": 1.6675,
+      "step": 204500
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00019454322827938672,
+      "loss": 1.6722,
+      "step": 205000
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00019452991908006813,
+      "loss": 1.6445,
+      "step": 205500
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 0.00019451660988074957,
+      "loss": 1.6236,
+      "step": 206000
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 0.000194503300681431,
+      "loss": 1.5769,
+      "step": 206500
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.00019448999148211245,
+      "loss": 1.6174,
+      "step": 207000
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 0.0001944766822827939,
+      "loss": 1.6012,
+      "step": 207500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.0001944633730834753,
+      "loss": 1.5659,
+      "step": 208000
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 0.00019445006388415674,
+      "loss": 1.6962,
+      "step": 208500
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00019443675468483817,
+      "loss": 1.6272,
+      "step": 209000
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.0001944234454855196,
+      "loss": 1.6406,
+      "step": 209500
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 0.00019441013628620105,
+      "loss": 1.5913,
+      "step": 210000
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 0.00019439682708688246,
+      "loss": 1.627,
+      "step": 210500
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0001943835178875639,
+      "loss": 1.6659,
+      "step": 211000
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 0.0001943702086882453,
+      "loss": 1.6649,
+      "step": 211500
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 0.00019435689948892675,
+      "loss": 1.5073,
+      "step": 212000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0001943435902896082,
+      "loss": 1.5495,
+      "step": 212500
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0001943302810902896,
+      "loss": 1.6255,
+      "step": 213000
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 0.00019431697189097104,
+      "loss": 1.6997,
+      "step": 213500
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.00019430366269165248,
+      "loss": 1.5781,
+      "step": 214000
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 0.0001942903534923339,
+      "loss": 1.5791,
+      "step": 214500
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 0.00019427704429301535,
+      "loss": 1.6821,
+      "step": 215000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.00019426373509369676,
+      "loss": 1.6547,
+      "step": 215500
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.0001942504258943782,
+      "loss": 1.65,
+      "step": 216000
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 0.00019423711669505964,
+      "loss": 1.591,
+      "step": 216500
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00019422380749574108,
+      "loss": 1.5678,
+      "step": 217000
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 0.00019421049829642252,
+      "loss": 1.6103,
+      "step": 217500
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 0.00019419718909710393,
+      "loss": 1.6536,
+      "step": 218000
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00019418387989778534,
+      "loss": 1.6381,
+      "step": 218500
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 0.00019417057069846678,
+      "loss": 1.7339,
+      "step": 219000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.0001941572614991482,
+      "loss": 1.5957,
+      "step": 219500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.00019414395229982965,
+      "loss": 1.6919,
+      "step": 220000
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 0.0001941306431005111,
+      "loss": 1.5275,
+      "step": 220500
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 0.0001941173339011925,
+      "loss": 1.5781,
+      "step": 221000
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00019410402470187394,
+      "loss": 1.5521,
+      "step": 221500
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 0.00019409071550255538,
+      "loss": 1.6003,
+      "step": 222000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00019407740630323682,
+      "loss": 1.555,
+      "step": 222500
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00019406409710391825,
+      "loss": 1.5257,
+      "step": 223000
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 0.00019405078790459966,
+      "loss": 1.6349,
+      "step": 223500
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 0.0001940374787052811,
+      "loss": 1.5243,
+      "step": 224000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00019402416950596254,
+      "loss": 1.599,
+      "step": 224500
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.00019401086030664398,
+      "loss": 1.5874,
+      "step": 225000
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.4729976654052734,
+      "eval_runtime": 154.2979,
+      "eval_samples_per_second": 60.876,
+      "eval_steps_per_second": 60.876,
+      "step": 225408
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0001939975511073254,
+      "loss": 1.5186,
+      "step": 225500
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.0001939842419080068,
+      "loss": 1.5337,
+      "step": 226000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00019397093270868824,
+      "loss": 1.5111,
+      "step": 226500
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00019395762350936968,
+      "loss": 1.4425,
+      "step": 227000
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00019394431431005112,
+      "loss": 1.4928,
+      "step": 227500
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00019393100511073255,
+      "loss": 1.4468,
+      "step": 228000
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00019391769591141396,
+      "loss": 1.5364,
+      "step": 228500
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.0001939043867120954,
+      "loss": 1.5078,
+      "step": 229000
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00019389107751277684,
+      "loss": 1.5033,
+      "step": 229500
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00019387776831345828,
+      "loss": 1.4433,
+      "step": 230000
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00019386445911413972,
+      "loss": 1.5558,
+      "step": 230500
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00019385114991482113,
+      "loss": 1.4858,
+      "step": 231000
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.00019383784071550257,
+      "loss": 1.4251,
+      "step": 231500
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.000193824531516184,
+      "loss": 1.5372,
+      "step": 232000
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00019381122231686544,
+      "loss": 1.4567,
+      "step": 232500
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00019379791311754685,
+      "loss": 1.5426,
+      "step": 233000
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.0001937846039182283,
+      "loss": 1.464,
+      "step": 233500
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.0001937712947189097,
+      "loss": 1.4851,
+      "step": 234000
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00019375798551959114,
+      "loss": 1.4391,
+      "step": 234500
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00019374467632027258,
+      "loss": 1.4509,
+      "step": 235000
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00019373136712095402,
+      "loss": 1.5808,
+      "step": 235500
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00019371805792163546,
+      "loss": 1.4845,
+      "step": 236000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00019370474872231687,
+      "loss": 1.3829,
+      "step": 236500
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.0001936914395229983,
+      "loss": 1.5426,
+      "step": 237000
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00019367813032367974,
+      "loss": 1.573,
+      "step": 237500
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00019366482112436118,
+      "loss": 1.5125,
+      "step": 238000
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.00019365151192504262,
+      "loss": 1.5845,
+      "step": 238500
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00019363820272572403,
+      "loss": 1.4887,
+      "step": 239000
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00019362489352640547,
+      "loss": 1.6045,
+      "step": 239500
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00019361158432708688,
+      "loss": 1.4978,
+      "step": 240000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00019359827512776832,
+      "loss": 1.4833,
+      "step": 240500
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00019358496592844976,
+      "loss": 1.6137,
+      "step": 241000
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00019357165672913117,
+      "loss": 1.4635,
+      "step": 241500
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.0001935583475298126,
+      "loss": 1.4702,
+      "step": 242000
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00019354503833049404,
+      "loss": 1.5509,
+      "step": 242500
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.00019353172913117548,
+      "loss": 1.4055,
+      "step": 243000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00019351841993185692,
+      "loss": 1.481,
+      "step": 243500
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00019350511073253833,
+      "loss": 1.475,
+      "step": 244000
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00019349180153321977,
+      "loss": 1.4319,
+      "step": 244500
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.0001934784923339012,
+      "loss": 1.4761,
+      "step": 245000
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00019346518313458264,
+      "loss": 1.5438,
+      "step": 245500
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.00019345187393526408,
+      "loss": 1.5407,
+      "step": 246000
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.0001934385647359455,
+      "loss": 1.5398,
+      "step": 246500
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.0001934252555366269,
+      "loss": 1.4815,
+      "step": 247000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00019341194633730834,
+      "loss": 1.5605,
+      "step": 247500
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00019339863713798978,
+      "loss": 1.4427,
+      "step": 248000
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00019338532793867122,
+      "loss": 1.5884,
+      "step": 248500
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00019337201873935266,
+      "loss": 1.4853,
+      "step": 249000
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00019335870954003407,
+      "loss": 1.4497,
+      "step": 249500
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.0001933454003407155,
+      "loss": 1.4862,
+      "step": 250000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00019333209114139695,
+      "loss": 1.4691,
+      "step": 250500
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00019331878194207838,
+      "loss": 1.4709,
+      "step": 251000
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.0001933054727427598,
+      "loss": 1.4811,
+      "step": 251500
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.00019329216354344123,
+      "loss": 1.4936,
+      "step": 252000
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.00019327885434412267,
+      "loss": 1.4381,
+      "step": 252500
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.0001932655451448041,
+      "loss": 1.5663,
+      "step": 253000
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.00019325223594548555,
+      "loss": 1.5392,
+      "step": 253500
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.00019323892674616696,
+      "loss": 1.455,
+      "step": 254000
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00019322561754684837,
+      "loss": 1.5874,
+      "step": 254500
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.0001932123083475298,
+      "loss": 1.5114,
+      "step": 255000
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00019319899914821125,
+      "loss": 1.5489,
+      "step": 255500
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00019318568994889268,
+      "loss": 1.4672,
+      "step": 256000
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00019317238074957412,
+      "loss": 1.5463,
+      "step": 256500
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00019315907155025553,
+      "loss": 1.4405,
+      "step": 257000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00019314576235093697,
+      "loss": 1.4718,
+      "step": 257500
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.0001931324531516184,
+      "loss": 1.5496,
+      "step": 258000
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00019311914395229985,
+      "loss": 1.4641,
+      "step": 258500
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.00019310583475298129,
+      "loss": 1.4768,
+      "step": 259000
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.0001930925255536627,
+      "loss": 1.4985,
+      "step": 259500
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.00019307921635434413,
+      "loss": 1.497,
+      "step": 260000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00019306590715502557,
+      "loss": 1.4342,
+      "step": 260500
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.000193052597955707,
+      "loss": 1.5205,
+      "step": 261000
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00019303928875638842,
+      "loss": 1.5208,
+      "step": 261500
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.00019302597955706986,
+      "loss": 1.4806,
+      "step": 262000
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.00019301267035775127,
+      "loss": 1.4631,
+      "step": 262500
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.0001929993611584327,
+      "loss": 1.4191,
+      "step": 263000
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00019298605195911415,
+      "loss": 1.5351,
+      "step": 263500
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00019297274275979559,
+      "loss": 1.4531,
+      "step": 264000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.000192959433560477,
+      "loss": 1.4089,
+      "step": 264500
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00019294612436115843,
+      "loss": 1.4426,
+      "step": 265000
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00019293281516183987,
+      "loss": 1.4957,
+      "step": 265500
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.0001929195059625213,
+      "loss": 1.6548,
+      "step": 266000
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.00019290619676320275,
+      "loss": 1.5277,
+      "step": 266500
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.00019289288756388416,
+      "loss": 1.5082,
+      "step": 267000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.0001928795783645656,
+      "loss": 1.4819,
+      "step": 267500
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00019286626916524704,
+      "loss": 1.4152,
+      "step": 268000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00019285295996592845,
+      "loss": 1.4693,
+      "step": 268500
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.00019283965076660989,
+      "loss": 1.5178,
+      "step": 269000
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00019282634156729132,
+      "loss": 1.4476,
+      "step": 269500
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00019281303236797273,
+      "loss": 1.4921,
+      "step": 270000
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00019279972316865417,
+      "loss": 1.4735,
+      "step": 270500
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.0001927864139693356,
+      "loss": 1.4243,
+      "step": 271000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00019277310477001705,
+      "loss": 1.4855,
+      "step": 271500
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.0001927597955706985,
+      "loss": 1.5052,
+      "step": 272000
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.0001927464863713799,
+      "loss": 1.4815,
+      "step": 272500
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.00019273317717206134,
+      "loss": 1.4478,
+      "step": 273000
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.00019271986797274277,
+      "loss": 1.5841,
+      "step": 273500
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.0001927065587734242,
+      "loss": 1.4793,
+      "step": 274000
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00019269324957410565,
+      "loss": 1.503,
+      "step": 274500
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.00019267994037478706,
+      "loss": 1.4324,
+      "step": 275000
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.0001926666311754685,
+      "loss": 1.471,
+      "step": 275500
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.0001926533219761499,
+      "loss": 1.4935,
+      "step": 276000
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.00019264001277683135,
+      "loss": 1.3821,
+      "step": 276500
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.0001926267035775128,
+      "loss": 1.4417,
+      "step": 277000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.0001926133943781942,
+      "loss": 1.4789,
+      "step": 277500
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00019260008517887564,
+      "loss": 1.5533,
+      "step": 278000
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00019258677597955707,
+      "loss": 1.5845,
+      "step": 278500
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.0001925734667802385,
+      "loss": 1.3698,
+      "step": 279000
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.00019256015758091995,
+      "loss": 1.5313,
+      "step": 279500
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.00019254684838160136,
+      "loss": 1.3552,
+      "step": 280000
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.0001925335391822828,
+      "loss": 1.3905,
+      "step": 280500
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00019252022998296424,
+      "loss": 1.5076,
+      "step": 281000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00019250692078364568,
+      "loss": 1.5277,
+      "step": 281500
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00019249361158432711,
+      "loss": 1.4739,
+      "step": 282000
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00019248030238500853,
+      "loss": 1.5227,
+      "step": 282500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00019246699318568994,
+      "loss": 1.4798,
+      "step": 283000
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.00019245368398637138,
+      "loss": 1.4321,
+      "step": 283500
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.0001924403747870528,
+      "loss": 1.5775,
+      "step": 284000
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00019242706558773425,
+      "loss": 1.5437,
+      "step": 284500
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.0001924137563884157,
+      "loss": 1.4227,
+      "step": 285000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.0001924004471890971,
+      "loss": 1.4115,
+      "step": 285500
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.00019238713798977854,
+      "loss": 1.5419,
+      "step": 286000
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.00019237382879045998,
+      "loss": 1.5094,
+      "step": 286500
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.00019236051959114142,
+      "loss": 1.4809,
+      "step": 287000
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00019234721039182285,
+      "loss": 1.5076,
+      "step": 287500
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00019233390119250426,
+      "loss": 1.4439,
+      "step": 288000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.0001923205919931857,
+      "loss": 1.4485,
+      "step": 288500
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00019230728279386714,
+      "loss": 1.5172,
+      "step": 289000
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00019229397359454858,
+      "loss": 1.4831,
+      "step": 289500
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.00019228066439523,
+      "loss": 1.494,
+      "step": 290000
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.0001922673551959114,
+      "loss": 1.4732,
+      "step": 290500
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.00019225404599659284,
+      "loss": 1.5131,
+      "step": 291000
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00019224073679727428,
+      "loss": 1.4725,
+      "step": 291500
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00019222742759795572,
+      "loss": 1.5446,
+      "step": 292000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00019221411839863715,
+      "loss": 1.508,
+      "step": 292500
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00019220080919931856,
+      "loss": 1.4328,
+      "step": 293000
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.0001921875,
+      "loss": 1.4796,
+      "step": 293500
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.00019217419080068144,
+      "loss": 1.4528,
+      "step": 294000
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00019216088160136288,
+      "loss": 1.5422,
+      "step": 294500
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.00019214757240204432,
+      "loss": 1.5243,
+      "step": 295000
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.00019213426320272573,
+      "loss": 1.452,
+      "step": 295500
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00019212095400340717,
+      "loss": 1.4417,
+      "step": 296000
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.0001921076448040886,
+      "loss": 1.4921,
+      "step": 296500
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00019209433560477004,
+      "loss": 1.4464,
+      "step": 297000
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00019208102640545145,
+      "loss": 1.5396,
+      "step": 297500
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.0001920677172061329,
+      "loss": 1.5577,
+      "step": 298000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.0001920544080068143,
+      "loss": 1.4185,
+      "step": 298500
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00019204109880749574,
+      "loss": 1.4259,
+      "step": 299000
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00019202778960817718,
+      "loss": 1.4868,
+      "step": 299500
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00019201448040885862,
+      "loss": 1.4656,
+      "step": 300000
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00019200117120954006,
+      "loss": 1.5002,
+      "step": 300500
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.3984296321868896,
+      "eval_runtime": 153.5164,
+      "eval_samples_per_second": 61.186,
+      "eval_steps_per_second": 61.186,
+      "step": 300544
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.00019198786201022147,
+      "loss": 1.3097,
+      "step": 301000
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.0001919745528109029,
+      "loss": 1.3614,
+      "step": 301500
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 0.00019196124361158434,
+      "loss": 1.3087,
+      "step": 302000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00019194793441226578,
+      "loss": 1.3089,
+      "step": 302500
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00019193462521294722,
+      "loss": 1.2949,
+      "step": 303000
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 0.00019192131601362863,
+      "loss": 1.4222,
+      "step": 303500
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 0.00019190800681431007,
+      "loss": 1.3124,
+      "step": 304000
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 0.00019189469761499148,
+      "loss": 1.4448,
+      "step": 304500
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00019188138841567292,
+      "loss": 1.3385,
+      "step": 305000
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 0.00019186807921635436,
+      "loss": 1.4459,
+      "step": 305500
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 0.00019185477001703577,
+      "loss": 1.3182,
+      "step": 306000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 0.0001918414608177172,
+      "loss": 1.4025,
+      "step": 306500
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 0.00019182815161839864,
+      "loss": 1.3668,
+      "step": 307000
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 0.00019181484241908008,
+      "loss": 1.3769,
+      "step": 307500
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.00019180153321976152,
+      "loss": 1.3798,
+      "step": 308000
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00019178822402044293,
+      "loss": 1.3717,
+      "step": 308500
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00019177491482112437,
+      "loss": 1.2967,
+      "step": 309000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 0.0001917616056218058,
+      "loss": 1.2975,
+      "step": 309500
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.00019174829642248724,
+      "loss": 1.3321,
+      "step": 310000
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.00019173498722316868,
+      "loss": 1.3499,
+      "step": 310500
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 0.0001917216780238501,
+      "loss": 1.3556,
+      "step": 311000
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 0.0001917083688245315,
+      "loss": 1.409,
+      "step": 311500
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 0.00019169505962521294,
+      "loss": 1.3666,
+      "step": 312000
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 0.00019168175042589438,
+      "loss": 1.4045,
+      "step": 312500
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.00019166844122657582,
+      "loss": 1.4168,
+      "step": 313000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.00019165513202725726,
+      "loss": 1.3961,
+      "step": 313500
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 0.00019164182282793867,
+      "loss": 1.3621,
+      "step": 314000
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 0.0001916285136286201,
+      "loss": 1.3018,
+      "step": 314500
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 0.00019161520442930154,
+      "loss": 1.341,
+      "step": 315000
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 0.00019160189522998298,
+      "loss": 1.4319,
+      "step": 315500
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.0001915885860306644,
+      "loss": 1.3935,
+      "step": 316000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.00019157527683134583,
+      "loss": 1.3445,
+      "step": 316500
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 0.00019156196763202727,
+      "loss": 1.4346,
+      "step": 317000
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.0001915486584327087,
+      "loss": 1.4014,
+      "step": 317500
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.00019153534923339015,
+      "loss": 1.4154,
+      "step": 318000
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 0.00019152204003407156,
+      "loss": 1.2909,
+      "step": 318500
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 0.00019150873083475297,
+      "loss": 1.352,
+      "step": 319000
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 0.0001914954216354344,
+      "loss": 1.3872,
+      "step": 319500
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 0.00019148211243611585,
+      "loss": 1.3701,
+      "step": 320000
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 0.00019146880323679728,
+      "loss": 1.4155,
+      "step": 320500
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 0.00019145549403747872,
+      "loss": 1.4313,
+      "step": 321000
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00019144218483816013,
+      "loss": 1.3959,
+      "step": 321500
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 0.00019142887563884157,
+      "loss": 1.3712,
+      "step": 322000
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 0.000191415566439523,
+      "loss": 1.2681,
+      "step": 322500
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 0.00019140225724020445,
+      "loss": 1.4065,
+      "step": 323000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00019138894804088589,
+      "loss": 1.4301,
+      "step": 323500
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.0001913756388415673,
+      "loss": 1.4,
+      "step": 324000
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 0.00019136232964224873,
+      "loss": 1.3669,
+      "step": 324500
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.00019134902044293017,
+      "loss": 1.4134,
+      "step": 325000
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.0001913357112436116,
+      "loss": 1.3596,
+      "step": 325500
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 0.00019132240204429302,
+      "loss": 1.325,
+      "step": 326000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 0.00019130909284497446,
+      "loss": 1.415,
+      "step": 326500
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 0.00019129578364565587,
+      "loss": 1.3677,
+      "step": 327000
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 0.0001912824744463373,
+      "loss": 1.4166,
+      "step": 327500
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 0.00019126916524701875,
+      "loss": 1.3797,
+      "step": 328000
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 0.00019125585604770019,
+      "loss": 1.4002,
+      "step": 328500
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 0.0001912425468483816,
+      "loss": 1.3016,
+      "step": 329000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00019122923764906303,
+      "loss": 1.2929,
+      "step": 329500
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00019121592844974447,
+      "loss": 1.303,
+      "step": 330000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.0001912026192504259,
+      "loss": 1.387,
+      "step": 330500
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.00019118931005110735,
+      "loss": 1.3859,
+      "step": 331000
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.00019117600085178876,
+      "loss": 1.3026,
+      "step": 331500
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 0.0001911626916524702,
+      "loss": 1.3718,
+      "step": 332000
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 0.00019114938245315164,
+      "loss": 1.3643,
+      "step": 332500
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 0.00019113607325383305,
+      "loss": 1.3631,
+      "step": 333000
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 0.00019112276405451449,
+      "loss": 1.3937,
+      "step": 333500
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.00019110945485519592,
+      "loss": 1.3724,
+      "step": 334000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.00019109614565587733,
+      "loss": 1.4076,
+      "step": 334500
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 0.00019108283645655877,
+      "loss": 1.4288,
+      "step": 335000
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 0.0001910695272572402,
+      "loss": 1.362,
+      "step": 335500
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 0.00019105621805792165,
+      "loss": 1.3388,
+      "step": 336000
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 0.0001910429088586031,
+      "loss": 1.3455,
+      "step": 336500
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.0001910295996592845,
+      "loss": 1.3507,
+      "step": 337000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.00019101629045996594,
+      "loss": 1.387,
+      "step": 337500
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 0.00019100298126064737,
+      "loss": 1.3889,
+      "step": 338000
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.0001909896720613288,
+      "loss": 1.4924,
+      "step": 338500
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.00019097636286201025,
+      "loss": 1.3709,
+      "step": 339000
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.00019096305366269166,
+      "loss": 1.3821,
+      "step": 339500
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 0.0001909497444633731,
+      "loss": 1.45,
+      "step": 340000
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 0.0001909364352640545,
+      "loss": 1.432,
+      "step": 340500
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.00019092312606473595,
+      "loss": 1.2902,
+      "step": 341000
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 0.0001909098168654174,
+      "loss": 1.393,
+      "step": 341500
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 0.0001908965076660988,
+      "loss": 1.382,
+      "step": 342000
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 0.00019088319846678024,
+      "loss": 1.4208,
+      "step": 342500
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 0.00019086988926746167,
+      "loss": 1.3832,
+      "step": 343000
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 0.0001908565800681431,
+      "loss": 1.3807,
+      "step": 343500
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 0.00019084327086882455,
+      "loss": 1.4388,
+      "step": 344000
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.00019082996166950596,
+      "loss": 1.3082,
+      "step": 344500
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.0001908166524701874,
+      "loss": 1.4276,
+      "step": 345000
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 0.00019080334327086884,
+      "loss": 1.4532,
+      "step": 345500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 0.00019079003407155028,
+      "loss": 1.3624,
+      "step": 346000
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 0.00019077672487223171,
+      "loss": 1.526,
+      "step": 346500
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.00019076341567291313,
+      "loss": 1.3321,
+      "step": 347000
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.00019075010647359454,
+      "loss": 1.4103,
+      "step": 347500
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 0.00019073679727427598,
+      "loss": 1.3803,
+      "step": 348000
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 0.0001907234880749574,
+      "loss": 1.3081,
+      "step": 348500
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 0.00019071017887563885,
+      "loss": 1.3709,
+      "step": 349000
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 0.0001906968696763203,
+      "loss": 1.4199,
+      "step": 349500
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 0.0001906835604770017,
+      "loss": 1.3941,
+      "step": 350000
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 0.00019067025127768314,
+      "loss": 1.3927,
+      "step": 350500
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 0.00019065694207836458,
+      "loss": 1.3579,
+      "step": 351000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.00019064363287904601,
+      "loss": 1.2922,
+      "step": 351500
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.00019063032367972745,
+      "loss": 1.3024,
+      "step": 352000
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.00019061701448040886,
+      "loss": 1.4132,
+      "step": 352500
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 0.0001906037052810903,
+      "loss": 1.5121,
+      "step": 353000
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 0.00019059039608177174,
+      "loss": 1.4538,
+      "step": 353500
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 0.00019057708688245318,
+      "loss": 1.5066,
+      "step": 354000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.0001905637776831346,
+      "loss": 1.3405,
+      "step": 354500
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.000190550468483816,
+      "loss": 1.33,
+      "step": 355000
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 0.00019053715928449744,
+      "loss": 1.2941,
+      "step": 355500
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.00019052385008517888,
+      "loss": 1.4025,
+      "step": 356000
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.00019051054088586032,
+      "loss": 1.3327,
+      "step": 356500
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 0.00019049723168654175,
+      "loss": 1.4117,
+      "step": 357000
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 0.00019048392248722316,
+      "loss": 1.3084,
+      "step": 357500
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 0.0001904706132879046,
+      "loss": 1.339,
+      "step": 358000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.00019045730408858604,
+      "loss": 1.4461,
+      "step": 358500
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 0.00019044399488926748,
+      "loss": 1.2968,
+      "step": 359000
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 0.00019043068568994892,
+      "loss": 1.312,
+      "step": 359500
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 0.00019041737649063033,
+      "loss": 1.3812,
+      "step": 360000
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.00019040406729131177,
+      "loss": 1.4189,
+      "step": 360500
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.0001903907580919932,
+      "loss": 1.43,
+      "step": 361000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 0.00019037744889267464,
+      "loss": 1.382,
+      "step": 361500
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.00019036413969335605,
+      "loss": 1.4704,
+      "step": 362000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.0001903508304940375,
+      "loss": 1.3853,
+      "step": 362500
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 0.0001903375212947189,
+      "loss": 1.3542,
+      "step": 363000
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.00019032421209540034,
+      "loss": 1.3554,
+      "step": 363500
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.00019031090289608178,
+      "loss": 1.528,
+      "step": 364000
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 0.00019029759369676322,
+      "loss": 1.4008,
+      "step": 364500
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.00019028428449744466,
+      "loss": 1.3436,
+      "step": 365000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.00019027097529812607,
+      "loss": 1.3648,
+      "step": 365500
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.0001902576660988075,
+      "loss": 1.491,
+      "step": 366000
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 0.00019024435689948894,
+      "loss": 1.3592,
+      "step": 366500
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 0.00019023104770017038,
+      "loss": 1.4573,
+      "step": 367000
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 0.0001902177385008518,
+      "loss": 1.4006,
+      "step": 367500
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.00019020442930153323,
+      "loss": 1.3101,
+      "step": 368000
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.00019019112010221467,
+      "loss": 1.3908,
+      "step": 368500
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 0.00019017781090289608,
+      "loss": 1.4466,
+      "step": 369000
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.00019016450170357752,
+      "loss": 1.3716,
+      "step": 369500
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.00019015119250425896,
+      "loss": 1.3971,
+      "step": 370000
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.00019013788330494037,
+      "loss": 1.438,
+      "step": 370500
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 0.0001901245741056218,
+      "loss": 1.3733,
+      "step": 371000
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 0.00019011126490630324,
+      "loss": 1.3873,
+      "step": 371500
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 0.00019009795570698468,
+      "loss": 1.3727,
+      "step": 372000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.00019008464650766612,
+      "loss": 1.3708,
+      "step": 372500
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.00019007133730834753,
+      "loss": 1.3796,
+      "step": 373000
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 0.00019005802810902897,
+      "loss": 1.3348,
+      "step": 373500
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.0001900447189097104,
+      "loss": 1.2869,
+      "step": 374000
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.00019003140971039184,
+      "loss": 1.4122,
+      "step": 374500
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 0.00019001810051107328,
+      "loss": 1.4599,
+      "step": 375000
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0001900047913117547,
+      "loss": 1.482,
+      "step": 375500
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 1.3631517887115479,
+      "eval_runtime": 151.4049,
+      "eval_samples_per_second": 62.039,
+      "eval_steps_per_second": 62.039,
+      "step": 375680
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0001899914821124361,
+      "loss": 1.2407,
+      "step": 376000
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 0.00018997817291311754,
+      "loss": 1.2129,
+      "step": 376500
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.00018996486371379898,
+      "loss": 1.3004,
+      "step": 377000
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 0.00018995155451448042,
+      "loss": 1.2392,
+      "step": 377500
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 0.00018993824531516186,
+      "loss": 1.353,
+      "step": 378000
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.00018992493611584327,
+      "loss": 1.2066,
+      "step": 378500
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 0.0001899116269165247,
+      "loss": 1.2607,
+      "step": 379000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.00018989831771720614,
+      "loss": 1.2476,
+      "step": 379500
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.00018988500851788758,
+      "loss": 1.2611,
+      "step": 380000
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 0.000189871699318569,
+      "loss": 1.3003,
+      "step": 380500
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 0.00018985839011925043,
+      "loss": 1.2301,
+      "step": 381000
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.00018984508091993187,
+      "loss": 1.1755,
+      "step": 381500
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 0.0001898317717206133,
+      "loss": 1.2484,
+      "step": 382000
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 0.00018981846252129475,
+      "loss": 1.3084,
+      "step": 382500
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.00018980515332197616,
+      "loss": 1.2391,
+      "step": 383000
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 0.00018979184412265757,
+      "loss": 1.2445,
+      "step": 383500
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 0.000189778534923339,
+      "loss": 1.3115,
+      "step": 384000
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 0.00018976522572402044,
+      "loss": 1.2671,
+      "step": 384500
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 0.00018975191652470188,
+      "loss": 1.3634,
+      "step": 385000
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 0.00018973860732538332,
+      "loss": 1.318,
+      "step": 385500
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00018972529812606473,
+      "loss": 1.3099,
+      "step": 386000
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00018971198892674617,
+      "loss": 1.2595,
+      "step": 386500
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 0.0001896986797274276,
+      "loss": 1.3852,
+      "step": 387000
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.00018968537052810905,
+      "loss": 1.2959,
+      "step": 387500
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 0.00018967206132879048,
+      "loss": 1.2599,
+      "step": 388000
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 0.0001896587521294719,
+      "loss": 1.2856,
+      "step": 388500
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.00018964544293015333,
+      "loss": 1.3148,
+      "step": 389000
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 0.00018963213373083477,
+      "loss": 1.2934,
+      "step": 389500
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.0001896188245315162,
+      "loss": 1.2729,
+      "step": 390000
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.00018960551533219762,
+      "loss": 1.3682,
+      "step": 390500
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 0.00018959220613287906,
+      "loss": 1.3046,
+      "step": 391000
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 0.00018957889693356047,
+      "loss": 1.2826,
+      "step": 391500
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.0001895655877342419,
+      "loss": 1.3338,
+      "step": 392000
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 0.00018955227853492335,
+      "loss": 1.3051,
+      "step": 392500
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.00018953896933560479,
+      "loss": 1.279,
+      "step": 393000
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 0.0001895256601362862,
+      "loss": 1.2571,
+      "step": 393500
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 0.00018951235093696763,
+      "loss": 1.2967,
+      "step": 394000
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 0.00018949904173764907,
+      "loss": 1.3146,
+      "step": 394500
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.0001894857325383305,
+      "loss": 1.3376,
+      "step": 395000
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 0.00018947242333901195,
+      "loss": 1.307,
+      "step": 395500
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 0.00018945911413969336,
+      "loss": 1.29,
+      "step": 396000
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.0001894458049403748,
+      "loss": 1.2576,
+      "step": 396500
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.00018943249574105624,
+      "loss": 1.2077,
+      "step": 397000
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 0.00018941918654173765,
+      "loss": 1.3651,
+      "step": 397500
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.00018940587734241909,
+      "loss": 1.3213,
+      "step": 398000
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 0.00018939256814310052,
+      "loss": 1.2586,
+      "step": 398500
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 0.00018937925894378193,
+      "loss": 1.3278,
+      "step": 399000
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.00018936594974446337,
+      "loss": 1.2745,
+      "step": 399500
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 0.0001893526405451448,
+      "loss": 1.2649,
+      "step": 400000
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 0.00018933933134582625,
+      "loss": 1.3187,
+      "step": 400500
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.0001893260221465077,
+      "loss": 1.2126,
+      "step": 401000
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 0.0001893127129471891,
+      "loss": 1.281,
+      "step": 401500
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 0.00018929940374787054,
+      "loss": 1.1913,
+      "step": 402000
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.00018928609454855197,
+      "loss": 1.3565,
+      "step": 402500
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 0.0001892727853492334,
+      "loss": 1.337,
+      "step": 403000
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.00018925947614991485,
+      "loss": 1.2379,
+      "step": 403500
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.00018924616695059626,
+      "loss": 1.2413,
+      "step": 404000
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 0.00018923285775127767,
+      "loss": 1.3328,
+      "step": 404500
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 0.0001892195485519591,
+      "loss": 1.2681,
+      "step": 405000
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 0.00018920623935264055,
+      "loss": 1.332,
+      "step": 405500
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 0.000189192930153322,
+      "loss": 1.3436,
+      "step": 406000
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 0.0001891796209540034,
+      "loss": 1.2292,
+      "step": 406500
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.00018916631175468484,
+      "loss": 1.3298,
+      "step": 407000
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.00018915300255536627,
+      "loss": 1.323,
+      "step": 407500
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 0.0001891396933560477,
+      "loss": 1.2168,
+      "step": 408000
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.00018912638415672915,
+      "loss": 1.2748,
+      "step": 408500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 0.00018911307495741056,
+      "loss": 1.2818,
+      "step": 409000
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 0.000189099765758092,
+      "loss": 1.3531,
+      "step": 409500
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.00018908645655877344,
+      "loss": 1.2883,
+      "step": 410000
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.00018907314735945488,
+      "loss": 1.2805,
+      "step": 410500
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 0.00018905983816013631,
+      "loss": 1.1821,
+      "step": 411000
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00018904652896081773,
+      "loss": 1.3083,
+      "step": 411500
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 0.00018903321976149914,
+      "loss": 1.2781,
+      "step": 412000
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 0.00018901991056218057,
+      "loss": 1.3032,
+      "step": 412500
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.000189006601362862,
+      "loss": 1.3755,
+      "step": 413000
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 0.00018899329216354345,
+      "loss": 1.3112,
+      "step": 413500
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.0001889799829642249,
+      "loss": 1.2157,
+      "step": 414000
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.0001889666737649063,
+      "loss": 1.3974,
+      "step": 414500
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 0.00018895336456558774,
+      "loss": 1.2776,
+      "step": 415000
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 0.00018894005536626918,
+      "loss": 1.3196,
+      "step": 415500
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00018892674616695061,
+      "loss": 1.2642,
+      "step": 416000
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 0.00018891343696763205,
+      "loss": 1.3229,
+      "step": 416500
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 0.00018890012776831346,
+      "loss": 1.3571,
+      "step": 417000
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.0001888868185689949,
+      "loss": 1.2576,
+      "step": 417500
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.00018887350936967634,
+      "loss": 1.2513,
+      "step": 418000
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 0.00018886020017035778,
+      "loss": 1.2904,
+      "step": 418500
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.0001888468909710392,
+      "loss": 1.2498,
+      "step": 419000
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 0.0001888335817717206,
+      "loss": 1.3343,
+      "step": 419500
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 0.00018882027257240204,
+      "loss": 1.287,
+      "step": 420000
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00018880696337308348,
+      "loss": 1.2978,
+      "step": 420500
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00018879365417376491,
+      "loss": 1.4191,
+      "step": 421000
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 0.00018878034497444635,
+      "loss": 1.2907,
+      "step": 421500
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.00018876703577512776,
+      "loss": 1.2843,
+      "step": 422000
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 0.0001887537265758092,
+      "loss": 1.2897,
+      "step": 422500
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 0.00018874041737649064,
+      "loss": 1.2947,
+      "step": 423000
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.00018872710817717208,
+      "loss": 1.2929,
+      "step": 423500
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 0.00018871379897785352,
+      "loss": 1.3531,
+      "step": 424000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.00018870048977853493,
+      "loss": 1.2122,
+      "step": 424500
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.00018868718057921637,
+      "loss": 1.3542,
+      "step": 425000
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 0.0001886738713798978,
+      "loss": 1.2293,
+      "step": 425500
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 0.00018866056218057922,
+      "loss": 1.3058,
+      "step": 426000
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 0.00018864725298126065,
+      "loss": 1.3417,
+      "step": 426500
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 0.0001886339437819421,
+      "loss": 1.3006,
+      "step": 427000
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 0.0001886206345826235,
+      "loss": 1.3495,
+      "step": 427500
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00018860732538330494,
+      "loss": 1.2705,
+      "step": 428000
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00018859401618398638,
+      "loss": 1.2444,
+      "step": 428500
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 0.00018858070698466782,
+      "loss": 1.3491,
+      "step": 429000
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00018856739778534926,
+      "loss": 1.3384,
+      "step": 429500
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 0.00018855408858603067,
+      "loss": 1.3292,
+      "step": 430000
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 0.0001885407793867121,
+      "loss": 1.3075,
+      "step": 430500
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.00018852747018739354,
+      "loss": 1.3825,
+      "step": 431000
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.00018851416098807498,
+      "loss": 1.3393,
+      "step": 431500
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 0.0001885008517887564,
+      "loss": 1.2794,
+      "step": 432000
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00018848754258943783,
+      "loss": 1.3184,
+      "step": 432500
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 0.00018847423339011927,
+      "loss": 1.312,
+      "step": 433000
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 0.00018846092419080068,
+      "loss": 1.2864,
+      "step": 433500
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.00018844761499148212,
+      "loss": 1.264,
+      "step": 434000
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 0.00018843430579216356,
+      "loss": 1.2549,
+      "step": 434500
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.00018842099659284497,
+      "loss": 1.397,
+      "step": 435000
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 0.0001884076873935264,
+      "loss": 1.4127,
+      "step": 435500
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 0.00018839437819420784,
+      "loss": 1.28,
+      "step": 436000
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 0.00018838106899488928,
+      "loss": 1.2579,
+      "step": 436500
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00018836775979557072,
+      "loss": 1.3153,
+      "step": 437000
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 0.00018835445059625213,
+      "loss": 1.2852,
+      "step": 437500
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 0.00018834114139693357,
+      "loss": 1.3078,
+      "step": 438000
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.000188327832197615,
+      "loss": 1.198,
+      "step": 438500
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00018831452299829644,
+      "loss": 1.294,
+      "step": 439000
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 0.00018830121379897788,
+      "loss": 1.3666,
+      "step": 439500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.0001882879045996593,
+      "loss": 1.2163,
+      "step": 440000
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 0.0001882745954003407,
+      "loss": 1.3198,
+      "step": 440500
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 0.00018826128620102214,
+      "loss": 1.2969,
+      "step": 441000
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00018824797700170358,
+      "loss": 1.2432,
+      "step": 441500
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00018823466780238502,
+      "loss": 1.2734,
+      "step": 442000
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 0.00018822135860306646,
+      "loss": 1.2472,
+      "step": 442500
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.00018820804940374787,
+      "loss": 1.3459,
+      "step": 443000
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 0.0001881947402044293,
+      "loss": 1.2772,
+      "step": 443500
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 0.00018818143100511074,
+      "loss": 1.2499,
+      "step": 444000
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 0.00018816812180579218,
+      "loss": 1.2926,
+      "step": 444500
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 0.0001881548126064736,
+      "loss": 1.3315,
+      "step": 445000
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.00018814150340715503,
+      "loss": 1.3512,
+      "step": 445500
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.00018812819420783647,
+      "loss": 1.309,
+      "step": 446000
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 0.0001881148850085179,
+      "loss": 1.3363,
+      "step": 446500
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 0.00018810157580919935,
+      "loss": 1.2856,
+      "step": 447000
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.00018808826660988076,
+      "loss": 1.302,
+      "step": 447500
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 0.00018807495741056217,
+      "loss": 1.2676,
+      "step": 448000
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.0001880616482112436,
+      "loss": 1.2811,
+      "step": 448500
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 0.00018804833901192504,
+      "loss": 1.3252,
+      "step": 449000
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 0.00018803502981260648,
+      "loss": 1.2677,
+      "step": 449500
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 0.00018802172061328792,
+      "loss": 1.367,
+      "step": 450000
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00018800841141396933,
+      "loss": 1.321,
+      "step": 450500
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.3407889604568481,
+      "eval_runtime": 150.7509,
+      "eval_samples_per_second": 62.308,
+      "eval_steps_per_second": 62.308,
+      "step": 450816
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 0.00018799510221465077,
+      "loss": 1.2992,
+      "step": 451000
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 0.0001879817930153322,
+      "loss": 1.2075,
+      "step": 451500
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00018796848381601365,
+      "loss": 1.211,
+      "step": 452000
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00018795517461669508,
+      "loss": 1.1365,
+      "step": 452500
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 0.0001879418654173765,
+      "loss": 1.1299,
+      "step": 453000
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00018792855621805793,
+      "loss": 1.233,
+      "step": 453500
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 0.00018791524701873937,
+      "loss": 1.1957,
+      "step": 454000
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 0.0001879019378194208,
+      "loss": 1.0896,
+      "step": 454500
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00018788862862010222,
+      "loss": 1.1323,
+      "step": 455000
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 0.00018787531942078366,
+      "loss": 1.1648,
+      "step": 455500
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.00018786201022146507,
+      "loss": 1.1893,
+      "step": 456000
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.0001878487010221465,
+      "loss": 1.2025,
+      "step": 456500
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 0.00018783539182282795,
+      "loss": 1.1842,
+      "step": 457000
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 0.00018782208262350938,
+      "loss": 1.2402,
+      "step": 457500
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 0.0001878087734241908,
+      "loss": 1.1911,
+      "step": 458000
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 0.00018779546422487223,
+      "loss": 1.1726,
+      "step": 458500
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.00018778215502555367,
+      "loss": 1.1387,
+      "step": 459000
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.0001877688458262351,
+      "loss": 1.1896,
+      "step": 459500
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 0.00018775553662691655,
+      "loss": 1.3012,
+      "step": 460000
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 0.00018774222742759796,
+      "loss": 1.2844,
+      "step": 460500
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.0001877289182282794,
+      "loss": 1.2092,
+      "step": 461000
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 0.00018771560902896084,
+      "loss": 1.2416,
+      "step": 461500
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 0.00018770229982964225,
+      "loss": 1.1934,
+      "step": 462000
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.00018768899063032369,
+      "loss": 1.2612,
+      "step": 462500
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.00018767568143100512,
+      "loss": 1.2074,
+      "step": 463000
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 0.00018766237223168653,
+      "loss": 1.1902,
+      "step": 463500
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.00018764906303236797,
+      "loss": 1.2334,
+      "step": 464000
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 0.0001876357538330494,
+      "loss": 1.2369,
+      "step": 464500
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 0.00018762244463373085,
+      "loss": 1.2007,
+      "step": 465000
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.0001876091354344123,
+      "loss": 1.2398,
+      "step": 465500
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 0.0001875958262350937,
+      "loss": 1.1333,
+      "step": 466000
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00018758251703577514,
+      "loss": 1.3337,
+      "step": 466500
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.00018756920783645657,
+      "loss": 1.2476,
+      "step": 467000
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 0.000187555898637138,
+      "loss": 1.2868,
+      "step": 467500
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 0.00018754258943781945,
+      "loss": 1.2082,
+      "step": 468000
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.00018752928023850086,
+      "loss": 1.2373,
+      "step": 468500
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 0.00018751597103918227,
+      "loss": 1.2823,
+      "step": 469000
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.0001875026618398637,
+      "loss": 1.1542,
+      "step": 469500
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.00018748935264054515,
+      "loss": 1.2151,
+      "step": 470000
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 0.0001874760434412266,
+      "loss": 1.145,
+      "step": 470500
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 0.000187462734241908,
+      "loss": 1.2274,
+      "step": 471000
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00018744942504258944,
+      "loss": 1.1398,
+      "step": 471500
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 0.00018743611584327087,
+      "loss": 1.2637,
+      "step": 472000
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 0.0001874228066439523,
+      "loss": 1.2776,
+      "step": 472500
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00018740949744463375,
+      "loss": 1.2042,
+      "step": 473000
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00018739618824531516,
+      "loss": 1.2968,
+      "step": 473500
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 0.0001873828790459966,
+      "loss": 1.2255,
+      "step": 474000
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00018736956984667804,
+      "loss": 1.1385,
+      "step": 474500
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 0.00018735626064735948,
+      "loss": 1.204,
+      "step": 475000
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 0.00018734295144804091,
+      "loss": 1.1504,
+      "step": 475500
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 0.00018732964224872233,
+      "loss": 1.2177,
+      "step": 476000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 0.00018731633304940374,
+      "loss": 1.2204,
+      "step": 476500
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 0.00018730302385008517,
+      "loss": 1.1733,
+      "step": 477000
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 0.0001872897146507666,
+      "loss": 1.1992,
+      "step": 477500
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 0.00018727640545144805,
+      "loss": 1.1848,
+      "step": 478000
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 0.0001872630962521295,
+      "loss": 1.2006,
+      "step": 478500
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.0001872497870528109,
+      "loss": 1.1762,
+      "step": 479000
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 0.00018723647785349234,
+      "loss": 1.326,
+      "step": 479500
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.00018722316865417378,
+      "loss": 1.2996,
+      "step": 480000
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00018720985945485521,
+      "loss": 1.2594,
+      "step": 480500
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 0.00018719655025553665,
+      "loss": 1.246,
+      "step": 481000
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 0.00018718324105621806,
+      "loss": 1.1864,
+      "step": 481500
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 0.0001871699318568995,
+      "loss": 1.2074,
+      "step": 482000
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 0.00018715662265758094,
+      "loss": 1.1846,
+      "step": 482500
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.00018714331345826238,
+      "loss": 1.2251,
+      "step": 483000
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 0.0001871300042589438,
+      "loss": 1.2194,
+      "step": 483500
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.0001871166950596252,
+      "loss": 1.1532,
+      "step": 484000
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.00018710338586030664,
+      "loss": 1.1033,
+      "step": 484500
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 0.00018709007666098808,
+      "loss": 1.2574,
+      "step": 485000
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 0.00018707676746166951,
+      "loss": 1.177,
+      "step": 485500
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.00018706345826235095,
+      "loss": 1.3113,
+      "step": 486000
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 0.00018705014906303236,
+      "loss": 1.2113,
+      "step": 486500
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.0001870368398637138,
+      "loss": 1.2341,
+      "step": 487000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.00018702353066439524,
+      "loss": 1.3178,
+      "step": 487500
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 0.00018701022146507668,
+      "loss": 1.1443,
+      "step": 488000
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 0.00018699691226575812,
+      "loss": 1.1875,
+      "step": 488500
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 0.00018698360306643953,
+      "loss": 1.2945,
+      "step": 489000
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 0.00018697029386712097,
+      "loss": 1.214,
+      "step": 489500
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 0.0001869569846678024,
+      "loss": 1.1998,
+      "step": 490000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00018694367546848382,
+      "loss": 1.2449,
+      "step": 490500
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00018693036626916525,
+      "loss": 1.2695,
+      "step": 491000
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 0.0001869170570698467,
+      "loss": 1.2479,
+      "step": 491500
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.0001869037478705281,
+      "loss": 1.1694,
+      "step": 492000
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 0.00018689043867120954,
+      "loss": 1.2422,
+      "step": 492500
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 0.00018687712947189098,
+      "loss": 1.2026,
+      "step": 493000
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.00018686382027257242,
+      "loss": 1.2459,
+      "step": 493500
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 0.00018685051107325385,
+      "loss": 1.1844,
+      "step": 494000
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00018683720187393527,
+      "loss": 1.2062,
+      "step": 494500
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.0001868238926746167,
+      "loss": 1.262,
+      "step": 495000
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 0.00018681058347529814,
+      "loss": 1.3093,
+      "step": 495500
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 0.00018679727427597958,
+      "loss": 1.2191,
+      "step": 496000
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.000186783965076661,
+      "loss": 1.2455,
+      "step": 496500
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 0.00018677065587734243,
+      "loss": 1.2894,
+      "step": 497000
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 0.00018675734667802387,
+      "loss": 1.2219,
+      "step": 497500
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.00018674403747870528,
+      "loss": 1.2323,
+      "step": 498000
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 0.00018673072827938672,
+      "loss": 1.2455,
+      "step": 498500
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 0.00018671741908006816,
+      "loss": 1.2717,
+      "step": 499000
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.00018670410988074957,
+      "loss": 1.3192,
+      "step": 499500
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 0.000186690800681431,
+      "loss": 1.1968,
+      "step": 500000
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 0.00018667749148211244,
+      "loss": 1.3435,
+      "step": 500500
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.00018666418228279388,
+      "loss": 1.1948,
+      "step": 501000
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.00018665087308347532,
+      "loss": 1.2478,
+      "step": 501500
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 0.00018663756388415673,
+      "loss": 1.2542,
+      "step": 502000
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 0.00018662425468483817,
+      "loss": 1.2587,
+      "step": 502500
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 0.0001866109454855196,
+      "loss": 1.1819,
+      "step": 503000
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 0.00018659763628620104,
+      "loss": 1.227,
+      "step": 503500
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.00018658432708688248,
+      "loss": 1.2652,
+      "step": 504000
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 0.0001865710178875639,
+      "loss": 1.2934,
+      "step": 504500
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 0.0001865577086882453,
+      "loss": 1.2275,
+      "step": 505000
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.00018654439948892674,
+      "loss": 1.2533,
+      "step": 505500
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 0.00018653109028960818,
+      "loss": 1.283,
+      "step": 506000
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 0.00018651778109028962,
+      "loss": 1.2534,
+      "step": 506500
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 0.00018650447189097106,
+      "loss": 1.2625,
+      "step": 507000
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 0.00018649116269165247,
+      "loss": 1.1477,
+      "step": 507500
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.0001864778534923339,
+      "loss": 1.2436,
+      "step": 508000
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.00018646454429301534,
+      "loss": 1.2876,
+      "step": 508500
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 0.00018645123509369678,
+      "loss": 1.1998,
+      "step": 509000
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 0.0001864379258943782,
+      "loss": 1.2604,
+      "step": 509500
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 0.00018642461669505963,
+      "loss": 1.2464,
+      "step": 510000
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 0.00018641130749574107,
+      "loss": 1.2474,
+      "step": 510500
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 0.0001863979982964225,
+      "loss": 1.2516,
+      "step": 511000
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.00018638468909710395,
+      "loss": 1.1846,
+      "step": 511500
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.00018637137989778536,
+      "loss": 1.2755,
+      "step": 512000
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 0.00018635807069846677,
+      "loss": 1.2636,
+      "step": 512500
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.0001863447614991482,
+      "loss": 1.2519,
+      "step": 513000
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 0.00018633145229982964,
+      "loss": 1.3264,
+      "step": 513500
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 0.00018631814310051108,
+      "loss": 1.2963,
+      "step": 514000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.00018630483390119252,
+      "loss": 1.3338,
+      "step": 514500
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.00018629152470187393,
+      "loss": 1.3028,
+      "step": 515000
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 0.00018627821550255537,
+      "loss": 1.2046,
+      "step": 515500
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.0001862649063032368,
+      "loss": 1.2223,
+      "step": 516000
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 0.00018625159710391825,
+      "loss": 1.257,
+      "step": 516500
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 0.00018623828790459968,
+      "loss": 1.2178,
+      "step": 517000
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 0.0001862249787052811,
+      "loss": 1.149,
+      "step": 517500
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 0.00018621166950596253,
+      "loss": 1.2528,
+      "step": 518000
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 0.00018619836030664397,
+      "loss": 1.2645,
+      "step": 518500
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.00018618505110732538,
+      "loss": 1.2341,
+      "step": 519000
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 0.00018617174190800682,
+      "loss": 1.2593,
+      "step": 519500
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 0.00018615843270868826,
+      "loss": 1.2828,
+      "step": 520000
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 0.00018614512350936967,
+      "loss": 1.2659,
+      "step": 520500
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 0.0001861318143100511,
+      "loss": 1.2177,
+      "step": 521000
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 0.00018611850511073255,
+      "loss": 1.2723,
+      "step": 521500
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00018610519591141398,
+      "loss": 1.2375,
+      "step": 522000
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.0001860918867120954,
+      "loss": 1.1982,
+      "step": 522500
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 0.00018607857751277683,
+      "loss": 1.288,
+      "step": 523000
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.00018606526831345827,
+      "loss": 1.2791,
+      "step": 523500
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 0.0001860519591141397,
+      "loss": 1.2297,
+      "step": 524000
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 0.00018603864991482115,
+      "loss": 1.2586,
+      "step": 524500
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.00018602534071550256,
+      "loss": 1.247,
+      "step": 525000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.000186012031516184,
+      "loss": 1.2443,
+      "step": 525500
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 1.3239233493804932,
+      "eval_runtime": 151.989,
+      "eval_samples_per_second": 61.801,
+      "eval_steps_per_second": 61.801,
+      "step": 525952
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 0.00018599872231686544,
+      "loss": 1.3257,
+      "step": 526000
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.00018598541311754685,
+      "loss": 1.1216,
+      "step": 526500
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 0.00018597210391822829,
+      "loss": 1.0827,
+      "step": 527000
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 0.00018595879471890972,
+      "loss": 1.1344,
+      "step": 527500
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.00018594548551959113,
+      "loss": 1.147,
+      "step": 528000
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 0.00018593217632027257,
+      "loss": 1.1411,
+      "step": 528500
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.000185918867120954,
+      "loss": 1.1783,
+      "step": 529000
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 0.00018590555792163545,
+      "loss": 1.199,
+      "step": 529500
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 0.0001858922487223169,
+      "loss": 1.1345,
+      "step": 530000
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 0.0001858789395229983,
+      "loss": 1.1102,
+      "step": 530500
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 0.00018586563032367974,
+      "loss": 1.19,
+      "step": 531000
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 0.00018585232112436117,
+      "loss": 1.1419,
+      "step": 531500
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 0.0001858390119250426,
+      "loss": 1.1731,
+      "step": 532000
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00018582570272572405,
+      "loss": 1.1318,
+      "step": 532500
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00018581239352640546,
+      "loss": 1.1764,
+      "step": 533000
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 0.00018579908432708687,
+      "loss": 1.1271,
+      "step": 533500
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 0.0001857857751277683,
+      "loss": 1.1208,
+      "step": 534000
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 0.00018577246592844975,
+      "loss": 1.1597,
+      "step": 534500
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 0.0001857591567291312,
+      "loss": 1.169,
+      "step": 535000
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.0001857458475298126,
+      "loss": 1.1914,
+      "step": 535500
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.00018573253833049404,
+      "loss": 1.1697,
+      "step": 536000
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 0.00018571922913117547,
+      "loss": 1.135,
+      "step": 536500
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.0001857059199318569,
+      "loss": 1.1092,
+      "step": 537000
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 0.00018569261073253835,
+      "loss": 1.2167,
+      "step": 537500
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 0.00018567930153321976,
+      "loss": 1.2271,
+      "step": 538000
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.0001856659923339012,
+      "loss": 1.0969,
+      "step": 538500
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 0.00018565268313458264,
+      "loss": 1.1361,
+      "step": 539000
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.00018563937393526408,
+      "loss": 1.1424,
+      "step": 539500
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.00018562606473594551,
+      "loss": 1.1361,
+      "step": 540000
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 0.00018561275553662693,
+      "loss": 1.1466,
+      "step": 540500
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 0.00018559944633730834,
+      "loss": 1.157,
+      "step": 541000
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 0.00018558613713798977,
+      "loss": 1.0685,
+      "step": 541500
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 0.0001855728279386712,
+      "loss": 1.0829,
+      "step": 542000
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00018555951873935265,
+      "loss": 1.1264,
+      "step": 542500
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 0.0001855462095400341,
+      "loss": 1.1427,
+      "step": 543000
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 0.0001855329003407155,
+      "loss": 1.2047,
+      "step": 543500
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 0.00018551959114139694,
+      "loss": 1.0471,
+      "step": 544000
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.00018550628194207838,
+      "loss": 1.2338,
+      "step": 544500
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 0.00018549297274275981,
+      "loss": 1.1665,
+      "step": 545000
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 0.00018547966354344125,
+      "loss": 1.1577,
+      "step": 545500
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00018546635434412266,
+      "loss": 1.2127,
+      "step": 546000
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.0001854530451448041,
+      "loss": 1.1236,
+      "step": 546500
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 0.00018543973594548554,
+      "loss": 1.1113,
+      "step": 547000
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.00018542642674616698,
+      "loss": 1.1382,
+      "step": 547500
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 0.0001854131175468484,
+      "loss": 1.1848,
+      "step": 548000
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 0.0001853998083475298,
+      "loss": 1.2489,
+      "step": 548500
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 0.00018538649914821124,
+      "loss": 1.1431,
+      "step": 549000
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 0.00018537318994889268,
+      "loss": 1.1791,
+      "step": 549500
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.00018535988074957411,
+      "loss": 1.102,
+      "step": 550000
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.00018534657155025555,
+      "loss": 1.1928,
+      "step": 550500
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 0.00018533326235093696,
+      "loss": 1.2256,
+      "step": 551000
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 0.0001853199531516184,
+      "loss": 1.1329,
+      "step": 551500
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00018530664395229984,
+      "loss": 1.1367,
+      "step": 552000
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 0.00018529333475298128,
+      "loss": 1.1584,
+      "step": 552500
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00018528002555366272,
+      "loss": 1.2093,
+      "step": 553000
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 0.00018526671635434413,
+      "loss": 1.1427,
+      "step": 553500
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 0.00018525340715502557,
+      "loss": 1.0839,
+      "step": 554000
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 0.000185240097955707,
+      "loss": 1.1644,
+      "step": 554500
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00018522678875638841,
+      "loss": 1.1239,
+      "step": 555000
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 0.00018521347955706985,
+      "loss": 1.2292,
+      "step": 555500
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 0.0001852001703577513,
+      "loss": 1.1276,
+      "step": 556000
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.0001851868611584327,
+      "loss": 1.1925,
+      "step": 556500
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00018517355195911414,
+      "loss": 1.1839,
+      "step": 557000
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 0.00018516024275979558,
+      "loss": 1.1241,
+      "step": 557500
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 0.00018514693356047702,
+      "loss": 1.1849,
+      "step": 558000
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 0.00018513362436115845,
+      "loss": 1.1465,
+      "step": 558500
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 0.00018512031516183987,
+      "loss": 1.2153,
+      "step": 559000
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 0.0001851070059625213,
+      "loss": 1.2451,
+      "step": 559500
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 0.00018509369676320274,
+      "loss": 1.1751,
+      "step": 560000
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 0.00018508038756388418,
+      "loss": 1.111,
+      "step": 560500
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 0.0001850670783645656,
+      "loss": 1.1836,
+      "step": 561000
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 0.00018505376916524703,
+      "loss": 1.2,
+      "step": 561500
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 0.00018504045996592844,
+      "loss": 1.1223,
+      "step": 562000
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 0.00018502715076660988,
+      "loss": 1.1108,
+      "step": 562500
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 0.00018501384156729132,
+      "loss": 1.2332,
+      "step": 563000
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 0.00018500053236797275,
+      "loss": 1.2405,
+      "step": 563500
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 0.00018498722316865417,
+      "loss": 1.1162,
+      "step": 564000
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 0.0001849739139693356,
+      "loss": 1.1503,
+      "step": 564500
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 0.00018496060477001704,
+      "loss": 1.1573,
+      "step": 565000
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 0.00018494729557069848,
+      "loss": 1.1674,
+      "step": 565500
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 0.00018493398637137992,
+      "loss": 1.2043,
+      "step": 566000
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 0.00018492067717206133,
+      "loss": 1.0957,
+      "step": 566500
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 0.00018490736797274277,
+      "loss": 1.1712,
+      "step": 567000
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 0.0001848940587734242,
+      "loss": 1.2128,
+      "step": 567500
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 0.00018488074957410564,
+      "loss": 1.147,
+      "step": 568000
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 0.00018486744037478708,
+      "loss": 1.2303,
+      "step": 568500
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 0.0001848541311754685,
+      "loss": 1.2177,
+      "step": 569000
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 0.0001848408219761499,
+      "loss": 1.1713,
+      "step": 569500
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 0.00018482751277683134,
+      "loss": 1.2166,
+      "step": 570000
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 0.00018481420357751278,
+      "loss": 1.1373,
+      "step": 570500
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 0.00018480089437819422,
+      "loss": 1.1508,
+      "step": 571000
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 0.00018478758517887566,
+      "loss": 1.2145,
+      "step": 571500
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 0.00018477427597955707,
+      "loss": 1.1326,
+      "step": 572000
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 0.0001847609667802385,
+      "loss": 1.1433,
+      "step": 572500
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 0.00018474765758091994,
+      "loss": 1.2325,
+      "step": 573000
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 0.00018473434838160138,
+      "loss": 1.1889,
+      "step": 573500
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 0.0001847210391822828,
+      "loss": 1.2153,
+      "step": 574000
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 0.00018470772998296423,
+      "loss": 1.254,
+      "step": 574500
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 0.00018469442078364567,
+      "loss": 1.194,
+      "step": 575000
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 0.0001846811115843271,
+      "loss": 1.1518,
+      "step": 575500
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 0.00018466780238500855,
+      "loss": 1.1641,
+      "step": 576000
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 0.00018465449318568996,
+      "loss": 1.1449,
+      "step": 576500
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 0.00018464118398637137,
+      "loss": 1.2422,
+      "step": 577000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 0.0001846278747870528,
+      "loss": 1.1855,
+      "step": 577500
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 0.00018461456558773424,
+      "loss": 1.2294,
+      "step": 578000
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 0.00018460125638841568,
+      "loss": 1.1216,
+      "step": 578500
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 0.00018458794718909712,
+      "loss": 1.1729,
+      "step": 579000
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 0.00018457463798977853,
+      "loss": 1.2155,
+      "step": 579500
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 0.00018456132879045997,
+      "loss": 1.1955,
+      "step": 580000
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 0.0001845480195911414,
+      "loss": 1.1139,
+      "step": 580500
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 0.00018453471039182285,
+      "loss": 1.2305,
+      "step": 581000
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 0.00018452140119250428,
+      "loss": 1.1756,
+      "step": 581500
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 0.0001845080919931857,
+      "loss": 1.121,
+      "step": 582000
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 0.00018449478279386713,
+      "loss": 1.2492,
+      "step": 582500
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 0.00018448147359454857,
+      "loss": 1.1613,
+      "step": 583000
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 0.00018446816439522998,
+      "loss": 1.1859,
+      "step": 583500
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 0.00018445485519591142,
+      "loss": 1.116,
+      "step": 584000
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 0.00018444154599659286,
+      "loss": 1.1553,
+      "step": 584500
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 0.00018442823679727427,
+      "loss": 1.1895,
+      "step": 585000
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 0.0001844149275979557,
+      "loss": 1.2087,
+      "step": 585500
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 0.00018440161839863715,
+      "loss": 1.12,
+      "step": 586000
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 0.00018438830919931858,
+      "loss": 1.1421,
+      "step": 586500
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 0.000184375,
+      "loss": 1.2359,
+      "step": 587000
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 0.00018436169080068143,
+      "loss": 1.2278,
+      "step": 587500
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 0.00018434838160136287,
+      "loss": 1.1655,
+      "step": 588000
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 0.0001843350724020443,
+      "loss": 1.1725,
+      "step": 588500
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 0.00018432176320272575,
+      "loss": 1.1489,
+      "step": 589000
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 0.00018430845400340716,
+      "loss": 1.1555,
+      "step": 589500
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 0.0001842951448040886,
+      "loss": 1.1856,
+      "step": 590000
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 0.00018428183560477004,
+      "loss": 1.1403,
+      "step": 590500
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 0.00018426852640545145,
+      "loss": 1.1823,
+      "step": 591000
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 0.00018425521720613288,
+      "loss": 1.217,
+      "step": 591500
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 0.00018424190800681432,
+      "loss": 1.1805,
+      "step": 592000
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 0.00018422859880749573,
+      "loss": 1.157,
+      "step": 592500
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 0.00018421528960817717,
+      "loss": 1.2218,
+      "step": 593000
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 0.0001842019804088586,
+      "loss": 1.142,
+      "step": 593500
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 0.00018418867120954005,
+      "loss": 1.2264,
+      "step": 594000
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 0.0001841753620102215,
+      "loss": 1.3256,
+      "step": 594500
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 0.0001841620528109029,
+      "loss": 1.27,
+      "step": 595000
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 0.00018414874361158434,
+      "loss": 1.1317,
+      "step": 595500
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 0.00018413543441226577,
+      "loss": 1.2121,
+      "step": 596000
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 0.0001841221252129472,
+      "loss": 1.1319,
+      "step": 596500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 0.00018410881601362865,
+      "loss": 1.233,
+      "step": 597000
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 0.00018409550681431006,
+      "loss": 1.2116,
+      "step": 597500
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 0.00018408219761499147,
+      "loss": 1.2027,
+      "step": 598000
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 0.0001840688884156729,
+      "loss": 1.2243,
+      "step": 598500
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 0.00018405557921635435,
+      "loss": 1.1878,
+      "step": 599000
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 0.0001840422700170358,
+      "loss": 1.261,
+      "step": 599500
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 0.0001840289608177172,
+      "loss": 1.1661,
+      "step": 600000
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 0.00018401565161839864,
+      "loss": 1.1935,
+      "step": 600500
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 0.00018400234241908007,
+      "loss": 1.2027,
+      "step": 601000
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 1.3234707117080688,
+      "eval_runtime": 151.4681,
+      "eval_samples_per_second": 62.013,
+      "eval_steps_per_second": 62.013,
+      "step": 601088
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 0.0001839890332197615,
+      "loss": 1.0817,
+      "step": 601500
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 0.00018397572402044295,
+      "loss": 1.0331,
+      "step": 602000
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 0.00018396241482112436,
+      "loss": 1.0431,
+      "step": 602500
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 0.0001839491056218058,
+      "loss": 1.0799,
+      "step": 603000
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 0.00018393579642248724,
+      "loss": 1.0249,
+      "step": 603500
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 0.00018392248722316868,
+      "loss": 1.0811,
+      "step": 604000
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 0.00018390917802385011,
+      "loss": 1.0711,
+      "step": 604500
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 0.00018389586882453153,
+      "loss": 1.1798,
+      "step": 605000
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 0.00018388255962521294,
+      "loss": 1.0025,
+      "step": 605500
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 0.00018386925042589437,
+      "loss": 1.1001,
+      "step": 606000
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 0.0001838559412265758,
+      "loss": 1.0006,
+      "step": 606500
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 0.00018384263202725725,
+      "loss": 1.0731,
+      "step": 607000
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 0.0001838293228279387,
+      "loss": 1.0955,
+      "step": 607500
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 0.0001838160136286201,
+      "loss": 1.066,
+      "step": 608000
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 0.00018380270442930154,
+      "loss": 1.1392,
+      "step": 608500
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 0.00018378939522998298,
+      "loss": 1.1349,
+      "step": 609000
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 0.00018377608603066441,
+      "loss": 1.0779,
+      "step": 609500
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 0.00018376277683134585,
+      "loss": 1.0935,
+      "step": 610000
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 0.00018374946763202726,
+      "loss": 1.0643,
+      "step": 610500
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 0.0001837361584327087,
+      "loss": 1.1145,
+      "step": 611000
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 0.00018372284923339014,
+      "loss": 1.0131,
+      "step": 611500
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 0.00018370954003407158,
+      "loss": 1.014,
+      "step": 612000
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 0.000183696230834753,
+      "loss": 1.0756,
+      "step": 612500
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 0.0001836829216354344,
+      "loss": 1.1036,
+      "step": 613000
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 0.00018366961243611584,
+      "loss": 1.0628,
+      "step": 613500
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 0.00018365630323679728,
+      "loss": 1.1678,
+      "step": 614000
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 0.00018364299403747871,
+      "loss": 1.1626,
+      "step": 614500
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 0.00018362968483816015,
+      "loss": 1.0559,
+      "step": 615000
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 0.00018361637563884156,
+      "loss": 1.135,
+      "step": 615500
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 0.000183603066439523,
+      "loss": 1.0733,
+      "step": 616000
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 0.00018358975724020444,
+      "loss": 1.1718,
+      "step": 616500
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 0.00018357644804088588,
+      "loss": 1.0518,
+      "step": 617000
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 0.00018356313884156732,
+      "loss": 1.0494,
+      "step": 617500
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 0.00018354982964224873,
+      "loss": 1.1376,
+      "step": 618000
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 0.00018353652044293017,
+      "loss": 1.1724,
+      "step": 618500
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 0.0001835232112436116,
+      "loss": 1.1045,
+      "step": 619000
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 0.00018350990204429301,
+      "loss": 1.1085,
+      "step": 619500
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 0.00018349659284497445,
+      "loss": 1.0354,
+      "step": 620000
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 0.0001834832836456559,
+      "loss": 1.0804,
+      "step": 620500
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 0.0001834699744463373,
+      "loss": 1.1119,
+      "step": 621000
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 0.00018345666524701874,
+      "loss": 1.0813,
+      "step": 621500
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 0.00018344335604770018,
+      "loss": 1.0846,
+      "step": 622000
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 0.00018343004684838162,
+      "loss": 1.2261,
+      "step": 622500
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 0.00018341673764906305,
+      "loss": 1.1004,
+      "step": 623000
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 0.00018340342844974447,
+      "loss": 1.0597,
+      "step": 623500
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 0.0001833901192504259,
+      "loss": 1.0623,
+      "step": 624000
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 0.00018337681005110734,
+      "loss": 1.0565,
+      "step": 624500
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 0.00018336350085178878,
+      "loss": 1.1088,
+      "step": 625000
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 0.0001833501916524702,
+      "loss": 1.054,
+      "step": 625500
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 0.00018333688245315163,
+      "loss": 1.0979,
+      "step": 626000
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 0.00018332357325383304,
+      "loss": 1.0508,
+      "step": 626500
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 0.00018331026405451448,
+      "loss": 1.0952,
+      "step": 627000
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 0.00018329695485519592,
+      "loss": 1.1436,
+      "step": 627500
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 0.00018328364565587735,
+      "loss": 1.1392,
+      "step": 628000
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 0.00018327033645655877,
+      "loss": 1.1703,
+      "step": 628500
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 0.0001832570272572402,
+      "loss": 1.0527,
+      "step": 629000
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 0.00018324371805792164,
+      "loss": 1.0601,
+      "step": 629500
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 0.00018323040885860308,
+      "loss": 1.1344,
+      "step": 630000
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 0.00018321709965928452,
+      "loss": 1.0683,
+      "step": 630500
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 0.00018320379045996593,
+      "loss": 1.0863,
+      "step": 631000
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 0.00018319048126064737,
+      "loss": 1.0342,
+      "step": 631500
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 0.0001831771720613288,
+      "loss": 1.0712,
+      "step": 632000
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 0.00018316386286201024,
+      "loss": 1.1242,
+      "step": 632500
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 0.00018315055366269168,
+      "loss": 1.1044,
+      "step": 633000
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 0.0001831372444633731,
+      "loss": 1.0876,
+      "step": 633500
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 0.0001831239352640545,
+      "loss": 1.1795,
+      "step": 634000
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 0.00018311062606473594,
+      "loss": 1.0908,
+      "step": 634500
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 0.00018309731686541738,
+      "loss": 1.1993,
+      "step": 635000
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 0.00018308400766609882,
+      "loss": 1.243,
+      "step": 635500
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 0.00018307069846678026,
+      "loss": 1.1208,
+      "step": 636000
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 0.00018305738926746167,
+      "loss": 1.1218,
+      "step": 636500
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 0.0001830440800681431,
+      "loss": 1.1829,
+      "step": 637000
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 0.00018303077086882454,
+      "loss": 1.1467,
+      "step": 637500
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 0.00018301746166950598,
+      "loss": 1.0848,
+      "step": 638000
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 0.0001830041524701874,
+      "loss": 1.1364,
+      "step": 638500
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 0.00018299084327086883,
+      "loss": 1.1178,
+      "step": 639000
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 0.00018297753407155027,
+      "loss": 1.118,
+      "step": 639500
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 0.0001829642248722317,
+      "loss": 1.1277,
+      "step": 640000
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 0.00018295091567291315,
+      "loss": 1.1548,
+      "step": 640500
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 0.00018293760647359456,
+      "loss": 1.1264,
+      "step": 641000
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 0.00018292429727427597,
+      "loss": 1.1071,
+      "step": 641500
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 0.0001829109880749574,
+      "loss": 1.0937,
+      "step": 642000
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 0.00018289767887563884,
+      "loss": 1.1876,
+      "step": 642500
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 0.00018288436967632028,
+      "loss": 1.1334,
+      "step": 643000
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 0.00018287106047700172,
+      "loss": 1.2755,
+      "step": 643500
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 0.00018285775127768313,
+      "loss": 1.1289,
+      "step": 644000
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 0.00018284444207836457,
+      "loss": 1.0956,
+      "step": 644500
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 0.000182831132879046,
+      "loss": 1.1023,
+      "step": 645000
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 0.00018281782367972745,
+      "loss": 1.0352,
+      "step": 645500
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 0.00018280451448040888,
+      "loss": 1.0956,
+      "step": 646000
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 0.0001827912052810903,
+      "loss": 1.0834,
+      "step": 646500
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 0.00018277789608177173,
+      "loss": 1.148,
+      "step": 647000
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 0.00018276458688245317,
+      "loss": 1.125,
+      "step": 647500
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 0.00018275127768313458,
+      "loss": 1.0742,
+      "step": 648000
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 0.00018273796848381602,
+      "loss": 1.1637,
+      "step": 648500
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 0.00018272465928449746,
+      "loss": 1.2304,
+      "step": 649000
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 0.00018271135008517887,
+      "loss": 1.1194,
+      "step": 649500
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 0.0001826980408858603,
+      "loss": 1.0773,
+      "step": 650000
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 0.00018268473168654175,
+      "loss": 1.0997,
+      "step": 650500
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 0.00018267142248722318,
+      "loss": 1.1384,
+      "step": 651000
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 0.0001826581132879046,
+      "loss": 1.0972,
+      "step": 651500
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 0.00018264480408858603,
+      "loss": 1.1346,
+      "step": 652000
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 0.00018263149488926747,
+      "loss": 1.1175,
+      "step": 652500
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 0.0001826181856899489,
+      "loss": 1.0364,
+      "step": 653000
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 0.00018260487649063035,
+      "loss": 1.1752,
+      "step": 653500
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 0.00018259156729131176,
+      "loss": 1.0964,
+      "step": 654000
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 0.0001825782580919932,
+      "loss": 1.0879,
+      "step": 654500
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 0.00018256494889267464,
+      "loss": 1.1167,
+      "step": 655000
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 0.00018255163969335605,
+      "loss": 1.1088,
+      "step": 655500
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 0.00018253833049403748,
+      "loss": 1.1252,
+      "step": 656000
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 0.00018252502129471892,
+      "loss": 1.1303,
+      "step": 656500
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 0.00018251171209540033,
+      "loss": 1.0708,
+      "step": 657000
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 0.00018249840289608177,
+      "loss": 1.1427,
+      "step": 657500
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 0.0001824850936967632,
+      "loss": 1.0791,
+      "step": 658000
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 0.00018247178449744465,
+      "loss": 1.0627,
+      "step": 658500
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 0.0001824584752981261,
+      "loss": 1.1436,
+      "step": 659000
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 0.0001824451660988075,
+      "loss": 1.1278,
+      "step": 659500
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 0.00018243185689948894,
+      "loss": 1.1135,
+      "step": 660000
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 0.00018241854770017037,
+      "loss": 1.1699,
+      "step": 660500
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 0.0001824052385008518,
+      "loss": 1.0952,
+      "step": 661000
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 0.00018239192930153325,
+      "loss": 1.1151,
+      "step": 661500
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 0.00018237862010221466,
+      "loss": 1.1177,
+      "step": 662000
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 0.00018236531090289607,
+      "loss": 1.0539,
+      "step": 662500
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 0.0001823520017035775,
+      "loss": 1.244,
+      "step": 663000
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 0.00018233869250425895,
+      "loss": 1.1707,
+      "step": 663500
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 0.0001823253833049404,
+      "loss": 1.1065,
+      "step": 664000
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 0.0001823120741056218,
+      "loss": 1.1775,
+      "step": 664500
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 0.00018229876490630324,
+      "loss": 1.1488,
+      "step": 665000
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 0.00018228545570698467,
+      "loss": 1.1673,
+      "step": 665500
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 0.0001822721465076661,
+      "loss": 1.1918,
+      "step": 666000
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 0.00018225883730834755,
+      "loss": 1.1079,
+      "step": 666500
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 0.00018224552810902896,
+      "loss": 1.1284,
+      "step": 667000
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 0.0001822322189097104,
+      "loss": 1.2182,
+      "step": 667500
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 0.00018221890971039184,
+      "loss": 1.1137,
+      "step": 668000
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 0.00018220560051107328,
+      "loss": 1.216,
+      "step": 668500
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 0.00018219229131175471,
+      "loss": 1.0904,
+      "step": 669000
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 0.00018217898211243613,
+      "loss": 1.2131,
+      "step": 669500
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 0.00018216567291311754,
+      "loss": 1.1682,
+      "step": 670000
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 0.00018215236371379897,
+      "loss": 1.0603,
+      "step": 670500
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 0.0001821390545144804,
+      "loss": 1.0959,
+      "step": 671000
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 0.00018212574531516185,
+      "loss": 1.2124,
+      "step": 671500
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 0.0001821124361158433,
+      "loss": 1.0974,
+      "step": 672000
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 0.0001820991269165247,
+      "loss": 1.1375,
+      "step": 672500
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 0.00018208581771720614,
+      "loss": 1.1494,
+      "step": 673000
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 0.00018207250851788758,
+      "loss": 1.0756,
+      "step": 673500
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 0.00018205919931856901,
+      "loss": 1.139,
+      "step": 674000
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 0.00018204589011925045,
+      "loss": 1.0752,
+      "step": 674500
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 0.00018203258091993186,
+      "loss": 1.192,
+      "step": 675000
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 0.0001820192717206133,
+      "loss": 1.1568,
+      "step": 675500
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 0.00018200596252129474,
+      "loss": 1.2032,
+      "step": 676000
+    },
+    {
+      "epoch": 9.0,
+      "eval_loss": 1.3050340414047241,
+      "eval_runtime": 151.9979,
+      "eval_samples_per_second": 61.797,
+      "eval_steps_per_second": 61.797,
+      "step": 676224
+    }
+  ],
+  "max_steps": 7513600,
+  "num_train_epochs": 100,
+  "total_flos": 6240548899325952.0,
+  "trial_name": null,
+  "trial_params": null
+}