diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,4799 +1,3843 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 4.999841596705211,
-  "global_step": 3945,
+  "epoch": 3.9998415967052114,
+  "global_step": 3156,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.01,
-      "learning_rate": 6.329113924050633e-05,
-      "loss": 5.9056,
+      "learning_rate": 9.374999999999999e-05,
+      "loss": 1.8139,
       "step": 5
     },
     {
       "epoch": 0.01,
-      "learning_rate": 0.00012658227848101267,
-      "loss": 4.729,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 1.7876,
       "step": 10
     },
     {
       "epoch": 0.02,
-      "learning_rate": 0.000189873417721519,
-      "loss": 4.2747,
+      "learning_rate": 0.00028125,
+      "loss": 1.8662,
       "step": 15
     },
     {
       "epoch": 0.03,
-      "learning_rate": 0.00025316455696202533,
-      "loss": 3.9938,
+      "learning_rate": 0.00037499999999999995,
+      "loss": 1.8047,
       "step": 20
     },
     {
       "epoch": 0.03,
-      "learning_rate": 0.00031645569620253165,
-      "loss": 3.7136,
+      "learning_rate": 0.00046875,
+      "loss": 1.9482,
       "step": 25
     },
     {
       "epoch": 0.04,
-      "learning_rate": 0.000379746835443038,
-      "loss": 3.6955,
+      "learning_rate": 0.0005625,
+      "loss": 1.9624,
       "step": 30
     },
     {
       "epoch": 0.04,
-      "learning_rate": 0.0004430379746835443,
-      "loss": 3.4653,
+      "learning_rate": 0.0005999986347533043,
+      "loss": 1.8496,
       "step": 35
     },
     {
       "epoch": 0.05,
-      "learning_rate": 0.0005063291139240507,
-      "loss": 3.3515,
+      "learning_rate": 0.0005999902916240525,
+      "loss": 1.8962,
       "step": 40
     },
     {
       "epoch": 0.06,
-      "learning_rate": 0.000569620253164557,
-      "loss": 3.2781,
+      "learning_rate": 0.0005999743640466132,
+      "loss": 1.9655,
       "step": 45
     },
     {
       "epoch": 0.06,
-      "learning_rate": 0.0006329113924050633,
-      "loss": 3.2616,
+      "learning_rate": 0.0005999508524236727,
+      "loss": 1.9308,
       "step": 50
     },
     {
       "epoch": 0.07,
-      "learning_rate": 0.0006962025316455697,
-      "loss": 3.2314,
+      "learning_rate": 0.000599919757349659,
+      "loss": 2.0406,
       "step": 55
     },
     {
       "epoch": 0.08,
-      "learning_rate": 0.000759493670886076,
-      "loss": 3.1758,
+      "learning_rate": 0.0005998810796107273,
+      "loss": 1.8524,
       "step": 60
     },
     {
       "epoch": 0.08,
-      "learning_rate": 0.0008227848101265824,
-      "loss": 3.1956,
+      "learning_rate": 0.0005998348201847405,
+      "loss": 1.9677,
       "step": 65
     },
     {
       "epoch": 0.09,
-      "learning_rate": 0.0008860759493670886,
-      "loss": 3.1717,
+      "learning_rate": 0.0005997809802412437,
+      "loss": 1.963,
       "step": 70
     },
     {
       "epoch": 0.1,
-      "learning_rate": 0.0009493670886075949,
-      "loss": 3.0096,
+      "learning_rate": 0.0005997195611414353,
+      "loss": 1.7783,
       "step": 75
     },
     {
       "epoch": 0.1,
-      "learning_rate": 0.0009999998349118006,
-      "loss": 3.0798,
+      "learning_rate": 0.000599650564438132,
+      "loss": 1.9602,
       "step": 80
     },
     {
       "epoch": 0.11,
-      "learning_rate": 0.000999994056836268,
-      "loss": 3.0737,
+      "learning_rate": 0.0005995739918757302,
+      "loss": 1.941,
       "step": 85
     },
     {
       "epoch": 0.11,
-      "learning_rate": 0.000999980024459781,
-      "loss": 3.1079,
+      "learning_rate": 0.000599489845390161,
+      "loss": 1.908,
       "step": 90
     },
     {
       "epoch": 0.12,
-      "learning_rate": 0.0009999577380139975,
-      "loss": 2.9692,
+      "learning_rate": 0.0005993981271088423,
+      "loss": 1.8707,
       "step": 95
     },
     {
       "epoch": 0.13,
-      "learning_rate": 0.0009999271978668395,
-      "loss": 3.0226,
+      "learning_rate": 0.0005992988393506244,
+      "loss": 1.9848,
       "step": 100
     },
     {
       "epoch": 0.13,
-      "learning_rate": 0.0009998884045224886,
-      "loss": 2.8905,
+      "learning_rate": 0.0005991919846257313,
+      "loss": 1.9445,
       "step": 105
     },
     {
       "epoch": 0.14,
-      "learning_rate": 0.0009998413586213759,
-      "loss": 2.9354,
+      "learning_rate": 0.0005990775656356976,
+      "loss": 1.8844,
       "step": 110
     },
     {
       "epoch": 0.15,
-      "learning_rate": 0.0009997860609401732,
-      "loss": 2.9668,
+      "learning_rate": 0.0005989555852732997,
+      "loss": 1.9902,
       "step": 115
     },
     {
       "epoch": 0.15,
-      "learning_rate": 0.000999722512391778,
-      "loss": 2.939,
+      "learning_rate": 0.0005988260466224834,
+      "loss": 1.8382,
       "step": 120
     },
     {
       "epoch": 0.16,
-      "learning_rate": 0.0009996507140253013,
-      "loss": 2.9213,
+      "learning_rate": 0.0005986889529582852,
+      "loss": 2.0264,
       "step": 125
     },
     {
       "epoch": 0.16,
-      "learning_rate": 0.000999570667026047,
-      "loss": 2.9371,
+      "learning_rate": 0.00059854430774675,
+      "loss": 1.9612,
       "step": 130
     },
     {
       "epoch": 0.17,
-      "learning_rate": 0.0009994823727154955,
-      "loss": 2.9326,
+      "learning_rate": 0.000598392114644843,
+      "loss": 2.0919,
       "step": 135
     },
     {
       "epoch": 0.18,
-      "learning_rate": 0.0009993858325512794,
-      "loss": 2.9915,
+      "learning_rate": 0.0005982323775003579,
+      "loss": 1.9855,
       "step": 140
     },
     {
       "epoch": 0.18,
-      "learning_rate": 0.000999281048127161,
-      "loss": 2.9352,
+      "learning_rate": 0.0005980651003518189,
+      "loss": 1.8857,
       "step": 145
     },
     {
       "epoch": 0.19,
-      "learning_rate": 0.0009991680211730051,
-      "loss": 2.8284,
+      "learning_rate": 0.0005978902874283793,
+      "loss": 1.972,
       "step": 150
     },
     {
       "epoch": 0.2,
-      "learning_rate": 0.000999046753554751,
-      "loss": 2.8232,
+      "learning_rate": 0.0005977079431497138,
+      "loss": 1.8702,
       "step": 155
     },
     {
       "epoch": 0.2,
-      "learning_rate": 0.000998917247274381,
-      "loss": 2.8925,
+      "learning_rate": 0.0005975180721259074,
+      "loss": 1.886,
       "step": 160
     },
     {
       "epoch": 0.21,
-      "learning_rate": 0.0009987795044698884,
-      "loss": 2.8381,
+      "learning_rate": 0.0005973206791573386,
+      "loss": 1.9137,
       "step": 165
     },
     {
       "epoch": 0.22,
-      "learning_rate": 0.0009986335274152413,
-      "loss": 2.8939,
+      "learning_rate": 0.0005971157692345581,
+      "loss": 1.9225,
       "step": 170
     },
     {
       "epoch": 0.22,
-      "learning_rate": 0.0009984793185203455,
-      "loss": 2.9206,
+      "learning_rate": 0.0005969033475381623,
+      "loss": 1.8239,
       "step": 175
     },
     {
       "epoch": 0.23,
-      "learning_rate": 0.0009983168803310042,
-      "loss": 2.733,
+      "learning_rate": 0.000596683419438663,
+      "loss": 1.9465,
       "step": 180
     },
     {
       "epoch": 0.23,
-      "learning_rate": 0.0009981462155288774,
-      "loss": 2.8812,
+      "learning_rate": 0.0005964559904963512,
+      "loss": 1.8873,
       "step": 185
     },
     {
       "epoch": 0.24,
-      "learning_rate": 0.000997967326931435,
-      "loss": 2.7199,
+      "learning_rate": 0.0005962210664611563,
+      "loss": 1.9938,
       "step": 190
     },
     {
       "epoch": 0.25,
-      "learning_rate": 0.0009977802174919133,
-      "loss": 2.8825,
+      "learning_rate": 0.000595978653272501,
+      "loss": 1.9666,
       "step": 195
     },
     {
       "epoch": 0.25,
-      "learning_rate": 0.0009975848902992641,
-      "loss": 2.8283,
+      "learning_rate": 0.0005957287570591513,
+      "loss": 1.8345,
       "step": 200
     },
     {
       "epoch": 0.26,
-      "learning_rate": 0.0009973813485781044,
-      "loss": 2.8447,
+      "learning_rate": 0.000595471384139061,
+      "loss": 1.983,
       "step": 205
     },
     {
       "epoch": 0.27,
-      "learning_rate": 0.000997169595688664,
-      "loss": 2.7746,
+      "learning_rate": 0.000595206541019213,
+      "loss": 1.9706,
       "step": 210
     },
     {
       "epoch": 0.27,
-      "learning_rate": 0.0009969496351267277,
-      "loss": 2.6779,
+      "learning_rate": 0.0005949342343954535,
+      "loss": 1.9829,
       "step": 215
     },
     {
       "epoch": 0.28,
-      "learning_rate": 0.000996721470523581,
-      "loss": 2.8015,
+      "learning_rate": 0.0005946544711523236,
+      "loss": 1.8883,
       "step": 220
     },
     {
       "epoch": 0.29,
-      "learning_rate": 0.0009964851056459465,
-      "loss": 2.8746,
+      "learning_rate": 0.0005943672583628848,
+      "loss": 2.0109,
       "step": 225
     },
     {
       "epoch": 0.29,
-      "learning_rate": 0.0009962405443959244,
-      "loss": 2.7497,
+      "learning_rate": 0.0005940726032885408,
+      "loss": 1.9654,
       "step": 230
     },
     {
       "epoch": 0.3,
-      "learning_rate": 0.0009959877908109274,
-      "loss": 2.8197,
+      "learning_rate": 0.0005937705133788529,
+      "loss": 1.8733,
       "step": 235
     },
     {
       "epoch": 0.3,
-      "learning_rate": 0.0009957268490636132,
-      "loss": 2.7054,
+      "learning_rate": 0.0005934609962713525,
+      "loss": 1.9419,
       "step": 240
     },
     {
       "epoch": 0.31,
-      "learning_rate": 0.000995457723461816,
-      "loss": 2.7127,
+      "learning_rate": 0.0005931440597913479,
+      "loss": 1.9287,
       "step": 245
     },
     {
       "epoch": 0.32,
-      "learning_rate": 0.0009951804184484766,
-      "loss": 2.7058,
+      "learning_rate": 0.0005928197119517257,
+      "loss": 1.9507,
       "step": 250
     },
     {
       "epoch": 0.32,
-      "learning_rate": 0.0009948949386015676,
-      "loss": 2.7129,
+      "learning_rate": 0.0005924879609527498,
+      "loss": 1.9224,
       "step": 255
     },
     {
       "epoch": 0.33,
-      "learning_rate": 0.0009946012886340176,
-      "loss": 2.7425,
+      "learning_rate": 0.0005921488151818521,
+      "loss": 1.9124,
       "step": 260
     },
     {
       "epoch": 0.34,
-      "learning_rate": 0.000994299473393635,
-      "loss": 2.7676,
+      "learning_rate": 0.000591802283213422,
+      "loss": 1.9667,
       "step": 265
     },
     {
       "epoch": 0.34,
-      "learning_rate": 0.0009939894978630267,
-      "loss": 2.6728,
+      "learning_rate": 0.0005914483738085887,
+      "loss": 1.955,
       "step": 270
     },
     {
       "epoch": 0.35,
-      "learning_rate": 0.0009936713671595157,
-      "loss": 2.7671,
+      "learning_rate": 0.0005910870959150005,
+      "loss": 1.949,
       "step": 275
     },
     {
       "epoch": 0.35,
-      "learning_rate": 0.0009933450865350583,
-      "loss": 2.6662,
+      "learning_rate": 0.0005907184586665977,
+      "loss": 1.7752,
       "step": 280
     },
     {
       "epoch": 0.36,
-      "learning_rate": 0.000993010661376155,
-      "loss": 2.7334,
+      "learning_rate": 0.0005903424713833823,
+      "loss": 1.8766,
       "step": 285
     },
     {
       "epoch": 0.37,
-      "learning_rate": 0.000992668097203763,
-      "loss": 2.7489,
+      "learning_rate": 0.0005899591435711821,
+      "loss": 2.0573,
       "step": 290
     },
     {
       "epoch": 0.37,
-      "learning_rate": 0.0009923173996732057,
-      "loss": 2.6428,
+      "learning_rate": 0.0005895684849214106,
+      "loss": 1.9837,
       "step": 295
     },
     {
       "epoch": 0.38,
-      "learning_rate": 0.0009919585745740776,
-      "loss": 2.7059,
+      "learning_rate": 0.0005891705053108217,
+      "loss": 2.0551,
       "step": 300
     },
     {
       "epoch": 0.39,
-      "learning_rate": 0.0009915916278301496,
-      "loss": 2.6961,
+      "learning_rate": 0.0005887652148012601,
+      "loss": 2.0254,
       "step": 305
     },
     {
       "epoch": 0.39,
-      "learning_rate": 0.0009912165654992716,
-      "loss": 2.8165,
+      "learning_rate": 0.000588352623639407,
+      "loss": 2.0329,
       "step": 310
     },
     {
       "epoch": 0.4,
-      "learning_rate": 0.0009908333937732718,
-      "loss": 2.5846,
+      "learning_rate": 0.0005879327422565209,
+      "loss": 1.8451,
       "step": 315
     },
     {
       "epoch": 0.41,
-      "learning_rate": 0.0009904421189778551,
-      "loss": 2.6058,
+      "learning_rate": 0.000587505581268174,
+      "loss": 1.9613,
       "step": 320
     },
     {
       "epoch": 0.41,
-      "learning_rate": 0.000990042747572498,
-      "loss": 2.6858,
+      "learning_rate": 0.0005870711514739836,
+      "loss": 1.9022,
       "step": 325
     },
     {
       "epoch": 0.42,
-      "learning_rate": 0.0009896352861503426,
-      "loss": 2.6797,
+      "learning_rate": 0.0005866294638573397,
+      "loss": 2.0333,
       "step": 330
     },
     {
       "epoch": 0.42,
-      "learning_rate": 0.0009892197414380869,
-      "loss": 2.7078,
+      "learning_rate": 0.0005861805295851261,
+      "loss": 1.8614,
       "step": 335
     },
     {
       "epoch": 0.43,
-      "learning_rate": 0.0009887961202958747,
-      "loss": 2.5945,
+      "learning_rate": 0.0005857243600074394,
+      "loss": 1.937,
       "step": 340
     },
     {
       "epoch": 0.44,
-      "learning_rate": 0.000988364429717182,
-      "loss": 2.6061,
+      "learning_rate": 0.0005852609666573011,
+      "loss": 1.8581,
       "step": 345
     },
     {
       "epoch": 0.44,
-      "learning_rate": 0.0009879246768287019,
-      "loss": 2.6557,
+      "learning_rate": 0.0005847903612503668,
+      "loss": 1.9423,
       "step": 350
     },
     {
       "epoch": 0.45,
-      "learning_rate": 0.0009874768688902251,
-      "loss": 2.6387,
+      "learning_rate": 0.000584312555684629,
+      "loss": 1.9807,
       "step": 355
     },
     {
       "epoch": 0.46,
-      "learning_rate": 0.0009870210132945235,
-      "loss": 2.6668,
+      "learning_rate": 0.0005838275620401171,
+      "loss": 1.9902,
       "step": 360
     },
     {
       "epoch": 0.46,
-      "learning_rate": 0.0009865571175672243,
-      "loss": 2.7257,
+      "learning_rate": 0.0005833353925785922,
+      "loss": 1.8792,
       "step": 365
     },
     {
       "epoch": 0.47,
-      "learning_rate": 0.0009860851893666888,
-      "loss": 2.7604,
+      "learning_rate": 0.0005828360597432358,
+      "loss": 1.9335,
       "step": 370
     },
     {
       "epoch": 0.48,
-      "learning_rate": 0.0009856052364838846,
-      "loss": 2.6524,
+      "learning_rate": 0.0005823295761583368,
+      "loss": 1.9518,
       "step": 375
     },
     {
       "epoch": 0.48,
-      "learning_rate": 0.0009851172668422564,
-      "loss": 2.5279,
+      "learning_rate": 0.0005818159546289711,
+      "loss": 2.0066,
       "step": 380
     },
     {
       "epoch": 0.49,
-      "learning_rate": 0.0009846212884975969,
-      "loss": 2.5394,
+      "learning_rate": 0.0005812952081406786,
+      "loss": 2.0164,
       "step": 385
     },
     {
       "epoch": 0.49,
-      "learning_rate": 0.0009841173096379124,
-      "loss": 2.6585,
+      "learning_rate": 0.0005807673498591347,
+      "loss": 2.0275,
       "step": 390
     },
     {
       "epoch": 0.5,
-      "learning_rate": 0.000983605338583288,
-      "loss": 2.6494,
+      "learning_rate": 0.0005802323931298169,
+      "loss": 1.983,
       "step": 395
     },
     {
       "epoch": 0.51,
-      "learning_rate": 0.0009830853837857507,
-      "loss": 2.6424,
+      "learning_rate": 0.0005796903514776685,
+      "loss": 1.9721,
       "step": 400
     },
     {
       "epoch": 0.51,
-      "learning_rate": 0.0009825574538291292,
-      "loss": 2.6346,
+      "learning_rate": 0.0005791412386067555,
+      "loss": 1.8981,
       "step": 405
     },
     {
       "epoch": 0.52,
-      "learning_rate": 0.000982021557428913,
-      "loss": 2.6241,
+      "learning_rate": 0.000578585068399921,
+      "loss": 2.073,
       "step": 410
     },
     {
       "epoch": 0.53,
-      "learning_rate": 0.0009814777034321067,
-      "loss": 2.5922,
+      "learning_rate": 0.0005780218549184339,
+      "loss": 1.9388,
       "step": 415
     },
     {
       "epoch": 0.53,
-      "learning_rate": 0.0009809259008170868,
-      "loss": 2.6258,
+      "learning_rate": 0.000577451612401633,
+      "loss": 2.0487,
       "step": 420
     },
     {
       "epoch": 0.54,
-      "learning_rate": 0.0009803661586934513,
-      "loss": 2.6141,
+      "learning_rate": 0.0005768743552665678,
+      "loss": 2.0537,
       "step": 425
     },
     {
       "epoch": 0.54,
-      "learning_rate": 0.0009797984863018697,
-      "loss": 2.624,
+      "learning_rate": 0.0005762900981076332,
+      "loss": 1.9692,
       "step": 430
     },
     {
       "epoch": 0.55,
-      "learning_rate": 0.000979222893013931,
-      "loss": 2.6261,
+      "learning_rate": 0.000575698855696201,
+      "loss": 1.9719,
       "step": 435
     },
     {
       "epoch": 0.56,
-      "learning_rate": 0.0009786393883319884,
-      "loss": 2.4675,
+      "learning_rate": 0.0005751006429802467,
+      "loss": 1.9929,
       "step": 440
     },
     {
       "epoch": 0.56,
-      "learning_rate": 0.0009780479818890032,
-      "loss": 2.6471,
+      "learning_rate": 0.0005744954750839706,
+      "loss": 2.0479,
       "step": 445
     },
     {
       "epoch": 0.57,
-      "learning_rate": 0.0009774486834483848,
-      "loss": 2.6596,
+      "learning_rate": 0.0005738833673074165,
+      "loss": 1.8216,
       "step": 450
     },
     {
       "epoch": 0.58,
-      "learning_rate": 0.0009768415029038306,
-      "loss": 2.6316,
+      "learning_rate": 0.0005732643351260843,
+      "loss": 1.8762,
       "step": 455
     },
     {
       "epoch": 0.58,
-      "learning_rate": 0.0009762264502791609,
-      "loss": 2.6561,
+      "learning_rate": 0.0005726383941905386,
+      "loss": 1.9418,
       "step": 460
     },
     {
       "epoch": 0.59,
-      "learning_rate": 0.0009756035357281559,
-      "loss": 2.6207,
+      "learning_rate": 0.0005720055603260138,
+      "loss": 2.0108,
       "step": 465
     },
     {
       "epoch": 0.6,
-      "learning_rate": 0.0009749727695343859,
-      "loss": 2.4761,
+      "learning_rate": 0.0005713658495320128,
+      "loss": 1.9163,
       "step": 470
     },
     {
       "epoch": 0.6,
-      "learning_rate": 0.0009743341621110429,
-      "loss": 2.5767,
+      "learning_rate": 0.000570719277981904,
+      "loss": 1.9502,
       "step": 475
     },
     {
       "epoch": 0.61,
-      "learning_rate": 0.0009736877240007677,
-      "loss": 2.5302,
+      "learning_rate": 0.0005700658620225108,
+      "loss": 2.1221,
       "step": 480
     },
     {
       "epoch": 0.61,
-      "learning_rate": 0.0009730334658754766,
-      "loss": 2.5426,
+      "learning_rate": 0.0005694056181736996,
+      "loss": 1.9281,
       "step": 485
     },
     {
       "epoch": 0.62,
-      "learning_rate": 0.0009723713985361853,
-      "loss": 2.5384,
+      "learning_rate": 0.0005687385631279615,
+      "loss": 2.0308,
       "step": 490
     },
     {
       "epoch": 0.63,
-      "learning_rate": 0.0009717015329128294,
-      "loss": 2.4947,
+      "learning_rate": 0.0005680647137499902,
+      "loss": 1.9564,
       "step": 495
     },
     {
       "epoch": 0.63,
-      "learning_rate": 0.0009710238800640854,
-      "loss": 2.5495,
+      "learning_rate": 0.0005673840870762561,
+      "loss": 1.9229,
       "step": 500
     },
     {
       "epoch": 0.64,
-      "learning_rate": 0.0009703384511771874,
-      "loss": 2.5251,
+      "learning_rate": 0.0005666967003145753,
+      "loss": 2.0006,
       "step": 505
     },
     {
       "epoch": 0.65,
-      "learning_rate": 0.0009696452575677426,
-      "loss": 2.4983,
+      "learning_rate": 0.0005660025708436743,
+      "loss": 2.0517,
       "step": 510
     },
     {
       "epoch": 0.65,
-      "learning_rate": 0.0009689443106795442,
-      "loss": 2.4939,
+      "learning_rate": 0.0005653017162127511,
+      "loss": 1.9226,
       "step": 515
     },
     {
       "epoch": 0.66,
-      "learning_rate": 0.000968235622084383,
-      "loss": 2.3895,
+      "learning_rate": 0.0005645941541410314,
+      "loss": 1.9131,
       "step": 520
     },
     {
       "epoch": 0.67,
-      "learning_rate": 0.0009675192034818561,
-      "loss": 2.4947,
+      "learning_rate": 0.00056387990251732,
+      "loss": 2.0067,
       "step": 525
     },
     {
       "epoch": 0.67,
-      "learning_rate": 0.0009667950666991733,
-      "loss": 2.4988,
+      "learning_rate": 0.0005631589793995495,
+      "loss": 1.7818,
       "step": 530
     },
     {
       "epoch": 0.68,
-      "learning_rate": 0.0009660632236909627,
-      "loss": 2.573,
+      "learning_rate": 0.0005624314030143236,
+      "loss": 1.9343,
       "step": 535
     },
     {
       "epoch": 0.68,
-      "learning_rate": 0.0009653236865390729,
-      "loss": 2.6162,
+      "learning_rate": 0.0005616971917564549,
+      "loss": 1.9357,
       "step": 540
     },
     {
       "epoch": 0.69,
-      "learning_rate": 0.0009645764674523731,
-      "loss": 2.613,
+      "learning_rate": 0.000560956364188502,
+      "loss": 1.8905,
       "step": 545
     },
     {
       "epoch": 0.7,
-      "learning_rate": 0.0009638215787665529,
-      "loss": 2.508,
+      "learning_rate": 0.0005602089390402986,
+      "loss": 1.9669,
       "step": 550
     },
     {
       "epoch": 0.7,
-      "learning_rate": 0.0009630590329439169,
-      "loss": 2.5019,
+      "learning_rate": 0.0005594549352084807,
+      "loss": 1.9771,
       "step": 555
     },
     {
       "epoch": 0.71,
-      "learning_rate": 0.0009622888425731801,
-      "loss": 2.622,
+      "learning_rate": 0.0005586943717560083,
+      "loss": 1.9627,
       "step": 560
     },
     {
       "epoch": 0.72,
-      "learning_rate": 0.0009615110203692601,
-      "loss": 2.4998,
+      "learning_rate": 0.0005579272679116841,
+      "loss": 1.9329,
       "step": 565
     },
     {
       "epoch": 0.72,
-      "learning_rate": 0.0009607255791730664,
-      "loss": 2.467,
+      "learning_rate": 0.000557153643069667,
+      "loss": 1.9314,
       "step": 570
     },
     {
       "epoch": 0.73,
-      "learning_rate": 0.0009599325319512893,
-      "loss": 2.6073,
+      "learning_rate": 0.0005563735167889817,
+      "loss": 1.8892,
       "step": 575
     },
     {
       "epoch": 0.73,
-      "learning_rate": 0.0009591318917961853,
-      "loss": 2.6101,
+      "learning_rate": 0.0005555869087930242,
+      "loss": 2.0062,
       "step": 580
     },
     {
       "epoch": 0.74,
-      "learning_rate": 0.000958323671925361,
-      "loss": 2.4277,
+      "learning_rate": 0.0005547938389690635,
+      "loss": 1.9624,
       "step": 585
     },
     {
       "epoch": 0.75,
-      "learning_rate": 0.0009575078856815554,
-      "loss": 2.5171,
+      "learning_rate": 0.0005539943273677384,
+      "loss": 2.0077,
       "step": 590
     },
     {
       "epoch": 0.75,
-      "learning_rate": 0.0009566845465324185,
-      "loss": 2.5376,
+      "learning_rate": 0.0005531883942025507,
+      "loss": 2.0083,
       "step": 595
     },
     {
       "epoch": 0.76,
-      "learning_rate": 0.0009558536680702903,
-      "loss": 2.5084,
+      "learning_rate": 0.0005523760598493544,
+      "loss": 2.0212,
       "step": 600
     },
     {
       "epoch": 0.77,
-      "learning_rate": 0.0009550152640119756,
-      "loss": 2.532,
+      "learning_rate": 0.0005515573448458399,
+      "loss": 1.9307,
       "step": 605
     },
     {
       "epoch": 0.77,
-      "learning_rate": 0.0009541693481985178,
-      "loss": 2.5169,
+      "learning_rate": 0.0005507322698910156,
+      "loss": 1.7577,
       "step": 610
     },
     {
       "epoch": 0.78,
-      "learning_rate": 0.0009533159345949704,
-      "loss": 2.6213,
+      "learning_rate": 0.000549900855844684,
+      "loss": 1.9654,
       "step": 615
     },
     {
       "epoch": 0.79,
-      "learning_rate": 0.000952455037290166,
-      "loss": 2.4695,
+      "learning_rate": 0.0005490631237269147,
+      "loss": 1.7376,
       "step": 620
     },
     {
       "epoch": 0.79,
-      "learning_rate": 0.0009515866704964846,
-      "loss": 2.4018,
+      "learning_rate": 0.0005482190947175124,
+      "loss": 1.9602,
       "step": 625
     },
     {
       "epoch": 0.8,
-      "learning_rate": 0.0009507108485496185,
-      "loss": 2.5084,
+      "learning_rate": 0.0005473687901554821,
+      "loss": 1.9872,
       "step": 630
     },
     {
       "epoch": 0.8,
-      "learning_rate": 0.0009498275859083352,
-      "loss": 2.5441,
+      "learning_rate": 0.0005465122315384889,
+      "loss": 1.9741,
       "step": 635
     },
     {
       "epoch": 0.81,
-      "learning_rate": 0.0009489368971542393,
-      "loss": 2.4815,
+      "learning_rate": 0.0005456494405223152,
+      "loss": 1.9905,
       "step": 640
     },
     {
       "epoch": 0.82,
-      "learning_rate": 0.0009480387969915317,
-      "loss": 2.468,
+      "learning_rate": 0.0005447804389203129,
+      "loss": 1.8507,
       "step": 645
     },
     {
       "epoch": 0.82,
-      "learning_rate": 0.0009471333002467668,
-      "loss": 2.5972,
+      "learning_rate": 0.0005439052487028517,
+      "loss": 1.987,
       "step": 650
     },
     {
       "epoch": 0.83,
-      "learning_rate": 0.0009462204218686074,
-      "loss": 2.4825,
+      "learning_rate": 0.0005430238919967637,
+      "loss": 1.9029,
       "step": 655
     },
     {
       "epoch": 0.84,
-      "learning_rate": 0.0009453001769275786,
-      "loss": 2.5862,
+      "learning_rate": 0.0005421363910847844,
+      "loss": 1.9159,
       "step": 660
     },
     {
       "epoch": 0.84,
-      "learning_rate": 0.0009443725806158182,
-      "loss": 2.5174,
+      "learning_rate": 0.0005412427684049888,
+      "loss": 1.914,
       "step": 665
     },
     {
       "epoch": 0.85,
-      "learning_rate": 0.0009434376482468265,
-      "loss": 2.3676,
+      "learning_rate": 0.0005403430465502244,
+      "loss": 1.9955,
       "step": 670
     },
     {
       "epoch": 0.86,
-      "learning_rate": 0.0009424953952552133,
-      "loss": 2.4715,
+      "learning_rate": 0.00053943724826754,
+      "loss": 1.9325,
       "step": 675
     },
     {
       "epoch": 0.86,
-      "learning_rate": 0.0009415458371964429,
-      "loss": 2.5297,
+      "learning_rate": 0.0005385253964576104,
+      "loss": 1.8784,
       "step": 680
     },
     {
       "epoch": 0.87,
-      "learning_rate": 0.0009405889897465778,
-      "loss": 2.5442,
+      "learning_rate": 0.000537607514174158,
+      "loss": 1.8416,
       "step": 685
     },
     {
       "epoch": 0.87,
-      "learning_rate": 0.0009396248687020191,
-      "loss": 2.5398,
+      "learning_rate": 0.0005366836246233687,
+      "loss": 1.9717,
       "step": 690
     },
     {
       "epoch": 0.88,
-      "learning_rate": 0.000938653489979247,
-      "loss": 2.6478,
+      "learning_rate": 0.0005357537511633068,
+      "loss": 1.8637,
       "step": 695
     },
     {
       "epoch": 0.89,
-      "learning_rate": 0.000937674869614556,
-      "loss": 2.4237,
+      "learning_rate": 0.0005348179173033233,
+      "loss": 1.804,
       "step": 700
     },
     {
       "epoch": 0.89,
-      "learning_rate": 0.000936689023763793,
-      "loss": 2.5133,
+      "learning_rate": 0.000533876146703462,
+      "loss": 1.9949,
       "step": 705
     },
     {
       "epoch": 0.9,
-      "learning_rate": 0.0009356959687020876,
-      "loss": 2.3837,
+      "learning_rate": 0.0005329284631738609,
+      "loss": 1.9731,
       "step": 710
     },
     {
       "epoch": 0.91,
-      "learning_rate": 0.0009346957208235855,
-      "loss": 2.5308,
+      "learning_rate": 0.0005319748906741511,
+      "loss": 1.8443,
       "step": 715
     },
     {
       "epoch": 0.91,
-      "learning_rate": 0.000933688296641177,
-      "loss": 2.4463,
+      "learning_rate": 0.0005310154533128501,
+      "loss": 1.8656,
       "step": 720
     },
     {
       "epoch": 0.92,
-      "learning_rate": 0.0009326737127862249,
-      "loss": 2.4815,
+      "learning_rate": 0.0005300501753467525,
+      "loss": 1.9067,
       "step": 725
     },
     {
       "epoch": 0.93,
-      "learning_rate": 0.0009316519860082886,
-      "loss": 2.5258,
+      "learning_rate": 0.0005290790811803176,
+      "loss": 1.9235,
       "step": 730
     },
     {
       "epoch": 0.93,
-      "learning_rate": 0.0009306231331748496,
-      "loss": 2.4161,
+      "learning_rate": 0.0005281021953650507,
+      "loss": 2.0954,
       "step": 735
     },
     {
       "epoch": 0.94,
-      "learning_rate": 0.0009295871712710317,
-      "loss": 2.5333,
+      "learning_rate": 0.0005271195425988842,
+      "loss": 1.9317,
       "step": 740
     },
     {
       "epoch": 0.94,
-      "learning_rate": 0.0009285441173993206,
-      "loss": 2.3897,
+      "learning_rate": 0.000526131147725552,
+      "loss": 1.9577,
       "step": 745
     },
     {
       "epoch": 0.95,
-      "learning_rate": 0.0009274939887792823,
-      "loss": 2.5545,
+      "learning_rate": 0.0005251370357339616,
+      "loss": 1.8044,
       "step": 750
     },
     {
       "epoch": 0.96,
-      "learning_rate": 0.0009264368027472784,
-      "loss": 2.3463,
+      "learning_rate": 0.0005241372317575629,
+      "loss": 2.0009,
       "step": 755
     },
     {
       "epoch": 0.96,
-      "learning_rate": 0.0009253725767561793,
-      "loss": 2.4669,
+      "learning_rate": 0.0005231317610737121,
+      "loss": 1.875,
       "step": 760
     },
     {
       "epoch": 0.97,
-      "learning_rate": 0.0009243013283750773,
-      "loss": 2.331,
+      "learning_rate": 0.0005221206491030327,
+      "loss": 1.9521,
       "step": 765
     },
     {
       "epoch": 0.98,
-      "learning_rate": 0.0009232230752889957,
-      "loss": 2.2515,
+      "learning_rate": 0.0005211039214087733,
+      "loss": 2.0129,
       "step": 770
     },
     {
       "epoch": 0.98,
-      "learning_rate": 0.0009221378352985967,
-      "loss": 2.462,
+      "learning_rate": 0.0005200816036961606,
+      "loss": 1.9002,
       "step": 775
     },
     {
       "epoch": 0.99,
-      "learning_rate": 0.0009210456263198886,
-      "loss": 2.3609,
+      "learning_rate": 0.0005190537218117502,
+      "loss": 1.9327,
       "step": 780
     },
     {
       "epoch": 0.99,
-      "learning_rate": 0.0009199464663839289,
-      "loss": 2.4393,
+      "learning_rate": 0.0005180203017427728,
+      "loss": 1.9045,
       "step": 785
     },
     {
       "epoch": 1.0,
-      "eval_loss": 2.382097005844116,
-      "eval_runtime": 95.3367,
-      "eval_samples_per_second": 134.345,
-      "eval_steps_per_second": 16.793,
+      "eval_loss": 2.0005805492401123,
+      "eval_runtime": 92.7966,
+      "eval_samples_per_second": 138.022,
+      "eval_steps_per_second": 8.632,
       "step": 789
     },
     {
       "epoch": 1.0,
-      "learning_rate": 0.000918840373636527,
-      "loss": 2.5088,
+      "learning_rate": 0.0005169813696164772,
+      "loss": 2.0109,
       "step": 790
     },
     {
       "epoch": 1.01,
-      "learning_rate": 0.0009177273663379448,
-      "loss": 2.1884,
+      "learning_rate": 0.0005159369516994695,
+      "loss": 1.7846,
       "step": 795
     },
     {
       "epoch": 1.01,
-      "learning_rate": 0.0009166074628625952,
-      "loss": 2.4646,
+      "learning_rate": 0.0005148870743970495,
+      "loss": 1.8012,
       "step": 800
     },
     {
       "epoch": 1.02,
-      "learning_rate": 0.0009154806816987386,
-      "loss": 2.4983,
+      "learning_rate": 0.000513831764252543,
+      "loss": 1.9483,
       "step": 805
     },
     {
       "epoch": 1.03,
-      "learning_rate": 0.0009143470414481776,
-      "loss": 2.3381,
+      "learning_rate": 0.0005127710479466303,
+      "loss": 1.9163,
       "step": 810
     },
     {
       "epoch": 1.03,
-      "learning_rate": 0.0009132065608259505,
-      "loss": 2.3715,
+      "learning_rate": 0.0005117049522966723,
+      "loss": 1.8544,
       "step": 815
     },
     {
       "epoch": 1.04,
-      "learning_rate": 0.0009120592586600216,
-      "loss": 2.4095,
+      "learning_rate": 0.0005106335042560317,
+      "loss": 1.7812,
       "step": 820
     },
     {
       "epoch": 1.05,
-      "learning_rate": 0.0009109051538909707,
-      "loss": 2.3306,
+      "learning_rate": 0.0005095567309133926,
+      "loss": 1.8897,
       "step": 825
     },
     {
       "epoch": 1.05,
-      "learning_rate": 0.0009097442655716803,
-      "loss": 2.3049,
+      "learning_rate": 0.0005084746594920745,
+      "loss": 1.8461,
       "step": 830
     },
     {
       "epoch": 1.06,
-      "learning_rate": 0.0009085766128670217,
-      "loss": 2.3163,
+      "learning_rate": 0.000507387317349345,
+      "loss": 1.8396,
       "step": 835
     },
     {
       "epoch": 1.06,
-      "learning_rate": 0.0009074022150535374,
-      "loss": 2.3609,
+      "learning_rate": 0.0005062947319757274,
+      "loss": 1.8366,
       "step": 840
     },
     {
       "epoch": 1.07,
-      "learning_rate": 0.000906221091519124,
-      "loss": 2.1964,
+      "learning_rate": 0.0005051969309943062,
+      "loss": 1.8389,
       "step": 845
     },
     {
       "epoch": 1.08,
-      "learning_rate": 0.0009050332617627115,
-      "loss": 2.3304,
+      "learning_rate": 0.0005040939421600284,
+      "loss": 1.7767,
       "step": 850
     },
     {
       "epoch": 1.08,
-      "learning_rate": 0.0009038387453939415,
-      "loss": 2.3855,
+      "learning_rate": 0.0005029857933590021,
+      "loss": 1.8001,
       "step": 855
     },
     {
       "epoch": 1.09,
-      "learning_rate": 0.0009026375621328436,
-      "loss": 2.2852,
+      "learning_rate": 0.0005018725126077912,
+      "loss": 1.8145,
       "step": 860
     },
     {
       "epoch": 1.1,
-      "learning_rate": 0.0009014297318095099,
-      "loss": 2.2837,
+      "learning_rate": 0.0005007541280527068,
+      "loss": 1.8317,
       "step": 865
     },
     {
       "epoch": 1.1,
-      "learning_rate": 0.0009002152743637673,
-      "loss": 2.2163,
+      "learning_rate": 0.0004996306679690967,
+      "loss": 1.8302,
       "step": 870
     },
     {
       "epoch": 1.11,
-      "learning_rate": 0.0008989942098448485,
-      "loss": 2.2896,
+      "learning_rate": 0.0004985021607606293,
+      "loss": 1.9682,
       "step": 875
     },
     {
       "epoch": 1.12,
-      "learning_rate": 0.0008977665584110612,
-      "loss": 2.3854,
+      "learning_rate": 0.0004973686349585758,
+      "loss": 1.841,
       "step": 880
     },
     {
       "epoch": 1.12,
-      "learning_rate": 0.0008965323403294553,
-      "loss": 2.3061,
+      "learning_rate": 0.0004962301192210897,
+      "loss": 1.8658,
       "step": 885
     },
     {
       "epoch": 1.13,
-      "learning_rate": 0.0008952915759754878,
-      "loss": 2.3625,
+      "learning_rate": 0.000495086642332481,
+      "loss": 1.8247,
       "step": 890
     },
     {
       "epoch": 1.13,
-      "learning_rate": 0.000894044285832687,
-      "loss": 2.4091,
+      "learning_rate": 0.0004939382332024895,
+      "loss": 1.8037,
       "step": 895
     },
     {
       "epoch": 1.14,
-      "learning_rate": 0.0008927904904923142,
-      "loss": 2.3056,
+      "learning_rate": 0.000492784920865553,
+      "loss": 1.9416,
       "step": 900
     },
     {
       "epoch": 1.15,
-      "learning_rate": 0.0008915302106530234,
-      "loss": 2.3323,
+      "learning_rate": 0.0004916267344800742,
+      "loss": 1.8733,
       "step": 905
     },
     {
       "epoch": 1.15,
-      "learning_rate": 0.0008902634671205201,
-      "loss": 2.3386,
+      "learning_rate": 0.0004904637033276827,
+      "loss": 1.8063,
       "step": 910
     },
     {
       "epoch": 1.16,
-      "learning_rate": 0.0008889902808072176,
-      "loss": 2.3661,
+      "learning_rate": 0.000489295856812495,
+      "loss": 1.9041,
       "step": 915
     },
     {
       "epoch": 1.17,
-      "learning_rate": 0.0008877106727318918,
-      "loss": 2.3923,
+      "learning_rate": 0.0004881232244603716,
+      "loss": 1.8182,
       "step": 920
     },
     {
       "epoch": 1.17,
-      "learning_rate": 0.0008864246640193341,
-      "loss": 2.2784,
+      "learning_rate": 0.0004869458359181693,
+      "loss": 1.7881,
       "step": 925
     },
     {
       "epoch": 1.18,
-      "learning_rate": 0.0008851322759000026,
-      "loss": 2.3131,
+      "learning_rate": 0.00048576372095299273,
+      "loss": 1.7624,
       "step": 930
     },
     {
       "epoch": 1.19,
-      "learning_rate": 0.0008838335297096721,
-      "loss": 2.4255,
+      "learning_rate": 0.00048457690945144147,
+      "loss": 1.9506,
       "step": 935
     },
     {
       "epoch": 1.19,
-      "learning_rate": 0.0008825284468890808,
-      "loss": 2.4198,
+      "learning_rate": 0.00048338543141885444,
+      "loss": 1.8866,
       "step": 940
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.0008812170489835783,
-      "loss": 2.3279,
+      "learning_rate": 0.000482189316978551,
+      "loss": 1.8682,
       "step": 945
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.0008798993576427671,
-      "loss": 2.3672,
+      "learning_rate": 0.0004809885963710698,
+      "loss": 1.8928,
       "step": 950
     },
     {
       "epoch": 1.21,
-      "learning_rate": 0.0008785753946201483,
-      "loss": 2.3138,
+      "learning_rate": 0.00047978329995340403,
+      "loss": 1.8434,
       "step": 955
     },
     {
       "epoch": 1.22,
-      "learning_rate": 0.0008772451817727598,
-      "loss": 2.2753,
+      "learning_rate": 0.000478573458198234,
+      "loss": 1.8086,
       "step": 960
     },
     {
       "epoch": 1.22,
-      "learning_rate": 0.0008759087410608169,
-      "loss": 2.2821,
+      "learning_rate": 0.00047735910169315645,
+      "loss": 1.9049,
       "step": 965
     },
     {
       "epoch": 1.23,
-      "learning_rate": 0.0008745660945473495,
-      "loss": 2.2495,
+      "learning_rate": 0.0004761402611399116,
+      "loss": 1.7393,
       "step": 970
     },
     {
       "epoch": 1.24,
-      "learning_rate": 0.0008732172643978382,
-      "loss": 2.47,
+      "learning_rate": 0.0004749169673536068,
+      "loss": 1.7668,
       "step": 975
     },
     {
       "epoch": 1.24,
-      "learning_rate": 0.0008718622728798476,
-      "loss": 2.3261,
+      "learning_rate": 0.0004736892512619375,
+      "loss": 1.8751,
       "step": 980
     },
     {
       "epoch": 1.25,
-      "learning_rate": 0.0008705011423626587,
-      "loss": 2.3,
+      "learning_rate": 0.0004724571439044048,
+      "loss": 1.7533,
       "step": 985
     },
     {
       "epoch": 1.25,
-      "learning_rate": 0.0008691338953169009,
-      "loss": 2.3209,
+      "learning_rate": 0.0004712206764315317,
+      "loss": 1.7789,
       "step": 990
     },
     {
       "epoch": 1.26,
-      "learning_rate": 0.0008677605543141796,
-      "loss": 2.3039,
+      "learning_rate": 0.00046997988010407483,
+      "loss": 1.9063,
       "step": 995
     },
     {
       "epoch": 1.27,
-      "learning_rate": 0.000866381142026704,
-      "loss": 2.2033,
+      "learning_rate": 0.0004687347862922342,
+      "loss": 1.7936,
       "step": 1000
     },
     {
       "epoch": 1.27,
-      "learning_rate": 0.0008649956812269133,
-      "loss": 2.3297,
+      "learning_rate": 0.00046748542647486,
+      "loss": 1.784,
       "step": 1005
     },
     {
       "epoch": 1.28,
-      "learning_rate": 0.0008636041947871001,
-      "loss": 2.273,
+      "learning_rate": 0.0004662318322386573,
+      "loss": 1.8396,
       "step": 1010
     },
     {
       "epoch": 1.29,
-      "learning_rate": 0.0008622067056790332,
-      "loss": 2.2885,
+      "learning_rate": 0.0004649740352773868,
+      "loss": 1.8145,
       "step": 1015
     },
     {
       "epoch": 1.29,
-      "learning_rate": 0.0008608032369735781,
-      "loss": 2.3532,
+      "learning_rate": 0.00046371206739106383,
+      "loss": 1.7536,
       "step": 1020
     },
     {
       "epoch": 1.3,
-      "learning_rate": 0.0008593938118403164,
-      "loss": 2.3094,
+      "learning_rate": 0.00046244596048515443,
+      "loss": 1.7931,
       "step": 1025
     },
     {
       "epoch": 1.31,
-      "learning_rate": 0.0008579784535471631,
-      "loss": 2.3238,
+      "learning_rate": 0.00046117574656976855,
+      "loss": 1.8734,
       "step": 1030
     },
     {
       "epoch": 1.31,
-      "learning_rate": 0.0008565571854599825,
-      "loss": 2.2605,
+      "learning_rate": 0.0004599014577588509,
+      "loss": 1.8342,
       "step": 1035
     },
     {
       "epoch": 1.32,
-      "learning_rate": 0.0008551300310422027,
-      "loss": 2.2964,
+      "learning_rate": 0.00045862312626936884,
+      "loss": 1.8973,
       "step": 1040
     },
     {
       "epoch": 1.32,
-      "learning_rate": 0.0008536970138544277,
-      "loss": 2.2659,
+      "learning_rate": 0.00045734078442049826,
+      "loss": 1.7734,
       "step": 1045
     },
     {
       "epoch": 1.33,
-      "learning_rate": 0.0008522581575540491,
-      "loss": 2.3661,
+      "learning_rate": 0.00045605446463280576,
+      "loss": 1.7644,
       "step": 1050
     },
     {
       "epoch": 1.34,
-      "learning_rate": 0.0008508134858948553,
-      "loss": 2.3099,
+      "learning_rate": 0.00045476419942742994,
+      "loss": 1.8709,
       "step": 1055
     },
     {
       "epoch": 1.34,
-      "learning_rate": 0.0008493630227266387,
-      "loss": 2.291,
+      "learning_rate": 0.0004534700214252584,
+      "loss": 1.7612,
       "step": 1060
     },
     {
       "epoch": 1.35,
-      "learning_rate": 0.0008479067919948031,
-      "loss": 2.4099,
+      "learning_rate": 0.0004521719633461034,
+      "loss": 1.9362,
       "step": 1065
     },
     {
       "epoch": 1.36,
-      "learning_rate": 0.000846444817739968,
-      "loss": 2.2842,
+      "learning_rate": 0.0004508700580078742,
+      "loss": 1.8653,
       "step": 1070
     },
     {
       "epoch": 1.36,
-      "learning_rate": 0.0008449771240975707,
-      "loss": 2.3229,
+      "learning_rate": 0.0004495643383257482,
+      "loss": 1.767,
       "step": 1075
     },
     {
       "epoch": 1.37,
-      "learning_rate": 0.0008435037352974696,
-      "loss": 2.2471,
+      "learning_rate": 0.0004482548373113377,
+      "loss": 1.829,
       "step": 1080
     },
     {
       "epoch": 1.38,
-      "learning_rate": 0.000842024675663543,
-      "loss": 2.3572,
+      "learning_rate": 0.00044694158807185636,
+      "loss": 1.8861,
       "step": 1085
     },
     {
       "epoch": 1.38,
-      "learning_rate": 0.0008405399696132878,
-      "loss": 2.1652,
+      "learning_rate": 0.0004456246238092812,
+      "loss": 1.7582,
       "step": 1090
     },
     {
       "epoch": 1.39,
-      "learning_rate": 0.0008390496416574165,
-      "loss": 2.3283,
+      "learning_rate": 0.0004443039778195137,
+      "loss": 1.8504,
       "step": 1095
     },
     {
       "epoch": 1.39,
-      "learning_rate": 0.0008375537163994522,
-      "loss": 2.2474,
+      "learning_rate": 0.0004429796834915381,
+      "loss": 1.8637,
       "step": 1100
     },
     {
       "epoch": 1.4,
-      "learning_rate": 0.0008360522185353234,
-      "loss": 2.3934,
+      "learning_rate": 0.00044165177430657685,
+      "loss": 1.8534,
       "step": 1105
     },
     {
       "epoch": 1.41,
-      "learning_rate": 0.0008345451728529552,
-      "loss": 2.2949,
+      "learning_rate": 0.00044032028383724446,
+      "loss": 1.7965,
       "step": 1110
     },
     {
       "epoch": 1.41,
-      "learning_rate": 0.0008330326042318603,
-      "loss": 2.2382,
+      "learning_rate": 0.00043898524574669873,
+      "loss": 1.8788,
       "step": 1115
     },
     {
       "epoch": 1.42,
-      "learning_rate": 0.0008315145376427288,
-      "loss": 2.356,
+      "learning_rate": 0.00043764669378778913,
+      "loss": 1.8786,
       "step": 1120
     },
     {
       "epoch": 1.43,
-      "learning_rate": 0.0008299909981470158,
-      "loss": 2.2622,
+      "learning_rate": 0.0004363046618022041,
+      "loss": 1.7909,
       "step": 1125
     },
     {
       "epoch": 1.43,
-      "learning_rate": 0.0008284620108965271,
-      "loss": 2.2402,
+      "learning_rate": 0.0004349591837196154,
+      "loss": 1.8828,
       "step": 1130
     },
     {
       "epoch": 1.44,
-      "learning_rate": 0.0008269276011330047,
-      "loss": 2.2855,
+      "learning_rate": 0.0004336102935568196,
+      "loss": 1.9064,
       "step": 1135
     },
     {
       "epoch": 1.44,
-      "learning_rate": 0.0008253877941877095,
-      "loss": 2.1968,
+      "learning_rate": 0.0004322580254168791,
+      "loss": 1.7983,
       "step": 1140
     },
     {
       "epoch": 1.45,
-      "learning_rate": 0.0008238426154810035,
-      "loss": 2.2679,
+      "learning_rate": 0.00043090241348825876,
+      "loss": 1.8791,
       "step": 1145
     },
     {
       "epoch": 1.46,
-      "learning_rate": 0.0008222920905219302,
-      "loss": 2.2169,
+      "learning_rate": 0.0004295434920439628,
+      "loss": 1.8287,
       "step": 1150
     },
     {
       "epoch": 1.46,
-      "learning_rate": 0.0008207362449077932,
-      "loss": 2.1951,
+      "learning_rate": 0.00042818129544066713,
+      "loss": 1.8337,
       "step": 1155
     },
     {
       "epoch": 1.47,
-      "learning_rate": 0.0008191751043237336,
-      "loss": 2.3119,
+      "learning_rate": 0.0004268158581178515,
+      "loss": 1.8558,
       "step": 1160
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.000817608694542306,
-      "loss": 2.408,
+      "learning_rate": 0.00042544721459692846,
+      "loss": 1.8345,
       "step": 1165
     },
     {
       "epoch": 1.48,
-      "learning_rate": 0.0008160370414230535,
-      "loss": 2.1998,
+      "learning_rate": 0.0004240753994803708,
+      "loss": 1.8341,
       "step": 1170
     },
     {
       "epoch": 1.49,
-      "learning_rate": 0.00081446017091208,
-      "loss": 2.3845,
+      "learning_rate": 0.00042270044745083645,
+      "loss": 1.8161,
       "step": 1175
     },
     {
       "epoch": 1.5,
-      "learning_rate": 0.0008128781090416228,
-      "loss": 2.2619,
+      "learning_rate": 0.00042132239327029173,
+      "loss": 1.6921,
       "step": 1180
     },
     {
       "epoch": 1.5,
-      "learning_rate": 0.0008112908819296217,
-      "loss": 2.3258,
+      "learning_rate": 0.00041994127177913247,
+      "loss": 1.7944,
       "step": 1185
     },
     {
       "epoch": 1.51,
-      "learning_rate": 0.0008096985157792886,
-      "loss": 2.3633,
+      "learning_rate": 0.00041855711789530345,
+      "loss": 1.8489,
       "step": 1190
     },
     {
       "epoch": 1.51,
-      "learning_rate": 0.0008081010368786751,
-      "loss": 2.2268,
+      "learning_rate": 0.00041716996661341507,
+      "loss": 1.8568,
       "step": 1195
     },
     {
       "epoch": 1.52,
-      "learning_rate": 0.0008064984716002377,
-      "loss": 2.2499,
+      "learning_rate": 0.00041577985300385913,
+      "loss": 1.883,
       "step": 1200
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.0008048908464004032,
-      "loss": 2.258,
+      "learning_rate": 0.0004143868122119217,
+      "loss": 1.9306,
       "step": 1205
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.0008032781878191313,
-      "loss": 2.2347,
+      "learning_rate": 0.00041299087945689496,
+      "loss": 1.8549,
       "step": 1210
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.0008016605224794773,
-      "loss": 2.2717,
+      "learning_rate": 0.0004115920900311864,
+      "loss": 1.7132,
       "step": 1215
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0008000378770871515,
-      "loss": 2.1741,
+      "learning_rate": 0.000410190479299427,
+      "loss": 1.8957,
       "step": 1220
     },
     {
       "epoch": 1.55,
-      "learning_rate": 0.0007984102784300793,
-      "loss": 2.1593,
+      "learning_rate": 0.0004087860826975767,
+      "loss": 1.7544,
       "step": 1225
     },
     {
       "epoch": 1.56,
-      "learning_rate": 0.0007967777533779582,
-      "loss": 2.2707,
+      "learning_rate": 0.00040737893573202877,
+      "loss": 1.8321,
       "step": 1230
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.0007951403288818149,
-      "loss": 2.2641,
+      "learning_rate": 0.00040596907397871205,
+      "loss": 1.8146,
       "step": 1235
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.0007934980319735599,
-      "loss": 2.2677,
+      "learning_rate": 0.00040455653308219127,
+      "loss": 1.7742,
       "step": 1240
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.000791850889765541,
-      "loss": 2.2982,
+      "learning_rate": 0.0004031413487547665,
+      "loss": 1.8268,
       "step": 1245
     },
     {
       "epoch": 1.58,
-      "learning_rate": 0.0007901989294500968,
-      "loss": 2.1954,
+      "learning_rate": 0.0004017235567755694,
+      "loss": 1.8419,
       "step": 1250
     },
     {
       "epoch": 1.59,
-      "learning_rate": 0.0007885421782991064,
-      "loss": 2.2439,
+      "learning_rate": 0.0004003031929896594,
+      "loss": 1.839,
       "step": 1255
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.0007868806636635405,
-      "loss": 2.2238,
+      "learning_rate": 0.00039888029330711704,
+      "loss": 1.8722,
       "step": 1260
     },
     {
       "epoch": 1.6,
-      "learning_rate": 0.0007852144129730086,
-      "loss": 2.234,
+      "learning_rate": 0.0003974548937021363,
+      "loss": 1.9201,
       "step": 1265
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.0007835434537353067,
-      "loss": 2.31,
+      "learning_rate": 0.00039602703021211484,
+      "loss": 1.9658,
       "step": 1270
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.0007818678135359641,
-      "loss": 2.2452,
+      "learning_rate": 0.0003945967389367431,
+      "loss": 1.8658,
       "step": 1275
     },
     {
       "epoch": 1.62,
-      "learning_rate": 0.0007801875200377862,
-      "loss": 2.1722,
+      "learning_rate": 0.00039316405603709157,
+      "loss": 1.8716,
       "step": 1280
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0007785026009803993,
-      "loss": 2.1681,
+      "learning_rate": 0.0003917290177346964,
+      "loss": 1.7848,
       "step": 1285
     },
     {
       "epoch": 1.63,
-      "learning_rate": 0.0007768130841797919,
-      "loss": 2.2128,
+      "learning_rate": 0.0003902916603106439,
+      "loss": 2.0245,
       "step": 1290
     },
     {
       "epoch": 1.64,
-      "learning_rate": 0.0007751189975278561,
-      "loss": 2.121,
+      "learning_rate": 0.0003888520201046532,
+      "loss": 1.7693,
       "step": 1295
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0007734203689919261,
-      "loss": 2.2649,
+      "learning_rate": 0.0003874101335141571,
+      "loss": 1.8287,
       "step": 1300
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0007717172266143178,
-      "loss": 2.1976,
+      "learning_rate": 0.00038596603699338276,
+      "loss": 1.8367,
       "step": 1305
     },
     {
       "epoch": 1.66,
-      "learning_rate": 0.000770009598511865,
-      "loss": 2.3699,
+      "learning_rate": 0.0003845197670524289,
+      "loss": 1.9042,
       "step": 1310
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0007682975128754548,
-      "loss": 2.1847,
+      "learning_rate": 0.0003830713602563437,
+      "loss": 1.7802,
       "step": 1315
     },
     {
       "epoch": 1.67,
-      "learning_rate": 0.0007665809979695639,
-      "loss": 2.2154,
+      "learning_rate": 0.00038162085322419973,
+      "loss": 1.7705,
       "step": 1320
     },
     {
       "epoch": 1.68,
-      "learning_rate": 0.00076486008213179,
-      "loss": 2.2729,
+      "learning_rate": 0.0003801682826281687,
+      "loss": 1.8283,
       "step": 1325
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.0007631347937723856,
-      "loss": 2.3637,
+      "learning_rate": 0.0003787136851925937,
+      "loss": 1.9271,
       "step": 1330
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.0007614051613737878,
-      "loss": 2.244,
+      "learning_rate": 0.00037725709769306114,
+      "loss": 1.8433,
       "step": 1335
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0007596712134901487,
-      "loss": 2.2387,
+      "learning_rate": 0.0003757985569554707,
+      "loss": 1.8718,
       "step": 1340
     },
     {
       "epoch": 1.7,
-      "learning_rate": 0.0007579329787468638,
-      "loss": 2.1745,
+      "learning_rate": 0.00037433809985510486,
+      "loss": 1.842,
       "step": 1345
     },
     {
       "epoch": 1.71,
-      "learning_rate": 0.0007561904858400999,
-      "loss": 2.1605,
+      "learning_rate": 0.00037287576331569564,
+      "loss": 1.9048,
       "step": 1350
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.0007544437635363206,
-      "loss": 2.2522,
+      "learning_rate": 0.0003714115843084918,
+      "loss": 1.8235,
       "step": 1355
     },
     {
       "epoch": 1.72,
-      "learning_rate": 0.0007526928406718115,
-      "loss": 2.2604,
+      "learning_rate": 0.00036994559985132415,
+      "loss": 1.8576,
       "step": 1360
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.0007509377461522049,
-      "loss": 2.2935,
+      "learning_rate": 0.00036847784700766933,
+      "loss": 1.836,
       "step": 1365
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.000749178508952002,
-      "loss": 2.2513,
+      "learning_rate": 0.0003670083628857128,
+      "loss": 1.8414,
       "step": 1370
     },
     {
       "epoch": 1.74,
-      "learning_rate": 0.0007474151581140947,
-      "loss": 2.1089,
+      "learning_rate": 0.0003655371846374107,
+      "loss": 1.942,
       "step": 1375
     },
     {
       "epoch": 1.75,
-      "learning_rate": 0.0007456477227492858,
-      "loss": 2.2211,
+      "learning_rate": 0.00036406434945755083,
+      "loss": 1.8823,
       "step": 1380
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.0007438762320358088,
-      "loss": 2.2009,
+      "learning_rate": 0.0003625898945828119,
+      "loss": 1.898,
       "step": 1385
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.0007421007152188468,
-      "loss": 2.1967,
+      "learning_rate": 0.00036111385729082224,
+      "loss": 1.8972,
       "step": 1390
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.0007403212016100484,
-      "loss": 2.3052,
+      "learning_rate": 0.0003596362748992173,
+      "loss": 1.7432,
       "step": 1395
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.0007385377205870442,
-      "loss": 2.2927,
+      "learning_rate": 0.0003581571847646965,
+      "loss": 1.8568,
       "step": 1400
     },
     {
       "epoch": 1.78,
-      "learning_rate": 0.0007367503015929626,
-      "loss": 2.1667,
+      "learning_rate": 0.000356676624282078,
+      "loss": 1.8891,
       "step": 1405
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.0007349589741359431,
-      "loss": 2.2273,
+      "learning_rate": 0.00035519463088335427,
+      "loss": 1.8463,
       "step": 1410
     },
     {
       "epoch": 1.79,
-      "learning_rate": 0.000733163767788649,
-      "loss": 2.3456,
+      "learning_rate": 0.00035371124203674486,
+      "loss": 1.759,
       "step": 1415
     },
     {
       "epoch": 1.8,
-      "learning_rate": 0.0007313647121877793,
-      "loss": 2.279,
+      "learning_rate": 0.0003522264952457498,
+      "loss": 1.7689,
       "step": 1420
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0007295618370335799,
-      "loss": 2.3189,
+      "learning_rate": 0.0003507404280482005,
+      "loss": 1.7517,
       "step": 1425
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.0007277551720893528,
-      "loss": 2.1636,
+      "learning_rate": 0.0003492530780153121,
+      "loss": 1.8594,
       "step": 1430
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.0007259447471809651,
-      "loss": 2.1341,
+      "learning_rate": 0.00034776448275073205,
+      "loss": 1.6709,
       "step": 1435
     },
     {
       "epoch": 1.82,
-      "learning_rate": 0.0007241305921963557,
-      "loss": 2.2994,
+      "learning_rate": 0.00034627467988959063,
+      "loss": 1.8854,
       "step": 1440
     },
     {
       "epoch": 1.83,
-      "learning_rate": 0.0007223127370850433,
-      "loss": 2.2897,
+      "learning_rate": 0.0003447837070975488,
+      "loss": 1.7318,
       "step": 1445
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.000720491211857631,
-      "loss": 2.2286,
+      "learning_rate": 0.000343291602069846,
+      "loss": 1.9173,
       "step": 1450
     },
     {
       "epoch": 1.84,
-      "learning_rate": 0.000718666046585311,
-      "loss": 2.2796,
+      "learning_rate": 0.0003417984025303472,
+      "loss": 1.7616,
       "step": 1455
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.0007168372713993683,
-      "loss": 2.2812,
+      "learning_rate": 0.0003403041462305894,
+      "loss": 1.9281,
       "step": 1460
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.0007150049164906835,
-      "loss": 2.1562,
+      "learning_rate": 0.0003388088709488265,
+      "loss": 1.8446,
       "step": 1465
     },
     {
       "epoch": 1.86,
-      "learning_rate": 0.0007131690121092336,
-      "loss": 2.305,
+      "learning_rate": 0.00033731261448907517,
+      "loss": 1.9168,
       "step": 1470
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.0007113295885635935,
-      "loss": 2.2369,
+      "learning_rate": 0.00033581541468015805,
+      "loss": 1.8417,
       "step": 1475
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.0007094866762204354,
-      "loss": 2.2044,
+      "learning_rate": 0.00033431730937474814,
+      "loss": 1.9331,
       "step": 1480
     },
     {
       "epoch": 1.88,
-      "learning_rate": 0.0007076403055040271,
-      "loss": 2.1396,
+      "learning_rate": 0.00033281833644841136,
+      "loss": 1.8163,
       "step": 1485
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.0007057905068957299,
-      "loss": 2.3368,
+      "learning_rate": 0.000331318533798649,
+      "loss": 1.8459,
       "step": 1490
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.0007039373109334957,
-      "loss": 2.2138,
+      "learning_rate": 0.00032981793934393975,
+      "loss": 1.7809,
       "step": 1495
     },
     {
       "epoch": 1.9,
-      "learning_rate": 0.0007020807482113627,
-      "loss": 2.1798,
+      "learning_rate": 0.0003283165910227812,
+      "loss": 1.8496,
       "step": 1500
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.00070022084937895,
-      "loss": 2.1819,
+      "learning_rate": 0.0003268145267927301,
+      "loss": 1.8601,
       "step": 1505
     },
     {
       "epoch": 1.91,
-      "learning_rate": 0.0006983576451409517,
-      "loss": 2.1694,
+      "learning_rate": 0.000325311784629443,
+      "loss": 1.7749,
       "step": 1510
     },
     {
       "epoch": 1.92,
-      "learning_rate": 0.0006964911662566308,
-      "loss": 2.1964,
+      "learning_rate": 0.00032380840252571656,
+      "loss": 1.844,
       "step": 1515
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0006946214435393103,
-      "loss": 2.0662,
+      "learning_rate": 0.0003223044184905262,
+      "loss": 1.8449,
       "step": 1520
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.0006927485078558652,
-      "loss": 2.1458,
+      "learning_rate": 0.00032079987054806585,
+      "loss": 1.871,
       "step": 1525
     },
     {
       "epoch": 1.94,
-      "learning_rate": 0.0006908723901262126,
-      "loss": 2.2772,
+      "learning_rate": 0.0003192947967367861,
+      "loss": 1.8817,
       "step": 1530
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.0006889931213228015,
-      "loss": 2.1833,
+      "learning_rate": 0.00031778923510843285,
+      "loss": 1.7883,
       "step": 1535
     },
     {
       "epoch": 1.95,
-      "learning_rate": 0.0006871107324701008,
-      "loss": 2.2241,
+      "learning_rate": 0.00031628322372708514,
+      "loss": 1.9115,
       "step": 1540
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.0006852252546440885,
-      "loss": 2.1994,
+      "learning_rate": 0.0003147768006681927,
+      "loss": 1.8596,
       "step": 1545
     },
     {
       "epoch": 1.96,
-      "learning_rate": 0.0006833367189717372,
-      "loss": 2.229,
+      "learning_rate": 0.0003132700040176136,
+      "loss": 1.8591,
       "step": 1550
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.0006814451566305014,
-      "loss": 2.2222,
+      "learning_rate": 0.000311762871870651,
+      "loss": 1.7955,
       "step": 1555
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.0006795505988478015,
-      "loss": 2.2849,
+      "learning_rate": 0.0003102554423310904,
+      "loss": 1.7796,
       "step": 1560
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.0006776530769005098,
-      "loss": 2.11,
+      "learning_rate": 0.0003087477535102357,
+      "loss": 1.7754,
       "step": 1565
     },
     {
       "epoch": 1.99,
-      "learning_rate": 0.0006757526221144329,
-      "loss": 2.1437,
+      "learning_rate": 0.0003072398435259463,
+      "loss": 1.8625,
       "step": 1570
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0006738492658637956,
-      "loss": 2.1549,
+      "learning_rate": 0.0003057317505016733,
+      "loss": 1.8115,
       "step": 1575
     },
     {
       "epoch": 2.0,
-      "eval_loss": 2.1981911659240723,
-      "eval_runtime": 95.327,
-      "eval_samples_per_second": 134.359,
-      "eval_steps_per_second": 16.795,
+      "eval_loss": 1.9557033777236938,
+      "eval_runtime": 92.8676,
+      "eval_samples_per_second": 137.917,
+      "eval_steps_per_second": 8.625,
       "step": 1578
     },
     {
       "epoch": 2.0,
-      "learning_rate": 0.0006719430395707215,
-      "loss": 2.3153,
+      "learning_rate": 0.00030422351256549493,
+      "loss": 1.8282,
       "step": 1580
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.0006700339747047161,
-      "loss": 2.0107,
+      "learning_rate": 0.0003027151678491537,
+      "loss": 1.8332,
       "step": 1585
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.0006681221027821458,
-      "loss": 2.1278,
+      "learning_rate": 0.00030120675448709134,
+      "loss": 1.7881,
       "step": 1590
     },
     {
       "epoch": 2.02,
-      "learning_rate": 0.000666207455365718,
-      "loss": 2.0655,
+      "learning_rate": 0.0002996983106154856,
+      "loss": 1.6833,
       "step": 1595
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.0006642900640639608,
-      "loss": 2.0664,
+      "learning_rate": 0.0002981898743712848,
+      "loss": 1.7887,
       "step": 1600
     },
     {
       "epoch": 2.03,
-      "learning_rate": 0.0006623699605306999,
-      "loss": 1.9791,
+      "learning_rate": 0.0002966814838912453,
+      "loss": 1.7439,
       "step": 1605
     },
     {
       "epoch": 2.04,
-      "learning_rate": 0.0006604471764645369,
-      "loss": 2.0518,
+      "learning_rate": 0.0002951731773109657,
+      "loss": 1.6378,
       "step": 1610
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.000658521743608326,
-      "loss": 2.0644,
+      "learning_rate": 0.00029366499276392403,
+      "loss": 1.7603,
       "step": 1615
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.0006565936937486492,
-      "loss": 2.0737,
+      "learning_rate": 0.00029215696838051256,
+      "loss": 1.7664,
       "step": 1620
     },
     {
       "epoch": 2.06,
-      "learning_rate": 0.0006546630587152929,
-      "loss": 2.0784,
+      "learning_rate": 0.0002906491422870745,
+      "loss": 1.7565,
       "step": 1625
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.000652729870380721,
-      "loss": 2.1277,
+      "learning_rate": 0.0002891415526049398,
+      "loss": 1.748,
       "step": 1630
     },
     {
       "epoch": 2.07,
-      "learning_rate": 0.0006507941606595491,
-      "loss": 2.1686,
+      "learning_rate": 0.0002876342374494613,
+      "loss": 1.7335,
       "step": 1635
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.0006488559615080188,
-      "loss": 2.1398,
+      "learning_rate": 0.0002861272349290513,
+      "loss": 1.8072,
       "step": 1640
     },
     {
       "epoch": 2.08,
-      "learning_rate": 0.0006469153049234683,
-      "loss": 2.1453,
+      "learning_rate": 0.00028462058314421794,
+      "loss": 1.7107,
       "step": 1645
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.0006449722229438053,
-      "loss": 2.0425,
+      "learning_rate": 0.00028311432018660173,
+      "loss": 1.6844,
       "step": 1650
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.0006430267476469783,
-      "loss": 2.0705,
+      "learning_rate": 0.0002816084841380131,
+      "loss": 1.729,
       "step": 1655
     },
     {
       "epoch": 2.1,
-      "learning_rate": 0.0006410789111504461,
-      "loss": 2.1157,
+      "learning_rate": 0.00028010311306946885,
+      "loss": 1.7699,
       "step": 1660
     },
     {
       "epoch": 2.11,
-      "learning_rate": 0.0006391287456106483,
-      "loss": 2.1243,
+      "learning_rate": 0.0002785982450402303,
+      "loss": 1.7964,
       "step": 1665
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.000637176283222474,
-      "loss": 2.0428,
+      "learning_rate": 0.00027709391809684075,
+      "loss": 1.6819,
       "step": 1670
     },
     {
       "epoch": 2.12,
-      "learning_rate": 0.0006352215562187306,
-      "loss": 2.0638,
+      "learning_rate": 0.0002755901702721633,
+      "loss": 1.7015,
       "step": 1675
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.000633264596869612,
-      "loss": 2.0722,
+      "learning_rate": 0.0002740870395844201,
+      "loss": 1.7115,
       "step": 1680
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.0006313054374821646,
-      "loss": 2.1089,
+      "learning_rate": 0.00027258456403623013,
+      "loss": 1.8647,
       "step": 1685
     },
     {
       "epoch": 2.14,
-      "learning_rate": 0.0006293441103997555,
-      "loss": 2.0444,
+      "learning_rate": 0.0002710827816136493,
+      "loss": 1.828,
       "step": 1690
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.0006273806480015374,
-      "loss": 2.0067,
+      "learning_rate": 0.00026958173028520944,
+      "loss": 1.8245,
       "step": 1695
     },
     {
       "epoch": 2.15,
-      "learning_rate": 0.0006254150827019148,
-      "loss": 2.2493,
+      "learning_rate": 0.0002680814480009589,
+      "loss": 1.837,
       "step": 1700
     },
     {
       "epoch": 2.16,
-      "learning_rate": 0.0006234474469500085,
-      "loss": 2.0435,
+      "learning_rate": 0.0002665819726915026,
+      "loss": 1.765,
       "step": 1705
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00062147777322912,
-      "loss": 2.1426,
+      "learning_rate": 0.0002650833422670432,
+      "loss": 1.933,
       "step": 1710
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.0006195060940561949,
-      "loss": 2.0512,
+      "learning_rate": 0.0002635855946164232,
+      "loss": 1.7883,
       "step": 1715
     },
     {
       "epoch": 2.18,
-      "learning_rate": 0.0006175324419812869,
-      "loss": 2.1365,
+      "learning_rate": 0.00026208876760616586,
+      "loss": 1.8351,
       "step": 1720
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.0006155568495870196,
-      "loss": 2.137,
+      "learning_rate": 0.0002605928990795188,
+      "loss": 1.7143,
       "step": 1725
     },
     {
       "epoch": 2.19,
-      "learning_rate": 0.000613579349488049,
-      "loss": 1.9907,
+      "learning_rate": 0.00025909802685549735,
+      "loss": 1.6925,
       "step": 1730
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.0006115999743305252,
-      "loss": 2.085,
+      "learning_rate": 0.00025760418872792737,
+      "loss": 1.8673,
       "step": 1735
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.0006096187567915528,
-      "loss": 2.1276,
+      "learning_rate": 0.0002561114224644908,
+      "loss": 1.7264,
       "step": 1740
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.0006076357295786525,
-      "loss": 1.9774,
+      "learning_rate": 0.0002546197658057702,
+      "loss": 1.7266,
       "step": 1745
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.0006056509254292202,
-      "loss": 2.0054,
+      "learning_rate": 0.0002531292564642949,
+      "loss": 1.6868,
       "step": 1750
     },
     {
       "epoch": 2.22,
-      "learning_rate": 0.0006036643771099871,
-      "loss": 2.0158,
+      "learning_rate": 0.00025163993212358715,
+      "loss": 1.6325,
       "step": 1755
     },
     {
       "epoch": 2.23,
-      "learning_rate": 0.000601676117416478,
-      "loss": 1.9952,
+      "learning_rate": 0.0002501518304372098,
+      "loss": 1.692,
       "step": 1760
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.0005996861791724712,
-      "loss": 2.0267,
+      "learning_rate": 0.0002486649890278143,
+      "loss": 1.7267,
       "step": 1765
     },
     {
       "epoch": 2.24,
-      "learning_rate": 0.0005976945952294553,
-      "loss": 1.9985,
+      "learning_rate": 0.0002471794454861891,
+      "loss": 1.6774,
       "step": 1770
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.0005957013984660875,
-      "loss": 2.0334,
+      "learning_rate": 0.0002456952373703098,
+      "loss": 1.7112,
       "step": 1775
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.0005937066217876511,
-      "loss": 2.0078,
+      "learning_rate": 0.0002442124022043891,
+      "loss": 1.765,
       "step": 1780
     },
     {
       "epoch": 2.26,
-      "learning_rate": 0.0005917102981255113,
-      "loss": 2.0976,
+      "learning_rate": 0.00024273097747792876,
+      "loss": 1.7441,
       "step": 1785
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.0005897124604365727,
-      "loss": 2.0843,
+      "learning_rate": 0.00024125100064477083,
+      "loss": 1.6806,
       "step": 1790
     },
     {
       "epoch": 2.27,
-      "learning_rate": 0.0005877131417027343,
-      "loss": 1.9916,
+      "learning_rate": 0.00023977250912215166,
+      "loss": 1.6897,
       "step": 1795
     },
     {
       "epoch": 2.28,
-      "learning_rate": 0.0005857123749303454,
-      "loss": 2.0005,
+      "learning_rate": 0.0002382955402897553,
+      "loss": 1.7494,
       "step": 1800
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.000583710193149661,
-      "loss": 2.134,
+      "learning_rate": 0.00023682013148876898,
+      "loss": 1.8519,
       "step": 1805
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.0005817066294142957,
-      "loss": 2.1503,
+      "learning_rate": 0.0002353463200209383,
+      "loss": 1.7991,
       "step": 1810
     },
     {
       "epoch": 2.3,
-      "learning_rate": 0.0005797017168006791,
-      "loss": 2.033,
+      "learning_rate": 0.00023387414314762496,
+      "loss": 1.7614,
       "step": 1815
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.0005776954884075085,
-      "loss": 2.0627,
+      "learning_rate": 0.00023240363808886403,
+      "loss": 1.6892,
       "step": 1820
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.0005756879773552036,
-      "loss": 1.985,
+      "learning_rate": 0.00023093484202242354,
+      "loss": 1.7324,
       "step": 1825
     },
     {
       "epoch": 2.32,
-      "learning_rate": 0.000573679216785359,
-      "loss": 2.0933,
+      "learning_rate": 0.00022946779208286397,
+      "loss": 1.7889,
       "step": 1830
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.0005716692398601975,
-      "loss": 2.1007,
+      "learning_rate": 0.00022800252536059988,
+      "loss": 1.7436,
       "step": 1835
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.0005696580797620226,
-      "loss": 2.0164,
+      "learning_rate": 0.0002265390789009619,
+      "loss": 1.7507,
       "step": 1840
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.0005676457696926703,
-      "loss": 2.0588,
+      "learning_rate": 0.00022507748970326005,
+      "loss": 1.796,
       "step": 1845
     },
     {
       "epoch": 2.34,
-      "learning_rate": 0.0005656323428729612,
-      "loss": 2.0087,
+      "learning_rate": 0.00022361779471984883,
+      "loss": 1.7975,
       "step": 1850
     },
     {
       "epoch": 2.35,
-      "learning_rate": 0.0005636178325421524,
-      "loss": 2.0798,
+      "learning_rate": 0.00022216003085519237,
+      "loss": 1.775,
       "step": 1855
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.000561602271957388,
-      "loss": 2.1189,
+      "learning_rate": 0.00022070423496493152,
+      "loss": 1.7634,
       "step": 1860
     },
     {
       "epoch": 2.36,
-      "learning_rate": 0.0005595856943931512,
-      "loss": 2.0921,
+      "learning_rate": 0.00021925044385495254,
+      "loss": 1.8228,
       "step": 1865
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.0005575681331407137,
-      "loss": 2.1177,
+      "learning_rate": 0.000217798694280456,
+      "loss": 1.7155,
       "step": 1870
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.0005555496215075871,
-      "loss": 2.1912,
+      "learning_rate": 0.00021634902294502778,
+      "loss": 1.6878,
       "step": 1875
     },
     {
       "epoch": 2.38,
-      "learning_rate": 0.0005535301928169722,
-      "loss": 2.1851,
+      "learning_rate": 0.00021490146649971097,
+      "loss": 1.7402,
       "step": 1880
     },
     {
       "epoch": 2.39,
-      "learning_rate": 0.00055150988040721,
-      "loss": 2.1525,
+      "learning_rate": 0.0002134560615420796,
+      "loss": 1.7177,
       "step": 1885
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.00054948871763123,
-      "loss": 2.0721,
+      "learning_rate": 0.00021201284461531289,
+      "loss": 1.6694,
       "step": 1890
     },
     {
       "epoch": 2.4,
-      "learning_rate": 0.0005474667378560006,
-      "loss": 2.1164,
+      "learning_rate": 0.00021057185220727152,
+      "loss": 1.7822,
       "step": 1895
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.0005454439744619776,
-      "loss": 2.1189,
+      "learning_rate": 0.00020913312074957556,
+      "loss": 1.6492,
       "step": 1900
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.0005434204608425538,
-      "loss": 2.0945,
+      "learning_rate": 0.00020769668661668254,
+      "loss": 1.7165,
       "step": 1905
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.0005413962304035068,
-      "loss": 2.2036,
+      "learning_rate": 0.00020626258612496876,
+      "loss": 1.5924,
       "step": 1910
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0005393713165624483,
-      "loss": 2.1643,
+      "learning_rate": 0.00020483085553181033,
+      "loss": 1.8017,
       "step": 1915
     },
     {
       "epoch": 2.43,
-      "learning_rate": 0.0005373457527482725,
-      "loss": 2.0757,
+      "learning_rate": 0.00020340153103466709,
+      "loss": 1.7278,
       "step": 1920
     },
     {
       "epoch": 2.44,
-      "learning_rate": 0.0005353195724006031,
-      "loss": 2.0375,
+      "learning_rate": 0.0002019746487701671,
+      "loss": 1.5994,
       "step": 1925
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.0005332928089692425,
-      "loss": 2.0028,
+      "learning_rate": 0.00020055024481319307,
+      "loss": 1.6479,
       "step": 1930
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.0005312654959136194,
-      "loss": 2.08,
+      "learning_rate": 0.00019912835517597036,
+      "loss": 1.6717,
       "step": 1935
     },
     {
       "epoch": 2.46,
-      "learning_rate": 0.0005292376667022355,
-      "loss": 2.0521,
+      "learning_rate": 0.00019770901580715673,
+      "loss": 1.7909,
       "step": 1940
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.0005272093548121141,
-      "loss": 1.9639,
+      "learning_rate": 0.00019629226259093303,
+      "loss": 1.7646,
       "step": 1945
     },
     {
       "epoch": 2.47,
-      "learning_rate": 0.0005251805937282467,
-      "loss": 2.1265,
+      "learning_rate": 0.00019487813134609635,
+      "loss": 1.7472,
       "step": 1950
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.0005231514169430402,
-      "loss": 2.0656,
+      "learning_rate": 0.00019346665782515415,
+      "loss": 1.6367,
       "step": 1955
     },
     {
       "epoch": 2.48,
-      "learning_rate": 0.000521121857955765,
-      "loss": 2.1371,
+      "learning_rate": 0.00019205787771342062,
+      "loss": 1.7175,
       "step": 1960
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.0005190919502720005,
-      "loss": 2.0298,
+      "learning_rate": 0.0001906518266281141,
+      "loss": 1.8483,
       "step": 1965
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.0005170617274030827,
-      "loss": 2.166,
+      "learning_rate": 0.00018924854011745718,
+      "loss": 1.7751,
       "step": 1970
     },
     {
       "epoch": 2.5,
-      "learning_rate": 0.0005150312228655515,
-      "loss": 2.0785,
+      "learning_rate": 0.00018784805365977734,
+      "loss": 1.7383,
       "step": 1975
     },
     {
       "epoch": 2.51,
-      "learning_rate": 0.0005130004701805963,
-      "loss": 2.0836,
+      "learning_rate": 0.0001864504026626103,
+      "loss": 1.7633,
       "step": 1980
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.0005109695028735029,
-      "loss": 2.1391,
+      "learning_rate": 0.00018505562246180486,
+      "loss": 1.8172,
       "step": 1985
     },
     {
       "epoch": 2.52,
-      "learning_rate": 0.0005089383544731012,
-      "loss": 1.9789,
+      "learning_rate": 0.0001836637483206293,
+      "loss": 1.7102,
       "step": 1990
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.0005069070585112097,
-      "loss": 2.0498,
+      "learning_rate": 0.00018227481542888033,
+      "loss": 1.7455,
       "step": 1995
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.0005048756485220836,
-      "loss": 2.1021,
+      "learning_rate": 0.00018088885890199267,
+      "loss": 1.6092,
       "step": 2000
     },
     {
       "epoch": 2.54,
-      "learning_rate": 0.0005028441580418602,
-      "loss": 2.0545,
+      "learning_rate": 0.00017950591378015175,
+      "loss": 1.7731,
       "step": 2005
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.0005008126206080059,
-      "loss": 2.0408,
+      "learning_rate": 0.00017812601502740797,
+      "loss": 1.8558,
       "step": 2010
     },
     {
       "epoch": 2.55,
-      "learning_rate": 0.0004987810697587617,
-      "loss": 2.1183,
+      "learning_rate": 0.00017674919753079218,
+      "loss": 1.7243,
       "step": 2015
     },
     {
       "epoch": 2.56,
-      "learning_rate": 0.0004967495390325911,
-      "loss": 2.1223,
+      "learning_rate": 0.0001753754960994341,
+      "loss": 1.6603,
       "step": 2020
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.0004947180619676244,
-      "loss": 1.9974,
+      "learning_rate": 0.0001740049454636821,
+      "loss": 1.783,
       "step": 2025
     },
     {
       "epoch": 2.57,
-      "learning_rate": 0.0004926866721011064,
-      "loss": 2.1279,
+      "learning_rate": 0.00017263758027422508,
+      "loss": 1.6785,
       "step": 2030
     },
     {
       "epoch": 2.58,
-      "learning_rate": 0.0004906554029688425,
-      "loss": 1.9737,
+      "learning_rate": 0.0001712734351012166,
+      "loss": 1.798,
       "step": 2035
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.0004886242881046447,
-      "loss": 1.9904,
+      "learning_rate": 0.00016991254443340082,
+      "loss": 1.7397,
       "step": 2040
     },
     {
       "epoch": 2.59,
-      "learning_rate": 0.00048659336103977805,
-      "loss": 2.0324,
+      "learning_rate": 0.00016855494267724018,
+      "loss": 1.7338,
       "step": 2045
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.00048456265530240784,
-      "loss": 2.1492,
+      "learning_rate": 0.00016720066415604605,
+      "loss": 1.8307,
       "step": 2050
     },
     {
       "epoch": 2.6,
-      "learning_rate": 0.00048253220441704507,
-      "loss": 1.9715,
+      "learning_rate": 0.00016584974310911082,
+      "loss": 1.6881,
       "step": 2055
     },
     {
       "epoch": 2.61,
-      "learning_rate": 0.00048050204190399335,
-      "loss": 2.0,
+      "learning_rate": 0.00016450221369084203,
+      "loss": 1.806,
       "step": 2060
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.00047847220127879606,
-      "loss": 2.1284,
+      "learning_rate": 0.00016315810996989904,
+      "loss": 1.7203,
       "step": 2065
     },
     {
       "epoch": 2.62,
-      "learning_rate": 0.00047644271605168213,
-      "loss": 2.0984,
+      "learning_rate": 0.00016181746592833145,
+      "loss": 1.784,
       "step": 2070
     },
     {
       "epoch": 2.63,
-      "learning_rate": 0.0004744136197270135,
-      "loss": 2.057,
+      "learning_rate": 0.00016048031546072054,
+      "loss": 1.885,
       "step": 2075
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00047238494580273193,
-      "loss": 2.0703,
+      "learning_rate": 0.00015914669237332188,
+      "loss": 1.6406,
       "step": 2080
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00047035672776980574,
-      "loss": 2.154,
+      "learning_rate": 0.0001578166303832104,
+      "loss": 1.6898,
       "step": 2085
     },
     {
       "epoch": 2.65,
-      "learning_rate": 0.00046832899911167697,
-      "loss": 2.154,
+      "learning_rate": 0.0001564901631174286,
+      "loss": 1.7131,
       "step": 2090
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.00046630179330370866,
-      "loss": 2.0377,
+      "learning_rate": 0.00015516732411213592,
+      "loss": 1.7048,
       "step": 2095
     },
     {
       "epoch": 2.66,
-      "learning_rate": 0.00046427514381263205,
-      "loss": 2.1521,
+      "learning_rate": 0.0001538481468117611,
+      "loss": 1.6594,
       "step": 2100
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.00046224908409599495,
-      "loss": 2.159,
+      "learning_rate": 0.00015253266456815613,
+      "loss": 1.7222,
       "step": 2105
     },
     {
       "epoch": 2.67,
-      "learning_rate": 0.000460223647601608,
-      "loss": 2.0178,
+      "learning_rate": 0.00015122091063975382,
+      "loss": 1.6652,
       "step": 2110
     },
     {
       "epoch": 2.68,
-      "learning_rate": 0.0004581988677669935,
-      "loss": 2.0769,
+      "learning_rate": 0.00014991291819072656,
+      "loss": 1.7111,
       "step": 2115
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00045617477801883295,
-      "loss": 2.1209,
+      "learning_rate": 0.00014860872029014746,
+      "loss": 1.7738,
       "step": 2120
     },
     {
       "epoch": 2.69,
-      "learning_rate": 0.00045415141177241545,
-      "loss": 1.9985,
+      "learning_rate": 0.00014730834991115503,
+      "loss": 1.8192,
       "step": 2125
     },
     {
       "epoch": 2.7,
-      "learning_rate": 0.0004521288024310856,
-      "loss": 1.9867,
+      "learning_rate": 0.0001460118399301191,
+      "loss": 1.6937,
       "step": 2130
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.00045010698338569333,
-      "loss": 2.022,
+      "learning_rate": 0.00014471922312580968,
+      "loss": 1.8442,
       "step": 2135
     },
     {
       "epoch": 2.71,
-      "learning_rate": 0.0004480859880140408,
-      "loss": 2.0363,
+      "learning_rate": 0.00014343053217856833,
+      "loss": 1.8353,
       "step": 2140
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.00044606584968033265,
-      "loss": 2.1197,
+      "learning_rate": 0.00014214579966948158,
+      "loss": 1.7505,
       "step": 2145
     },
     {
       "epoch": 2.72,
-      "learning_rate": 0.00044404660173462477,
-      "loss": 2.0195,
+      "learning_rate": 0.0001408650580795579,
+      "loss": 1.7784,
       "step": 2150
     },
     {
       "epoch": 2.73,
-      "learning_rate": 0.0004420282775122739,
-      "loss": 2.0232,
+      "learning_rate": 0.00013958833978890598,
+      "loss": 1.7562,
       "step": 2155
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.00044001091033338735,
-      "loss": 2.0646,
+      "learning_rate": 0.00013831567707591592,
+      "loss": 1.8261,
       "step": 2160
     },
     {
       "epoch": 2.74,
-      "learning_rate": 0.0004379945335022726,
-      "loss": 1.9648,
+      "learning_rate": 0.0001370471021164438,
+      "loss": 1.8067,
       "step": 2165
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.00043597918030688777,
-      "loss": 2.1262,
+      "learning_rate": 0.0001357826469829977,
+      "loss": 1.766,
       "step": 2170
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.0004339648840182919,
-      "loss": 2.0872,
+      "learning_rate": 0.0001345223436439271,
+      "loss": 1.7806,
       "step": 2175
     },
     {
       "epoch": 2.76,
-      "learning_rate": 0.0004319516778900957,
-      "loss": 2.0648,
+      "learning_rate": 0.0001332662239626141,
+      "loss": 1.6875,
       "step": 2180
     },
     {
       "epoch": 2.77,
-      "learning_rate": 0.0004299395951579126,
-      "loss": 1.9803,
+      "learning_rate": 0.0001320143196966687,
+      "loss": 1.7399,
       "step": 2185
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.0004279286690388107,
-      "loss": 2.0624,
+      "learning_rate": 0.00013076666249712532,
+      "loss": 1.6684,
       "step": 2190
     },
     {
       "epoch": 2.78,
-      "learning_rate": 0.00042591893273076313,
-      "loss": 1.989,
+      "learning_rate": 0.0001295232839076428,
+      "loss": 1.7932,
       "step": 2195
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00042391041941210086,
-      "loss": 2.0994,
+      "learning_rate": 0.00012828421536370652,
+      "loss": 1.7512,
       "step": 2200
     },
     {
       "epoch": 2.79,
-      "learning_rate": 0.00042190316224096497,
-      "loss": 2.1275,
+      "learning_rate": 0.00012704948819183433,
+      "loss": 1.7321,
       "step": 2205
     },
     {
       "epoch": 2.8,
-      "learning_rate": 0.00041989719435475865,
-      "loss": 1.8627,
+      "learning_rate": 0.00012581913360878378,
+      "loss": 1.6867,
       "step": 2210
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.0004178925488696012,
-      "loss": 2.0948,
+      "learning_rate": 0.00012459318272076372,
+      "loss": 1.7741,
       "step": 2215
     },
     {
       "epoch": 2.81,
-      "learning_rate": 0.0004158892588797801,
-      "loss": 2.1067,
+      "learning_rate": 0.0001233716665226469,
+      "loss": 1.7314,
       "step": 2220
     },
     {
       "epoch": 2.82,
-      "learning_rate": 0.0004138873574572053,
-      "loss": 2.0034,
+      "learning_rate": 0.00012215461589718722,
+      "loss": 1.7479,
       "step": 2225
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.0004118868776508633,
-      "loss": 2.0439,
+      "learning_rate": 0.00012094206161423849,
+      "loss": 1.7112,
       "step": 2230
     },
     {
       "epoch": 2.83,
-      "learning_rate": 0.0004098878524862715,
-      "loss": 2.0353,
+      "learning_rate": 0.00011973403432997661,
+      "loss": 1.5846,
       "step": 2235
     },
     {
       "epoch": 2.84,
-      "learning_rate": 0.00040789031496493255,
-      "loss": 2.0757,
+      "learning_rate": 0.00011853056458612428,
+      "loss": 1.8141,
       "step": 2240
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.00040589429806379057,
-      "loss": 2.0462,
+      "learning_rate": 0.00011733168280917927,
+      "loss": 1.7156,
       "step": 2245
     },
     {
       "epoch": 2.85,
-      "learning_rate": 0.0004038998347346854,
-      "loss": 1.9894,
+      "learning_rate": 0.00011613741930964498,
+      "loss": 1.7403,
       "step": 2250
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.0004019069579038096,
-      "loss": 2.0345,
+      "learning_rate": 0.00011494780428126375,
+      "loss": 1.6505,
       "step": 2255
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.00039991570047116456,
-      "loss": 2.044,
+      "learning_rate": 0.00011376286780025415,
+      "loss": 1.7534,
       "step": 2260
     },
     {
       "epoch": 2.87,
-      "learning_rate": 0.0003979260953100169,
-      "loss": 1.9671,
+      "learning_rate": 0.00011258263982455014,
+      "loss": 1.727,
       "step": 2265
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.0003959381752663572,
-      "loss": 1.9423,
+      "learning_rate": 0.00011140715019304389,
+      "loss": 1.813,
       "step": 2270
     },
     {
       "epoch": 2.88,
-      "learning_rate": 0.00039395197315835573,
-      "loss": 1.9895,
+      "learning_rate": 0.00011023642862483104,
+      "loss": 1.789,
       "step": 2275
     },
     {
       "epoch": 2.89,
-      "learning_rate": 0.0003919675217758224,
-      "loss": 1.9851,
+      "learning_rate": 0.00010907050471845962,
+      "loss": 1.8312,
       "step": 2280
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00038998485387966425,
-      "loss": 2.1416,
+      "learning_rate": 0.00010790940795118183,
+      "loss": 1.6525,
       "step": 2285
     },
     {
       "epoch": 2.9,
-      "learning_rate": 0.00038800400220134556,
-      "loss": 2.0666,
+      "learning_rate": 0.00010675316767820862,
+      "loss": 1.7615,
       "step": 2290
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.0003860249994423467,
-      "loss": 2.0336,
+      "learning_rate": 0.00010560181313196731,
+      "loss": 1.7096,
       "step": 2295
     },
     {
       "epoch": 2.91,
-      "learning_rate": 0.00038404787827362477,
-      "loss": 2.1442,
+      "learning_rate": 0.00010445537342136298,
+      "loss": 1.7094,
       "step": 2300
     },
     {
       "epoch": 2.92,
-      "learning_rate": 0.0003820726713350742,
-      "loss": 2.0041,
+      "learning_rate": 0.00010331387753104222,
+      "loss": 1.7123,
       "step": 2305
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.00038009941123498737,
-      "loss": 2.0138,
+      "learning_rate": 0.00010217735432066055,
+      "loss": 1.6466,
       "step": 2310
     },
     {
       "epoch": 2.93,
-      "learning_rate": 0.0003781281305495171,
-      "loss": 2.1307,
+      "learning_rate": 0.00010104583252415227,
+      "loss": 1.723,
       "step": 2315
     },
     {
       "epoch": 2.94,
-      "learning_rate": 0.00037615886182213826,
-      "loss": 1.9921,
+      "learning_rate": 9.991934074900475e-05,
+      "loss": 1.727,
       "step": 2320
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.0003741916375631105,
-      "loss": 2.1046,
+      "learning_rate": 9.879790747553463e-05,
+      "loss": 1.7698,
       "step": 2325
     },
     {
       "epoch": 2.95,
-      "learning_rate": 0.0003722264902489425,
-      "loss": 2.0436,
+      "learning_rate": 9.768156105616809e-05,
+      "loss": 1.6874,
       "step": 2330
     },
     {
       "epoch": 2.96,
-      "learning_rate": 0.00037026345232185414,
-      "loss": 1.9607,
+      "learning_rate": 9.657032971472353e-05,
+      "loss": 1.7301,
       "step": 2335
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00036830255618924265,
-      "loss": 1.9253,
+      "learning_rate": 9.546424154569864e-05,
+      "loss": 1.7576,
       "step": 2340
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.0003663438342231462,
-      "loss": 2.127,
+      "learning_rate": 9.436332451355987e-05,
+      "loss": 1.6674,
       "step": 2345
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.00036438731875971034,
-      "loss": 2.1264,
+      "learning_rate": 9.326760645203518e-05,
+      "loss": 1.6413,
       "step": 2350
     },
     {
       "epoch": 2.98,
-      "learning_rate": 0.00036243304209865405,
-      "loss": 2.038,
+      "learning_rate": 9.217711506341044e-05,
+      "loss": 1.7491,
       "step": 2355
     },
     {
       "epoch": 2.99,
-      "learning_rate": 0.0003604810365027362,
-      "loss": 2.1015,
+      "learning_rate": 9.109187791782946e-05,
+      "loss": 1.6581,
       "step": 2360
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.00035853133419722316,
-      "loss": 2.1424,
+      "learning_rate": 9.001192245259649e-05,
+      "loss": 1.8501,
       "step": 2365
     },
     {
       "epoch": 3.0,
-      "eval_loss": 2.1064605712890625,
-      "eval_runtime": 95.2849,
-      "eval_samples_per_second": 134.418,
-      "eval_steps_per_second": 16.802,
+      "eval_loss": 1.9110476970672607,
+      "eval_runtime": 92.8315,
+      "eval_samples_per_second": 137.97,
+      "eval_steps_per_second": 8.629,
       "step": 2367
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.0003565839673693566,
-      "loss": 2.0409,
+      "learning_rate": 8.89372759714827e-05,
+      "loss": 1.7392,
       "step": 2370
     },
     {
       "epoch": 3.01,
-      "learning_rate": 0.0003546389681678224,
-      "loss": 2.023,
+      "learning_rate": 8.786796564403575e-05,
+      "loss": 1.7441,
       "step": 2375
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.0003526963687022198,
-      "loss": 1.8723,
+      "learning_rate": 8.680401850489317e-05,
+      "loss": 1.6942,
       "step": 2380
     },
     {
       "epoch": 3.02,
-      "learning_rate": 0.0003507562010425312,
-      "loss": 1.852,
+      "learning_rate": 8.574546145309867e-05,
+      "loss": 1.6711,
       "step": 2385
     },
     {
       "epoch": 3.03,
-      "learning_rate": 0.000348818497218593,
-      "loss": 2.0228,
+      "learning_rate": 8.469232125142186e-05,
+      "loss": 1.6821,
       "step": 2390
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.00034688328921956635,
-      "loss": 1.8949,
+      "learning_rate": 8.364462452568217e-05,
+      "loss": 1.7513,
       "step": 2395
     },
     {
       "epoch": 3.04,
-      "learning_rate": 0.0003449506089934095,
-      "loss": 1.911,
+      "learning_rate": 8.260239776407518e-05,
+      "loss": 1.7362,
       "step": 2400
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.0003430204884463499,
-      "loss": 1.9915,
+      "learning_rate": 8.156566731650332e-05,
+      "loss": 1.6774,
       "step": 2405
     },
     {
       "epoch": 3.05,
-      "learning_rate": 0.0003410929594423585,
-      "loss": 1.8748,
+      "learning_rate": 8.053445939390922e-05,
+      "loss": 1.7459,
       "step": 2410
     },
     {
       "epoch": 3.06,
-      "learning_rate": 0.0003391680538026224,
-      "loss": 2.0819,
+      "learning_rate": 7.950880006761338e-05,
+      "loss": 1.6401,
       "step": 2415
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00033724580330502007,
-      "loss": 1.9166,
+      "learning_rate": 7.84887152686551e-05,
+      "loss": 1.6756,
       "step": 2420
     },
     {
       "epoch": 3.07,
-      "learning_rate": 0.00033532623968359697,
-      "loss": 1.9435,
+      "learning_rate": 7.747423078713675e-05,
+      "loss": 1.6506,
       "step": 2425
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.0003334093946280417,
-      "loss": 1.9159,
+      "learning_rate": 7.64653722715715e-05,
+      "loss": 1.7614,
       "step": 2430
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.0003314952997831618,
-      "loss": 2.0227,
+      "learning_rate": 7.546216522823525e-05,
+      "loss": 1.6238,
       "step": 2435
     },
     {
       "epoch": 3.09,
-      "learning_rate": 0.00032958398674836324,
-      "loss": 1.9393,
+      "learning_rate": 7.446463502052165e-05,
+      "loss": 1.7027,
       "step": 2440
     },
     {
       "epoch": 3.1,
-      "learning_rate": 0.00032767548707712693,
-      "loss": 1.9019,
+      "learning_rate": 7.347280686830084e-05,
+      "loss": 1.6646,
       "step": 2445
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.0003257698322764886,
-      "loss": 1.9114,
+      "learning_rate": 7.248670584728154e-05,
+      "loss": 1.7557,
       "step": 2450
     },
     {
       "epoch": 3.11,
-      "learning_rate": 0.00032386705380651874,
-      "loss": 1.8678,
+      "learning_rate": 7.150635688837767e-05,
+      "loss": 1.6833,
       "step": 2455
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.00032196718307980277,
-      "loss": 1.9498,
+      "learning_rate": 7.053178477707758e-05,
+      "loss": 1.5925,
       "step": 2460
     },
     {
       "epoch": 3.12,
-      "learning_rate": 0.00032007025146092347,
-      "loss": 1.9621,
+      "learning_rate": 6.95630141528177e-05,
+      "loss": 1.6888,
       "step": 2465
     },
     {
       "epoch": 3.13,
-      "learning_rate": 0.00031817629026594164,
-      "loss": 1.9599,
+      "learning_rate": 6.86000695083591e-05,
+      "loss": 1.7231,
       "step": 2470
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.0003162853307618805,
-      "loss": 1.8853,
+      "learning_rate": 6.764297518916897e-05,
+      "loss": 1.6539,
       "step": 2475
     },
     {
       "epoch": 3.14,
-      "learning_rate": 0.0003143974041662087,
-      "loss": 1.9151,
+      "learning_rate": 6.669175539280469e-05,
+      "loss": 1.6521,
       "step": 2480
     },
     {
       "epoch": 3.15,
-      "learning_rate": 0.00031251254164632537,
-      "loss": 1.8902,
+      "learning_rate": 6.574643416830197e-05,
+      "loss": 1.7747,
       "step": 2485
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.0003106307743190452,
-      "loss": 1.9937,
+      "learning_rate": 6.480703541556694e-05,
+      "loss": 1.6329,
       "step": 2490
     },
     {
       "epoch": 3.16,
-      "learning_rate": 0.0003087521332500854,
-      "loss": 2.0179,
+      "learning_rate": 6.387358288477226e-05,
+      "loss": 1.6779,
       "step": 2495
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.0003068766494535522,
-      "loss": 2.0089,
+      "learning_rate": 6.294610017575608e-05,
+      "loss": 1.6526,
       "step": 2500
     },
     {
       "epoch": 3.17,
-      "learning_rate": 0.0003050043538914287,
-      "loss": 1.9892,
+      "learning_rate": 6.202461073742586e-05,
+      "loss": 1.6946,
       "step": 2505
     },
     {
       "epoch": 3.18,
-      "learning_rate": 0.00030313527747306444,
-      "loss": 1.8901,
+      "learning_rate": 6.1109137867165e-05,
+      "loss": 1.7617,
       "step": 2510
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.00030126945105466487,
-      "loss": 1.9044,
+      "learning_rate": 6.019970471024448e-05,
+      "loss": 1.6621,
       "step": 2515
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.0002994069054387817,
-      "loss": 1.949,
+      "learning_rate": 5.929633425923729e-05,
+      "loss": 1.6993,
       "step": 2520
     },
     {
       "epoch": 3.2,
-      "learning_rate": 0.00029754767137380426,
-      "loss": 1.8189,
+      "learning_rate": 5.839904935343699e-05,
+      "loss": 1.7324,
       "step": 2525
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.0002956917795534528,
-      "loss": 1.988,
+      "learning_rate": 5.7507872678280676e-05,
+      "loss": 1.6393,
       "step": 2530
     },
     {
       "epoch": 3.21,
-      "learning_rate": 0.00029383926061627054,
-      "loss": 2.0487,
+      "learning_rate": 5.6622826764775207e-05,
+      "loss": 1.763,
       "step": 2535
     },
     {
       "epoch": 3.22,
-      "learning_rate": 0.0002919901451451187,
-      "loss": 1.9252,
+      "learning_rate": 5.574393398892777e-05,
+      "loss": 1.7249,
       "step": 2540
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.00029014446366667113,
-      "loss": 1.9607,
+      "learning_rate": 5.487121657117961e-05,
+      "loss": 1.6655,
       "step": 2545
     },
     {
       "epoch": 3.23,
-      "learning_rate": 0.00028830224665091106,
-      "loss": 2.0455,
+      "learning_rate": 5.400469657584504e-05,
+      "loss": 1.6952,
       "step": 2550
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.0002864635245106272,
-      "loss": 1.9129,
+      "learning_rate": 5.314439591055294e-05,
+      "loss": 1.6554,
       "step": 2555
     },
     {
       "epoch": 3.24,
-      "learning_rate": 0.0002846283276009124,
-      "loss": 2.0086,
+      "learning_rate": 5.229033632569339e-05,
+      "loss": 1.7214,
       "step": 2560
     },
     {
       "epoch": 3.25,
-      "learning_rate": 0.0002827966862186616,
-      "loss": 1.9384,
+      "learning_rate": 5.144253941386721e-05,
+      "loss": 1.6063,
       "step": 2565
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.000280968630602073,
-      "loss": 1.991,
+      "learning_rate": 5.0601026609340625e-05,
+      "loss": 1.62,
       "step": 2570
     },
     {
       "epoch": 3.26,
-      "learning_rate": 0.00027914419093014734,
-      "loss": 1.8516,
+      "learning_rate": 4.9765819187503044e-05,
+      "loss": 1.6993,
       "step": 2575
     },
     {
       "epoch": 3.27,
-      "learning_rate": 0.00027732339732219137,
-      "loss": 1.966,
+      "learning_rate": 4.893693826432921e-05,
+      "loss": 1.7051,
       "step": 2580
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.0002755062798373189,
-      "loss": 1.9472,
+      "learning_rate": 4.811440479584526e-05,
+      "loss": 1.6912,
       "step": 2585
     },
     {
       "epoch": 3.28,
-      "learning_rate": 0.0002736928684739558,
-      "loss": 1.8879,
+      "learning_rate": 4.7298239577599126e-05,
+      "loss": 1.748,
       "step": 2590
     },
     {
       "epoch": 3.29,
-      "learning_rate": 0.0002718831931693443,
-      "loss": 1.9771,
+      "learning_rate": 4.648846324413458e-05,
+      "loss": 1.5512,
       "step": 2595
     },
     {
       "epoch": 3.3,
-      "learning_rate": 0.00027007728379904814,
-      "loss": 1.8985,
+      "learning_rate": 4.5685096268469725e-05,
+      "loss": 1.5986,
       "step": 2600
     },
     {
       "epoch": 3.3,
-      "learning_rate": 0.0002682751701764615,
-      "loss": 2.0464,
+      "learning_rate": 4.488815896157907e-05,
+      "loss": 1.7881,
       "step": 2605
     },
     {
       "epoch": 3.31,
-      "learning_rate": 0.0002664768820523137,
-      "loss": 1.8301,
+      "learning_rate": 4.409767147188038e-05,
+      "loss": 1.6585,
       "step": 2610
     },
     {
       "epoch": 3.31,
-      "learning_rate": 0.0002646824491141807,
-      "loss": 1.8646,
+      "learning_rate": 4.33136537847252e-05,
+      "loss": 1.6925,
       "step": 2615
     },
     {
       "epoch": 3.32,
-      "learning_rate": 0.0002628919009859934,
-      "loss": 1.9277,
+      "learning_rate": 4.253612572189331e-05,
+      "loss": 1.6719,
       "step": 2620
     },
     {
       "epoch": 3.33,
-      "learning_rate": 0.0002611052672275495,
-      "loss": 1.8923,
+      "learning_rate": 4.176510694109188e-05,
+      "loss": 1.628,
       "step": 2625
     },
     {
       "epoch": 3.33,
-      "learning_rate": 0.0002593225773340251,
-      "loss": 1.8747,
+      "learning_rate": 4.100061693545836e-05,
+      "loss": 1.7002,
       "step": 2630
     },
     {
       "epoch": 3.34,
-      "learning_rate": 0.00025754386073548776,
-      "loss": 1.9835,
+      "learning_rate": 4.024267503306772e-05,
+      "loss": 1.7025,
       "step": 2635
     },
     {
       "epoch": 3.35,
-      "learning_rate": 0.000255769146796411,
-      "loss": 1.9034,
+      "learning_rate": 3.94913003964438e-05,
+      "loss": 1.6637,
       "step": 2640
     },
     {
       "epoch": 3.35,
-      "learning_rate": 0.0002539984648151886,
-      "loss": 1.9571,
+      "learning_rate": 3.874651202207457e-05,
+      "loss": 1.6331,
       "step": 2645
     },
     {
       "epoch": 3.36,
-      "learning_rate": 0.00025223184402365224,
-      "loss": 1.9356,
+      "learning_rate": 3.8008328739932173e-05,
+      "loss": 1.8004,
       "step": 2650
     },
     {
       "epoch": 3.36,
-      "learning_rate": 0.0002504693135865875,
-      "loss": 1.9416,
+      "learning_rate": 3.727676921299686e-05,
+      "loss": 1.5581,
       "step": 2655
     },
     {
       "epoch": 3.37,
-      "learning_rate": 0.00024871090260125423,
-      "loss": 2.0262,
+      "learning_rate": 3.655185193678483e-05,
+      "loss": 1.7183,
       "step": 2660
     },
     {
       "epoch": 3.38,
-      "learning_rate": 0.000246956640096904,
-      "loss": 2.0165,
+      "learning_rate": 3.583359523888095e-05,
+      "loss": 1.8335,
       "step": 2665
     },
     {
       "epoch": 3.38,
-      "learning_rate": 0.0002452065550343026,
-      "loss": 2.0138,
+      "learning_rate": 3.512201727847522e-05,
+      "loss": 1.6968,
       "step": 2670
     },
     {
       "epoch": 3.39,
-      "learning_rate": 0.00024346067630525082,
-      "loss": 2.0022,
+      "learning_rate": 3.441713604590386e-05,
+      "loss": 1.6507,
       "step": 2675
     },
     {
       "epoch": 3.4,
-      "learning_rate": 0.00024171903273210828,
-      "loss": 1.8902,
+      "learning_rate": 3.371896936219405e-05,
+      "loss": 1.7317,
       "step": 2680
     },
     {
       "epoch": 3.4,
-      "learning_rate": 0.0002399816530673171,
-      "loss": 2.0161,
+      "learning_rate": 3.302753487861384e-05,
+      "loss": 1.6244,
       "step": 2685
     },
     {
       "epoch": 3.41,
-      "learning_rate": 0.00023824856599292762,
-      "loss": 1.973,
+      "learning_rate": 3.234285007622576e-05,
+      "loss": 1.7972,
       "step": 2690
     },
     {
       "epoch": 3.42,
-      "learning_rate": 0.00023651980012012454,
-      "loss": 1.8932,
+      "learning_rate": 3.166493226544461e-05,
+      "loss": 1.6254,
       "step": 2695
     },
     {
       "epoch": 3.42,
-      "learning_rate": 0.00023479538398875433,
-      "loss": 1.9342,
+      "learning_rate": 3.099379858560002e-05,
+      "loss": 1.684,
       "step": 2700
     },
     {
       "epoch": 3.43,
-      "learning_rate": 0.00023307534606685525,
-      "loss": 1.8307,
+      "learning_rate": 3.0329466004503232e-05,
+      "loss": 1.6569,
       "step": 2705
     },
     {
       "epoch": 3.43,
-      "learning_rate": 0.00023135971475018557,
-      "loss": 1.9087,
+      "learning_rate": 2.9671951318017896e-05,
+      "loss": 1.6789,
       "step": 2710
     },
     {
       "epoch": 3.44,
-      "learning_rate": 0.00022964851836175705,
-      "loss": 1.8213,
+      "learning_rate": 2.9021271149635617e-05,
+      "loss": 1.6185,
       "step": 2715
     },
     {
       "epoch": 3.45,
-      "learning_rate": 0.00022794178515136494,
-      "loss": 1.9313,
+      "learning_rate": 2.837744195005537e-05,
+      "loss": 1.6109,
       "step": 2720
     },
     {
       "epoch": 3.45,
-      "learning_rate": 0.00022623954329512348,
-      "loss": 2.0212,
+      "learning_rate": 2.774047999676804e-05,
+      "loss": 1.6916,
       "step": 2725
     },
     {
       "epoch": 3.46,
-      "learning_rate": 0.00022454182089500004,
-      "loss": 1.8266,
+      "learning_rate": 2.7110401393644464e-05,
+      "loss": 1.7584,
       "step": 2730
     },
     {
       "epoch": 3.47,
-      "learning_rate": 0.00022284864597835058,
-      "loss": 1.9772,
+      "learning_rate": 2.648722207052869e-05,
+      "loss": 1.7248,
       "step": 2735
     },
     {
       "epoch": 3.47,
-      "learning_rate": 0.00022116004649745826,
-      "loss": 1.9601,
+      "learning_rate": 2.587095778283469e-05,
+      "loss": 1.6746,
       "step": 2740
     },
     {
       "epoch": 3.48,
-      "learning_rate": 0.00021947605032907099,
-      "loss": 1.8989,
+      "learning_rate": 2.5261624111148647e-05,
+      "loss": 1.528,
       "step": 2745
     },
     {
       "epoch": 3.49,
-      "learning_rate": 0.00021779668527394187,
-      "loss": 1.9805,
+      "learning_rate": 2.4659236460834653e-05,
+      "loss": 1.6895,
       "step": 2750
     },
     {
       "epoch": 3.49,
-      "learning_rate": 0.0002161219790563691,
-      "loss": 1.8144,
+      "learning_rate": 2.4063810061645227e-05,
+      "loss": 1.6019,
       "step": 2755
     },
     {
       "epoch": 3.5,
-      "learning_rate": 0.00021445195932373986,
-      "loss": 1.9025,
+      "learning_rate": 2.3475359967336515e-05,
+      "loss": 1.722,
       "step": 2760
     },
     {
       "epoch": 3.5,
-      "learning_rate": 0.0002127866536460727,
-      "loss": 1.8811,
+      "learning_rate": 2.289390105528741e-05,
+      "loss": 1.6991,
       "step": 2765
     },
     {
       "epoch": 3.51,
-      "learning_rate": 0.00021112608951556316,
-      "loss": 1.998,
+      "learning_rate": 2.231944802612363e-05,
+      "loss": 1.5819,
       "step": 2770
     },
     {
       "epoch": 3.52,
-      "learning_rate": 0.00020947029434612886,
-      "loss": 1.8758,
+      "learning_rate": 2.1752015403346012e-05,
+      "loss": 1.6904,
       "step": 2775
     },
     {
       "epoch": 3.52,
-      "learning_rate": 0.00020781929547295815,
-      "loss": 1.8745,
+      "learning_rate": 2.1191617532963134e-05,
+      "loss": 1.7386,
       "step": 2780
     },
     {
       "epoch": 3.53,
-      "learning_rate": 0.00020617312015205842,
-      "loss": 1.8758,
+      "learning_rate": 2.0638268583128948e-05,
+      "loss": 1.6986,
       "step": 2785
     },
     {
       "epoch": 3.54,
-      "learning_rate": 0.00020453179555980533,
-      "loss": 1.9658,
+      "learning_rate": 2.009198254378436e-05,
+      "loss": 1.5248,
       "step": 2790
     },
     {
       "epoch": 3.54,
-      "learning_rate": 0.00020289534879249543,
-      "loss": 1.7781,
+      "learning_rate": 1.955277322630341e-05,
+      "loss": 1.6078,
       "step": 2795
     },
     {
       "epoch": 3.55,
-      "learning_rate": 0.00020126380686589807,
-      "loss": 1.9071,
+      "learning_rate": 1.90206542631444e-05,
+      "loss": 1.7261,
       "step": 2800
     },
     {
       "epoch": 3.56,
-      "learning_rate": 0.0001996371967148098,
-      "loss": 1.813,
+      "learning_rate": 1.849563910750508e-05,
+      "loss": 1.7388,
       "step": 2805
     },
     {
       "epoch": 3.56,
-      "learning_rate": 0.00019801554519260884,
-      "loss": 1.9647,
+      "learning_rate": 1.7977741032982505e-05,
+      "loss": 1.5929,
       "step": 2810
     },
     {
       "epoch": 3.57,
-      "learning_rate": 0.00019639887907081294,
-      "loss": 1.9145,
+      "learning_rate": 1.746697313323737e-05,
+      "loss": 1.6728,
       "step": 2815
     },
     {
       "epoch": 3.57,
-      "learning_rate": 0.00019478722503863706,
-      "loss": 1.9741,
+      "learning_rate": 1.6963348321663196e-05,
+      "loss": 1.6515,
       "step": 2820
     },
     {
       "epoch": 3.58,
-      "learning_rate": 0.0001931806097025517,
-      "loss": 1.9005,
+      "learning_rate": 1.6466879331059625e-05,
+      "loss": 1.7021,
       "step": 2825
     },
     {
       "epoch": 3.59,
-      "learning_rate": 0.00019157905958584582,
-      "loss": 1.9434,
+      "learning_rate": 1.5977578713310756e-05,
+      "loss": 1.6253,
       "step": 2830
     },
     {
       "epoch": 3.59,
-      "learning_rate": 0.0001899826011281865,
-      "loss": 1.8557,
+      "learning_rate": 1.549545883906733e-05,
+      "loss": 1.6577,
       "step": 2835
     },
     {
       "epoch": 3.6,
-      "learning_rate": 0.00018839126068518448,
-      "loss": 1.9641,
+      "learning_rate": 1.5020531897434607e-05,
+      "loss": 1.706,
       "step": 2840
     },
     {
       "epoch": 3.61,
-      "learning_rate": 0.0001868050645279576,
-      "loss": 1.9206,
+      "learning_rate": 1.4552809895663708e-05,
+      "loss": 1.5845,
       "step": 2845
     },
     {
       "epoch": 3.61,
-      "learning_rate": 0.00018522403884269813,
-      "loss": 1.8437,
+      "learning_rate": 1.4092304658848408e-05,
+      "loss": 1.6619,
       "step": 2850
     },
     {
       "epoch": 3.62,
-      "learning_rate": 0.00018364820973024,
-      "loss": 1.9311,
+      "learning_rate": 1.363902782962566e-05,
+      "loss": 1.6819,
       "step": 2855
     },
     {
       "epoch": 3.62,
-      "learning_rate": 0.00018207760320562794,
-      "loss": 1.9442,
+      "learning_rate": 1.3192990867881836e-05,
+      "loss": 1.7691,
       "step": 2860
     },
     {
       "epoch": 3.63,
-      "learning_rate": 0.00018051224519768816,
-      "loss": 1.8921,
+      "learning_rate": 1.2754205050462595e-05,
+      "loss": 1.7589,
       "step": 2865
     },
     {
       "epoch": 3.64,
-      "learning_rate": 0.00017895216154859966,
-      "loss": 1.8873,
+      "learning_rate": 1.2322681470887975e-05,
+      "loss": 1.5541,
       "step": 2870
     },
     {
       "epoch": 3.64,
-      "learning_rate": 0.00017739737801346895,
-      "loss": 1.9562,
+      "learning_rate": 1.1898431039071744e-05,
+      "loss": 1.6272,
       "step": 2875
     },
     {
       "epoch": 3.65,
-      "learning_rate": 0.00017584792025990287,
-      "loss": 1.9437,
+      "learning_rate": 1.1481464481045766e-05,
+      "loss": 1.7351,
       "step": 2880
     },
     {
       "epoch": 3.66,
-      "learning_rate": 0.00017430381386758747,
-      "loss": 1.8848,
+      "learning_rate": 1.1071792338688712e-05,
+      "loss": 1.7503,
       "step": 2885
     },
     {
       "epoch": 3.66,
-      "learning_rate": 0.00017276508432786304,
-      "loss": 1.9044,
+      "learning_rate": 1.066942496945954e-05,
+      "loss": 1.6801,
       "step": 2890
     },
     {
       "epoch": 3.67,
-      "learning_rate": 0.00017123175704330512,
-      "loss": 1.8581,
+      "learning_rate": 1.0274372546135668e-05,
+      "loss": 1.6642,
       "step": 2895
     },
     {
       "epoch": 3.68,
-      "learning_rate": 0.00016970385732730414,
-      "loss": 1.9017,
+      "learning_rate": 9.886645056555786e-06,
+      "loss": 1.683,
       "step": 2900
     },
     {
       "epoch": 3.68,
-      "learning_rate": 0.00016818141040364815,
-      "loss": 1.7544,
+      "learning_rate": 9.50625230336729e-06,
+      "loss": 1.6882,
       "step": 2905
     },
     {
       "epoch": 3.69,
-      "learning_rate": 0.00016666444140610615,
-      "loss": 1.92,
+      "learning_rate": 9.133203903778507e-06,
+      "loss": 1.6786,
       "step": 2910
     },
     {
       "epoch": 3.69,
-      "learning_rate": 0.00016515297537801304,
-      "loss": 1.8981,
+      "learning_rate": 8.767509289315499e-06,
+      "loss": 1.6582,
       "step": 2915
     },
     {
       "epoch": 3.7,
-      "learning_rate": 0.00016364703727185664,
-      "loss": 1.97,
+      "learning_rate": 8.409177705583647e-06,
+      "loss": 1.7466,
       "step": 2920
     },
     {
       "epoch": 3.71,
-      "learning_rate": 0.00016214665194886474,
-      "loss": 1.8704,
+      "learning_rate": 8.05821821203394e-06,
+      "loss": 1.6726,
       "step": 2925
     },
     {
       "epoch": 3.71,
-      "learning_rate": 0.00016065184417859613,
-      "loss": 1.9395,
+      "learning_rate": 7.714639681733758e-06,
+      "loss": 1.7564,
       "step": 2930
     },
     {
       "epoch": 3.72,
-      "learning_rate": 0.00015916263863853,
-      "loss": 1.9092,
+      "learning_rate": 7.378450801142821e-06,
+      "loss": 1.6652,
       "step": 2935
     },
     {
       "epoch": 3.73,
-      "learning_rate": 0.0001576790599136607,
-      "loss": 1.8171,
+      "learning_rate": 7.049660069893292e-06,
+      "loss": 1.6479,
       "step": 2940
     },
     {
       "epoch": 3.73,
-      "learning_rate": 0.00015620113249608942,
-      "loss": 1.8367,
+      "learning_rate": 6.7282758005750895e-06,
+      "loss": 1.5413,
       "step": 2945
     },
     {
       "epoch": 3.74,
-      "learning_rate": 0.00015472888078462172,
-      "loss": 1.9402,
+      "learning_rate": 6.414306118525614e-06,
+      "loss": 1.6467,
       "step": 2950
     },
     {
       "epoch": 3.75,
-      "learning_rate": 0.00015326232908436405,
-      "loss": 1.9839,
+      "learning_rate": 6.107758961624354e-06,
+      "loss": 1.667,
       "step": 2955
     },
     {
       "epoch": 3.75,
-      "learning_rate": 0.00015180150160632195,
-      "loss": 1.8956,
+      "learning_rate": 5.808642080092207e-06,
+      "loss": 1.798,
       "step": 2960
     },
     {
       "epoch": 3.76,
-      "learning_rate": 0.00015034642246700204,
-      "loss": 1.8861,
+      "learning_rate": 5.516963036295608e-06,
+      "loss": 1.5463,
       "step": 2965
     },
     {
       "epoch": 3.76,
-      "learning_rate": 0.00014889711568801152,
-      "loss": 2.0093,
+      "learning_rate": 5.2327292045551085e-06,
+      "loss": 1.6894,
       "step": 2970
     },
     {
       "epoch": 3.77,
-      "learning_rate": 0.00014745360519566382,
-      "loss": 1.9242,
+      "learning_rate": 4.955947770959134e-06,
+      "loss": 1.7042,
       "step": 2975
     },
     {
       "epoch": 3.78,
-      "learning_rate": 0.0001460159148205819,
-      "loss": 1.9005,
+      "learning_rate": 4.686625733182326e-06,
+      "loss": 1.6668,
       "step": 2980
     },
     {
       "epoch": 3.78,
-      "learning_rate": 0.0001445840682973062,
-      "loss": 1.9462,
+      "learning_rate": 4.4247699003085115e-06,
+      "loss": 1.6509,
       "step": 2985
     },
     {
       "epoch": 3.79,
-      "learning_rate": 0.00014315808926390188,
-      "loss": 2.0249,
+      "learning_rate": 4.170386892658584e-06,
+      "loss": 1.6335,
       "step": 2990
     },
     {
       "epoch": 3.8,
-      "learning_rate": 0.00014173800126156917,
-      "loss": 1.9317,
+      "learning_rate": 3.9234831416230964e-06,
+      "loss": 1.6733,
       "step": 2995
     },
     {
       "epoch": 3.8,
-      "learning_rate": 0.00014032382773425384,
-      "loss": 2.0039,
+      "learning_rate": 3.684064889499794e-06,
+      "loss": 1.5888,
       "step": 3000
     },
     {
       "epoch": 3.81,
-      "learning_rate": 0.00013891559202826133,
-      "loss": 1.9658,
+      "learning_rate": 3.452138189335707e-06,
+      "loss": 1.6766,
       "step": 3005
     },
     {
       "epoch": 3.81,
-      "learning_rate": 0.0001375133173918709,
-      "loss": 1.9448,
+      "learning_rate": 3.22770890477394e-06,
+      "loss": 1.7595,
       "step": 3010
     },
     {
       "epoch": 3.82,
-      "learning_rate": 0.00013611702697495086,
-      "loss": 1.8653,
+      "learning_rate": 3.0107827099057902e-06,
+      "loss": 1.6369,
       "step": 3015
     },
     {
       "epoch": 3.83,
-      "learning_rate": 0.0001347267438285786,
-      "loss": 1.9222,
+      "learning_rate": 2.8013650891269947e-06,
+      "loss": 1.7237,
       "step": 3020
     },
     {
       "epoch": 3.83,
-      "learning_rate": 0.0001333424909046574,
-      "loss": 1.8796,
+      "learning_rate": 2.5994613369992424e-06,
+      "loss": 1.6978,
       "step": 3025
     },
     {
       "epoch": 3.84,
-      "learning_rate": 0.0001319642910555396,
-      "loss": 1.8942,
+      "learning_rate": 2.405076558116181e-06,
+      "loss": 1.6101,
       "step": 3030
     },
     {
       "epoch": 3.85,
-      "learning_rate": 0.00013059216703364813,
-      "loss": 1.9026,
+      "learning_rate": 2.2182156669744543e-06,
+      "loss": 1.7091,
       "step": 3035
     },
     {
       "epoch": 3.85,
-      "learning_rate": 0.00012922614149110152,
-      "loss": 1.9212,
+      "learning_rate": 2.038883387849499e-06,
+      "loss": 1.7317,
       "step": 3040
     },
     {
       "epoch": 3.86,
-      "learning_rate": 0.0001278662369793398,
-      "loss": 1.9068,
+      "learning_rate": 1.8670842546759102e-06,
+      "loss": 1.6832,
       "step": 3045
     },
     {
       "epoch": 3.87,
-      "learning_rate": 0.000126512475948752,
-      "loss": 1.9956,
+      "learning_rate": 1.7028226109329967e-06,
+      "loss": 1.7484,
       "step": 3050
     },
     {
       "epoch": 3.87,
-      "learning_rate": 0.00012516488074830584,
-      "loss": 1.8775,
+      "learning_rate": 1.5461026095349048e-06,
+      "loss": 1.7254,
       "step": 3055
     },
     {
       "epoch": 3.88,
-      "learning_rate": 0.00012382347362517815,
-      "loss": 1.9005,
+      "learning_rate": 1.396928212725701e-06,
+      "loss": 1.6582,
       "step": 3060
     },
     {
       "epoch": 3.88,
-      "learning_rate": 0.00012248827672438868,
-      "loss": 1.9538,
+      "learning_rate": 1.255303191978918e-06,
+      "loss": 1.6625,
       "step": 3065
     },
     {
       "epoch": 3.89,
-      "learning_rate": 0.00012115931208843318,
-      "loss": 1.8309,
+      "learning_rate": 1.1212311279026331e-06,
+      "loss": 1.7218,
       "step": 3070
     },
     {
       "epoch": 3.9,
-      "learning_rate": 0.00011983660165692079,
-      "loss": 2.0001,
+      "learning_rate": 9.947154101486056e-07,
+      "loss": 1.7083,
       "step": 3075
     },
     {
       "epoch": 3.9,
-      "learning_rate": 0.00011852016726621096,
-      "loss": 1.8293,
+      "learning_rate": 8.757592373266787e-07,
+      "loss": 1.6781,
       "step": 3080
     },
     {
       "epoch": 3.91,
-      "learning_rate": 0.00011721003064905328,
-      "loss": 1.9337,
+      "learning_rate": 7.643656169239454e-07,
+      "loss": 1.6367,
       "step": 3085
     },
     {
       "epoch": 3.92,
-      "learning_rate": 0.0001159062134342289,
-      "loss": 1.912,
+      "learning_rate": 6.60537365228675e-07,
+      "loss": 1.7258,
       "step": 3090
     },
     {
       "epoch": 3.92,
-      "learning_rate": 0.00011460873714619274,
-      "loss": 1.8585,
+      "learning_rate": 5.642771072591701e-07,
+      "loss": 1.6879,
       "step": 3095
     },
     {
       "epoch": 3.93,
-      "learning_rate": 0.00011331762320471906,
-      "loss": 1.9535,
+      "learning_rate": 4.755872766972868e-07,
+      "loss": 1.6689,
       "step": 3100
     },
     {
       "epoch": 3.94,
-      "learning_rate": 0.00011203289292454726,
-      "loss": 1.9018,
+      "learning_rate": 3.9447011582698363e-07,
+      "loss": 1.6646,
       "step": 3105
     },
     {
       "epoch": 3.94,
-      "learning_rate": 0.00011075456751503043,
-      "loss": 1.9418,
+      "learning_rate": 3.2092767547770015e-07,
+      "loss": 1.738,
       "step": 3110
     },
     {
       "epoch": 3.95,
-      "learning_rate": 0.0001094826680797843,
-      "loss": 1.9581,
+      "learning_rate": 2.5496181497233206e-07,
+      "loss": 1.6647,
       "step": 3115
     },
     {
       "epoch": 3.95,
-      "learning_rate": 0.00010821721561634024,
-      "loss": 1.9045,
+      "learning_rate": 1.9657420208036846e-07,
+      "loss": 1.6639,
       "step": 3120
     },
     {
       "epoch": 3.96,
-      "learning_rate": 0.00010695823101579727,
-      "loss": 1.8314,
+      "learning_rate": 1.4576631297562592e-07,
+      "loss": 1.7068,
       "step": 3125
     },
     {
       "epoch": 3.97,
-      "learning_rate": 0.00010570573506247811,
-      "loss": 1.9717,
+      "learning_rate": 1.025394321990447e-07,
+      "loss": 1.6258,
       "step": 3130
     },
     {
       "epoch": 3.97,
-      "learning_rate": 0.00010445974843358564,
-      "loss": 1.9037,
+      "learning_rate": 6.689465262614823e-08,
+      "loss": 1.6464,
       "step": 3135
     },
     {
       "epoch": 3.98,
-      "learning_rate": 0.0001032202916988616,
-      "loss": 1.8119,
+      "learning_rate": 3.8832875439331887e-08,
+      "loss": 1.6809,
       "step": 3140
     },
     {
       "epoch": 3.99,
-      "learning_rate": 0.00010198738532024716,
-      "loss": 1.8549,
+      "learning_rate": 1.8354810105214446e-08,
+      "loss": 1.7999,
       "step": 3145
     },
     {
       "epoch": 3.99,
-      "learning_rate": 0.00010076104965154459,
-      "loss": 1.8318,
+      "learning_rate": 5.4609743566191854e-09,
+      "loss": 1.6112,
       "step": 3150
     },
     {
       "epoch": 4.0,
-      "learning_rate": 9.9541304938082e-05,
-      "loss": 1.9885,
+      "learning_rate": 1.516941795509652e-10,
+      "loss": 1.7376,
       "step": 3155
     },
     {
       "epoch": 4.0,
-      "eval_loss": 2.0513949394226074,
-      "eval_runtime": 95.3176,
-      "eval_samples_per_second": 134.372,
-      "eval_steps_per_second": 16.796,
+      "eval_loss": 1.912619948387146,
+      "eval_runtime": 92.8221,
+      "eval_samples_per_second": 137.984,
+      "eval_steps_per_second": 8.629,
       "step": 3156
     },
     {
-      "epoch": 4.01,
-      "learning_rate": 9.832817131637867e-05,
-      "loss": 1.9155,
-      "step": 3160
-    },
-    {
-      "epoch": 4.01,
-      "learning_rate": 9.71216688138128e-05,
-      "loss": 1.8082,
-      "step": 3165
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 9.592181734829036e-05,
-      "loss": 1.8647,
-      "step": 3170
-    },
-    {
-      "epoch": 4.02,
-      "learning_rate": 9.472863672791721e-05,
-      "loss": 1.8277,
-      "step": 3175
-    },
-    {
-      "epoch": 4.03,
-      "learning_rate": 9.35421466506714e-05,
-      "loss": 1.9529,
-      "step": 3180
-    },
-    {
-      "epoch": 4.04,
-      "learning_rate": 9.236236670407771e-05,
-      "loss": 1.8167,
-      "step": 3185
-    },
-    {
-      "epoch": 4.04,
-      "learning_rate": 9.11893163648857e-05,
-      "loss": 1.9159,
-      "step": 3190
-    },
-    {
-      "epoch": 4.05,
-      "learning_rate": 9.002301499874621e-05,
-      "loss": 1.7958,
-      "step": 3195
-    },
-    {
-      "epoch": 4.06,
-      "learning_rate": 8.886348185989334e-05,
-      "loss": 1.865,
-      "step": 3200
-    },
-    {
-      "epoch": 4.06,
-      "learning_rate": 8.771073609082531e-05,
-      "loss": 1.859,
-      "step": 3205
-    },
-    {
-      "epoch": 4.07,
-      "learning_rate": 8.656479672198947e-05,
-      "loss": 1.8602,
-      "step": 3210
-    },
-    {
-      "epoch": 4.07,
-      "learning_rate": 8.54256826714676e-05,
-      "loss": 1.8442,
-      "step": 3215
-    },
-    {
-      "epoch": 4.08,
-      "learning_rate": 8.429341274466368e-05,
-      "loss": 1.8399,
-      "step": 3220
-    },
-    {
-      "epoch": 4.09,
-      "learning_rate": 8.316800563399307e-05,
-      "loss": 1.895,
-      "step": 3225
-    },
-    {
-      "epoch": 4.09,
-      "learning_rate": 8.204947991857486e-05,
-      "loss": 1.8071,
-      "step": 3230
-    },
-    {
-      "epoch": 4.1,
-      "learning_rate": 8.093785406392429e-05,
-      "loss": 1.9761,
-      "step": 3235
-    },
-    {
-      "epoch": 4.11,
-      "learning_rate": 7.983314642164786e-05,
-      "loss": 1.8828,
-      "step": 3240
-    },
-    {
-      "epoch": 4.11,
-      "learning_rate": 7.873537522914154e-05,
-      "loss": 1.9077,
-      "step": 3245
-    },
-    {
-      "epoch": 4.12,
-      "learning_rate": 7.764455860928793e-05,
-      "loss": 1.867,
-      "step": 3250
-    },
-    {
-      "epoch": 4.13,
-      "learning_rate": 7.656071457015878e-05,
-      "loss": 1.8226,
-      "step": 3255
-    },
-    {
-      "epoch": 4.13,
-      "learning_rate": 7.548386100471621e-05,
-      "loss": 1.821,
-      "step": 3260
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 7.441401569051847e-05,
-      "loss": 1.8415,
-      "step": 3265
-    },
-    {
-      "epoch": 4.14,
-      "learning_rate": 7.335119628942594e-05,
-      "loss": 1.8206,
-      "step": 3270
-    },
-    {
-      "epoch": 4.15,
-      "learning_rate": 7.229542034730952e-05,
-      "loss": 1.8202,
-      "step": 3275
-    },
-    {
-      "epoch": 4.16,
-      "learning_rate": 7.124670529376126e-05,
-      "loss": 1.9373,
-      "step": 3280
-    },
-    {
-      "epoch": 4.16,
-      "learning_rate": 7.020506844180608e-05,
-      "loss": 1.8372,
-      "step": 3285
-    },
-    {
-      "epoch": 4.17,
-      "learning_rate": 6.917052698761666e-05,
-      "loss": 1.8923,
-      "step": 3290
-    },
-    {
-      "epoch": 4.18,
-      "learning_rate": 6.814309801022872e-05,
-      "loss": 1.8606,
-      "step": 3295
-    },
-    {
-      "epoch": 4.18,
-      "learning_rate": 6.712279847126035e-05,
-      "loss": 1.8864,
-      "step": 3300
-    },
-    {
-      "epoch": 4.19,
-      "learning_rate": 6.610964521463031e-05,
-      "loss": 1.8178,
-      "step": 3305
-    },
-    {
-      "epoch": 4.2,
-      "learning_rate": 6.510365496628129e-05,
-      "loss": 1.8438,
-      "step": 3310
-    },
-    {
-      "epoch": 4.2,
-      "learning_rate": 6.410484433390334e-05,
-      "loss": 1.79,
-      "step": 3315
-    },
-    {
-      "epoch": 4.21,
-      "learning_rate": 6.311322980665934e-05,
-      "loss": 1.8641,
-      "step": 3320
-    },
-    {
-      "epoch": 4.21,
-      "learning_rate": 6.212882775491352e-05,
-      "loss": 1.8712,
-      "step": 3325
-    },
-    {
-      "epoch": 4.22,
-      "learning_rate": 6.115165442996056e-05,
-      "loss": 1.863,
-      "step": 3330
-    },
-    {
-      "epoch": 4.23,
-      "learning_rate": 6.018172596375776e-05,
-      "loss": 1.8744,
-      "step": 3335
-    },
-    {
-      "epoch": 4.23,
-      "learning_rate": 5.921905836865821e-05,
-      "loss": 1.8524,
-      "step": 3340
-    },
-    {
-      "epoch": 4.24,
-      "learning_rate": 5.826366753714707e-05,
-      "loss": 1.9031,
-      "step": 3345
-    },
-    {
-      "epoch": 4.25,
-      "learning_rate": 5.731556924157844e-05,
-      "loss": 1.784,
-      "step": 3350
-    },
-    {
-      "epoch": 4.25,
-      "learning_rate": 5.6374779133916035e-05,
-      "loss": 1.8172,
-      "step": 3355
-    },
-    {
-      "epoch": 4.26,
-      "learning_rate": 5.544131274547365e-05,
-      "loss": 1.8779,
-      "step": 3360
-    },
-    {
-      "epoch": 4.26,
-      "learning_rate": 5.4515185486659454e-05,
-      "loss": 1.7402,
-      "step": 3365
-    },
-    {
-      "epoch": 4.27,
-      "learning_rate": 5.359641264672155e-05,
-      "loss": 1.9493,
-      "step": 3370
-    },
-    {
-      "epoch": 4.28,
-      "learning_rate": 5.268500939349513e-05,
-      "loss": 1.8279,
-      "step": 3375
-    },
-    {
-      "epoch": 4.28,
-      "learning_rate": 5.178099077315257e-05,
-      "loss": 1.7599,
-      "step": 3380
-    },
-    {
-      "epoch": 4.29,
-      "learning_rate": 5.0884371709954804e-05,
-      "loss": 1.7905,
-      "step": 3385
-    },
-    {
-      "epoch": 4.3,
-      "learning_rate": 4.999516700600504e-05,
-      "loss": 1.8888,
-      "step": 3390
-    },
-    {
-      "epoch": 4.3,
-      "learning_rate": 4.911339134100401e-05,
-      "loss": 1.8539,
-      "step": 3395
-    },
-    {
-      "epoch": 4.31,
-      "learning_rate": 4.823905927200833e-05,
-      "loss": 1.8406,
-      "step": 3400
-    },
-    {
-      "epoch": 4.32,
-      "learning_rate": 4.737218523318965e-05,
-      "loss": 1.8983,
-      "step": 3405
-    },
-    {
-      "epoch": 4.32,
-      "learning_rate": 4.65127835355964e-05,
-      "loss": 1.7739,
-      "step": 3410
-    },
-    {
-      "epoch": 4.33,
-      "learning_rate": 4.566086836691791e-05,
-      "loss": 1.8877,
-      "step": 3415
-    },
-    {
-      "epoch": 4.33,
-      "learning_rate": 4.4816453791249436e-05,
-      "loss": 1.8403,
-      "step": 3420
-    },
-    {
-      "epoch": 4.34,
-      "learning_rate": 4.3979553748861035e-05,
-      "loss": 1.9217,
-      "step": 3425
-    },
-    {
-      "epoch": 4.35,
-      "learning_rate": 4.3150182055966384e-05,
-      "loss": 1.8122,
-      "step": 3430
-    },
-    {
-      "epoch": 4.35,
-      "learning_rate": 4.232835240449534e-05,
-      "loss": 1.8977,
-      "step": 3435
-    },
-    {
-      "epoch": 4.36,
-      "learning_rate": 4.1514078361867826e-05,
-      "loss": 1.9035,
-      "step": 3440
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 4.070737337076963e-05,
-      "loss": 1.8811,
-      "step": 3445
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 3.990825074893056e-05,
-      "loss": 1.9786,
-      "step": 3450
-    },
-    {
-      "epoch": 4.38,
-      "learning_rate": 3.911672368890462e-05,
-      "loss": 1.9312,
-      "step": 3455
-    },
-    {
-      "epoch": 4.39,
-      "learning_rate": 3.8332805257852374e-05,
-      "loss": 1.8757,
-      "step": 3460
-    },
-    {
-      "epoch": 4.39,
-      "learning_rate": 3.755650839732489e-05,
-      "loss": 1.8269,
-      "step": 3465
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 3.67878459230504e-05,
-      "loss": 1.7784,
-      "step": 3470
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 3.6026830524722444e-05,
-      "loss": 1.8365,
-      "step": 3475
-    },
-    {
-      "epoch": 4.41,
-      "learning_rate": 3.527347476579068e-05,
-      "loss": 1.9105,
-      "step": 3480
-    },
-    {
-      "epoch": 4.42,
-      "learning_rate": 3.4527791083253154e-05,
-      "loss": 1.8909,
-      "step": 3485
-    },
-    {
-      "epoch": 4.42,
-      "learning_rate": 3.3789791787451344e-05,
-      "loss": 1.9213,
-      "step": 3490
-    },
-    {
-      "epoch": 4.43,
-      "learning_rate": 3.305948906186662e-05,
-      "loss": 1.7467,
-      "step": 3495
-    },
-    {
-      "epoch": 4.44,
-      "learning_rate": 3.233689496291931e-05,
-      "loss": 1.8005,
-      "step": 3500
-    },
-    {
-      "epoch": 4.44,
-      "learning_rate": 3.162202141976955e-05,
-      "loss": 1.9142,
-      "step": 3505
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 3.09148802341202e-05,
-      "loss": 1.8589,
-      "step": 3510
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 3.0215483080022478e-05,
-      "loss": 1.9262,
-      "step": 3515
-    },
-    {
-      "epoch": 4.46,
-      "learning_rate": 2.9523841503682757e-05,
-      "loss": 1.8654,
-      "step": 3520
-    },
-    {
-      "epoch": 4.47,
-      "learning_rate": 2.8839966923272288e-05,
-      "loss": 1.8914,
-      "step": 3525
-    },
-    {
-      "epoch": 4.47,
-      "learning_rate": 2.8163870628738375e-05,
-      "loss": 1.8989,
-      "step": 3530
-    },
-    {
-      "epoch": 4.48,
-      "learning_rate": 2.7495563781618328e-05,
-      "loss": 1.8157,
-      "step": 3535
-    },
-    {
-      "epoch": 4.49,
-      "learning_rate": 2.6835057414855036e-05,
-      "loss": 1.8452,
-      "step": 3540
-    },
-    {
-      "epoch": 4.49,
-      "learning_rate": 2.6182362432614513e-05,
-      "loss": 1.8583,
-      "step": 3545
-    },
-    {
-      "epoch": 4.5,
-      "learning_rate": 2.5537489610106824e-05,
-      "loss": 1.7927,
-      "step": 3550
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 2.4900449593406984e-05,
-      "loss": 1.928,
-      "step": 3555
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 2.4271252899280173e-05,
-      "loss": 1.8615,
-      "step": 3560
-    },
-    {
-      "epoch": 4.52,
-      "learning_rate": 2.364990991500743e-05,
-      "loss": 1.8917,
-      "step": 3565
-    },
-    {
-      "epoch": 4.52,
-      "learning_rate": 2.3036430898214832e-05,
-      "loss": 1.7866,
-      "step": 3570
-    },
-    {
-      "epoch": 4.53,
-      "learning_rate": 2.2430825976703484e-05,
-      "loss": 1.768,
-      "step": 3575
-    },
-    {
-      "epoch": 4.54,
-      "learning_rate": 2.1833105148282905e-05,
-      "loss": 1.7868,
-      "step": 3580
-    },
-    {
-      "epoch": 4.54,
-      "learning_rate": 2.1243278280605515e-05,
-      "loss": 1.8383,
-      "step": 3585
-    },
-    {
-      "epoch": 4.55,
-      "learning_rate": 2.0661355111004032e-05,
-      "loss": 1.8518,
-      "step": 3590
-    },
-    {
-      "epoch": 4.56,
-      "learning_rate": 2.0087345246330713e-05,
-      "loss": 1.939,
-      "step": 3595
-    },
-    {
-      "epoch": 4.56,
-      "learning_rate": 1.9521258162798263e-05,
-      "loss": 1.8259,
-      "step": 3600
-    },
-    {
-      "epoch": 4.57,
-      "learning_rate": 1.8963103205824395e-05,
-      "loss": 1.9559,
-      "step": 3605
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 1.8412889589876358e-05,
-      "loss": 1.8916,
-      "step": 3610
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 1.787062639831988e-05,
-      "loss": 1.8576,
-      "step": 3615
-    },
-    {
-      "epoch": 4.59,
-      "learning_rate": 1.7336322583268336e-05,
-      "loss": 1.8634,
-      "step": 3620
-    },
-    {
-      "epoch": 4.59,
-      "learning_rate": 1.6809986965435677e-05,
-      "loss": 1.9129,
-      "step": 3625
-    },
-    {
-      "epoch": 4.6,
-      "learning_rate": 1.6291628233990387e-05,
-      "loss": 1.7839,
-      "step": 3630
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 1.5781254946412028e-05,
-      "loss": 1.8129,
-      "step": 3635
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 1.5278875528350277e-05,
-      "loss": 1.8645,
-      "step": 3640
-    },
-    {
-      "epoch": 4.62,
-      "learning_rate": 1.4784498273485436e-05,
-      "loss": 1.7878,
-      "step": 3645
-    },
-    {
-      "epoch": 4.63,
-      "learning_rate": 1.4298131343391918e-05,
-      "loss": 1.8806,
-      "step": 3650
-    },
-    {
-      "epoch": 4.63,
-      "learning_rate": 1.3819782767403033e-05,
-      "loss": 1.8544,
-      "step": 3655
-    },
-    {
-      "epoch": 4.64,
-      "learning_rate": 1.3349460442479089e-05,
-      "loss": 1.9462,
-      "step": 3660
-    },
-    {
-      "epoch": 4.65,
-      "learning_rate": 1.2887172133076219e-05,
-      "loss": 1.8129,
-      "step": 3665
-    },
-    {
-      "epoch": 4.65,
-      "learning_rate": 1.243292547101893e-05,
-      "loss": 1.8055,
-      "step": 3670
-    },
-    {
-      "epoch": 4.66,
-      "learning_rate": 1.1986727955373588e-05,
-      "loss": 1.8297,
-      "step": 3675
-    },
-    {
-      "epoch": 4.66,
-      "learning_rate": 1.1548586952324912e-05,
-      "loss": 1.7961,
-      "step": 3680
-    },
-    {
-      "epoch": 4.67,
-      "learning_rate": 1.1118509695054235e-05,
-      "loss": 1.9098,
-      "step": 3685
-    },
-    {
-      "epoch": 4.68,
-      "learning_rate": 1.0696503283620319e-05,
-      "loss": 1.7803,
-      "step": 3690
-    },
-    {
-      "epoch": 4.68,
-      "learning_rate": 1.0282574684841784e-05,
-      "loss": 1.9406,
-      "step": 3695
-    },
-    {
-      "epoch": 4.69,
-      "learning_rate": 9.876730732182203e-06,
-      "loss": 1.8219,
-      "step": 3700
-    },
-    {
-      "epoch": 4.7,
-      "learning_rate": 9.478978125637583e-06,
-      "loss": 1.9694,
-      "step": 3705
-    },
-    {
-      "epoch": 4.7,
-      "learning_rate": 9.089323431625273e-06,
-      "loss": 1.8647,
-      "step": 3710
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 8.70777308287618e-06,
-      "loss": 1.8981,
-      "step": 3715
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 8.334333378327785e-06,
-      "loss": 1.8218,
-      "step": 3720
-    },
-    {
-      "epoch": 4.72,
-      "learning_rate": 7.969010483020845e-06,
-      "loss": 1.8954,
-      "step": 3725
-    },
-    {
-      "epoch": 4.73,
-      "learning_rate": 7.611810427997412e-06,
-      "loss": 1.7638,
-      "step": 3730
-    },
-    {
-      "epoch": 4.73,
-      "learning_rate": 7.262739110200922e-06,
-      "loss": 1.7664,
-      "step": 3735
-    },
-    {
-      "epoch": 4.74,
-      "learning_rate": 6.921802292379376e-06,
-      "loss": 1.8825,
-      "step": 3740
-    },
-    {
-      "epoch": 4.75,
-      "learning_rate": 6.589005602989862e-06,
-      "loss": 1.7941,
-      "step": 3745
-    },
-    {
-      "epoch": 4.75,
-      "learning_rate": 6.264354536105799e-06,
-      "loss": 1.8809,
-      "step": 3750
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 5.947854451326007e-06,
-      "loss": 1.7973,
-      "step": 3755
-    },
-    {
-      "epoch": 4.77,
-      "learning_rate": 5.639510573686613e-06,
-      "loss": 1.9447,
-      "step": 3760
-    },
-    {
-      "epoch": 4.77,
-      "learning_rate": 5.339327993574339e-06,
-      "loss": 1.8651,
-      "step": 3765
-    },
-    {
-      "epoch": 4.78,
-      "learning_rate": 5.047311666642851e-06,
-      "loss": 1.8057,
-      "step": 3770
-    },
-    {
-      "epoch": 4.78,
-      "learning_rate": 4.763466413730821e-06,
-      "loss": 1.8872,
-      "step": 3775
-    },
-    {
-      "epoch": 4.79,
-      "learning_rate": 4.4877969207821615e-06,
-      "loss": 1.7565,
-      "step": 3780
-    },
-    {
-      "epoch": 4.8,
-      "learning_rate": 4.2203077387688586e-06,
-      "loss": 1.9462,
-      "step": 3785
-    },
-    {
-      "epoch": 4.8,
-      "learning_rate": 3.961003283615816e-06,
-      "loss": 1.8327,
-      "step": 3790
-    },
-    {
-      "epoch": 4.81,
-      "learning_rate": 3.709887836128023e-06,
-      "loss": 1.8064,
-      "step": 3795
-    },
-    {
-      "epoch": 4.82,
-      "learning_rate": 3.4669655419196087e-06,
-      "loss": 1.8423,
-      "step": 3800
-    },
-    {
-      "epoch": 4.82,
-      "learning_rate": 3.232240411345788e-06,
-      "loss": 1.887,
-      "step": 3805
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 3.0057163194363023e-06,
-      "loss": 1.7267,
-      "step": 3810
-    },
-    {
-      "epoch": 4.84,
-      "learning_rate": 2.7873970058316934e-06,
-      "loss": 1.8499,
-      "step": 3815
-    },
-    {
-      "epoch": 4.84,
-      "learning_rate": 2.5772860747214078e-06,
-      "loss": 1.8719,
-      "step": 3820
-    },
-    {
-      "epoch": 4.85,
-      "learning_rate": 2.3753869947843455e-06,
-      "loss": 1.8829,
-      "step": 3825
-    },
-    {
-      "epoch": 4.85,
-      "learning_rate": 2.1817030991317934e-06,
-      "loss": 1.8648,
-      "step": 3830
-    },
-    {
-      "epoch": 4.86,
-      "learning_rate": 1.9962375852519698e-06,
-      "loss": 1.8077,
-      "step": 3835
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 1.8189935149576786e-06,
-      "loss": 1.7893,
-      "step": 3840
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 1.6499738143354593e-06,
-      "loss": 1.8117,
-      "step": 3845
-    },
-    {
-      "epoch": 4.88,
-      "learning_rate": 1.4891812736974596e-06,
-      "loss": 1.8163,
-      "step": 3850
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 1.3366185475351955e-06,
-      "loss": 1.8515,
-      "step": 3855
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 1.192288154476029e-06,
-      "loss": 1.8055,
-      "step": 3860
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 1.0561924772412024e-06,
-      "loss": 1.8637,
-      "step": 3865
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 9.283337626068144e-07,
-      "loss": 1.8176,
-      "step": 3870
-    },
-    {
-      "epoch": 4.91,
-      "learning_rate": 8.087141213665717e-07,
-      "loss": 1.9019,
-      "step": 3875
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 6.973355282969274e-07,
-      "loss": 1.8888,
-      "step": 3880
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 5.941998221247191e-07,
-      "loss": 1.8405,
-      "step": 3885
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 4.993087054963597e-07,
-      "loss": 1.8748,
-      "step": 3890
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 4.1266374495024795e-07,
-      "loss": 1.7451,
-      "step": 3895
-    },
-    {
-      "epoch": 4.94,
-      "learning_rate": 3.342663708905125e-07,
-      "loss": 1.8469,
-      "step": 3900
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 2.6411787756353e-07,
-      "loss": 1.9005,
-      "step": 3905
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 2.0221942303666474e-07,
-      "loss": 1.8063,
-      "step": 3910
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 1.4857202917900603e-07,
-      "loss": 1.8288,
-      "step": 3915
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 1.0317658164454846e-07,
-      "loss": 1.8788,
-      "step": 3920
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 6.603382985759243e-08,
-      "loss": 1.937,
-      "step": 3925
-    },
-    {
-      "epoch": 4.98,
-      "learning_rate": 3.714438700025413e-08,
-      "loss": 1.7552,
-      "step": 3930
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 1.650873000258457e-08,
-      "loss": 1.803,
-      "step": 3935
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 4.1271995344094316e-09,
-      "loss": 1.8769,
-      "step": 3940
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 0.0,
-      "loss": 1.806,
-      "step": 3945
-    },
-    {
-      "epoch": 5.0,
-      "eval_loss": 2.0400655269622803,
-      "eval_runtime": 95.5556,
-      "eval_samples_per_second": 134.037,
-      "eval_steps_per_second": 16.755,
-      "step": 3945
-    },
-    {
-      "epoch": 5.0,
-      "step": 3945,
-      "total_flos": 1.3195381386195763e+17,
-      "train_loss": 2.171095475799986,
-      "train_runtime": 13808.1422,
-      "train_samples_per_second": 36.573,
-      "train_steps_per_second": 0.286
+      "epoch": 4.0,
+      "step": 3156,
+      "total_flos": 1.0556101300007731e+17,
+      "train_loss": 1.8003431212237277,
+      "train_runtime": 10681.1524,
+      "train_samples_per_second": 37.823,
+      "train_steps_per_second": 0.295
     }
   ],
-  "max_steps": 3945,
-  "num_train_epochs": 5,
-  "total_flos": 1.3195381386195763e+17,
+  "max_steps": 3156,
+  "num_train_epochs": 4,
+  "total_flos": 1.0556101300007731e+17,
   "trial_name": null,
   "trial_params": null
 }