diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5224 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 39.99994211019902,
+  "eval_steps": 500,
+  "global_step": 1727400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.6310589688175354e-05,
+      "loss": 4.5055,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 9.262117937635071e-05,
+      "loss": 1.4851,
+      "step": 4000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00013893176906452608,
+      "loss": 1.1808,
+      "step": 6000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00018524235875270141,
+      "loss": 1.132,
+      "step": 8000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0002315529484408768,
+      "loss": 1.1349,
+      "step": 10000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00027786353812905216,
+      "loss": 1.1422,
+      "step": 12000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0002997558073716234,
+      "loss": 1.1513,
+      "step": 14000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002992880054015534,
+      "loss": 1.142,
+      "step": 16000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00029882020343148334,
+      "loss": 1.1303,
+      "step": 18000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002983524014614133,
+      "loss": 1.1215,
+      "step": 20000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002978845994913433,
+      "loss": 1.109,
+      "step": 22000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00029741679752127325,
+      "loss": 1.1034,
+      "step": 24000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0002969489955512032,
+      "loss": 1.0998,
+      "step": 26000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00029648119358113324,
+      "loss": 1.0979,
+      "step": 28000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 0.00029601339161106316,
+      "loss": 1.083,
+      "step": 30000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 0.0002955455896409932,
+      "loss": 1.0741,
+      "step": 32000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 0.0002950777876709231,
+      "loss": 1.0741,
+      "step": 34000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 0.0002946099857008531,
+      "loss": 1.0707,
+      "step": 36000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 0.00029414218373078303,
+      "loss": 1.0689,
+      "step": 38000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 0.00029367438176071306,
+      "loss": 1.0536,
+      "step": 40000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 0.00029320657979064297,
+      "loss": 1.0556,
+      "step": 42000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.000292738777820573,
+      "loss": 1.0358,
+      "step": 44000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 0.0002922709758505029,
+      "loss": 1.018,
+      "step": 46000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 0.00029180317388043293,
+      "loss": 1.0162,
+      "step": 48000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 0.00029133537191036285,
+      "loss": 1.017,
+      "step": 50000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 0.00029086756994029287,
+      "loss": 1.0145,
+      "step": 52000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 0.0002903997679702228,
+      "loss": 1.0103,
+      "step": 54000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 0.0002899319660001528,
+      "loss": 1.0046,
+      "step": 56000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 0.0002894641640300827,
+      "loss": 1.01,
+      "step": 58000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 0.00028899636206001274,
+      "loss": 1.0146,
+      "step": 60000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 0.00028852856008994266,
+      "loss": 1.0122,
+      "step": 62000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 0.0002880607581198727,
+      "loss": 0.9961,
+      "step": 64000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 0.00028759295614980265,
+      "loss": 0.9914,
+      "step": 66000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 0.0002871251541797326,
+      "loss": 0.9936,
+      "step": 68000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 0.0002866573522096626,
+      "loss": 0.9998,
+      "step": 70000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 0.00028618955023959256,
+      "loss": 0.9833,
+      "step": 72000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 0.0002857217482695225,
+      "loss": 0.9901,
+      "step": 74000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 0.0002852539462994525,
+      "loss": 0.9905,
+      "step": 76000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 0.00028478614432938246,
+      "loss": 0.9845,
+      "step": 78000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 0.00028431834235931243,
+      "loss": 0.9886,
+      "step": 80000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 0.0002838505403892424,
+      "loss": 0.9848,
+      "step": 82000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 0.00028338273841917237,
+      "loss": 0.9827,
+      "step": 84000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 0.00028291493644910234,
+      "loss": 0.9764,
+      "step": 86000
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 0.0002824471344790323,
+      "loss": 0.9546,
+      "step": 88000
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 0.0002819793325089623,
+      "loss": 0.9388,
+      "step": 90000
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 0.00028151153053889225,
+      "loss": 0.9432,
+      "step": 92000
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 0.0002810437285688222,
+      "loss": 0.9591,
+      "step": 94000
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 0.0002805759265987522,
+      "loss": 0.9431,
+      "step": 96000
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 0.00028010812462868215,
+      "loss": 0.9388,
+      "step": 98000
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 0.0002796403226586121,
+      "loss": 0.9421,
+      "step": 100000
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 0.0002791725206885421,
+      "loss": 0.9452,
+      "step": 102000
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 0.0002787047187184721,
+      "loss": 0.9535,
+      "step": 104000
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 0.00027823691674840203,
+      "loss": 0.9373,
+      "step": 106000
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 0.00027776911477833205,
+      "loss": 0.9345,
+      "step": 108000
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 0.00027730131280826197,
+      "loss": 0.9478,
+      "step": 110000
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 0.000276833510838192,
+      "loss": 0.9445,
+      "step": 112000
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 0.0002763657088681219,
+      "loss": 0.9297,
+      "step": 114000
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 0.0002758979068980519,
+      "loss": 0.9294,
+      "step": 116000
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 0.00027543010492798184,
+      "loss": 0.9273,
+      "step": 118000
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 0.00027496230295791186,
+      "loss": 0.9242,
+      "step": 120000
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 0.0002744945009878418,
+      "loss": 0.9228,
+      "step": 122000
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 0.0002740266990177718,
+      "loss": 0.9316,
+      "step": 124000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 0.0002735588970477017,
+      "loss": 0.9448,
+      "step": 126000
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 0.00027309109507763174,
+      "loss": 0.9253,
+      "step": 128000
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 0.00027262329310756165,
+      "loss": 0.9197,
+      "step": 130000
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.0002721554911374917,
+      "loss": 0.9025,
+      "step": 132000
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.0002716876891674216,
+      "loss": 0.8983,
+      "step": 134000
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.0002712198871973516,
+      "loss": 0.9013,
+      "step": 136000
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.0002707520852272816,
+      "loss": 0.9019,
+      "step": 138000
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00027028428325721155,
+      "loss": 0.9016,
+      "step": 140000
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.0002698164812871415,
+      "loss": 0.8979,
+      "step": 142000
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.0002693486793170715,
+      "loss": 0.8961,
+      "step": 144000
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.00026888087734700146,
+      "loss": 0.9007,
+      "step": 146000
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00026841307537693143,
+      "loss": 0.8951,
+      "step": 148000
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.0002679452734068614,
+      "loss": 0.8926,
+      "step": 150000
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00026747747143679137,
+      "loss": 0.8924,
+      "step": 152000
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00026700966946672134,
+      "loss": 0.9044,
+      "step": 154000
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.0002665418674966513,
+      "loss": 0.8952,
+      "step": 156000
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.00026607406552658127,
+      "loss": 0.9001,
+      "step": 158000
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00026560626355651124,
+      "loss": 0.8898,
+      "step": 160000
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.0002651384615864412,
+      "loss": 0.895,
+      "step": 162000
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.0002646706596163712,
+      "loss": 0.9015,
+      "step": 164000
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.00026420285764630115,
+      "loss": 0.892,
+      "step": 166000
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.0002637350556762311,
+      "loss": 0.8903,
+      "step": 168000
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.0002632672537061611,
+      "loss": 0.8916,
+      "step": 170000
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00026279945173609105,
+      "loss": 0.8941,
+      "step": 172000
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.000262331649766021,
+      "loss": 0.8771,
+      "step": 174000
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 0.000261863847795951,
+      "loss": 0.8716,
+      "step": 176000
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 0.00026139604582588096,
+      "loss": 0.8632,
+      "step": 178000
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.00026092824385581093,
+      "loss": 0.8573,
+      "step": 180000
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.0002604604418857409,
+      "loss": 0.8642,
+      "step": 182000
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 0.00025999263991567087,
+      "loss": 0.8642,
+      "step": 184000
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00025952483794560084,
+      "loss": 0.8617,
+      "step": 186000
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 0.0002590570359755308,
+      "loss": 0.8574,
+      "step": 188000
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.0002585892340054608,
+      "loss": 0.8612,
+      "step": 190000
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.00025812143203539074,
+      "loss": 0.8706,
+      "step": 192000
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.0002576536300653207,
+      "loss": 0.8605,
+      "step": 194000
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.0002571858280952507,
+      "loss": 0.8703,
+      "step": 196000
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 0.00025671802612518065,
+      "loss": 0.8691,
+      "step": 198000
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 0.0002562502241551106,
+      "loss": 0.873,
+      "step": 200000
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.0002557824221850406,
+      "loss": 0.857,
+      "step": 202000
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.00025531462021497056,
+      "loss": 0.8686,
+      "step": 204000
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.0002548468182449005,
+      "loss": 0.868,
+      "step": 206000
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.0002543790162748305,
+      "loss": 0.8756,
+      "step": 208000
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.0002539112143047605,
+      "loss": 0.8656,
+      "step": 210000
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 0.00025344341233469043,
+      "loss": 0.8725,
+      "step": 212000
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.00025297561036462045,
+      "loss": 0.8675,
+      "step": 214000
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.00025250780839455037,
+      "loss": 0.8644,
+      "step": 216000
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 0.0002520400064244804,
+      "loss": 0.8397,
+      "step": 218000
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 0.0002515722044544103,
+      "loss": 0.8374,
+      "step": 220000
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 0.00025110440248434033,
+      "loss": 0.8364,
+      "step": 222000
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 0.00025063660051427025,
+      "loss": 0.8454,
+      "step": 224000
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 0.00025016879854420027,
+      "loss": 0.8367,
+      "step": 226000
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 0.0002497009965741302,
+      "loss": 0.8275,
+      "step": 228000
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 0.0002492331946040602,
+      "loss": 0.8398,
+      "step": 230000
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 0.0002487653926339901,
+      "loss": 0.8403,
+      "step": 232000
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 0.00024829759066392014,
+      "loss": 0.8409,
+      "step": 234000
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 0.00024782978869385006,
+      "loss": 0.8366,
+      "step": 236000
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 0.0002473619867237801,
+      "loss": 0.8381,
+      "step": 238000
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 0.00024689418475371,
+      "loss": 0.842,
+      "step": 240000
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 0.00024642638278364,
+      "loss": 0.843,
+      "step": 242000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 0.00024595858081357,
+      "loss": 0.8489,
+      "step": 244000
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 0.00024549077884349996,
+      "loss": 0.8313,
+      "step": 246000
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 0.0002450229768734299,
+      "loss": 0.8468,
+      "step": 248000
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 0.0002445551749033599,
+      "loss": 0.8446,
+      "step": 250000
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 0.00024408737293328986,
+      "loss": 0.8329,
+      "step": 252000
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 0.00024361957096321983,
+      "loss": 0.8402,
+      "step": 254000
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 0.0002431517689931498,
+      "loss": 0.8502,
+      "step": 256000
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 0.00024268396702307977,
+      "loss": 0.8437,
+      "step": 258000
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 0.00024221616505300974,
+      "loss": 0.8341,
+      "step": 260000
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 0.0002417483630829397,
+      "loss": 0.8152,
+      "step": 262000
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 0.00024128056111286968,
+      "loss": 0.8189,
+      "step": 264000
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 0.00024081275914279965,
+      "loss": 0.8193,
+      "step": 266000
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 0.00024034495717272961,
+      "loss": 0.825,
+      "step": 268000
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 0.00023987715520265958,
+      "loss": 0.818,
+      "step": 270000
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 0.00023940935323258955,
+      "loss": 0.8204,
+      "step": 272000
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 0.00023894155126251952,
+      "loss": 0.823,
+      "step": 274000
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 0.0002384737492924495,
+      "loss": 0.8179,
+      "step": 276000
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 0.00023800594732237946,
+      "loss": 0.8152,
+      "step": 278000
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 0.00023753814535230943,
+      "loss": 0.8178,
+      "step": 280000
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 0.00023707034338223942,
+      "loss": 0.8212,
+      "step": 282000
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 0.00023660254141216937,
+      "loss": 0.8161,
+      "step": 284000
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 0.00023613473944209936,
+      "loss": 0.8162,
+      "step": 286000
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 0.0002356669374720293,
+      "loss": 0.828,
+      "step": 288000
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 0.0002351991355019593,
+      "loss": 0.823,
+      "step": 290000
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 0.00023473133353188927,
+      "loss": 0.8156,
+      "step": 292000
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 0.00023426353156181924,
+      "loss": 0.8176,
+      "step": 294000
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 0.0002337957295917492,
+      "loss": 0.8226,
+      "step": 296000
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 0.00023332792762167917,
+      "loss": 0.8189,
+      "step": 298000
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 0.00023286012565160914,
+      "loss": 0.8082,
+      "step": 300000
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 0.0002323923236815391,
+      "loss": 0.8257,
+      "step": 302000
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 0.00023192452171146908,
+      "loss": 0.7925,
+      "step": 304000
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 0.00023145671974139905,
+      "loss": 0.7902,
+      "step": 306000
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 0.00023098891777132902,
+      "loss": 0.7994,
+      "step": 308000
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 0.000230521115801259,
+      "loss": 0.8029,
+      "step": 310000
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 0.00023005331383118896,
+      "loss": 0.7983,
+      "step": 312000
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 0.00022958551186111893,
+      "loss": 0.797,
+      "step": 314000
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 0.00022911770989104892,
+      "loss": 0.8009,
+      "step": 316000
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 0.00022864990792097886,
+      "loss": 0.7979,
+      "step": 318000
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 0.00022818210595090886,
+      "loss": 0.8055,
+      "step": 320000
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 0.0002277143039808388,
+      "loss": 0.7971,
+      "step": 322000
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 0.0002272465020107688,
+      "loss": 0.8066,
+      "step": 324000
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 0.00022677870004069874,
+      "loss": 0.7975,
+      "step": 326000
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 0.00022631089807062873,
+      "loss": 0.7912,
+      "step": 328000
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 0.00022584309610055868,
+      "loss": 0.7988,
+      "step": 330000
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 0.00022537529413048867,
+      "loss": 0.7999,
+      "step": 332000
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 0.00022490749216041861,
+      "loss": 0.8019,
+      "step": 334000
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 0.0002244396901903486,
+      "loss": 0.8108,
+      "step": 336000
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 0.00022397188822027855,
+      "loss": 0.8075,
+      "step": 338000
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 0.00022350408625020855,
+      "loss": 0.7995,
+      "step": 340000
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 0.0002230362842801385,
+      "loss": 0.802,
+      "step": 342000
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 0.00022256848231006848,
+      "loss": 0.7998,
+      "step": 344000
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 0.00022210068033999843,
+      "loss": 0.8023,
+      "step": 346000
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 0.00022163287836992842,
+      "loss": 0.7748,
+      "step": 348000
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 0.00022116507639985836,
+      "loss": 0.7839,
+      "step": 350000
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 0.00022069727442978836,
+      "loss": 0.7806,
+      "step": 352000
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 0.00022022947245971836,
+      "loss": 0.7775,
+      "step": 354000
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 0.0002197616704896483,
+      "loss": 0.7734,
+      "step": 356000
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 0.0002192938685195783,
+      "loss": 0.7728,
+      "step": 358000
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 0.00021882606654950824,
+      "loss": 0.7879,
+      "step": 360000
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 0.00021835826457943823,
+      "loss": 0.7891,
+      "step": 362000
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 0.00021789046260936817,
+      "loss": 0.7922,
+      "step": 364000
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 0.00021742266063929817,
+      "loss": 0.7837,
+      "step": 366000
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 0.0002169548586692281,
+      "loss": 0.7838,
+      "step": 368000
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 0.0002164870566991581,
+      "loss": 0.7797,
+      "step": 370000
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 0.00021601925472908805,
+      "loss": 0.7818,
+      "step": 372000
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 0.00021555145275901804,
+      "loss": 0.7838,
+      "step": 374000
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 0.000215083650788948,
+      "loss": 0.7828,
+      "step": 376000
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 0.00021461584881887798,
+      "loss": 0.7822,
+      "step": 378000
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 0.00021414804684880792,
+      "loss": 0.7952,
+      "step": 380000
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 0.00021368024487873792,
+      "loss": 0.7888,
+      "step": 382000
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 0.00021321244290866786,
+      "loss": 0.7813,
+      "step": 384000
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 0.00021274464093859786,
+      "loss": 0.7784,
+      "step": 386000
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 0.00021227683896852783,
+      "loss": 0.776,
+      "step": 388000
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 0.0002118090369984578,
+      "loss": 0.7686,
+      "step": 390000
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 0.0002113412350283878,
+      "loss": 0.7576,
+      "step": 392000
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 0.00021087343305831773,
+      "loss": 0.7611,
+      "step": 394000
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 0.00021040563108824773,
+      "loss": 0.7698,
+      "step": 396000
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 0.00020993782911817767,
+      "loss": 0.7632,
+      "step": 398000
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 0.00020947002714810767,
+      "loss": 0.7725,
+      "step": 400000
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 0.0002090022251780376,
+      "loss": 0.7706,
+      "step": 402000
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 0.0002085344232079676,
+      "loss": 0.7709,
+      "step": 404000
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 0.00020806662123789755,
+      "loss": 0.7651,
+      "step": 406000
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 0.00020759881926782754,
+      "loss": 0.7657,
+      "step": 408000
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 0.00020713101729775748,
+      "loss": 0.7589,
+      "step": 410000
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 0.00020666321532768748,
+      "loss": 0.7683,
+      "step": 412000
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 0.00020619541335761742,
+      "loss": 0.7684,
+      "step": 414000
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 0.00020572761138754742,
+      "loss": 0.7756,
+      "step": 416000
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 0.00020525980941747736,
+      "loss": 0.7653,
+      "step": 418000
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 0.00020479200744740736,
+      "loss": 0.7718,
+      "step": 420000
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 0.0002043242054773373,
+      "loss": 0.7676,
+      "step": 422000
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 0.0002038564035072673,
+      "loss": 0.772,
+      "step": 424000
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 0.00020338860153719726,
+      "loss": 0.766,
+      "step": 426000
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 0.00020292079956712723,
+      "loss": 0.7739,
+      "step": 428000
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 0.0002024529975970572,
+      "loss": 0.7743,
+      "step": 430000
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.00020198519562698717,
+      "loss": 0.7719,
+      "step": 432000
+    },
+    {
+      "epoch": 10.05,
+      "learning_rate": 0.00020151739365691714,
+      "loss": 0.7443,
+      "step": 434000
+    },
+    {
+      "epoch": 10.1,
+      "learning_rate": 0.0002010495916868471,
+      "loss": 0.7573,
+      "step": 436000
+    },
+    {
+      "epoch": 10.14,
+      "learning_rate": 0.00020058178971677708,
+      "loss": 0.7546,
+      "step": 438000
+    },
+    {
+      "epoch": 10.19,
+      "learning_rate": 0.00020011398774670704,
+      "loss": 0.7516,
+      "step": 440000
+    },
+    {
+      "epoch": 10.23,
+      "learning_rate": 0.000199646185776637,
+      "loss": 0.7444,
+      "step": 442000
+    },
+    {
+      "epoch": 10.28,
+      "learning_rate": 0.00019917838380656698,
+      "loss": 0.7656,
+      "step": 444000
+    },
+    {
+      "epoch": 10.33,
+      "learning_rate": 0.00019871058183649695,
+      "loss": 0.7452,
+      "step": 446000
+    },
+    {
+      "epoch": 10.37,
+      "learning_rate": 0.00019824277986642692,
+      "loss": 0.7555,
+      "step": 448000
+    },
+    {
+      "epoch": 10.42,
+      "learning_rate": 0.0001977749778963569,
+      "loss": 0.7486,
+      "step": 450000
+    },
+    {
+      "epoch": 10.47,
+      "learning_rate": 0.00019730717592628686,
+      "loss": 0.7509,
+      "step": 452000
+    },
+    {
+      "epoch": 10.51,
+      "learning_rate": 0.00019683937395621683,
+      "loss": 0.7484,
+      "step": 454000
+    },
+    {
+      "epoch": 10.56,
+      "learning_rate": 0.0001963715719861468,
+      "loss": 0.7554,
+      "step": 456000
+    },
+    {
+      "epoch": 10.61,
+      "learning_rate": 0.00019590377001607676,
+      "loss": 0.7557,
+      "step": 458000
+    },
+    {
+      "epoch": 10.65,
+      "learning_rate": 0.00019543596804600676,
+      "loss": 0.7603,
+      "step": 460000
+    },
+    {
+      "epoch": 10.7,
+      "learning_rate": 0.00019496816607593673,
+      "loss": 0.7577,
+      "step": 462000
+    },
+    {
+      "epoch": 10.74,
+      "learning_rate": 0.0001945003641058667,
+      "loss": 0.7641,
+      "step": 464000
+    },
+    {
+      "epoch": 10.79,
+      "learning_rate": 0.00019403256213579667,
+      "loss": 0.7648,
+      "step": 466000
+    },
+    {
+      "epoch": 10.84,
+      "learning_rate": 0.00019356476016572664,
+      "loss": 0.755,
+      "step": 468000
+    },
+    {
+      "epoch": 10.88,
+      "learning_rate": 0.0001930969581956566,
+      "loss": 0.7445,
+      "step": 470000
+    },
+    {
+      "epoch": 10.93,
+      "learning_rate": 0.00019262915622558657,
+      "loss": 0.7614,
+      "step": 472000
+    },
+    {
+      "epoch": 10.98,
+      "learning_rate": 0.00019216135425551654,
+      "loss": 0.7526,
+      "step": 474000
+    },
+    {
+      "epoch": 11.02,
+      "learning_rate": 0.0001916935522854465,
+      "loss": 0.7493,
+      "step": 476000
+    },
+    {
+      "epoch": 11.07,
+      "learning_rate": 0.00019122575031537648,
+      "loss": 0.7299,
+      "step": 478000
+    },
+    {
+      "epoch": 11.11,
+      "learning_rate": 0.00019075794834530645,
+      "loss": 0.7379,
+      "step": 480000
+    },
+    {
+      "epoch": 11.16,
+      "learning_rate": 0.00019029014637523642,
+      "loss": 0.7365,
+      "step": 482000
+    },
+    {
+      "epoch": 11.21,
+      "learning_rate": 0.00018982234440516639,
+      "loss": 0.7402,
+      "step": 484000
+    },
+    {
+      "epoch": 11.25,
+      "learning_rate": 0.00018935454243509636,
+      "loss": 0.7409,
+      "step": 486000
+    },
+    {
+      "epoch": 11.3,
+      "learning_rate": 0.00018888674046502632,
+      "loss": 0.7294,
+      "step": 488000
+    },
+    {
+      "epoch": 11.35,
+      "learning_rate": 0.0001884189384949563,
+      "loss": 0.7467,
+      "step": 490000
+    },
+    {
+      "epoch": 11.39,
+      "learning_rate": 0.00018795113652488626,
+      "loss": 0.7357,
+      "step": 492000
+    },
+    {
+      "epoch": 11.44,
+      "learning_rate": 0.00018748333455481623,
+      "loss": 0.744,
+      "step": 494000
+    },
+    {
+      "epoch": 11.49,
+      "learning_rate": 0.0001870155325847462,
+      "loss": 0.741,
+      "step": 496000
+    },
+    {
+      "epoch": 11.53,
+      "learning_rate": 0.0001865477306146762,
+      "loss": 0.7404,
+      "step": 498000
+    },
+    {
+      "epoch": 11.58,
+      "learning_rate": 0.00018607992864460614,
+      "loss": 0.749,
+      "step": 500000
+    },
+    {
+      "epoch": 11.62,
+      "learning_rate": 0.00018561212667453613,
+      "loss": 0.7388,
+      "step": 502000
+    },
+    {
+      "epoch": 11.67,
+      "learning_rate": 0.00018514432470446607,
+      "loss": 0.742,
+      "step": 504000
+    },
+    {
+      "epoch": 11.72,
+      "learning_rate": 0.00018467652273439607,
+      "loss": 0.7481,
+      "step": 506000
+    },
+    {
+      "epoch": 11.76,
+      "learning_rate": 0.000184208720764326,
+      "loss": 0.7553,
+      "step": 508000
+    },
+    {
+      "epoch": 11.81,
+      "learning_rate": 0.000183740918794256,
+      "loss": 0.7457,
+      "step": 510000
+    },
+    {
+      "epoch": 11.86,
+      "learning_rate": 0.00018327311682418595,
+      "loss": 0.7447,
+      "step": 512000
+    },
+    {
+      "epoch": 11.9,
+      "learning_rate": 0.00018280531485411595,
+      "loss": 0.752,
+      "step": 514000
+    },
+    {
+      "epoch": 11.95,
+      "learning_rate": 0.0001823375128840459,
+      "loss": 0.7419,
+      "step": 516000
+    },
+    {
+      "epoch": 11.99,
+      "learning_rate": 0.00018186971091397588,
+      "loss": 0.7412,
+      "step": 518000
+    },
+    {
+      "epoch": 12.04,
+      "learning_rate": 0.00018140190894390583,
+      "loss": 0.7262,
+      "step": 520000
+    },
+    {
+      "epoch": 12.09,
+      "learning_rate": 0.00018093410697383582,
+      "loss": 0.7299,
+      "step": 522000
+    },
+    {
+      "epoch": 12.13,
+      "learning_rate": 0.00018046630500376576,
+      "loss": 0.7279,
+      "step": 524000
+    },
+    {
+      "epoch": 12.18,
+      "learning_rate": 0.00017999850303369576,
+      "loss": 0.7308,
+      "step": 526000
+    },
+    {
+      "epoch": 12.23,
+      "learning_rate": 0.0001795307010636257,
+      "loss": 0.7305,
+      "step": 528000
+    },
+    {
+      "epoch": 12.27,
+      "learning_rate": 0.0001790628990935557,
+      "loss": 0.7348,
+      "step": 530000
+    },
+    {
+      "epoch": 12.32,
+      "learning_rate": 0.0001785950971234857,
+      "loss": 0.7312,
+      "step": 532000
+    },
+    {
+      "epoch": 12.37,
+      "learning_rate": 0.00017812729515341563,
+      "loss": 0.7275,
+      "step": 534000
+    },
+    {
+      "epoch": 12.41,
+      "learning_rate": 0.00017765949318334563,
+      "loss": 0.7291,
+      "step": 536000
+    },
+    {
+      "epoch": 12.46,
+      "learning_rate": 0.00017719169121327557,
+      "loss": 0.7265,
+      "step": 538000
+    },
+    {
+      "epoch": 12.5,
+      "learning_rate": 0.00017672388924320557,
+      "loss": 0.7224,
+      "step": 540000
+    },
+    {
+      "epoch": 12.55,
+      "learning_rate": 0.0001762560872731355,
+      "loss": 0.7232,
+      "step": 542000
+    },
+    {
+      "epoch": 12.6,
+      "learning_rate": 0.0001757882853030655,
+      "loss": 0.7272,
+      "step": 544000
+    },
+    {
+      "epoch": 12.64,
+      "learning_rate": 0.00017532048333299545,
+      "loss": 0.7305,
+      "step": 546000
+    },
+    {
+      "epoch": 12.69,
+      "learning_rate": 0.00017485268136292544,
+      "loss": 0.7375,
+      "step": 548000
+    },
+    {
+      "epoch": 12.74,
+      "learning_rate": 0.00017438487939285539,
+      "loss": 0.7377,
+      "step": 550000
+    },
+    {
+      "epoch": 12.78,
+      "learning_rate": 0.00017391707742278538,
+      "loss": 0.7278,
+      "step": 552000
+    },
+    {
+      "epoch": 12.83,
+      "learning_rate": 0.00017344927545271532,
+      "loss": 0.7369,
+      "step": 554000
+    },
+    {
+      "epoch": 12.87,
+      "learning_rate": 0.00017298147348264532,
+      "loss": 0.7366,
+      "step": 556000
+    },
+    {
+      "epoch": 12.92,
+      "learning_rate": 0.00017251367151257526,
+      "loss": 0.736,
+      "step": 558000
+    },
+    {
+      "epoch": 12.97,
+      "learning_rate": 0.00017204586954250526,
+      "loss": 0.737,
+      "step": 560000
+    },
+    {
+      "epoch": 13.01,
+      "learning_rate": 0.0001715780675724352,
+      "loss": 0.7301,
+      "step": 562000
+    },
+    {
+      "epoch": 13.06,
+      "learning_rate": 0.0001711102656023652,
+      "loss": 0.7003,
+      "step": 564000
+    },
+    {
+      "epoch": 13.11,
+      "learning_rate": 0.00017064246363229514,
+      "loss": 0.7132,
+      "step": 566000
+    },
+    {
+      "epoch": 13.15,
+      "learning_rate": 0.00017017466166222513,
+      "loss": 0.7178,
+      "step": 568000
+    },
+    {
+      "epoch": 13.2,
+      "learning_rate": 0.00016970685969215513,
+      "loss": 0.7187,
+      "step": 570000
+    },
+    {
+      "epoch": 13.25,
+      "learning_rate": 0.00016923905772208507,
+      "loss": 0.7239,
+      "step": 572000
+    },
+    {
+      "epoch": 13.29,
+      "learning_rate": 0.00016877125575201507,
+      "loss": 0.7271,
+      "step": 574000
+    },
+    {
+      "epoch": 13.34,
+      "learning_rate": 0.000168303453781945,
+      "loss": 0.7208,
+      "step": 576000
+    },
+    {
+      "epoch": 13.38,
+      "learning_rate": 0.000167835651811875,
+      "loss": 0.7199,
+      "step": 578000
+    },
+    {
+      "epoch": 13.43,
+      "learning_rate": 0.00016736784984180495,
+      "loss": 0.7094,
+      "step": 580000
+    },
+    {
+      "epoch": 13.48,
+      "learning_rate": 0.00016690004787173494,
+      "loss": 0.7114,
+      "step": 582000
+    },
+    {
+      "epoch": 13.52,
+      "learning_rate": 0.00016643224590166488,
+      "loss": 0.7196,
+      "step": 584000
+    },
+    {
+      "epoch": 13.57,
+      "learning_rate": 0.00016596444393159488,
+      "loss": 0.7222,
+      "step": 586000
+    },
+    {
+      "epoch": 13.62,
+      "learning_rate": 0.00016549664196152482,
+      "loss": 0.7345,
+      "step": 588000
+    },
+    {
+      "epoch": 13.66,
+      "learning_rate": 0.00016502883999145482,
+      "loss": 0.7208,
+      "step": 590000
+    },
+    {
+      "epoch": 13.71,
+      "learning_rate": 0.00016456103802138476,
+      "loss": 0.7298,
+      "step": 592000
+    },
+    {
+      "epoch": 13.75,
+      "learning_rate": 0.00016409323605131475,
+      "loss": 0.7324,
+      "step": 594000
+    },
+    {
+      "epoch": 13.8,
+      "learning_rate": 0.0001636254340812447,
+      "loss": 0.7243,
+      "step": 596000
+    },
+    {
+      "epoch": 13.85,
+      "learning_rate": 0.0001631576321111747,
+      "loss": 0.7215,
+      "step": 598000
+    },
+    {
+      "epoch": 13.89,
+      "learning_rate": 0.00016268983014110463,
+      "loss": 0.7246,
+      "step": 600000
+    },
+    {
+      "epoch": 13.94,
+      "learning_rate": 0.00016222202817103463,
+      "loss": 0.7219,
+      "step": 602000
+    },
+    {
+      "epoch": 13.99,
+      "learning_rate": 0.0001617542262009646,
+      "loss": 0.7248,
+      "step": 604000
+    },
+    {
+      "epoch": 14.03,
+      "learning_rate": 0.00016128642423089457,
+      "loss": 0.7139,
+      "step": 606000
+    },
+    {
+      "epoch": 14.08,
+      "learning_rate": 0.00016081862226082454,
+      "loss": 0.7026,
+      "step": 608000
+    },
+    {
+      "epoch": 14.13,
+      "learning_rate": 0.0001603508202907545,
+      "loss": 0.7107,
+      "step": 610000
+    },
+    {
+      "epoch": 14.17,
+      "learning_rate": 0.00015988301832068447,
+      "loss": 0.7037,
+      "step": 612000
+    },
+    {
+      "epoch": 14.22,
+      "learning_rate": 0.00015941521635061444,
+      "loss": 0.7145,
+      "step": 614000
+    },
+    {
+      "epoch": 14.26,
+      "learning_rate": 0.0001589474143805444,
+      "loss": 0.7181,
+      "step": 616000
+    },
+    {
+      "epoch": 14.31,
+      "learning_rate": 0.00015847961241047438,
+      "loss": 0.7026,
+      "step": 618000
+    },
+    {
+      "epoch": 14.36,
+      "learning_rate": 0.00015801181044040435,
+      "loss": 0.7142,
+      "step": 620000
+    },
+    {
+      "epoch": 14.4,
+      "learning_rate": 0.00015754400847033432,
+      "loss": 0.7087,
+      "step": 622000
+    },
+    {
+      "epoch": 14.45,
+      "learning_rate": 0.0001570762065002643,
+      "loss": 0.7109,
+      "step": 624000
+    },
+    {
+      "epoch": 14.5,
+      "learning_rate": 0.00015660840453019426,
+      "loss": 0.7031,
+      "step": 626000
+    },
+    {
+      "epoch": 14.54,
+      "learning_rate": 0.00015614060256012425,
+      "loss": 0.7101,
+      "step": 628000
+    },
+    {
+      "epoch": 14.59,
+      "learning_rate": 0.0001556728005900542,
+      "loss": 0.7152,
+      "step": 630000
+    },
+    {
+      "epoch": 14.63,
+      "learning_rate": 0.0001552049986199842,
+      "loss": 0.7147,
+      "step": 632000
+    },
+    {
+      "epoch": 14.68,
+      "learning_rate": 0.00015473719664991413,
+      "loss": 0.7144,
+      "step": 634000
+    },
+    {
+      "epoch": 14.73,
+      "learning_rate": 0.00015426939467984413,
+      "loss": 0.7113,
+      "step": 636000
+    },
+    {
+      "epoch": 14.77,
+      "learning_rate": 0.00015380159270977407,
+      "loss": 0.7071,
+      "step": 638000
+    },
+    {
+      "epoch": 14.82,
+      "learning_rate": 0.00015333379073970407,
+      "loss": 0.7118,
+      "step": 640000
+    },
+    {
+      "epoch": 14.87,
+      "learning_rate": 0.00015286598876963403,
+      "loss": 0.7098,
+      "step": 642000
+    },
+    {
+      "epoch": 14.91,
+      "learning_rate": 0.000152398186799564,
+      "loss": 0.706,
+      "step": 644000
+    },
+    {
+      "epoch": 14.96,
+      "learning_rate": 0.00015193038482949397,
+      "loss": 0.709,
+      "step": 646000
+    },
+    {
+      "epoch": 15.01,
+      "learning_rate": 0.00015146258285942394,
+      "loss": 0.7087,
+      "step": 648000
+    },
+    {
+      "epoch": 15.05,
+      "learning_rate": 0.0001509947808893539,
+      "loss": 0.6983,
+      "step": 650000
+    },
+    {
+      "epoch": 15.1,
+      "learning_rate": 0.00015052697891928388,
+      "loss": 0.693,
+      "step": 652000
+    },
+    {
+      "epoch": 15.14,
+      "learning_rate": 0.00015005917694921385,
+      "loss": 0.6953,
+      "step": 654000
+    },
+    {
+      "epoch": 15.19,
+      "learning_rate": 0.00014959137497914382,
+      "loss": 0.6994,
+      "step": 656000
+    },
+    {
+      "epoch": 15.24,
+      "learning_rate": 0.00014912357300907379,
+      "loss": 0.6975,
+      "step": 658000
+    },
+    {
+      "epoch": 15.28,
+      "learning_rate": 0.00014865577103900375,
+      "loss": 0.7047,
+      "step": 660000
+    },
+    {
+      "epoch": 15.33,
+      "learning_rate": 0.00014818796906893372,
+      "loss": 0.6975,
+      "step": 662000
+    },
+    {
+      "epoch": 15.38,
+      "learning_rate": 0.0001477201670988637,
+      "loss": 0.704,
+      "step": 664000
+    },
+    {
+      "epoch": 15.42,
+      "learning_rate": 0.00014725236512879366,
+      "loss": 0.7042,
+      "step": 666000
+    },
+    {
+      "epoch": 15.47,
+      "learning_rate": 0.00014678456315872363,
+      "loss": 0.6917,
+      "step": 668000
+    },
+    {
+      "epoch": 15.51,
+      "learning_rate": 0.0001463167611886536,
+      "loss": 0.6914,
+      "step": 670000
+    },
+    {
+      "epoch": 15.56,
+      "learning_rate": 0.00014584895921858357,
+      "loss": 0.7018,
+      "step": 672000
+    },
+    {
+      "epoch": 15.61,
+      "learning_rate": 0.00014538115724851354,
+      "loss": 0.7016,
+      "step": 674000
+    },
+    {
+      "epoch": 15.65,
+      "learning_rate": 0.0001449133552784435,
+      "loss": 0.7078,
+      "step": 676000
+    },
+    {
+      "epoch": 15.7,
+      "learning_rate": 0.00014444555330837347,
+      "loss": 0.6932,
+      "step": 678000
+    },
+    {
+      "epoch": 15.75,
+      "learning_rate": 0.00014397775133830344,
+      "loss": 0.6964,
+      "step": 680000
+    },
+    {
+      "epoch": 15.79,
+      "learning_rate": 0.0001435099493682334,
+      "loss": 0.6997,
+      "step": 682000
+    },
+    {
+      "epoch": 15.84,
+      "learning_rate": 0.00014304214739816338,
+      "loss": 0.7065,
+      "step": 684000
+    },
+    {
+      "epoch": 15.88,
+      "learning_rate": 0.00014257434542809335,
+      "loss": 0.7047,
+      "step": 686000
+    },
+    {
+      "epoch": 15.93,
+      "learning_rate": 0.00014210654345802332,
+      "loss": 0.7154,
+      "step": 688000
+    },
+    {
+      "epoch": 15.98,
+      "learning_rate": 0.0001416387414879533,
+      "loss": 0.6993,
+      "step": 690000
+    },
+    {
+      "epoch": 16.02,
+      "learning_rate": 0.00014117093951788326,
+      "loss": 0.6969,
+      "step": 692000
+    },
+    {
+      "epoch": 16.07,
+      "learning_rate": 0.00014070313754781325,
+      "loss": 0.689,
+      "step": 694000
+    },
+    {
+      "epoch": 16.12,
+      "learning_rate": 0.00014023533557774322,
+      "loss": 0.6888,
+      "step": 696000
+    },
+    {
+      "epoch": 16.16,
+      "learning_rate": 0.0001397675336076732,
+      "loss": 0.6818,
+      "step": 698000
+    },
+    {
+      "epoch": 16.21,
+      "learning_rate": 0.00013929973163760316,
+      "loss": 0.693,
+      "step": 700000
+    },
+    {
+      "epoch": 16.26,
+      "learning_rate": 0.00013883192966753313,
+      "loss": 0.6909,
+      "step": 702000
+    },
+    {
+      "epoch": 16.3,
+      "learning_rate": 0.0001383641276974631,
+      "loss": 0.6873,
+      "step": 704000
+    },
+    {
+      "epoch": 16.35,
+      "learning_rate": 0.00013789632572739307,
+      "loss": 0.6906,
+      "step": 706000
+    },
+    {
+      "epoch": 16.39,
+      "learning_rate": 0.00013742852375732303,
+      "loss": 0.6866,
+      "step": 708000
+    },
+    {
+      "epoch": 16.44,
+      "learning_rate": 0.000136960721787253,
+      "loss": 0.701,
+      "step": 710000
+    },
+    {
+      "epoch": 16.49,
+      "learning_rate": 0.00013649291981718297,
+      "loss": 0.6937,
+      "step": 712000
+    },
+    {
+      "epoch": 16.53,
+      "learning_rate": 0.00013602511784711294,
+      "loss": 0.6907,
+      "step": 714000
+    },
+    {
+      "epoch": 16.58,
+      "learning_rate": 0.0001355573158770429,
+      "loss": 0.6897,
+      "step": 716000
+    },
+    {
+      "epoch": 16.63,
+      "learning_rate": 0.00013508951390697288,
+      "loss": 0.6952,
+      "step": 718000
+    },
+    {
+      "epoch": 16.67,
+      "learning_rate": 0.00013462171193690285,
+      "loss": 0.6865,
+      "step": 720000
+    },
+    {
+      "epoch": 16.72,
+      "learning_rate": 0.00013415390996683282,
+      "loss": 0.6935,
+      "step": 722000
+    },
+    {
+      "epoch": 16.76,
+      "learning_rate": 0.00013368610799676278,
+      "loss": 0.6919,
+      "step": 724000
+    },
+    {
+      "epoch": 16.81,
+      "learning_rate": 0.00013321830602669275,
+      "loss": 0.6904,
+      "step": 726000
+    },
+    {
+      "epoch": 16.86,
+      "learning_rate": 0.00013275050405662272,
+      "loss": 0.6964,
+      "step": 728000
+    },
+    {
+      "epoch": 16.9,
+      "learning_rate": 0.00013228270208655272,
+      "loss": 0.6943,
+      "step": 730000
+    },
+    {
+      "epoch": 16.95,
+      "learning_rate": 0.0001318149001164827,
+      "loss": 0.6949,
+      "step": 732000
+    },
+    {
+      "epoch": 17.0,
+      "learning_rate": 0.00013134709814641266,
+      "loss": 0.6943,
+      "step": 734000
+    },
+    {
+      "epoch": 17.04,
+      "learning_rate": 0.00013087929617634263,
+      "loss": 0.6851,
+      "step": 736000
+    },
+    {
+      "epoch": 17.09,
+      "learning_rate": 0.0001304114942062726,
+      "loss": 0.6802,
+      "step": 738000
+    },
+    {
+      "epoch": 17.14,
+      "learning_rate": 0.00012994369223620256,
+      "loss": 0.6801,
+      "step": 740000
+    },
+    {
+      "epoch": 17.18,
+      "learning_rate": 0.00012947589026613253,
+      "loss": 0.6756,
+      "step": 742000
+    },
+    {
+      "epoch": 17.23,
+      "learning_rate": 0.0001290080882960625,
+      "loss": 0.6824,
+      "step": 744000
+    },
+    {
+      "epoch": 17.27,
+      "learning_rate": 0.00012854028632599247,
+      "loss": 0.6894,
+      "step": 746000
+    },
+    {
+      "epoch": 17.32,
+      "learning_rate": 0.00012807248435592244,
+      "loss": 0.682,
+      "step": 748000
+    },
+    {
+      "epoch": 17.37,
+      "learning_rate": 0.0001276046823858524,
+      "loss": 0.6814,
+      "step": 750000
+    },
+    {
+      "epoch": 17.41,
+      "learning_rate": 0.00012713688041578238,
+      "loss": 0.6737,
+      "step": 752000
+    },
+    {
+      "epoch": 17.46,
+      "learning_rate": 0.00012666907844571234,
+      "loss": 0.6874,
+      "step": 754000
+    },
+    {
+      "epoch": 17.51,
+      "learning_rate": 0.00012620127647564231,
+      "loss": 0.6842,
+      "step": 756000
+    },
+    {
+      "epoch": 17.55,
+      "learning_rate": 0.00012573347450557228,
+      "loss": 0.6871,
+      "step": 758000
+    },
+    {
+      "epoch": 17.6,
+      "learning_rate": 0.00012526567253550225,
+      "loss": 0.6833,
+      "step": 760000
+    },
+    {
+      "epoch": 17.64,
+      "learning_rate": 0.00012479787056543222,
+      "loss": 0.6818,
+      "step": 762000
+    },
+    {
+      "epoch": 17.69,
+      "learning_rate": 0.0001243300685953622,
+      "loss": 0.6824,
+      "step": 764000
+    },
+    {
+      "epoch": 17.74,
+      "learning_rate": 0.00012386226662529219,
+      "loss": 0.684,
+      "step": 766000
+    },
+    {
+      "epoch": 17.78,
+      "learning_rate": 0.00012339446465522215,
+      "loss": 0.6822,
+      "step": 768000
+    },
+    {
+      "epoch": 17.83,
+      "learning_rate": 0.00012292666268515212,
+      "loss": 0.68,
+      "step": 770000
+    },
+    {
+      "epoch": 17.88,
+      "learning_rate": 0.0001224588607150821,
+      "loss": 0.6842,
+      "step": 772000
+    },
+    {
+      "epoch": 17.92,
+      "learning_rate": 0.00012199105874501206,
+      "loss": 0.6827,
+      "step": 774000
+    },
+    {
+      "epoch": 17.97,
+      "learning_rate": 0.00012152325677494203,
+      "loss": 0.6901,
+      "step": 776000
+    },
+    {
+      "epoch": 18.02,
+      "learning_rate": 0.000121055454804872,
+      "loss": 0.6781,
+      "step": 778000
+    },
+    {
+      "epoch": 18.06,
+      "learning_rate": 0.00012058765283480197,
+      "loss": 0.6768,
+      "step": 780000
+    },
+    {
+      "epoch": 18.11,
+      "learning_rate": 0.00012011985086473194,
+      "loss": 0.6704,
+      "step": 782000
+    },
+    {
+      "epoch": 18.15,
+      "learning_rate": 0.0001196520488946619,
+      "loss": 0.6767,
+      "step": 784000
+    },
+    {
+      "epoch": 18.2,
+      "learning_rate": 0.00011918424692459187,
+      "loss": 0.6696,
+      "step": 786000
+    },
+    {
+      "epoch": 18.25,
+      "learning_rate": 0.00011871644495452184,
+      "loss": 0.6717,
+      "step": 788000
+    },
+    {
+      "epoch": 18.29,
+      "learning_rate": 0.00011824864298445181,
+      "loss": 0.6666,
+      "step": 790000
+    },
+    {
+      "epoch": 18.34,
+      "learning_rate": 0.00011778084101438178,
+      "loss": 0.6681,
+      "step": 792000
+    },
+    {
+      "epoch": 18.39,
+      "learning_rate": 0.00011731303904431175,
+      "loss": 0.6688,
+      "step": 794000
+    },
+    {
+      "epoch": 18.43,
+      "learning_rate": 0.00011684523707424172,
+      "loss": 0.6809,
+      "step": 796000
+    },
+    {
+      "epoch": 18.48,
+      "learning_rate": 0.00011637743510417169,
+      "loss": 0.6704,
+      "step": 798000
+    },
+    {
+      "epoch": 18.52,
+      "learning_rate": 0.00011590963313410166,
+      "loss": 0.6732,
+      "step": 800000
+    },
+    {
+      "epoch": 18.57,
+      "learning_rate": 0.00011544183116403164,
+      "loss": 0.6688,
+      "step": 802000
+    },
+    {
+      "epoch": 18.62,
+      "learning_rate": 0.00011497402919396161,
+      "loss": 0.6767,
+      "step": 804000
+    },
+    {
+      "epoch": 18.66,
+      "learning_rate": 0.00011450622722389158,
+      "loss": 0.6721,
+      "step": 806000
+    },
+    {
+      "epoch": 18.71,
+      "learning_rate": 0.00011403842525382154,
+      "loss": 0.6716,
+      "step": 808000
+    },
+    {
+      "epoch": 18.76,
+      "learning_rate": 0.00011357062328375151,
+      "loss": 0.673,
+      "step": 810000
+    },
+    {
+      "epoch": 18.8,
+      "learning_rate": 0.00011310282131368148,
+      "loss": 0.6717,
+      "step": 812000
+    },
+    {
+      "epoch": 18.85,
+      "learning_rate": 0.00011263501934361145,
+      "loss": 0.6607,
+      "step": 814000
+    },
+    {
+      "epoch": 18.9,
+      "learning_rate": 0.00011216721737354142,
+      "loss": 0.6732,
+      "step": 816000
+    },
+    {
+      "epoch": 18.94,
+      "learning_rate": 0.00011169941540347139,
+      "loss": 0.6715,
+      "step": 818000
+    },
+    {
+      "epoch": 18.99,
+      "learning_rate": 0.00011123161343340136,
+      "loss": 0.678,
+      "step": 820000
+    },
+    {
+      "epoch": 19.03,
+      "learning_rate": 0.00011076381146333133,
+      "loss": 0.6618,
+      "step": 822000
+    },
+    {
+      "epoch": 19.08,
+      "learning_rate": 0.0001102960094932613,
+      "loss": 0.6589,
+      "step": 824000
+    },
+    {
+      "epoch": 19.13,
+      "learning_rate": 0.00010982820752319126,
+      "loss": 0.6624,
+      "step": 826000
+    },
+    {
+      "epoch": 19.17,
+      "learning_rate": 0.00010936040555312123,
+      "loss": 0.6618,
+      "step": 828000
+    },
+    {
+      "epoch": 19.22,
+      "learning_rate": 0.0001088926035830512,
+      "loss": 0.6666,
+      "step": 830000
+    },
+    {
+      "epoch": 19.27,
+      "learning_rate": 0.00010842480161298117,
+      "loss": 0.6645,
+      "step": 832000
+    },
+    {
+      "epoch": 19.31,
+      "learning_rate": 0.00010795699964291114,
+      "loss": 0.6667,
+      "step": 834000
+    },
+    {
+      "epoch": 19.36,
+      "learning_rate": 0.00010748919767284111,
+      "loss": 0.6649,
+      "step": 836000
+    },
+    {
+      "epoch": 19.4,
+      "learning_rate": 0.0001070213957027711,
+      "loss": 0.659,
+      "step": 838000
+    },
+    {
+      "epoch": 19.45,
+      "learning_rate": 0.00010655359373270107,
+      "loss": 0.6611,
+      "step": 840000
+    },
+    {
+      "epoch": 19.5,
+      "learning_rate": 0.00010608579176263104,
+      "loss": 0.6565,
+      "step": 842000
+    },
+    {
+      "epoch": 19.54,
+      "learning_rate": 0.00010561798979256101,
+      "loss": 0.6631,
+      "step": 844000
+    },
+    {
+      "epoch": 19.59,
+      "learning_rate": 0.00010515018782249098,
+      "loss": 0.6593,
+      "step": 846000
+    },
+    {
+      "epoch": 19.64,
+      "learning_rate": 0.00010468238585242095,
+      "loss": 0.6654,
+      "step": 848000
+    },
+    {
+      "epoch": 19.68,
+      "learning_rate": 0.00010421458388235092,
+      "loss": 0.6621,
+      "step": 850000
+    },
+    {
+      "epoch": 19.73,
+      "learning_rate": 0.00010374678191228089,
+      "loss": 0.661,
+      "step": 852000
+    },
+    {
+      "epoch": 19.78,
+      "learning_rate": 0.00010327897994221086,
+      "loss": 0.6515,
+      "step": 854000
+    },
+    {
+      "epoch": 19.82,
+      "learning_rate": 0.00010281117797214082,
+      "loss": 0.6614,
+      "step": 856000
+    },
+    {
+      "epoch": 19.87,
+      "learning_rate": 0.0001023433760020708,
+      "loss": 0.6616,
+      "step": 858000
+    },
+    {
+      "epoch": 19.91,
+      "learning_rate": 0.00010187557403200076,
+      "loss": 0.6598,
+      "step": 860000
+    },
+    {
+      "epoch": 19.96,
+      "learning_rate": 0.00010140777206193073,
+      "loss": 0.6616,
+      "step": 862000
+    },
+    {
+      "epoch": 20.01,
+      "learning_rate": 0.0001009399700918607,
+      "loss": 0.6679,
+      "step": 864000
+    },
+    {
+      "epoch": 20.05,
+      "learning_rate": 0.00010047216812179067,
+      "loss": 0.6518,
+      "step": 866000
+    },
+    {
+      "epoch": 20.1,
+      "learning_rate": 0.00010000436615172064,
+      "loss": 0.6463,
+      "step": 868000
+    },
+    {
+      "epoch": 20.15,
+      "learning_rate": 9.95365641816506e-05,
+      "loss": 0.6529,
+      "step": 870000
+    },
+    {
+      "epoch": 20.19,
+      "learning_rate": 9.906876221158058e-05,
+      "loss": 0.6463,
+      "step": 872000
+    },
+    {
+      "epoch": 20.24,
+      "learning_rate": 9.860096024151056e-05,
+      "loss": 0.6545,
+      "step": 874000
+    },
+    {
+      "epoch": 20.28,
+      "learning_rate": 9.813315827144053e-05,
+      "loss": 0.6531,
+      "step": 876000
+    },
+    {
+      "epoch": 20.33,
+      "learning_rate": 9.76653563013705e-05,
+      "loss": 0.6442,
+      "step": 878000
+    },
+    {
+      "epoch": 20.38,
+      "learning_rate": 9.719755433130046e-05,
+      "loss": 0.65,
+      "step": 880000
+    },
+    {
+      "epoch": 20.42,
+      "learning_rate": 9.672975236123043e-05,
+      "loss": 0.6518,
+      "step": 882000
+    },
+    {
+      "epoch": 20.47,
+      "learning_rate": 9.62619503911604e-05,
+      "loss": 0.6546,
+      "step": 884000
+    },
+    {
+      "epoch": 20.52,
+      "learning_rate": 9.579414842109037e-05,
+      "loss": 0.6494,
+      "step": 886000
+    },
+    {
+      "epoch": 20.56,
+      "learning_rate": 9.532634645102035e-05,
+      "loss": 0.654,
+      "step": 888000
+    },
+    {
+      "epoch": 20.61,
+      "learning_rate": 9.485854448095032e-05,
+      "loss": 0.6536,
+      "step": 890000
+    },
+    {
+      "epoch": 20.66,
+      "learning_rate": 9.439074251088029e-05,
+      "loss": 0.6547,
+      "step": 892000
+    },
+    {
+      "epoch": 20.7,
+      "learning_rate": 9.392294054081026e-05,
+      "loss": 0.6421,
+      "step": 894000
+    },
+    {
+      "epoch": 20.75,
+      "learning_rate": 9.345513857074023e-05,
+      "loss": 0.6506,
+      "step": 896000
+    },
+    {
+      "epoch": 20.79,
+      "learning_rate": 9.29873366006702e-05,
+      "loss": 0.6551,
+      "step": 898000
+    },
+    {
+      "epoch": 20.84,
+      "learning_rate": 9.251953463060017e-05,
+      "loss": 0.6542,
+      "step": 900000
+    },
+    {
+      "epoch": 20.89,
+      "learning_rate": 9.205173266053014e-05,
+      "loss": 0.6398,
+      "step": 902000
+    },
+    {
+      "epoch": 20.93,
+      "learning_rate": 9.15839306904601e-05,
+      "loss": 0.653,
+      "step": 904000
+    },
+    {
+      "epoch": 20.98,
+      "learning_rate": 9.111612872039007e-05,
+      "loss": 0.6476,
+      "step": 906000
+    },
+    {
+      "epoch": 21.03,
+      "learning_rate": 9.064832675032004e-05,
+      "loss": 0.6378,
+      "step": 908000
+    },
+    {
+      "epoch": 21.07,
+      "learning_rate": 9.018052478025002e-05,
+      "loss": 0.6413,
+      "step": 910000
+    },
+    {
+      "epoch": 21.12,
+      "learning_rate": 8.971272281017999e-05,
+      "loss": 0.6368,
+      "step": 912000
+    },
+    {
+      "epoch": 21.16,
+      "learning_rate": 8.924492084010996e-05,
+      "loss": 0.6366,
+      "step": 914000
+    },
+    {
+      "epoch": 21.21,
+      "learning_rate": 8.877711887003993e-05,
+      "loss": 0.6455,
+      "step": 916000
+    },
+    {
+      "epoch": 21.26,
+      "learning_rate": 8.83093168999699e-05,
+      "loss": 0.6448,
+      "step": 918000
+    },
+    {
+      "epoch": 21.3,
+      "learning_rate": 8.784151492989987e-05,
+      "loss": 0.6371,
+      "step": 920000
+    },
+    {
+      "epoch": 21.35,
+      "learning_rate": 8.737371295982984e-05,
+      "loss": 0.6457,
+      "step": 922000
+    },
+    {
+      "epoch": 21.4,
+      "learning_rate": 8.69059109897598e-05,
+      "loss": 0.6399,
+      "step": 924000
+    },
+    {
+      "epoch": 21.44,
+      "learning_rate": 8.643810901968978e-05,
+      "loss": 0.6389,
+      "step": 926000
+    },
+    {
+      "epoch": 21.49,
+      "learning_rate": 8.597030704961974e-05,
+      "loss": 0.6444,
+      "step": 928000
+    },
+    {
+      "epoch": 21.54,
+      "learning_rate": 8.550250507954971e-05,
+      "loss": 0.6346,
+      "step": 930000
+    },
+    {
+      "epoch": 21.58,
+      "learning_rate": 8.503470310947968e-05,
+      "loss": 0.6394,
+      "step": 932000
+    },
+    {
+      "epoch": 21.63,
+      "learning_rate": 8.456690113940965e-05,
+      "loss": 0.6397,
+      "step": 934000
+    },
+    {
+      "epoch": 21.67,
+      "learning_rate": 8.409909916933962e-05,
+      "loss": 0.6411,
+      "step": 936000
+    },
+    {
+      "epoch": 21.72,
+      "learning_rate": 8.363129719926959e-05,
+      "loss": 0.6383,
+      "step": 938000
+    },
+    {
+      "epoch": 21.77,
+      "learning_rate": 8.316349522919956e-05,
+      "loss": 0.6416,
+      "step": 940000
+    },
+    {
+      "epoch": 21.81,
+      "learning_rate": 8.269569325912953e-05,
+      "loss": 0.635,
+      "step": 942000
+    },
+    {
+      "epoch": 21.86,
+      "learning_rate": 8.22278912890595e-05,
+      "loss": 0.6371,
+      "step": 944000
+    },
+    {
+      "epoch": 21.91,
+      "learning_rate": 8.176008931898949e-05,
+      "loss": 0.6412,
+      "step": 946000
+    },
+    {
+      "epoch": 21.95,
+      "learning_rate": 8.129228734891946e-05,
+      "loss": 0.6414,
+      "step": 948000
+    },
+    {
+      "epoch": 22.0,
+      "learning_rate": 8.082448537884943e-05,
+      "loss": 0.6285,
+      "step": 950000
+    },
+    {
+      "epoch": 22.04,
+      "learning_rate": 8.03566834087794e-05,
+      "loss": 0.6285,
+      "step": 952000
+    },
+    {
+      "epoch": 22.09,
+      "learning_rate": 7.988888143870937e-05,
+      "loss": 0.6268,
+      "step": 954000
+    },
+    {
+      "epoch": 22.14,
+      "learning_rate": 7.942107946863934e-05,
+      "loss": 0.6251,
+      "step": 956000
+    },
+    {
+      "epoch": 22.18,
+      "learning_rate": 7.89532774985693e-05,
+      "loss": 0.6306,
+      "step": 958000
+    },
+    {
+      "epoch": 22.23,
+      "learning_rate": 7.848547552849927e-05,
+      "loss": 0.6283,
+      "step": 960000
+    },
+    {
+      "epoch": 22.28,
+      "learning_rate": 7.801767355842924e-05,
+      "loss": 0.6264,
+      "step": 962000
+    },
+    {
+      "epoch": 22.32,
+      "learning_rate": 7.754987158835921e-05,
+      "loss": 0.6279,
+      "step": 964000
+    },
+    {
+      "epoch": 22.37,
+      "learning_rate": 7.708206961828918e-05,
+      "loss": 0.6272,
+      "step": 966000
+    },
+    {
+      "epoch": 22.41,
+      "learning_rate": 7.661426764821915e-05,
+      "loss": 0.6355,
+      "step": 968000
+    },
+    {
+      "epoch": 22.46,
+      "learning_rate": 7.614646567814912e-05,
+      "loss": 0.6349,
+      "step": 970000
+    },
+    {
+      "epoch": 22.51,
+      "learning_rate": 7.567866370807909e-05,
+      "loss": 0.6281,
+      "step": 972000
+    },
+    {
+      "epoch": 22.55,
+      "learning_rate": 7.521086173800905e-05,
+      "loss": 0.6269,
+      "step": 974000
+    },
+    {
+      "epoch": 22.6,
+      "learning_rate": 7.474305976793904e-05,
+      "loss": 0.6221,
+      "step": 976000
+    },
+    {
+      "epoch": 22.65,
+      "learning_rate": 7.4275257797869e-05,
+      "loss": 0.6295,
+      "step": 978000
+    },
+    {
+      "epoch": 22.69,
+      "learning_rate": 7.380745582779897e-05,
+      "loss": 0.6265,
+      "step": 980000
+    },
+    {
+      "epoch": 22.74,
+      "learning_rate": 7.333965385772894e-05,
+      "loss": 0.6203,
+      "step": 982000
+    },
+    {
+      "epoch": 22.79,
+      "learning_rate": 7.287185188765891e-05,
+      "loss": 0.6306,
+      "step": 984000
+    },
+    {
+      "epoch": 22.83,
+      "learning_rate": 7.240404991758888e-05,
+      "loss": 0.6319,
+      "step": 986000
+    },
+    {
+      "epoch": 22.88,
+      "learning_rate": 7.193624794751885e-05,
+      "loss": 0.6211,
+      "step": 988000
+    },
+    {
+      "epoch": 22.92,
+      "learning_rate": 7.146844597744882e-05,
+      "loss": 0.6244,
+      "step": 990000
+    },
+    {
+      "epoch": 22.97,
+      "learning_rate": 7.100064400737879e-05,
+      "loss": 0.6262,
+      "step": 992000
+    },
+    {
+      "epoch": 23.02,
+      "learning_rate": 7.053284203730876e-05,
+      "loss": 0.6166,
+      "step": 994000
+    },
+    {
+      "epoch": 23.06,
+      "learning_rate": 7.006504006723873e-05,
+      "loss": 0.6166,
+      "step": 996000
+    },
+    {
+      "epoch": 23.11,
+      "learning_rate": 6.95972380971687e-05,
+      "loss": 0.6175,
+      "step": 998000
+    },
+    {
+      "epoch": 23.16,
+      "learning_rate": 6.912943612709866e-05,
+      "loss": 0.6151,
+      "step": 1000000
+    },
+    {
+      "epoch": 23.2,
+      "learning_rate": 6.866163415702863e-05,
+      "loss": 0.6153,
+      "step": 1002000
+    },
+    {
+      "epoch": 23.25,
+      "learning_rate": 6.81938321869586e-05,
+      "loss": 0.6212,
+      "step": 1004000
+    },
+    {
+      "epoch": 23.29,
+      "learning_rate": 6.772603021688858e-05,
+      "loss": 0.6161,
+      "step": 1006000
+    },
+    {
+      "epoch": 23.34,
+      "learning_rate": 6.725822824681855e-05,
+      "loss": 0.6158,
+      "step": 1008000
+    },
+    {
+      "epoch": 23.39,
+      "learning_rate": 6.679042627674852e-05,
+      "loss": 0.6089,
+      "step": 1010000
+    },
+    {
+      "epoch": 23.43,
+      "learning_rate": 6.632262430667849e-05,
+      "loss": 0.6166,
+      "step": 1012000
+    },
+    {
+      "epoch": 23.48,
+      "learning_rate": 6.585482233660846e-05,
+      "loss": 0.6134,
+      "step": 1014000
+    },
+    {
+      "epoch": 23.53,
+      "learning_rate": 6.538702036653843e-05,
+      "loss": 0.6171,
+      "step": 1016000
+    },
+    {
+      "epoch": 23.57,
+      "learning_rate": 6.49192183964684e-05,
+      "loss": 0.6122,
+      "step": 1018000
+    },
+    {
+      "epoch": 23.62,
+      "learning_rate": 6.445141642639837e-05,
+      "loss": 0.6176,
+      "step": 1020000
+    },
+    {
+      "epoch": 23.67,
+      "learning_rate": 6.398361445632833e-05,
+      "loss": 0.6146,
+      "step": 1022000
+    },
+    {
+      "epoch": 23.71,
+      "learning_rate": 6.35158124862583e-05,
+      "loss": 0.6069,
+      "step": 1024000
+    },
+    {
+      "epoch": 23.76,
+      "learning_rate": 6.304801051618829e-05,
+      "loss": 0.6169,
+      "step": 1026000
+    },
+    {
+      "epoch": 23.8,
+      "learning_rate": 6.258020854611825e-05,
+      "loss": 0.6222,
+      "step": 1028000
+    },
+    {
+      "epoch": 23.85,
+      "learning_rate": 6.211240657604822e-05,
+      "loss": 0.6152,
+      "step": 1030000
+    },
+    {
+      "epoch": 23.9,
+      "learning_rate": 6.164460460597819e-05,
+      "loss": 0.6181,
+      "step": 1032000
+    },
+    {
+      "epoch": 23.94,
+      "learning_rate": 6.117680263590816e-05,
+      "loss": 0.6123,
+      "step": 1034000
+    },
+    {
+      "epoch": 23.99,
+      "learning_rate": 6.070900066583813e-05,
+      "loss": 0.619,
+      "step": 1036000
+    },
+    {
+      "epoch": 24.04,
+      "learning_rate": 6.02411986957681e-05,
+      "loss": 0.6099,
+      "step": 1038000
+    },
+    {
+      "epoch": 24.08,
+      "learning_rate": 5.977339672569807e-05,
+      "loss": 0.6098,
+      "step": 1040000
+    },
+    {
+      "epoch": 24.13,
+      "learning_rate": 5.930559475562804e-05,
+      "loss": 0.5965,
+      "step": 1042000
+    },
+    {
+      "epoch": 24.17,
+      "learning_rate": 5.883779278555802e-05,
+      "loss": 0.6059,
+      "step": 1044000
+    },
+    {
+      "epoch": 24.22,
+      "learning_rate": 5.836999081548799e-05,
+      "loss": 0.6021,
+      "step": 1046000
+    },
+    {
+      "epoch": 24.27,
+      "learning_rate": 5.790218884541796e-05,
+      "loss": 0.6093,
+      "step": 1048000
+    },
+    {
+      "epoch": 24.31,
+      "learning_rate": 5.7434386875347926e-05,
+      "loss": 0.6031,
+      "step": 1050000
+    },
+    {
+      "epoch": 24.36,
+      "learning_rate": 5.6966584905277895e-05,
+      "loss": 0.6053,
+      "step": 1052000
+    },
+    {
+      "epoch": 24.41,
+      "learning_rate": 5.6498782935207863e-05,
+      "loss": 0.6036,
+      "step": 1054000
+    },
+    {
+      "epoch": 24.45,
+      "learning_rate": 5.603098096513783e-05,
+      "loss": 0.6011,
+      "step": 1056000
+    },
+    {
+      "epoch": 24.5,
+      "learning_rate": 5.55631789950678e-05,
+      "loss": 0.6035,
+      "step": 1058000
+    },
+    {
+      "epoch": 24.55,
+      "learning_rate": 5.509537702499777e-05,
+      "loss": 0.6066,
+      "step": 1060000
+    },
+    {
+      "epoch": 24.59,
+      "learning_rate": 5.4627575054927746e-05,
+      "loss": 0.6061,
+      "step": 1062000
+    },
+    {
+      "epoch": 24.64,
+      "learning_rate": 5.4159773084857714e-05,
+      "loss": 0.6027,
+      "step": 1064000
+    },
+    {
+      "epoch": 24.68,
+      "learning_rate": 5.369197111478768e-05,
+      "loss": 0.6,
+      "step": 1066000
+    },
+    {
+      "epoch": 24.73,
+      "learning_rate": 5.322416914471765e-05,
+      "loss": 0.6062,
+      "step": 1068000
+    },
+    {
+      "epoch": 24.78,
+      "learning_rate": 5.275636717464762e-05,
+      "loss": 0.6003,
+      "step": 1070000
+    },
+    {
+      "epoch": 24.82,
+      "learning_rate": 5.228856520457759e-05,
+      "loss": 0.5988,
+      "step": 1072000
+    },
+    {
+      "epoch": 24.87,
+      "learning_rate": 5.182076323450756e-05,
+      "loss": 0.6096,
+      "step": 1074000
+    },
+    {
+      "epoch": 24.92,
+      "learning_rate": 5.135296126443753e-05,
+      "loss": 0.5988,
+      "step": 1076000
+    },
+    {
+      "epoch": 24.96,
+      "learning_rate": 5.08851592943675e-05,
+      "loss": 0.6086,
+      "step": 1078000
+    },
+    {
+      "epoch": 25.01,
+      "learning_rate": 5.041735732429748e-05,
+      "loss": 0.5942,
+      "step": 1080000
+    },
+    {
+      "epoch": 25.05,
+      "learning_rate": 4.994955535422745e-05,
+      "loss": 0.5954,
+      "step": 1082000
+    },
+    {
+      "epoch": 25.1,
+      "learning_rate": 4.9481753384157417e-05,
+      "loss": 0.5948,
+      "step": 1084000
+    },
+    {
+      "epoch": 25.15,
+      "learning_rate": 4.9013951414087385e-05,
+      "loss": 0.5946,
+      "step": 1086000
+    },
+    {
+      "epoch": 25.19,
+      "learning_rate": 4.8546149444017354e-05,
+      "loss": 0.5938,
+      "step": 1088000
+    },
+    {
+      "epoch": 25.24,
+      "learning_rate": 4.807834747394732e-05,
+      "loss": 0.5961,
+      "step": 1090000
+    },
+    {
+      "epoch": 25.29,
+      "learning_rate": 4.761054550387729e-05,
+      "loss": 0.5947,
+      "step": 1092000
+    },
+    {
+      "epoch": 25.33,
+      "learning_rate": 4.714274353380726e-05,
+      "loss": 0.6019,
+      "step": 1094000
+    },
+    {
+      "epoch": 25.38,
+      "learning_rate": 4.667494156373723e-05,
+      "loss": 0.5927,
+      "step": 1096000
+    },
+    {
+      "epoch": 25.43,
+      "learning_rate": 4.6207139593667205e-05,
+      "loss": 0.5921,
+      "step": 1098000
+    },
+    {
+      "epoch": 25.47,
+      "learning_rate": 4.5739337623597174e-05,
+      "loss": 0.5954,
+      "step": 1100000
+    },
+    {
+      "epoch": 25.52,
+      "learning_rate": 4.527153565352715e-05,
+      "loss": 0.5926,
+      "step": 1102000
+    },
+    {
+      "epoch": 25.56,
+      "learning_rate": 4.480373368345712e-05,
+      "loss": 0.5963,
+      "step": 1104000
+    },
+    {
+      "epoch": 25.61,
+      "learning_rate": 4.433593171338709e-05,
+      "loss": 0.5902,
+      "step": 1106000
+    },
+    {
+      "epoch": 25.66,
+      "learning_rate": 4.3868129743317056e-05,
+      "loss": 0.5952,
+      "step": 1108000
+    },
+    {
+      "epoch": 25.7,
+      "learning_rate": 4.3400327773247025e-05,
+      "loss": 0.5878,
+      "step": 1110000
+    },
+    {
+      "epoch": 25.75,
+      "learning_rate": 4.2932525803176994e-05,
+      "loss": 0.5926,
+      "step": 1112000
+    },
+    {
+      "epoch": 25.8,
+      "learning_rate": 4.246472383310696e-05,
+      "loss": 0.5854,
+      "step": 1114000
+    },
+    {
+      "epoch": 25.84,
+      "learning_rate": 4.199692186303694e-05,
+      "loss": 0.5916,
+      "step": 1116000
+    },
+    {
+      "epoch": 25.89,
+      "learning_rate": 4.152911989296691e-05,
+      "loss": 0.5869,
+      "step": 1118000
+    },
+    {
+      "epoch": 25.93,
+      "learning_rate": 4.1061317922896876e-05,
+      "loss": 0.5913,
+      "step": 1120000
+    },
+    {
+      "epoch": 25.98,
+      "learning_rate": 4.0593515952826845e-05,
+      "loss": 0.5822,
+      "step": 1122000
+    },
+    {
+      "epoch": 26.03,
+      "learning_rate": 4.0125713982756814e-05,
+      "loss": 0.5831,
+      "step": 1124000
+    },
+    {
+      "epoch": 26.07,
+      "learning_rate": 3.965791201268678e-05,
+      "loss": 0.5847,
+      "step": 1126000
+    },
+    {
+      "epoch": 26.12,
+      "learning_rate": 3.919011004261675e-05,
+      "loss": 0.5828,
+      "step": 1128000
+    },
+    {
+      "epoch": 26.17,
+      "learning_rate": 3.872230807254672e-05,
+      "loss": 0.5825,
+      "step": 1130000
+    },
+    {
+      "epoch": 26.21,
+      "learning_rate": 3.825450610247669e-05,
+      "loss": 0.5848,
+      "step": 1132000
+    },
+    {
+      "epoch": 26.26,
+      "learning_rate": 3.778670413240667e-05,
+      "loss": 0.5866,
+      "step": 1134000
+    },
+    {
+      "epoch": 26.31,
+      "learning_rate": 3.7318902162336634e-05,
+      "loss": 0.5832,
+      "step": 1136000
+    },
+    {
+      "epoch": 26.35,
+      "learning_rate": 3.685110019226661e-05,
+      "loss": 0.58,
+      "step": 1138000
+    },
+    {
+      "epoch": 26.4,
+      "learning_rate": 3.638329822219658e-05,
+      "loss": 0.5767,
+      "step": 1140000
+    },
+    {
+      "epoch": 26.44,
+      "learning_rate": 3.591549625212655e-05,
+      "loss": 0.5792,
+      "step": 1142000
+    },
+    {
+      "epoch": 26.49,
+      "learning_rate": 3.5447694282056516e-05,
+      "loss": 0.5764,
+      "step": 1144000
+    },
+    {
+      "epoch": 26.54,
+      "learning_rate": 3.4979892311986485e-05,
+      "loss": 0.5794,
+      "step": 1146000
+    },
+    {
+      "epoch": 26.58,
+      "learning_rate": 3.451209034191646e-05,
+      "loss": 0.5738,
+      "step": 1148000
+    },
+    {
+      "epoch": 26.63,
+      "learning_rate": 3.404428837184643e-05,
+      "loss": 0.5822,
+      "step": 1150000
+    },
+    {
+      "epoch": 26.68,
+      "learning_rate": 3.35764864017764e-05,
+      "loss": 0.5734,
+      "step": 1152000
+    },
+    {
+      "epoch": 26.72,
+      "learning_rate": 3.310868443170637e-05,
+      "loss": 0.5794,
+      "step": 1154000
+    },
+    {
+      "epoch": 26.77,
+      "learning_rate": 3.2640882461636336e-05,
+      "loss": 0.5853,
+      "step": 1156000
+    },
+    {
+      "epoch": 26.81,
+      "learning_rate": 3.2173080491566305e-05,
+      "loss": 0.5842,
+      "step": 1158000
+    },
+    {
+      "epoch": 26.86,
+      "learning_rate": 3.170527852149628e-05,
+      "loss": 0.5847,
+      "step": 1160000
+    },
+    {
+      "epoch": 26.91,
+      "learning_rate": 3.123747655142625e-05,
+      "loss": 0.5786,
+      "step": 1162000
+    },
+    {
+      "epoch": 26.95,
+      "learning_rate": 3.076967458135622e-05,
+      "loss": 0.5818,
+      "step": 1164000
+    },
+    {
+      "epoch": 27.0,
+      "learning_rate": 3.030187261128619e-05,
+      "loss": 0.5722,
+      "step": 1166000
+    },
+    {
+      "epoch": 27.05,
+      "learning_rate": 2.983407064121616e-05,
+      "loss": 0.5726,
+      "step": 1168000
+    },
+    {
+      "epoch": 27.09,
+      "learning_rate": 2.936626867114613e-05,
+      "loss": 0.5745,
+      "step": 1170000
+    },
+    {
+      "epoch": 27.14,
+      "learning_rate": 2.8898466701076097e-05,
+      "loss": 0.5655,
+      "step": 1172000
+    },
+    {
+      "epoch": 27.19,
+      "learning_rate": 2.843066473100607e-05,
+      "loss": 0.5747,
+      "step": 1174000
+    },
+    {
+      "epoch": 27.23,
+      "learning_rate": 2.7962862760936038e-05,
+      "loss": 0.5734,
+      "step": 1176000
+    },
+    {
+      "epoch": 27.28,
+      "learning_rate": 2.7495060790866007e-05,
+      "loss": 0.5752,
+      "step": 1178000
+    },
+    {
+      "epoch": 27.32,
+      "learning_rate": 2.7027258820795976e-05,
+      "loss": 0.5784,
+      "step": 1180000
+    },
+    {
+      "epoch": 27.37,
+      "learning_rate": 2.6559456850725945e-05,
+      "loss": 0.5667,
+      "step": 1182000
+    },
+    {
+      "epoch": 27.42,
+      "learning_rate": 2.609165488065592e-05,
+      "loss": 0.5748,
+      "step": 1184000
+    },
+    {
+      "epoch": 27.46,
+      "learning_rate": 2.562385291058589e-05,
+      "loss": 0.5762,
+      "step": 1186000
+    },
+    {
+      "epoch": 27.51,
+      "learning_rate": 2.5156050940515858e-05,
+      "loss": 0.5783,
+      "step": 1188000
+    },
+    {
+      "epoch": 27.56,
+      "learning_rate": 2.4688248970445827e-05,
+      "loss": 0.5668,
+      "step": 1190000
+    },
+    {
+      "epoch": 27.6,
+      "learning_rate": 2.42204470003758e-05,
+      "loss": 0.5671,
+      "step": 1192000
+    },
+    {
+      "epoch": 27.65,
+      "learning_rate": 2.3752645030305768e-05,
+      "loss": 0.5688,
+      "step": 1194000
+    },
+    {
+      "epoch": 27.69,
+      "learning_rate": 2.328484306023574e-05,
+      "loss": 0.5643,
+      "step": 1196000
+    },
+    {
+      "epoch": 27.74,
+      "learning_rate": 2.281704109016571e-05,
+      "loss": 0.5688,
+      "step": 1198000
+    },
+    {
+      "epoch": 27.79,
+      "learning_rate": 2.2349239120095678e-05,
+      "loss": 0.5651,
+      "step": 1200000
+    },
+    {
+      "epoch": 27.83,
+      "learning_rate": 2.188143715002565e-05,
+      "loss": 0.5705,
+      "step": 1202000
+    },
+    {
+      "epoch": 27.88,
+      "learning_rate": 2.141363517995562e-05,
+      "loss": 0.5684,
+      "step": 1204000
+    },
+    {
+      "epoch": 27.93,
+      "learning_rate": 2.0945833209885588e-05,
+      "loss": 0.567,
+      "step": 1206000
+    },
+    {
+      "epoch": 27.97,
+      "learning_rate": 2.0478031239815557e-05,
+      "loss": 0.5711,
+      "step": 1208000
+    },
+    {
+      "epoch": 28.02,
+      "learning_rate": 2.0010229269745533e-05,
+      "loss": 0.5684,
+      "step": 1210000
+    },
+    {
+      "epoch": 28.06,
+      "learning_rate": 1.95424272996755e-05,
+      "loss": 0.5604,
+      "step": 1212000
+    },
+    {
+      "epoch": 28.11,
+      "learning_rate": 1.907462532960547e-05,
+      "loss": 0.5615,
+      "step": 1214000
+    },
+    {
+      "epoch": 28.16,
+      "learning_rate": 1.860682335953544e-05,
+      "loss": 0.5679,
+      "step": 1216000
+    },
+    {
+      "epoch": 28.2,
+      "learning_rate": 1.813902138946541e-05,
+      "loss": 0.5644,
+      "step": 1218000
+    },
+    {
+      "epoch": 28.25,
+      "learning_rate": 1.767121941939538e-05,
+      "loss": 0.5663,
+      "step": 1220000
+    },
+    {
+      "epoch": 28.3,
+      "learning_rate": 1.720341744932535e-05,
+      "loss": 0.5584,
+      "step": 1222000
+    },
+    {
+      "epoch": 28.34,
+      "learning_rate": 1.6735615479255318e-05,
+      "loss": 0.558,
+      "step": 1224000
+    },
+    {
+      "epoch": 28.39,
+      "learning_rate": 1.626781350918529e-05,
+      "loss": 0.5575,
+      "step": 1226000
+    },
+    {
+      "epoch": 28.44,
+      "learning_rate": 1.580001153911526e-05,
+      "loss": 0.5728,
+      "step": 1228000
+    },
+    {
+      "epoch": 28.48,
+      "learning_rate": 1.533220956904523e-05,
+      "loss": 0.5653,
+      "step": 1230000
+    },
+    {
+      "epoch": 28.53,
+      "learning_rate": 1.48644075989752e-05,
+      "loss": 0.5603,
+      "step": 1232000
+    },
+    {
+      "epoch": 28.57,
+      "learning_rate": 1.4396605628905172e-05,
+      "loss": 0.5613,
+      "step": 1234000
+    },
+    {
+      "epoch": 28.62,
+      "learning_rate": 1.3928803658835141e-05,
+      "loss": 0.5563,
+      "step": 1236000
+    },
+    {
+      "epoch": 28.67,
+      "learning_rate": 1.346100168876511e-05,
+      "loss": 0.5705,
+      "step": 1238000
+    },
+    {
+      "epoch": 28.71,
+      "learning_rate": 1.299319971869508e-05,
+      "loss": 0.5568,
+      "step": 1240000
+    },
+    {
+      "epoch": 28.76,
+      "learning_rate": 1.252539774862505e-05,
+      "loss": 0.5517,
+      "step": 1242000
+    },
+    {
+      "epoch": 28.81,
+      "learning_rate": 1.2057595778555022e-05,
+      "loss": 0.5647,
+      "step": 1244000
+    },
+    {
+      "epoch": 28.85,
+      "learning_rate": 1.158979380848499e-05,
+      "loss": 0.5551,
+      "step": 1246000
+    },
+    {
+      "epoch": 28.9,
+      "learning_rate": 1.1121991838414961e-05,
+      "loss": 0.5598,
+      "step": 1248000
+    },
+    {
+      "epoch": 28.94,
+      "learning_rate": 1.0654189868344932e-05,
+      "loss": 0.562,
+      "step": 1250000
+    },
+    {
+      "epoch": 28.99,
+      "learning_rate": 1.0186387898274902e-05,
+      "loss": 0.5563,
+      "step": 1252000
+    },
+    {
+      "epoch": 29.04,
+      "learning_rate": 9.718585928204871e-06,
+      "loss": 0.5606,
+      "step": 1254000
+    },
+    {
+      "epoch": 29.08,
+      "learning_rate": 9.250783958134842e-06,
+      "loss": 0.5509,
+      "step": 1256000
+    },
+    {
+      "epoch": 29.13,
+      "learning_rate": 8.782981988064812e-06,
+      "loss": 0.551,
+      "step": 1258000
+    },
+    {
+      "epoch": 29.18,
+      "learning_rate": 8.315180017994783e-06,
+      "loss": 0.5548,
+      "step": 1260000
+    },
+    {
+      "epoch": 29.22,
+      "learning_rate": 7.847378047924752e-06,
+      "loss": 0.5562,
+      "step": 1262000
+    },
+    {
+      "epoch": 29.27,
+      "learning_rate": 7.3795760778547214e-06,
+      "loss": 0.5563,
+      "step": 1264000
+    },
+    {
+      "epoch": 29.32,
+      "learning_rate": 6.911774107784692e-06,
+      "loss": 0.5551,
+      "step": 1266000
+    },
+    {
+      "epoch": 29.36,
+      "learning_rate": 6.443972137714662e-06,
+      "loss": 0.555,
+      "step": 1268000
+    },
+    {
+      "epoch": 29.41,
+      "learning_rate": 5.976170167644632e-06,
+      "loss": 0.554,
+      "step": 1270000
+    },
+    {
+      "epoch": 29.45,
+      "learning_rate": 5.508368197574602e-06,
+      "loss": 0.5522,
+      "step": 1272000
+    },
+    {
+      "epoch": 29.5,
+      "learning_rate": 5.0405662275045725e-06,
+      "loss": 0.5522,
+      "step": 1274000
+    },
+    {
+      "epoch": 29.55,
+      "learning_rate": 4.572764257434542e-06,
+      "loss": 0.5601,
+      "step": 1276000
+    },
+    {
+      "epoch": 29.59,
+      "learning_rate": 4.104962287364513e-06,
+      "loss": 0.5578,
+      "step": 1278000
+    },
+    {
+      "epoch": 29.64,
+      "learning_rate": 3.6371603172944825e-06,
+      "loss": 0.5602,
+      "step": 1280000
+    },
+    {
+      "epoch": 29.69,
+      "learning_rate": 3.1693583472244526e-06,
+      "loss": 0.5517,
+      "step": 1282000
+    },
+    {
+      "epoch": 29.73,
+      "learning_rate": 2.701556377154423e-06,
+      "loss": 0.5538,
+      "step": 1284000
+    },
+    {
+      "epoch": 29.78,
+      "learning_rate": 2.233754407084393e-06,
+      "loss": 0.549,
+      "step": 1286000
+    },
+    {
+      "epoch": 29.82,
+      "learning_rate": 1.765952437014363e-06,
+      "loss": 0.5595,
+      "step": 1288000
+    },
+    {
+      "epoch": 29.87,
+      "learning_rate": 1.2981504669443331e-06,
+      "loss": 0.5546,
+      "step": 1290000
+    },
+    {
+      "epoch": 29.92,
+      "learning_rate": 8.303484968743031e-07,
+      "loss": 0.5547,
+      "step": 1292000
+    },
+    {
+      "epoch": 29.96,
+      "learning_rate": 3.6254652680427316e-07,
+      "loss": 0.5494,
+      "step": 1294000
+    },
+    {
+      "epoch": 30.0,
+      "step": 1295550,
+      "total_flos": 2.6480449905256835e+21,
+      "train_loss": 0.7350748663054241,
+      "train_runtime": 658563.1153,
+      "train_samples_per_second": 31.476,
+      "train_steps_per_second": 1.967
+    },
+    {
+      "epoch": 30.01,
+      "learning_rate": 3.624385885736958e-07,
+      "loss": 0.5685,
+      "step": 1296000
+    },
+    {
+      "epoch": 30.06,
+      "learning_rate": 3.619588631044631e-07,
+      "loss": 0.5522,
+      "step": 1298000
+    },
+    {
+      "epoch": 30.1,
+      "learning_rate": 3.6147913763523037e-07,
+      "loss": 0.5498,
+      "step": 1300000
+    },
+    {
+      "epoch": 30.15,
+      "learning_rate": 3.609994121659976e-07,
+      "loss": 0.5494,
+      "step": 1302000
+    },
+    {
+      "epoch": 30.2,
+      "learning_rate": 3.6051968669676483e-07,
+      "loss": 0.5514,
+      "step": 1304000
+    },
+    {
+      "epoch": 30.24,
+      "learning_rate": 3.600399612275321e-07,
+      "loss": 0.5472,
+      "step": 1306000
+    },
+    {
+      "epoch": 30.29,
+      "learning_rate": 3.595602357582994e-07,
+      "loss": 0.5596,
+      "step": 1308000
+    },
+    {
+      "epoch": 30.33,
+      "learning_rate": 3.590805102890667e-07,
+      "loss": 0.5525,
+      "step": 1310000
+    },
+    {
+      "epoch": 30.38,
+      "learning_rate": 3.586007848198339e-07,
+      "loss": 0.5601,
+      "step": 1312000
+    },
+    {
+      "epoch": 30.43,
+      "learning_rate": 3.581210593506012e-07,
+      "loss": 0.5526,
+      "step": 1314000
+    },
+    {
+      "epoch": 30.47,
+      "learning_rate": 3.5764133388136847e-07,
+      "loss": 0.551,
+      "step": 1316000
+    },
+    {
+      "epoch": 30.52,
+      "learning_rate": 3.5716160841213575e-07,
+      "loss": 0.5513,
+      "step": 1318000
+    },
+    {
+      "epoch": 30.57,
+      "learning_rate": 3.56681882942903e-07,
+      "loss": 0.5456,
+      "step": 1320000
+    },
+    {
+      "epoch": 30.61,
+      "learning_rate": 3.5620215747367026e-07,
+      "loss": 0.5563,
+      "step": 1322000
+    },
+    {
+      "epoch": 30.66,
+      "learning_rate": 3.5572243200443755e-07,
+      "loss": 0.5567,
+      "step": 1324000
+    },
+    {
+      "epoch": 30.71,
+      "learning_rate": 3.5524270653520483e-07,
+      "loss": 0.5573,
+      "step": 1326000
+    },
+    {
+      "epoch": 30.75,
+      "learning_rate": 3.5476298106597206e-07,
+      "loss": 0.5575,
+      "step": 1328000
+    },
+    {
+      "epoch": 30.8,
+      "learning_rate": 3.542832555967393e-07,
+      "loss": 0.5599,
+      "step": 1330000
+    },
+    {
+      "epoch": 30.84,
+      "learning_rate": 3.5380353012750657e-07,
+      "loss": 0.554,
+      "step": 1332000
+    },
+    {
+      "epoch": 30.89,
+      "learning_rate": 3.5332380465827385e-07,
+      "loss": 0.5507,
+      "step": 1334000
+    },
+    {
+      "epoch": 30.94,
+      "learning_rate": 3.5284407918904113e-07,
+      "loss": 0.5522,
+      "step": 1336000
+    },
+    {
+      "epoch": 30.98,
+      "learning_rate": 3.5236435371980836e-07,
+      "loss": 0.5439,
+      "step": 1338000
+    },
+    {
+      "epoch": 31.03,
+      "learning_rate": 3.5188462825057564e-07,
+      "loss": 0.554,
+      "step": 1340000
+    },
+    {
+      "epoch": 31.08,
+      "learning_rate": 3.514049027813429e-07,
+      "loss": 0.5477,
+      "step": 1342000
+    },
+    {
+      "epoch": 31.12,
+      "learning_rate": 3.509251773121102e-07,
+      "loss": 0.5565,
+      "step": 1344000
+    },
+    {
+      "epoch": 31.17,
+      "learning_rate": 3.5044545184287744e-07,
+      "loss": 0.5482,
+      "step": 1346000
+    },
+    {
+      "epoch": 31.21,
+      "learning_rate": 3.499657263736447e-07,
+      "loss": 0.5506,
+      "step": 1348000
+    },
+    {
+      "epoch": 31.26,
+      "learning_rate": 3.49486000904412e-07,
+      "loss": 0.5471,
+      "step": 1350000
+    },
+    {
+      "epoch": 31.31,
+      "learning_rate": 3.490062754351793e-07,
+      "loss": 0.5529,
+      "step": 1352000
+    },
+    {
+      "epoch": 31.35,
+      "learning_rate": 3.485265499659465e-07,
+      "loss": 0.5547,
+      "step": 1354000
+    },
+    {
+      "epoch": 31.4,
+      "learning_rate": 3.480468244967138e-07,
+      "loss": 0.5522,
+      "step": 1356000
+    },
+    {
+      "epoch": 31.45,
+      "learning_rate": 3.475670990274811e-07,
+      "loss": 0.5473,
+      "step": 1358000
+    },
+    {
+      "epoch": 31.49,
+      "learning_rate": 3.470873735582483e-07,
+      "loss": 0.5599,
+      "step": 1360000
+    },
+    {
+      "epoch": 31.54,
+      "learning_rate": 3.466076480890156e-07,
+      "loss": 0.5573,
+      "step": 1362000
+    },
+    {
+      "epoch": 31.59,
+      "learning_rate": 3.461279226197828e-07,
+      "loss": 0.553,
+      "step": 1364000
+    },
+    {
+      "epoch": 31.63,
+      "learning_rate": 3.456481971505501e-07,
+      "loss": 0.5506,
+      "step": 1366000
+    },
+    {
+      "epoch": 31.68,
+      "learning_rate": 3.451684716813174e-07,
+      "loss": 0.5537,
+      "step": 1368000
+    },
+    {
+      "epoch": 31.72,
+      "learning_rate": 3.4468874621208466e-07,
+      "loss": 0.547,
+      "step": 1370000
+    },
+    {
+      "epoch": 31.77,
+      "learning_rate": 3.442090207428519e-07,
+      "loss": 0.5579,
+      "step": 1372000
+    },
+    {
+      "epoch": 31.82,
+      "learning_rate": 3.437292952736192e-07,
+      "loss": 0.5446,
+      "step": 1374000
+    },
+    {
+      "epoch": 31.86,
+      "learning_rate": 3.4324956980438646e-07,
+      "loss": 0.5491,
+      "step": 1376000
+    },
+    {
+      "epoch": 31.91,
+      "learning_rate": 3.4276984433515374e-07,
+      "loss": 0.5489,
+      "step": 1378000
+    },
+    {
+      "epoch": 31.96,
+      "learning_rate": 3.4229011886592097e-07,
+      "loss": 0.553,
+      "step": 1380000
+    },
+    {
+      "epoch": 32.0,
+      "learning_rate": 3.4181039339668825e-07,
+      "loss": 0.549,
+      "step": 1382000
+    },
+    {
+      "epoch": 32.05,
+      "learning_rate": 3.4133066792745553e-07,
+      "loss": 0.5498,
+      "step": 1384000
+    },
+    {
+      "epoch": 32.09,
+      "learning_rate": 3.408509424582228e-07,
+      "loss": 0.5619,
+      "step": 1386000
+    },
+    {
+      "epoch": 32.14,
+      "learning_rate": 3.4037121698899004e-07,
+      "loss": 0.5461,
+      "step": 1388000
+    },
+    {
+      "epoch": 32.19,
+      "learning_rate": 3.398914915197573e-07,
+      "loss": 0.5501,
+      "step": 1390000
+    },
+    {
+      "epoch": 32.23,
+      "learning_rate": 3.3941176605052456e-07,
+      "loss": 0.5583,
+      "step": 1392000
+    },
+    {
+      "epoch": 32.28,
+      "learning_rate": 3.3893204058129184e-07,
+      "loss": 0.5597,
+      "step": 1394000
+    },
+    {
+      "epoch": 32.33,
+      "learning_rate": 3.384523151120591e-07,
+      "loss": 0.5482,
+      "step": 1396000
+    },
+    {
+      "epoch": 32.37,
+      "learning_rate": 3.3797258964282635e-07,
+      "loss": 0.5573,
+      "step": 1398000
+    },
+    {
+      "epoch": 32.42,
+      "learning_rate": 3.3749286417359363e-07,
+      "loss": 0.546,
+      "step": 1400000
+    },
+    {
+      "epoch": 32.46,
+      "learning_rate": 3.370131387043609e-07,
+      "loss": 0.5481,
+      "step": 1402000
+    },
+    {
+      "epoch": 32.51,
+      "learning_rate": 3.365334132351282e-07,
+      "loss": 0.5572,
+      "step": 1404000
+    },
+    {
+      "epoch": 32.56,
+      "learning_rate": 3.360536877658954e-07,
+      "loss": 0.5479,
+      "step": 1406000
+    },
+    {
+      "epoch": 32.6,
+      "learning_rate": 3.355739622966627e-07,
+      "loss": 0.5456,
+      "step": 1408000
+    },
+    {
+      "epoch": 32.65,
+      "learning_rate": 3.3509423682743e-07,
+      "loss": 0.5498,
+      "step": 1410000
+    },
+    {
+      "epoch": 32.7,
+      "learning_rate": 3.3461451135819727e-07,
+      "loss": 0.5519,
+      "step": 1412000
+    },
+    {
+      "epoch": 32.74,
+      "learning_rate": 3.341347858889645e-07,
+      "loss": 0.5533,
+      "step": 1414000
+    },
+    {
+      "epoch": 32.79,
+      "learning_rate": 3.336550604197318e-07,
+      "loss": 0.5455,
+      "step": 1416000
+    },
+    {
+      "epoch": 32.84,
+      "learning_rate": 3.33175334950499e-07,
+      "loss": 0.5535,
+      "step": 1418000
+    },
+    {
+      "epoch": 32.88,
+      "learning_rate": 3.326956094812663e-07,
+      "loss": 0.5512,
+      "step": 1420000
+    },
+    {
+      "epoch": 32.93,
+      "learning_rate": 3.322158840120336e-07,
+      "loss": 0.5496,
+      "step": 1422000
+    },
+    {
+      "epoch": 32.97,
+      "learning_rate": 3.317361585428008e-07,
+      "loss": 0.5505,
+      "step": 1424000
+    },
+    {
+      "epoch": 33.02,
+      "learning_rate": 3.312564330735681e-07,
+      "loss": 0.5561,
+      "step": 1426000
+    },
+    {
+      "epoch": 33.07,
+      "learning_rate": 3.3077670760433537e-07,
+      "loss": 0.5512,
+      "step": 1428000
+    },
+    {
+      "epoch": 33.11,
+      "learning_rate": 3.3029698213510265e-07,
+      "loss": 0.5533,
+      "step": 1430000
+    },
+    {
+      "epoch": 33.16,
+      "learning_rate": 3.298172566658699e-07,
+      "loss": 0.5554,
+      "step": 1432000
+    },
+    {
+      "epoch": 33.21,
+      "learning_rate": 3.2933753119663716e-07,
+      "loss": 0.5536,
+      "step": 1434000
+    },
+    {
+      "epoch": 33.25,
+      "learning_rate": 3.2885780572740445e-07,
+      "loss": 0.5553,
+      "step": 1436000
+    },
+    {
+      "epoch": 33.3,
+      "learning_rate": 3.2837808025817173e-07,
+      "loss": 0.5451,
+      "step": 1438000
+    },
+    {
+      "epoch": 33.34,
+      "learning_rate": 3.2789835478893896e-07,
+      "loss": 0.5468,
+      "step": 1440000
+    },
+    {
+      "epoch": 33.39,
+      "learning_rate": 3.2741862931970624e-07,
+      "loss": 0.5532,
+      "step": 1442000
+    },
+    {
+      "epoch": 33.44,
+      "learning_rate": 3.269389038504735e-07,
+      "loss": 0.5451,
+      "step": 1444000
+    },
+    {
+      "epoch": 33.48,
+      "learning_rate": 3.2645917838124075e-07,
+      "loss": 0.5482,
+      "step": 1446000
+    },
+    {
+      "epoch": 33.53,
+      "learning_rate": 3.2597945291200803e-07,
+      "loss": 0.553,
+      "step": 1448000
+    },
+    {
+      "epoch": 33.58,
+      "learning_rate": 3.2549972744277526e-07,
+      "loss": 0.5483,
+      "step": 1450000
+    },
+    {
+      "epoch": 33.62,
+      "learning_rate": 3.2502000197354254e-07,
+      "loss": 0.552,
+      "step": 1452000
+    },
+    {
+      "epoch": 33.67,
+      "learning_rate": 3.245402765043098e-07,
+      "loss": 0.5494,
+      "step": 1454000
+    },
+    {
+      "epoch": 33.72,
+      "learning_rate": 3.240605510350771e-07,
+      "loss": 0.5583,
+      "step": 1456000
+    },
+    {
+      "epoch": 33.76,
+      "learning_rate": 3.2358082556584434e-07,
+      "loss": 0.5506,
+      "step": 1458000
+    },
+    {
+      "epoch": 33.81,
+      "learning_rate": 3.231011000966116e-07,
+      "loss": 0.5524,
+      "step": 1460000
+    },
+    {
+      "epoch": 33.85,
+      "learning_rate": 3.226213746273789e-07,
+      "loss": 0.5529,
+      "step": 1462000
+    },
+    {
+      "epoch": 33.9,
+      "learning_rate": 3.221416491581462e-07,
+      "loss": 0.5563,
+      "step": 1464000
+    },
+    {
+      "epoch": 33.95,
+      "learning_rate": 3.216619236889134e-07,
+      "loss": 0.5493,
+      "step": 1466000
+    },
+    {
+      "epoch": 33.99,
+      "learning_rate": 3.211821982196807e-07,
+      "loss": 0.5478,
+      "step": 1468000
+    },
+    {
+      "epoch": 34.04,
+      "learning_rate": 3.20702472750448e-07,
+      "loss": 0.546,
+      "step": 1470000
+    },
+    {
+      "epoch": 34.09,
+      "learning_rate": 3.2022274728121526e-07,
+      "loss": 0.5556,
+      "step": 1472000
+    },
+    {
+      "epoch": 34.13,
+      "learning_rate": 3.197430218119825e-07,
+      "loss": 0.5511,
+      "step": 1474000
+    },
+    {
+      "epoch": 34.18,
+      "learning_rate": 3.192632963427497e-07,
+      "loss": 0.5467,
+      "step": 1476000
+    },
+    {
+      "epoch": 34.22,
+      "learning_rate": 3.18783570873517e-07,
+      "loss": 0.5534,
+      "step": 1478000
+    },
+    {
+      "epoch": 34.27,
+      "learning_rate": 3.183038454042843e-07,
+      "loss": 0.5558,
+      "step": 1480000
+    },
+    {
+      "epoch": 34.32,
+      "learning_rate": 3.1782411993505156e-07,
+      "loss": 0.5532,
+      "step": 1482000
+    },
+    {
+      "epoch": 34.36,
+      "learning_rate": 3.173443944658188e-07,
+      "loss": 0.5468,
+      "step": 1484000
+    },
+    {
+      "epoch": 34.41,
+      "learning_rate": 3.168646689965861e-07,
+      "loss": 0.5532,
+      "step": 1486000
+    },
+    {
+      "epoch": 34.46,
+      "learning_rate": 3.1638494352735336e-07,
+      "loss": 0.549,
+      "step": 1488000
+    },
+    {
+      "epoch": 34.5,
+      "learning_rate": 3.1590521805812064e-07,
+      "loss": 0.5499,
+      "step": 1490000
+    },
+    {
+      "epoch": 34.55,
+      "learning_rate": 3.1542549258888787e-07,
+      "loss": 0.5535,
+      "step": 1492000
+    },
+    {
+      "epoch": 34.6,
+      "learning_rate": 3.1494576711965515e-07,
+      "loss": 0.5503,
+      "step": 1494000
+    },
+    {
+      "epoch": 34.64,
+      "learning_rate": 3.1446604165042243e-07,
+      "loss": 0.5468,
+      "step": 1496000
+    },
+    {
+      "epoch": 34.69,
+      "learning_rate": 3.139863161811897e-07,
+      "loss": 0.5515,
+      "step": 1498000
+    },
+    {
+      "epoch": 34.73,
+      "learning_rate": 3.1350659071195695e-07,
+      "loss": 0.5507,
+      "step": 1500000
+    },
+    {
+      "epoch": 34.78,
+      "learning_rate": 3.1302686524272423e-07,
+      "loss": 0.5539,
+      "step": 1502000
+    },
+    {
+      "epoch": 34.83,
+      "learning_rate": 3.1254713977349146e-07,
+      "loss": 0.5451,
+      "step": 1504000
+    },
+    {
+      "epoch": 34.87,
+      "learning_rate": 3.1206741430425874e-07,
+      "loss": 0.5502,
+      "step": 1506000
+    },
+    {
+      "epoch": 34.92,
+      "learning_rate": 3.11587688835026e-07,
+      "loss": 0.5534,
+      "step": 1508000
+    },
+    {
+      "epoch": 34.97,
+      "learning_rate": 3.1110796336579325e-07,
+      "loss": 0.5511,
+      "step": 1510000
+    },
+    {
+      "epoch": 35.0,
+      "step": 1511475,
+      "total_flos": 3.089388519897025e+21,
+      "train_loss": 0.07881541758525794,
+      "train_runtime": 103159.2271,
+      "train_samples_per_second": 234.43,
+      "train_steps_per_second": 14.652
+    },
+    {
+      "epoch": 35.01,
+      "learning_rate": 3.6243633986055876e-07,
+      "loss": 0.5431,
+      "step": 1512000
+    },
+    {
+      "epoch": 35.06,
+      "learning_rate": 3.620165800749801e-07,
+      "loss": 0.5573,
+      "step": 1514000
+    },
+    {
+      "epoch": 35.1,
+      "learning_rate": 3.6159682028940147e-07,
+      "loss": 0.5447,
+      "step": 1516000
+    },
+    {
+      "epoch": 35.15,
+      "learning_rate": 3.611770605038228e-07,
+      "loss": 0.5497,
+      "step": 1518000
+    },
+    {
+      "epoch": 35.2,
+      "learning_rate": 3.607573007182442e-07,
+      "loss": 0.5369,
+      "step": 1520000
+    },
+    {
+      "epoch": 35.24,
+      "learning_rate": 3.603375409326656e-07,
+      "loss": 0.5425,
+      "step": 1522000
+    },
+    {
+      "epoch": 35.29,
+      "learning_rate": 3.599177811470869e-07,
+      "loss": 0.5472,
+      "step": 1524000
+    },
+    {
+      "epoch": 35.34,
+      "learning_rate": 3.594980213615083e-07,
+      "loss": 0.5528,
+      "step": 1526000
+    },
+    {
+      "epoch": 35.38,
+      "learning_rate": 3.590782615759296e-07,
+      "loss": 0.5469,
+      "step": 1528000
+    },
+    {
+      "epoch": 35.43,
+      "learning_rate": 3.58658501790351e-07,
+      "loss": 0.548,
+      "step": 1530000
+    },
+    {
+      "epoch": 35.48,
+      "learning_rate": 3.5823874200477234e-07,
+      "loss": 0.5515,
+      "step": 1532000
+    },
+    {
+      "epoch": 35.52,
+      "learning_rate": 3.5781898221919373e-07,
+      "loss": 0.5562,
+      "step": 1534000
+    },
+    {
+      "epoch": 35.57,
+      "learning_rate": 3.5739922243361506e-07,
+      "loss": 0.5544,
+      "step": 1536000
+    },
+    {
+      "epoch": 35.61,
+      "learning_rate": 3.569794626480364e-07,
+      "loss": 0.5488,
+      "step": 1538000
+    },
+    {
+      "epoch": 35.66,
+      "learning_rate": 3.565597028624578e-07,
+      "loss": 0.5498,
+      "step": 1540000
+    },
+    {
+      "epoch": 35.71,
+      "learning_rate": 3.561399430768791e-07,
+      "loss": 0.5545,
+      "step": 1542000
+    },
+    {
+      "epoch": 35.75,
+      "learning_rate": 3.557201832913005e-07,
+      "loss": 0.5547,
+      "step": 1544000
+    },
+    {
+      "epoch": 35.8,
+      "learning_rate": 3.553004235057218e-07,
+      "loss": 0.5455,
+      "step": 1546000
+    },
+    {
+      "epoch": 35.85,
+      "learning_rate": 3.548806637201432e-07,
+      "loss": 0.558,
+      "step": 1548000
+    },
+    {
+      "epoch": 35.89,
+      "learning_rate": 3.544609039345646e-07,
+      "loss": 0.551,
+      "step": 1550000
+    },
+    {
+      "epoch": 35.94,
+      "learning_rate": 3.5404114414898593e-07,
+      "loss": 0.5445,
+      "step": 1552000
+    },
+    {
+      "epoch": 35.98,
+      "learning_rate": 3.536213843634073e-07,
+      "loss": 0.5491,
+      "step": 1554000
+    },
+    {
+      "epoch": 36.03,
+      "learning_rate": 3.5320162457782865e-07,
+      "loss": 0.5408,
+      "step": 1556000
+    },
+    {
+      "epoch": 36.08,
+      "learning_rate": 3.5278186479225003e-07,
+      "loss": 0.5503,
+      "step": 1558000
+    },
+    {
+      "epoch": 36.12,
+      "learning_rate": 3.5236210500667136e-07,
+      "loss": 0.5569,
+      "step": 1560000
+    },
+    {
+      "epoch": 36.17,
+      "learning_rate": 3.519423452210927e-07,
+      "loss": 0.5472,
+      "step": 1562000
+    },
+    {
+      "epoch": 36.22,
+      "learning_rate": 3.5152258543551403e-07,
+      "loss": 0.5464,
+      "step": 1564000
+    },
+    {
+      "epoch": 36.26,
+      "learning_rate": 3.511028256499354e-07,
+      "loss": 0.5448,
+      "step": 1566000
+    },
+    {
+      "epoch": 36.31,
+      "learning_rate": 3.506830658643568e-07,
+      "loss": 0.5461,
+      "step": 1568000
+    },
+    {
+      "epoch": 36.36,
+      "learning_rate": 3.5026330607877813e-07,
+      "loss": 0.5453,
+      "step": 1570000
+    },
+    {
+      "epoch": 36.4,
+      "learning_rate": 3.498435462931995e-07,
+      "loss": 0.5527,
+      "step": 1572000
+    },
+    {
+      "epoch": 36.45,
+      "learning_rate": 3.4942378650762085e-07,
+      "loss": 0.549,
+      "step": 1574000
+    },
+    {
+      "epoch": 36.49,
+      "learning_rate": 3.4900402672204223e-07,
+      "loss": 0.5472,
+      "step": 1576000
+    },
+    {
+      "epoch": 36.54,
+      "learning_rate": 3.4858426693646356e-07,
+      "loss": 0.5522,
+      "step": 1578000
+    },
+    {
+      "epoch": 36.59,
+      "learning_rate": 3.4816450715088495e-07,
+      "loss": 0.5518,
+      "step": 1580000
+    },
+    {
+      "epoch": 36.63,
+      "learning_rate": 3.4774474736530633e-07,
+      "loss": 0.5552,
+      "step": 1582000
+    },
+    {
+      "epoch": 36.68,
+      "learning_rate": 3.4732498757972767e-07,
+      "loss": 0.55,
+      "step": 1584000
+    },
+    {
+      "epoch": 36.73,
+      "learning_rate": 3.46905227794149e-07,
+      "loss": 0.5487,
+      "step": 1586000
+    },
+    {
+      "epoch": 36.77,
+      "learning_rate": 3.464854680085704e-07,
+      "loss": 0.5579,
+      "step": 1588000
+    },
+    {
+      "epoch": 36.82,
+      "learning_rate": 3.460657082229917e-07,
+      "loss": 0.543,
+      "step": 1590000
+    },
+    {
+      "epoch": 36.86,
+      "learning_rate": 3.4564594843741305e-07,
+      "loss": 0.5513,
+      "step": 1592000
+    },
+    {
+      "epoch": 36.91,
+      "learning_rate": 3.4522618865183443e-07,
+      "loss": 0.5476,
+      "step": 1594000
+    },
+    {
+      "epoch": 36.96,
+      "learning_rate": 3.448064288662558e-07,
+      "loss": 0.5523,
+      "step": 1596000
+    },
+    {
+      "epoch": 37.0,
+      "learning_rate": 3.4438666908067715e-07,
+      "loss": 0.5498,
+      "step": 1598000
+    },
+    {
+      "epoch": 37.05,
+      "learning_rate": 3.4396690929509854e-07,
+      "loss": 0.5556,
+      "step": 1600000
+    },
+    {
+      "epoch": 37.1,
+      "learning_rate": 3.4354714950951987e-07,
+      "loss": 0.5466,
+      "step": 1602000
+    },
+    {
+      "epoch": 37.14,
+      "learning_rate": 3.4312738972394125e-07,
+      "loss": 0.5505,
+      "step": 1604000
+    },
+    {
+      "epoch": 37.19,
+      "learning_rate": 3.427076299383626e-07,
+      "loss": 0.5453,
+      "step": 1606000
+    },
+    {
+      "epoch": 37.24,
+      "learning_rate": 3.4228787015278397e-07,
+      "loss": 0.5552,
+      "step": 1608000
+    },
+    {
+      "epoch": 37.28,
+      "learning_rate": 3.418681103672053e-07,
+      "loss": 0.5526,
+      "step": 1610000
+    },
+    {
+      "epoch": 37.33,
+      "learning_rate": 3.414483505816267e-07,
+      "loss": 0.5522,
+      "step": 1612000
+    },
+    {
+      "epoch": 37.37,
+      "learning_rate": 3.41028590796048e-07,
+      "loss": 0.5523,
+      "step": 1614000
+    },
+    {
+      "epoch": 37.42,
+      "learning_rate": 3.4060883101046935e-07,
+      "loss": 0.5473,
+      "step": 1616000
+    },
+    {
+      "epoch": 37.47,
+      "learning_rate": 3.4018907122489074e-07,
+      "loss": 0.5514,
+      "step": 1618000
+    },
+    {
+      "epoch": 37.51,
+      "learning_rate": 3.3976931143931207e-07,
+      "loss": 0.5513,
+      "step": 1620000
+    },
+    {
+      "epoch": 37.56,
+      "learning_rate": 3.3934955165373345e-07,
+      "loss": 0.5409,
+      "step": 1622000
+    },
+    {
+      "epoch": 37.61,
+      "learning_rate": 3.389297918681548e-07,
+      "loss": 0.5521,
+      "step": 1624000
+    },
+    {
+      "epoch": 37.65,
+      "learning_rate": 3.3851003208257617e-07,
+      "loss": 0.5508,
+      "step": 1626000
+    },
+    {
+      "epoch": 37.7,
+      "learning_rate": 3.3809027229699756e-07,
+      "loss": 0.5425,
+      "step": 1628000
+    },
+    {
+      "epoch": 37.74,
+      "learning_rate": 3.376705125114189e-07,
+      "loss": 0.5453,
+      "step": 1630000
+    },
+    {
+      "epoch": 37.79,
+      "learning_rate": 3.372507527258403e-07,
+      "loss": 0.5409,
+      "step": 1632000
+    },
+    {
+      "epoch": 37.84,
+      "learning_rate": 3.368309929402616e-07,
+      "loss": 0.5393,
+      "step": 1634000
+    },
+    {
+      "epoch": 37.88,
+      "learning_rate": 3.36411233154683e-07,
+      "loss": 0.5566,
+      "step": 1636000
+    },
+    {
+      "epoch": 37.93,
+      "learning_rate": 3.359914733691043e-07,
+      "loss": 0.5505,
+      "step": 1638000
+    },
+    {
+      "epoch": 37.98,
+      "learning_rate": 3.3557171358352566e-07,
+      "loss": 0.552,
+      "step": 1640000
+    },
+    {
+      "epoch": 38.02,
+      "learning_rate": 3.35151953797947e-07,
+      "loss": 0.5508,
+      "step": 1642000
+    },
+    {
+      "epoch": 38.07,
+      "learning_rate": 3.3473219401236837e-07,
+      "loss": 0.5447,
+      "step": 1644000
+    },
+    {
+      "epoch": 38.12,
+      "learning_rate": 3.3431243422678976e-07,
+      "loss": 0.5525,
+      "step": 1646000
+    },
+    {
+      "epoch": 38.16,
+      "learning_rate": 3.338926744412111e-07,
+      "loss": 0.5497,
+      "step": 1648000
+    },
+    {
+      "epoch": 38.21,
+      "learning_rate": 3.334729146556325e-07,
+      "loss": 0.553,
+      "step": 1650000
+    },
+    {
+      "epoch": 38.25,
+      "learning_rate": 3.330531548700538e-07,
+      "loss": 0.5471,
+      "step": 1652000
+    },
+    {
+      "epoch": 38.3,
+      "learning_rate": 3.326333950844752e-07,
+      "loss": 0.5522,
+      "step": 1654000
+    },
+    {
+      "epoch": 38.35,
+      "learning_rate": 3.322136352988965e-07,
+      "loss": 0.5514,
+      "step": 1656000
+    },
+    {
+      "epoch": 38.39,
+      "learning_rate": 3.317938755133179e-07,
+      "loss": 0.5477,
+      "step": 1658000
+    },
+    {
+      "epoch": 38.44,
+      "learning_rate": 3.313741157277393e-07,
+      "loss": 0.5403,
+      "step": 1660000
+    },
+    {
+      "epoch": 38.49,
+      "learning_rate": 3.3095435594216063e-07,
+      "loss": 0.5448,
+      "step": 1662000
+    },
+    {
+      "epoch": 38.53,
+      "learning_rate": 3.3053459615658196e-07,
+      "loss": 0.5504,
+      "step": 1664000
+    },
+    {
+      "epoch": 38.58,
+      "learning_rate": 3.301148363710033e-07,
+      "loss": 0.5531,
+      "step": 1666000
+    },
+    {
+      "epoch": 38.62,
+      "learning_rate": 3.296950765854247e-07,
+      "loss": 0.5531,
+      "step": 1668000
+    },
+    {
+      "epoch": 38.67,
+      "learning_rate": 3.29275316799846e-07,
+      "loss": 0.5436,
+      "step": 1670000
+    },
+    {
+      "epoch": 38.72,
+      "learning_rate": 3.288555570142674e-07,
+      "loss": 0.5443,
+      "step": 1672000
+    },
+    {
+      "epoch": 38.76,
+      "learning_rate": 3.284357972286888e-07,
+      "loss": 0.5496,
+      "step": 1674000
+    },
+    {
+      "epoch": 38.81,
+      "learning_rate": 3.280160374431101e-07,
+      "loss": 0.5475,
+      "step": 1676000
+    },
+    {
+      "epoch": 38.86,
+      "learning_rate": 3.275962776575315e-07,
+      "loss": 0.5528,
+      "step": 1678000
+    },
+    {
+      "epoch": 38.9,
+      "learning_rate": 3.2717651787195283e-07,
+      "loss": 0.5414,
+      "step": 1680000
+    },
+    {
+      "epoch": 38.95,
+      "learning_rate": 3.267567580863742e-07,
+      "loss": 0.549,
+      "step": 1682000
+    },
+    {
+      "epoch": 38.99,
+      "learning_rate": 3.2633699830079555e-07,
+      "loss": 0.5523,
+      "step": 1684000
+    },
+    {
+      "epoch": 39.04,
+      "learning_rate": 3.2591723851521693e-07,
+      "loss": 0.5477,
+      "step": 1686000
+    },
+    {
+      "epoch": 39.09,
+      "learning_rate": 3.2549747872963826e-07,
+      "loss": 0.5512,
+      "step": 1688000
+    },
+    {
+      "epoch": 39.13,
+      "learning_rate": 3.250777189440596e-07,
+      "loss": 0.5459,
+      "step": 1690000
+    },
+    {
+      "epoch": 39.18,
+      "learning_rate": 3.24657959158481e-07,
+      "loss": 0.5448,
+      "step": 1692000
+    },
+    {
+      "epoch": 39.23,
+      "learning_rate": 3.242381993729023e-07,
+      "loss": 0.5425,
+      "step": 1694000
+    },
+    {
+      "epoch": 39.27,
+      "learning_rate": 3.238184395873237e-07,
+      "loss": 0.549,
+      "step": 1696000
+    },
+    {
+      "epoch": 39.32,
+      "learning_rate": 3.2339867980174503e-07,
+      "loss": 0.554,
+      "step": 1698000
+    },
+    {
+      "epoch": 39.37,
+      "learning_rate": 3.229789200161664e-07,
+      "loss": 0.5514,
+      "step": 1700000
+    },
+    {
+      "epoch": 39.41,
+      "learning_rate": 3.2255916023058775e-07,
+      "loss": 0.5526,
+      "step": 1702000
+    },
+    {
+      "epoch": 39.46,
+      "learning_rate": 3.2213940044500913e-07,
+      "loss": 0.5433,
+      "step": 1704000
+    },
+    {
+      "epoch": 39.5,
+      "learning_rate": 3.217196406594305e-07,
+      "loss": 0.5414,
+      "step": 1706000
+    },
+    {
+      "epoch": 39.55,
+      "learning_rate": 3.2129988087385185e-07,
+      "loss": 0.5546,
+      "step": 1708000
+    },
+    {
+      "epoch": 39.6,
+      "learning_rate": 3.2088012108827324e-07,
+      "loss": 0.5445,
+      "step": 1710000
+    },
+    {
+      "epoch": 39.64,
+      "learning_rate": 3.2046036130269457e-07,
+      "loss": 0.5568,
+      "step": 1712000
+    },
+    {
+      "epoch": 39.69,
+      "learning_rate": 3.200406015171159e-07,
+      "loss": 0.5403,
+      "step": 1714000
+    },
+    {
+      "epoch": 39.74,
+      "learning_rate": 3.196208417315373e-07,
+      "loss": 0.5463,
+      "step": 1716000
+    },
+    {
+      "epoch": 39.78,
+      "learning_rate": 3.192010819459586e-07,
+      "loss": 0.5515,
+      "step": 1718000
+    },
+    {
+      "epoch": 39.83,
+      "learning_rate": 3.1878132216038e-07,
+      "loss": 0.5464,
+      "step": 1720000
+    },
+    {
+      "epoch": 39.87,
+      "learning_rate": 3.1836156237480133e-07,
+      "loss": 0.546,
+      "step": 1722000
+    },
+    {
+      "epoch": 39.92,
+      "learning_rate": 3.179418025892227e-07,
+      "loss": 0.551,
+      "step": 1724000
+    },
+    {
+      "epoch": 39.97,
+      "learning_rate": 3.1752204280364405e-07,
+      "loss": 0.5533,
+      "step": 1726000
+    },
+    {
+      "epoch": 40.0,
+      "step": 1727400,
+      "total_flos": 3.530727795360958e+21,
+      "train_loss": 0.06863045204288915,
+      "train_runtime": 104273.7009,
+      "train_samples_per_second": 265.057,
+      "train_steps_per_second": 16.566
+    }
+  ],
+  "logging_steps": 2000,
+  "max_steps": 1727400,
+  "num_train_epochs": 40,
+  "save_steps": 500,
+  "total_flos": 3.530727795360958e+21,
+  "trial_name": null,
+  "trial_params": null
+}