diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,2023 +1,6685 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.0,
-  "global_step": 333,
+  "epoch": 10.0,
+  "global_step": 1110,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.01,
-      "learning_rate": 1.993993993993994e-05,
+      "learning_rate": 0.00019981981981981984,
       "loss": 9.3363,
       "step": 1
     },
     {
       "epoch": 0.02,
-      "learning_rate": 1.987987987987988e-05,
-      "loss": 6.3834,
+      "learning_rate": 0.00019963963963963967,
+      "loss": 5.0479,
       "step": 2
     },
     {
       "epoch": 0.03,
-      "learning_rate": 1.981981981981982e-05,
-      "loss": 5.1305,
+      "learning_rate": 0.00019945945945945947,
+      "loss": 4.3977,
       "step": 3
     },
     {
       "epoch": 0.04,
-      "learning_rate": 1.9759759759759763e-05,
-      "loss": 4.4537,
+      "learning_rate": 0.0001992792792792793,
+      "loss": 3.9062,
       "step": 4
     },
     {
       "epoch": 0.05,
-      "learning_rate": 1.9699699699699702e-05,
-      "loss": 4.0593,
+      "learning_rate": 0.00019909909909909912,
+      "loss": 3.6201,
       "step": 5
     },
     {
       "epoch": 0.05,
-      "learning_rate": 1.963963963963964e-05,
-      "loss": 3.8647,
+      "learning_rate": 0.00019891891891891895,
+      "loss": 3.2993,
       "step": 6
     },
     {
       "epoch": 0.06,
-      "learning_rate": 1.957957957957958e-05,
-      "loss": 3.5701,
+      "learning_rate": 0.00019873873873873873,
+      "loss": 2.7671,
       "step": 7
     },
     {
       "epoch": 0.07,
-      "learning_rate": 1.951951951951952e-05,
-      "loss": 3.594,
+      "learning_rate": 0.00019855855855855856,
+      "loss": 2.8142,
       "step": 8
     },
     {
       "epoch": 0.08,
-      "learning_rate": 1.9459459459459463e-05,
-      "loss": 3.5468,
+      "learning_rate": 0.00019837837837837838,
+      "loss": 2.6968,
       "step": 9
     },
     {
       "epoch": 0.09,
-      "learning_rate": 1.9399399399399402e-05,
-      "loss": 3.2149,
+      "learning_rate": 0.0001981981981981982,
+      "loss": 2.3751,
       "step": 10
     },
     {
       "epoch": 0.1,
-      "learning_rate": 1.9339339339339342e-05,
-      "loss": 3.2163,
+      "learning_rate": 0.000198018018018018,
+      "loss": 2.2539,
       "step": 11
     },
     {
       "epoch": 0.11,
-      "learning_rate": 1.927927927927928e-05,
-      "loss": 3.0711,
+      "learning_rate": 0.00019783783783783784,
+      "loss": 2.2155,
       "step": 12
     },
     {
       "epoch": 0.12,
-      "learning_rate": 1.921921921921922e-05,
-      "loss": 2.9672,
+      "learning_rate": 0.00019765765765765767,
+      "loss": 2.1593,
       "step": 13
     },
     {
       "epoch": 0.13,
-      "learning_rate": 1.915915915915916e-05,
-      "loss": 2.9493,
+      "learning_rate": 0.0001974774774774775,
+      "loss": 2.1809,
       "step": 14
     },
     {
       "epoch": 0.14,
-      "learning_rate": 1.90990990990991e-05,
-      "loss": 3.1754,
+      "learning_rate": 0.0001972972972972973,
+      "loss": 2.4524,
       "step": 15
     },
     {
       "epoch": 0.14,
-      "learning_rate": 1.903903903903904e-05,
-      "loss": 2.8729,
+      "learning_rate": 0.00019711711711711713,
+      "loss": 2.1628,
       "step": 16
     },
     {
       "epoch": 0.15,
-      "learning_rate": 1.8978978978978982e-05,
-      "loss": 2.8177,
+      "learning_rate": 0.00019693693693693696,
+      "loss": 2.1451,
       "step": 17
     },
     {
       "epoch": 0.16,
-      "learning_rate": 1.891891891891892e-05,
-      "loss": 2.7751,
+      "learning_rate": 0.00019675675675675678,
+      "loss": 2.1149,
       "step": 18
     },
     {
       "epoch": 0.17,
-      "learning_rate": 1.885885885885886e-05,
-      "loss": 2.7206,
+      "learning_rate": 0.00019657657657657659,
+      "loss": 1.8829,
       "step": 19
     },
     {
       "epoch": 0.18,
-      "learning_rate": 1.87987987987988e-05,
-      "loss": 2.6308,
+      "learning_rate": 0.0001963963963963964,
+      "loss": 1.752,
       "step": 20
     },
     {
       "epoch": 0.19,
-      "learning_rate": 1.873873873873874e-05,
-      "loss": 2.6611,
+      "learning_rate": 0.00019621621621621622,
+      "loss": 1.9402,
       "step": 21
     },
     {
       "epoch": 0.2,
-      "learning_rate": 1.8678678678678682e-05,
-      "loss": 2.6,
+      "learning_rate": 0.00019603603603603604,
+      "loss": 1.915,
       "step": 22
     },
     {
       "epoch": 0.21,
-      "learning_rate": 1.861861861861862e-05,
-      "loss": 2.3343,
+      "learning_rate": 0.00019585585585585587,
+      "loss": 1.5993,
       "step": 23
     },
     {
       "epoch": 0.22,
-      "learning_rate": 1.855855855855856e-05,
-      "loss": 2.5656,
+      "learning_rate": 0.00019567567567567567,
+      "loss": 1.8401,
       "step": 24
     },
     {
       "epoch": 0.23,
-      "learning_rate": 1.84984984984985e-05,
-      "loss": 2.6288,
+      "learning_rate": 0.0001954954954954955,
+      "loss": 2.0089,
       "step": 25
     },
     {
       "epoch": 0.23,
-      "learning_rate": 1.843843843843844e-05,
-      "loss": 2.6641,
+      "learning_rate": 0.00019531531531531533,
+      "loss": 1.9848,
       "step": 26
     },
     {
       "epoch": 0.24,
-      "learning_rate": 1.8378378378378383e-05,
-      "loss": 2.2362,
+      "learning_rate": 0.00019513513513513516,
+      "loss": 1.6511,
       "step": 27
     },
     {
       "epoch": 0.25,
-      "learning_rate": 1.831831831831832e-05,
-      "loss": 2.2673,
+      "learning_rate": 0.00019495495495495496,
+      "loss": 1.6407,
       "step": 28
     },
     {
       "epoch": 0.26,
-      "learning_rate": 1.8258258258258258e-05,
-      "loss": 2.5034,
+      "learning_rate": 0.0001947747747747748,
+      "loss": 1.8594,
       "step": 29
     },
     {
       "epoch": 0.27,
-      "learning_rate": 1.81981981981982e-05,
-      "loss": 2.4348,
+      "learning_rate": 0.00019459459459459462,
+      "loss": 1.8043,
       "step": 30
     },
     {
       "epoch": 0.28,
-      "learning_rate": 1.813813813813814e-05,
-      "loss": 2.3081,
+      "learning_rate": 0.00019441441441441442,
+      "loss": 1.6376,
       "step": 31
     },
     {
       "epoch": 0.29,
-      "learning_rate": 1.807807807807808e-05,
-      "loss": 2.2955,
+      "learning_rate": 0.00019423423423423425,
+      "loss": 1.6761,
       "step": 32
     },
     {
       "epoch": 0.3,
-      "learning_rate": 1.801801801801802e-05,
-      "loss": 2.2767,
+      "learning_rate": 0.00019405405405405405,
+      "loss": 1.5476,
       "step": 33
     },
     {
       "epoch": 0.31,
-      "learning_rate": 1.795795795795796e-05,
-      "loss": 2.3092,
+      "learning_rate": 0.00019387387387387388,
+      "loss": 1.6939,
       "step": 34
     },
     {
       "epoch": 0.32,
-      "learning_rate": 1.78978978978979e-05,
-      "loss": 2.3261,
+      "learning_rate": 0.0001936936936936937,
+      "loss": 1.7951,
       "step": 35
     },
     {
       "epoch": 0.32,
-      "learning_rate": 1.783783783783784e-05,
-      "loss": 2.4289,
+      "learning_rate": 0.00019351351351351353,
+      "loss": 1.8343,
       "step": 36
     },
     {
       "epoch": 0.33,
-      "learning_rate": 1.7777777777777777e-05,
-      "loss": 2.4217,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 1.8248,
       "step": 37
     },
     {
       "epoch": 0.34,
-      "learning_rate": 1.771771771771772e-05,
-      "loss": 2.3543,
+      "learning_rate": 0.00019315315315315316,
+      "loss": 1.7202,
       "step": 38
     },
     {
       "epoch": 0.35,
-      "learning_rate": 1.765765765765766e-05,
-      "loss": 2.2967,
+      "learning_rate": 0.000192972972972973,
+      "loss": 1.7938,
       "step": 39
     },
     {
       "epoch": 0.36,
-      "learning_rate": 1.7597597597597598e-05,
-      "loss": 2.3744,
+      "learning_rate": 0.00019279279279279282,
+      "loss": 1.7348,
       "step": 40
     },
     {
       "epoch": 0.37,
-      "learning_rate": 1.7537537537537538e-05,
-      "loss": 2.2404,
+      "learning_rate": 0.00019261261261261262,
+      "loss": 1.6271,
       "step": 41
     },
     {
       "epoch": 0.38,
-      "learning_rate": 1.7477477477477477e-05,
-      "loss": 2.0521,
+      "learning_rate": 0.00019243243243243245,
+      "loss": 1.5305,
       "step": 42
     },
     {
       "epoch": 0.39,
-      "learning_rate": 1.741741741741742e-05,
-      "loss": 2.1582,
+      "learning_rate": 0.00019225225225225225,
+      "loss": 1.6801,
       "step": 43
     },
     {
       "epoch": 0.4,
-      "learning_rate": 1.735735735735736e-05,
-      "loss": 2.2699,
+      "learning_rate": 0.00019207207207207208,
+      "loss": 1.6942,
       "step": 44
     },
     {
       "epoch": 0.41,
-      "learning_rate": 1.72972972972973e-05,
-      "loss": 2.1781,
+      "learning_rate": 0.0001918918918918919,
+      "loss": 1.7218,
       "step": 45
     },
     {
       "epoch": 0.41,
-      "learning_rate": 1.7237237237237238e-05,
-      "loss": 2.2529,
+      "learning_rate": 0.0001917117117117117,
+      "loss": 1.6827,
       "step": 46
     },
     {
       "epoch": 0.42,
-      "learning_rate": 1.7177177177177177e-05,
-      "loss": 2.0092,
+      "learning_rate": 0.00019153153153153154,
+      "loss": 1.4275,
       "step": 47
     },
     {
       "epoch": 0.43,
-      "learning_rate": 1.711711711711712e-05,
-      "loss": 2.2386,
+      "learning_rate": 0.00019135135135135137,
+      "loss": 1.605,
       "step": 48
     },
     {
       "epoch": 0.44,
-      "learning_rate": 1.705705705705706e-05,
-      "loss": 2.1246,
+      "learning_rate": 0.0001911711711711712,
+      "loss": 1.6862,
       "step": 49
     },
     {
       "epoch": 0.45,
-      "learning_rate": 1.6996996996997e-05,
-      "loss": 2.0728,
+      "learning_rate": 0.000190990990990991,
+      "loss": 1.529,
       "step": 50
     },
     {
       "epoch": 0.46,
-      "learning_rate": 1.693693693693694e-05,
-      "loss": 2.1956,
+      "learning_rate": 0.00019081081081081082,
+      "loss": 1.6986,
       "step": 51
     },
     {
       "epoch": 0.47,
-      "learning_rate": 1.6876876876876878e-05,
-      "loss": 2.1301,
+      "learning_rate": 0.00019063063063063065,
+      "loss": 1.6081,
       "step": 52
     },
     {
       "epoch": 0.48,
-      "learning_rate": 1.6816816816816817e-05,
-      "loss": 2.2538,
+      "learning_rate": 0.00019045045045045048,
+      "loss": 1.6748,
       "step": 53
     },
     {
       "epoch": 0.49,
-      "learning_rate": 1.6756756756756757e-05,
-      "loss": 2.0491,
+      "learning_rate": 0.00019027027027027028,
+      "loss": 1.3963,
       "step": 54
     },
     {
       "epoch": 0.5,
-      "learning_rate": 1.6696696696696696e-05,
-      "loss": 2.143,
+      "learning_rate": 0.00019009009009009008,
+      "loss": 1.6614,
       "step": 55
     },
     {
       "epoch": 0.5,
-      "learning_rate": 1.663663663663664e-05,
-      "loss": 2.0769,
+      "learning_rate": 0.0001899099099099099,
+      "loss": 1.4796,
       "step": 56
     },
     {
       "epoch": 0.51,
-      "learning_rate": 1.6576576576576578e-05,
-      "loss": 2.0398,
+      "learning_rate": 0.00018972972972972974,
+      "loss": 1.4965,
       "step": 57
     },
     {
       "epoch": 0.52,
-      "learning_rate": 1.6516516516516518e-05,
-      "loss": 2.1762,
+      "learning_rate": 0.00018954954954954957,
+      "loss": 1.5498,
       "step": 58
     },
     {
       "epoch": 0.53,
-      "learning_rate": 1.6456456456456457e-05,
-      "loss": 2.1344,
+      "learning_rate": 0.00018936936936936937,
+      "loss": 1.5678,
       "step": 59
     },
     {
       "epoch": 0.54,
-      "learning_rate": 1.6396396396396396e-05,
-      "loss": 1.8858,
+      "learning_rate": 0.0001891891891891892,
+      "loss": 1.4824,
       "step": 60
     },
     {
       "epoch": 0.55,
-      "learning_rate": 1.633633633633634e-05,
-      "loss": 2.0582,
+      "learning_rate": 0.00018900900900900903,
+      "loss": 1.4765,
       "step": 61
     },
     {
       "epoch": 0.56,
-      "learning_rate": 1.627627627627628e-05,
-      "loss": 2.0495,
+      "learning_rate": 0.00018882882882882885,
+      "loss": 1.5106,
       "step": 62
     },
     {
       "epoch": 0.57,
-      "learning_rate": 1.6216216216216218e-05,
-      "loss": 2.0432,
+      "learning_rate": 0.00018864864864864866,
+      "loss": 1.5217,
       "step": 63
     },
     {
       "epoch": 0.58,
-      "learning_rate": 1.6156156156156157e-05,
-      "loss": 2.0205,
+      "learning_rate": 0.00018846846846846848,
+      "loss": 1.5239,
       "step": 64
     },
     {
       "epoch": 0.59,
-      "learning_rate": 1.6096096096096097e-05,
-      "loss": 2.0229,
+      "learning_rate": 0.0001882882882882883,
+      "loss": 1.5392,
       "step": 65
     },
     {
       "epoch": 0.59,
-      "learning_rate": 1.6036036036036036e-05,
-      "loss": 1.9601,
+      "learning_rate": 0.00018810810810810814,
+      "loss": 1.4498,
       "step": 66
     },
     {
       "epoch": 0.6,
-      "learning_rate": 1.5975975975975976e-05,
-      "loss": 2.0695,
+      "learning_rate": 0.00018792792792792791,
+      "loss": 1.4997,
       "step": 67
     },
     {
       "epoch": 0.61,
-      "learning_rate": 1.591591591591592e-05,
-      "loss": 1.9346,
+      "learning_rate": 0.00018774774774774774,
+      "loss": 1.4124,
       "step": 68
     },
     {
       "epoch": 0.62,
-      "learning_rate": 1.5855855855855858e-05,
-      "loss": 2.11,
+      "learning_rate": 0.00018756756756756757,
+      "loss": 1.5549,
       "step": 69
     },
     {
       "epoch": 0.63,
-      "learning_rate": 1.5795795795795797e-05,
-      "loss": 2.0759,
+      "learning_rate": 0.0001873873873873874,
+      "loss": 1.431,
       "step": 70
     },
     {
       "epoch": 0.64,
-      "learning_rate": 1.5735735735735737e-05,
-      "loss": 1.9352,
+      "learning_rate": 0.0001872072072072072,
+      "loss": 1.3728,
       "step": 71
     },
     {
       "epoch": 0.65,
-      "learning_rate": 1.5675675675675676e-05,
-      "loss": 2.1723,
+      "learning_rate": 0.00018702702702702703,
+      "loss": 1.5376,
       "step": 72
     },
     {
       "epoch": 0.66,
-      "learning_rate": 1.5615615615615616e-05,
-      "loss": 2.0719,
+      "learning_rate": 0.00018684684684684686,
+      "loss": 1.6378,
       "step": 73
     },
     {
       "epoch": 0.67,
-      "learning_rate": 1.555555555555556e-05,
-      "loss": 1.9658,
+      "learning_rate": 0.0001866666666666667,
+      "loss": 1.4646,
       "step": 74
     },
     {
       "epoch": 0.68,
-      "learning_rate": 1.5495495495495498e-05,
-      "loss": 1.7428,
+      "learning_rate": 0.0001864864864864865,
+      "loss": 1.2362,
       "step": 75
     },
     {
       "epoch": 0.68,
-      "learning_rate": 1.5435435435435437e-05,
-      "loss": 1.8585,
+      "learning_rate": 0.00018630630630630632,
+      "loss": 1.3428,
       "step": 76
     },
     {
       "epoch": 0.69,
-      "learning_rate": 1.5375375375375377e-05,
-      "loss": 1.9221,
+      "learning_rate": 0.00018612612612612614,
+      "loss": 1.4464,
       "step": 77
     },
     {
       "epoch": 0.7,
-      "learning_rate": 1.5315315315315316e-05,
-      "loss": 1.9126,
+      "learning_rate": 0.00018594594594594597,
+      "loss": 1.4192,
       "step": 78
     },
     {
       "epoch": 0.71,
-      "learning_rate": 1.5255255255255257e-05,
-      "loss": 1.8657,
+      "learning_rate": 0.00018576576576576577,
+      "loss": 1.4174,
       "step": 79
     },
     {
       "epoch": 0.72,
-      "learning_rate": 1.5195195195195196e-05,
-      "loss": 1.9183,
+      "learning_rate": 0.00018558558558558558,
+      "loss": 1.5149,
       "step": 80
     },
     {
       "epoch": 0.73,
-      "learning_rate": 1.5135135135135138e-05,
-      "loss": 1.983,
+      "learning_rate": 0.0001854054054054054,
+      "loss": 1.504,
       "step": 81
     },
     {
       "epoch": 0.74,
-      "learning_rate": 1.5075075075075077e-05,
-      "loss": 1.9661,
+      "learning_rate": 0.00018522522522522523,
+      "loss": 1.4452,
       "step": 82
     },
     {
       "epoch": 0.75,
-      "learning_rate": 1.5015015015015015e-05,
-      "loss": 1.9937,
+      "learning_rate": 0.00018504504504504506,
+      "loss": 1.5763,
       "step": 83
     },
     {
       "epoch": 0.76,
-      "learning_rate": 1.4954954954954957e-05,
-      "loss": 1.8473,
+      "learning_rate": 0.00018486486486486486,
+      "loss": 1.4288,
       "step": 84
     },
     {
       "epoch": 0.77,
-      "learning_rate": 1.4894894894894895e-05,
-      "loss": 1.9417,
+      "learning_rate": 0.0001846846846846847,
+      "loss": 1.5056,
       "step": 85
     },
     {
       "epoch": 0.77,
-      "learning_rate": 1.4834834834834836e-05,
-      "loss": 1.8969,
+      "learning_rate": 0.00018450450450450452,
+      "loss": 1.3584,
       "step": 86
     },
     {
       "epoch": 0.78,
-      "learning_rate": 1.4774774774774776e-05,
-      "loss": 1.8054,
+      "learning_rate": 0.00018432432432432435,
+      "loss": 1.2679,
       "step": 87
     },
     {
       "epoch": 0.79,
-      "learning_rate": 1.4714714714714715e-05,
-      "loss": 1.7336,
+      "learning_rate": 0.00018414414414414415,
+      "loss": 1.2421,
       "step": 88
     },
     {
       "epoch": 0.8,
-      "learning_rate": 1.4654654654654656e-05,
-      "loss": 1.8264,
+      "learning_rate": 0.00018396396396396398,
+      "loss": 1.2359,
       "step": 89
     },
     {
       "epoch": 0.81,
-      "learning_rate": 1.4594594594594596e-05,
-      "loss": 1.8013,
+      "learning_rate": 0.0001837837837837838,
+      "loss": 1.2961,
       "step": 90
     },
     {
       "epoch": 0.82,
-      "learning_rate": 1.4534534534534537e-05,
-      "loss": 1.9437,
+      "learning_rate": 0.0001836036036036036,
+      "loss": 1.5433,
       "step": 91
     },
     {
       "epoch": 0.83,
-      "learning_rate": 1.4474474474474476e-05,
-      "loss": 1.893,
+      "learning_rate": 0.00018342342342342343,
+      "loss": 1.3349,
       "step": 92
     },
     {
       "epoch": 0.84,
-      "learning_rate": 1.4414414414414416e-05,
-      "loss": 1.83,
+      "learning_rate": 0.00018324324324324324,
+      "loss": 1.3233,
       "step": 93
     },
     {
       "epoch": 0.85,
-      "learning_rate": 1.4354354354354357e-05,
-      "loss": 1.966,
+      "learning_rate": 0.00018306306306306306,
+      "loss": 1.4935,
       "step": 94
     },
     {
       "epoch": 0.86,
-      "learning_rate": 1.4294294294294296e-05,
-      "loss": 1.9087,
+      "learning_rate": 0.0001828828828828829,
+      "loss": 1.4808,
       "step": 95
     },
     {
       "epoch": 0.86,
-      "learning_rate": 1.4234234234234234e-05,
-      "loss": 1.8374,
+      "learning_rate": 0.00018270270270270272,
+      "loss": 1.3105,
       "step": 96
     },
     {
       "epoch": 0.87,
-      "learning_rate": 1.4174174174174175e-05,
-      "loss": 2.1122,
+      "learning_rate": 0.00018252252252252252,
+      "loss": 1.6545,
       "step": 97
     },
     {
       "epoch": 0.88,
-      "learning_rate": 1.4114114114114114e-05,
-      "loss": 2.0771,
+      "learning_rate": 0.00018234234234234235,
+      "loss": 1.5198,
       "step": 98
     },
     {
       "epoch": 0.89,
-      "learning_rate": 1.4054054054054055e-05,
-      "loss": 1.8904,
+      "learning_rate": 0.00018216216216216218,
+      "loss": 1.2939,
       "step": 99
     },
     {
       "epoch": 0.9,
-      "learning_rate": 1.3993993993993995e-05,
-      "loss": 1.9853,
+      "learning_rate": 0.000181981981981982,
+      "loss": 1.4136,
       "step": 100
     },
     {
       "epoch": 0.91,
-      "learning_rate": 1.3933933933933934e-05,
-      "loss": 1.8186,
+      "learning_rate": 0.0001818018018018018,
+      "loss": 1.2831,
       "step": 101
     },
     {
       "epoch": 0.92,
-      "learning_rate": 1.3873873873873875e-05,
-      "loss": 1.8105,
+      "learning_rate": 0.00018162162162162164,
+      "loss": 1.2908,
       "step": 102
     },
     {
       "epoch": 0.93,
-      "learning_rate": 1.3813813813813815e-05,
-      "loss": 1.6838,
+      "learning_rate": 0.00018144144144144144,
+      "loss": 1.2553,
       "step": 103
     },
     {
       "epoch": 0.94,
-      "learning_rate": 1.3753753753753756e-05,
-      "loss": 1.7476,
+      "learning_rate": 0.00018126126126126127,
+      "loss": 1.2482,
       "step": 104
     },
     {
       "epoch": 0.95,
-      "learning_rate": 1.3693693693693695e-05,
-      "loss": 1.8371,
+      "learning_rate": 0.0001810810810810811,
+      "loss": 1.3024,
       "step": 105
     },
     {
       "epoch": 0.95,
-      "learning_rate": 1.3633633633633635e-05,
-      "loss": 1.8751,
+      "learning_rate": 0.0001809009009009009,
+      "loss": 1.3831,
       "step": 106
     },
     {
       "epoch": 0.96,
-      "learning_rate": 1.3573573573573576e-05,
-      "loss": 1.7548,
+      "learning_rate": 0.00018072072072072072,
+      "loss": 1.2204,
       "step": 107
     },
     {
       "epoch": 0.97,
-      "learning_rate": 1.3513513513513515e-05,
-      "loss": 1.6747,
+      "learning_rate": 0.00018054054054054055,
+      "loss": 1.193,
       "step": 108
     },
     {
       "epoch": 0.98,
-      "learning_rate": 1.3453453453453456e-05,
-      "loss": 1.7632,
+      "learning_rate": 0.00018036036036036038,
+      "loss": 1.2178,
       "step": 109
     },
     {
       "epoch": 0.99,
-      "learning_rate": 1.3393393393393394e-05,
-      "loss": 1.8086,
+      "learning_rate": 0.00018018018018018018,
+      "loss": 1.3477,
       "step": 110
     },
     {
       "epoch": 1.0,
-      "learning_rate": 1.3333333333333333e-05,
-      "loss": 2.2582,
+      "learning_rate": 0.00018,
+      "loss": 1.848,
       "step": 111
     },
     {
       "epoch": 1.01,
-      "learning_rate": 1.3273273273273274e-05,
-      "loss": 1.6199,
+      "learning_rate": 0.00017981981981981984,
+      "loss": 1.0224,
       "step": 112
     },
     {
       "epoch": 1.02,
-      "learning_rate": 1.3213213213213214e-05,
-      "loss": 1.6633,
+      "learning_rate": 0.00017963963963963967,
+      "loss": 1.0316,
       "step": 113
     },
     {
       "epoch": 1.03,
-      "learning_rate": 1.3153153153153155e-05,
-      "loss": 1.6855,
+      "learning_rate": 0.00017945945945945947,
+      "loss": 0.9298,
       "step": 114
     },
     {
       "epoch": 1.04,
-      "learning_rate": 1.3093093093093094e-05,
-      "loss": 1.7482,
+      "learning_rate": 0.00017927927927927927,
+      "loss": 0.9607,
       "step": 115
     },
     {
       "epoch": 1.05,
-      "learning_rate": 1.3033033033033034e-05,
-      "loss": 1.6648,
+      "learning_rate": 0.0001790990990990991,
+      "loss": 0.9851,
       "step": 116
     },
     {
       "epoch": 1.05,
-      "learning_rate": 1.2972972972972975e-05,
-      "loss": 1.7411,
+      "learning_rate": 0.00017891891891891893,
+      "loss": 1.1184,
       "step": 117
     },
     {
       "epoch": 1.06,
-      "learning_rate": 1.2912912912912914e-05,
-      "loss": 1.7804,
+      "learning_rate": 0.00017873873873873876,
+      "loss": 1.1724,
       "step": 118
     },
     {
       "epoch": 1.07,
-      "learning_rate": 1.2852852852852854e-05,
-      "loss": 1.6331,
+      "learning_rate": 0.00017855855855855856,
+      "loss": 0.9032,
       "step": 119
     },
     {
       "epoch": 1.08,
-      "learning_rate": 1.2792792792792795e-05,
-      "loss": 1.7026,
+      "learning_rate": 0.00017837837837837839,
+      "loss": 1.0655,
       "step": 120
     },
     {
       "epoch": 1.09,
-      "learning_rate": 1.2732732732732732e-05,
-      "loss": 1.5627,
+      "learning_rate": 0.00017819819819819821,
+      "loss": 0.9477,
       "step": 121
     },
     {
       "epoch": 1.1,
-      "learning_rate": 1.2672672672672675e-05,
-      "loss": 1.7268,
+      "learning_rate": 0.00017801801801801804,
+      "loss": 1.0931,
       "step": 122
     },
     {
       "epoch": 1.11,
-      "learning_rate": 1.2612612612612613e-05,
-      "loss": 1.7191,
+      "learning_rate": 0.00017783783783783784,
+      "loss": 1.1222,
       "step": 123
     },
     {
       "epoch": 1.12,
-      "learning_rate": 1.2552552552552552e-05,
-      "loss": 1.7727,
+      "learning_rate": 0.00017765765765765767,
+      "loss": 1.0831,
       "step": 124
     },
     {
       "epoch": 1.13,
-      "learning_rate": 1.2492492492492493e-05,
-      "loss": 1.6784,
+      "learning_rate": 0.0001774774774774775,
+      "loss": 0.996,
       "step": 125
     },
     {
       "epoch": 1.14,
-      "learning_rate": 1.2432432432432433e-05,
-      "loss": 1.7173,
+      "learning_rate": 0.00017729729729729733,
+      "loss": 1.1083,
       "step": 126
     },
     {
       "epoch": 1.14,
-      "learning_rate": 1.2372372372372374e-05,
-      "loss": 1.7755,
+      "learning_rate": 0.0001771171171171171,
+      "loss": 1.0783,
       "step": 127
     },
     {
       "epoch": 1.15,
-      "learning_rate": 1.2312312312312313e-05,
-      "loss": 1.5832,
+      "learning_rate": 0.00017693693693693693,
+      "loss": 1.0126,
       "step": 128
     },
     {
       "epoch": 1.16,
-      "learning_rate": 1.2252252252252253e-05,
-      "loss": 1.6627,
+      "learning_rate": 0.00017675675675675676,
+      "loss": 1.0566,
       "step": 129
     },
     {
       "epoch": 1.17,
-      "learning_rate": 1.2192192192192194e-05,
-      "loss": 1.7832,
+      "learning_rate": 0.0001765765765765766,
+      "loss": 1.1383,
       "step": 130
     },
     {
       "epoch": 1.18,
-      "learning_rate": 1.2132132132132133e-05,
-      "loss": 1.6711,
+      "learning_rate": 0.0001763963963963964,
+      "loss": 1.0675,
       "step": 131
     },
     {
       "epoch": 1.19,
-      "learning_rate": 1.2072072072072074e-05,
-      "loss": 1.6217,
+      "learning_rate": 0.00017621621621621622,
+      "loss": 0.9634,
       "step": 132
     },
     {
       "epoch": 1.2,
-      "learning_rate": 1.2012012012012014e-05,
-      "loss": 1.6746,
+      "learning_rate": 0.00017603603603603605,
+      "loss": 0.9781,
       "step": 133
     },
     {
       "epoch": 1.21,
-      "learning_rate": 1.1951951951951951e-05,
-      "loss": 1.602,
+      "learning_rate": 0.00017585585585585587,
+      "loss": 1.0291,
       "step": 134
     },
     {
       "epoch": 1.22,
-      "learning_rate": 1.1891891891891894e-05,
-      "loss": 1.7648,
+      "learning_rate": 0.00017567567567567568,
+      "loss": 1.0779,
       "step": 135
     },
     {
       "epoch": 1.23,
-      "learning_rate": 1.1831831831831832e-05,
-      "loss": 1.5384,
+      "learning_rate": 0.0001754954954954955,
+      "loss": 0.946,
       "step": 136
     },
     {
       "epoch": 1.23,
-      "learning_rate": 1.1771771771771771e-05,
-      "loss": 1.6899,
+      "learning_rate": 0.00017531531531531533,
+      "loss": 1.1033,
       "step": 137
     },
     {
       "epoch": 1.24,
-      "learning_rate": 1.1711711711711713e-05,
-      "loss": 1.7466,
+      "learning_rate": 0.00017513513513513516,
+      "loss": 1.1328,
       "step": 138
     },
     {
       "epoch": 1.25,
-      "learning_rate": 1.1651651651651652e-05,
-      "loss": 1.5775,
+      "learning_rate": 0.00017495495495495496,
+      "loss": 1.014,
       "step": 139
     },
     {
       "epoch": 1.26,
-      "learning_rate": 1.1591591591591593e-05,
-      "loss": 1.5667,
+      "learning_rate": 0.00017477477477477476,
+      "loss": 1.0098,
       "step": 140
     },
     {
       "epoch": 1.27,
-      "learning_rate": 1.1531531531531532e-05,
-      "loss": 1.6807,
+      "learning_rate": 0.0001745945945945946,
+      "loss": 1.0973,
       "step": 141
     },
     {
       "epoch": 1.28,
-      "learning_rate": 1.1471471471471472e-05,
-      "loss": 1.7766,
+      "learning_rate": 0.00017441441441441442,
+      "loss": 1.0991,
       "step": 142
     },
     {
       "epoch": 1.29,
-      "learning_rate": 1.1411411411411413e-05,
-      "loss": 1.7,
+      "learning_rate": 0.00017423423423423425,
+      "loss": 0.9668,
       "step": 143
     },
     {
       "epoch": 1.3,
-      "learning_rate": 1.1351351351351352e-05,
-      "loss": 1.5558,
+      "learning_rate": 0.00017405405405405405,
+      "loss": 0.8774,
       "step": 144
     },
     {
       "epoch": 1.31,
-      "learning_rate": 1.1291291291291293e-05,
-      "loss": 1.7148,
+      "learning_rate": 0.00017387387387387388,
+      "loss": 1.0744,
       "step": 145
     },
     {
       "epoch": 1.32,
-      "learning_rate": 1.1231231231231233e-05,
-      "loss": 1.5432,
+      "learning_rate": 0.0001736936936936937,
+      "loss": 0.8432,
       "step": 146
     },
     {
       "epoch": 1.32,
-      "learning_rate": 1.117117117117117e-05,
-      "loss": 1.5042,
+      "learning_rate": 0.00017351351351351353,
+      "loss": 0.9951,
       "step": 147
     },
     {
       "epoch": 1.33,
-      "learning_rate": 1.1111111111111113e-05,
-      "loss": 1.8052,
+      "learning_rate": 0.00017333333333333334,
+      "loss": 1.0802,
       "step": 148
     },
     {
       "epoch": 1.34,
-      "learning_rate": 1.1051051051051051e-05,
-      "loss": 1.8016,
+      "learning_rate": 0.00017315315315315316,
+      "loss": 1.1801,
       "step": 149
     },
     {
       "epoch": 1.35,
-      "learning_rate": 1.0990990990990992e-05,
-      "loss": 1.4395,
+      "learning_rate": 0.000172972972972973,
+      "loss": 0.9059,
       "step": 150
     },
     {
       "epoch": 1.36,
-      "learning_rate": 1.0930930930930932e-05,
-      "loss": 1.6007,
+      "learning_rate": 0.0001727927927927928,
+      "loss": 0.9663,
       "step": 151
     },
     {
       "epoch": 1.37,
-      "learning_rate": 1.0870870870870871e-05,
-      "loss": 1.6131,
+      "learning_rate": 0.00017261261261261262,
+      "loss": 1.1322,
       "step": 152
     },
     {
       "epoch": 1.38,
-      "learning_rate": 1.0810810810810812e-05,
-      "loss": 1.612,
+      "learning_rate": 0.00017243243243243242,
+      "loss": 1.082,
       "step": 153
     },
     {
       "epoch": 1.39,
-      "learning_rate": 1.0750750750750751e-05,
-      "loss": 1.5766,
+      "learning_rate": 0.00017225225225225225,
+      "loss": 0.9189,
       "step": 154
     },
     {
       "epoch": 1.4,
-      "learning_rate": 1.0690690690690693e-05,
-      "loss": 1.577,
+      "learning_rate": 0.00017207207207207208,
+      "loss": 1.0324,
       "step": 155
     },
     {
       "epoch": 1.41,
-      "learning_rate": 1.0630630630630632e-05,
-      "loss": 1.6121,
+      "learning_rate": 0.0001718918918918919,
+      "loss": 0.9904,
       "step": 156
     },
     {
       "epoch": 1.41,
-      "learning_rate": 1.0570570570570571e-05,
-      "loss": 1.7291,
+      "learning_rate": 0.0001717117117117117,
+      "loss": 1.1135,
       "step": 157
     },
     {
       "epoch": 1.42,
-      "learning_rate": 1.0510510510510512e-05,
-      "loss": 1.5699,
+      "learning_rate": 0.00017153153153153154,
+      "loss": 0.9649,
       "step": 158
     },
     {
       "epoch": 1.43,
-      "learning_rate": 1.0450450450450452e-05,
-      "loss": 1.564,
+      "learning_rate": 0.00017135135135135137,
+      "loss": 1.0241,
       "step": 159
     },
     {
       "epoch": 1.44,
-      "learning_rate": 1.039039039039039e-05,
-      "loss": 1.5114,
+      "learning_rate": 0.0001711711711711712,
+      "loss": 0.8952,
       "step": 160
     },
     {
       "epoch": 1.45,
-      "learning_rate": 1.033033033033033e-05,
-      "loss": 1.5204,
+      "learning_rate": 0.000170990990990991,
+      "loss": 0.9294,
       "step": 161
     },
     {
       "epoch": 1.46,
-      "learning_rate": 1.027027027027027e-05,
-      "loss": 1.5483,
+      "learning_rate": 0.00017081081081081083,
+      "loss": 1.0734,
       "step": 162
     },
     {
       "epoch": 1.47,
-      "learning_rate": 1.0210210210210211e-05,
-      "loss": 1.7533,
+      "learning_rate": 0.00017063063063063063,
+      "loss": 1.075,
       "step": 163
     },
     {
       "epoch": 1.48,
-      "learning_rate": 1.015015015015015e-05,
-      "loss": 1.7088,
+      "learning_rate": 0.00017045045045045045,
+      "loss": 0.9994,
       "step": 164
     },
     {
       "epoch": 1.49,
-      "learning_rate": 1.009009009009009e-05,
-      "loss": 1.3951,
+      "learning_rate": 0.00017027027027027028,
+      "loss": 0.8595,
       "step": 165
     },
     {
       "epoch": 1.5,
-      "learning_rate": 1.0030030030030031e-05,
-      "loss": 1.533,
+      "learning_rate": 0.00017009009009009008,
+      "loss": 0.9343,
       "step": 166
     },
     {
       "epoch": 1.5,
-      "learning_rate": 9.96996996996997e-06,
-      "loss": 1.6597,
+      "learning_rate": 0.0001699099099099099,
+      "loss": 1.0496,
       "step": 167
     },
     {
       "epoch": 1.51,
-      "learning_rate": 9.90990990990991e-06,
-      "loss": 1.6057,
+      "learning_rate": 0.00016972972972972974,
+      "loss": 0.9934,
       "step": 168
     },
     {
       "epoch": 1.52,
-      "learning_rate": 9.849849849849851e-06,
-      "loss": 1.6473,
+      "learning_rate": 0.00016954954954954957,
+      "loss": 0.9766,
       "step": 169
     },
     {
       "epoch": 1.53,
-      "learning_rate": 9.78978978978979e-06,
-      "loss": 1.3704,
+      "learning_rate": 0.00016936936936936937,
+      "loss": 0.8376,
       "step": 170
     },
     {
       "epoch": 1.54,
-      "learning_rate": 9.729729729729732e-06,
-      "loss": 1.5277,
+      "learning_rate": 0.0001691891891891892,
+      "loss": 0.9374,
       "step": 171
     },
     {
       "epoch": 1.55,
-      "learning_rate": 9.669669669669671e-06,
-      "loss": 1.4736,
+      "learning_rate": 0.00016900900900900903,
+      "loss": 1.0476,
       "step": 172
     },
     {
       "epoch": 1.56,
-      "learning_rate": 9.60960960960961e-06,
-      "loss": 1.5006,
+      "learning_rate": 0.00016882882882882886,
+      "loss": 0.9219,
       "step": 173
     },
     {
       "epoch": 1.57,
-      "learning_rate": 9.54954954954955e-06,
-      "loss": 1.4437,
+      "learning_rate": 0.00016864864864864866,
+      "loss": 0.9586,
       "step": 174
     },
     {
       "epoch": 1.58,
-      "learning_rate": 9.489489489489491e-06,
-      "loss": 1.553,
+      "learning_rate": 0.00016846846846846846,
+      "loss": 1.0288,
       "step": 175
     },
     {
       "epoch": 1.59,
-      "learning_rate": 9.42942942942943e-06,
-      "loss": 1.5653,
+      "learning_rate": 0.0001682882882882883,
+      "loss": 0.9829,
       "step": 176
     },
     {
       "epoch": 1.59,
-      "learning_rate": 9.36936936936937e-06,
-      "loss": 1.4045,
+      "learning_rate": 0.00016810810810810812,
+      "loss": 0.8158,
       "step": 177
     },
     {
       "epoch": 1.6,
-      "learning_rate": 9.30930930930931e-06,
-      "loss": 1.6598,
+      "learning_rate": 0.00016792792792792794,
+      "loss": 1.0338,
       "step": 178
     },
     {
       "epoch": 1.61,
-      "learning_rate": 9.24924924924925e-06,
-      "loss": 1.4961,
+      "learning_rate": 0.00016774774774774775,
+      "loss": 1.0119,
       "step": 179
     },
     {
       "epoch": 1.62,
-      "learning_rate": 9.189189189189191e-06,
-      "loss": 1.4487,
+      "learning_rate": 0.00016756756756756757,
+      "loss": 0.7975,
       "step": 180
     },
     {
       "epoch": 1.63,
-      "learning_rate": 9.129129129129129e-06,
-      "loss": 1.4329,
+      "learning_rate": 0.0001673873873873874,
+      "loss": 0.8601,
       "step": 181
     },
     {
       "epoch": 1.64,
-      "learning_rate": 9.06906906906907e-06,
-      "loss": 1.5354,
+      "learning_rate": 0.00016720720720720723,
+      "loss": 1.0333,
       "step": 182
     },
     {
       "epoch": 1.65,
-      "learning_rate": 9.00900900900901e-06,
-      "loss": 1.7155,
+      "learning_rate": 0.00016702702702702703,
+      "loss": 1.0862,
       "step": 183
     },
     {
       "epoch": 1.66,
-      "learning_rate": 8.94894894894895e-06,
-      "loss": 1.6296,
+      "learning_rate": 0.00016684684684684686,
+      "loss": 0.8784,
       "step": 184
     },
     {
       "epoch": 1.67,
-      "learning_rate": 8.888888888888888e-06,
-      "loss": 1.5776,
+      "learning_rate": 0.0001666666666666667,
+      "loss": 1.0995,
       "step": 185
     },
     {
       "epoch": 1.68,
-      "learning_rate": 8.82882882882883e-06,
-      "loss": 1.5153,
+      "learning_rate": 0.00016648648648648652,
+      "loss": 1.0452,
       "step": 186
     },
     {
       "epoch": 1.68,
-      "learning_rate": 8.768768768768769e-06,
-      "loss": 1.353,
+      "learning_rate": 0.00016630630630630632,
+      "loss": 0.8971,
       "step": 187
     },
     {
       "epoch": 1.69,
-      "learning_rate": 8.70870870870871e-06,
-      "loss": 1.5484,
+      "learning_rate": 0.00016612612612612612,
+      "loss": 0.9339,
       "step": 188
     },
     {
       "epoch": 1.7,
-      "learning_rate": 8.64864864864865e-06,
-      "loss": 1.3934,
+      "learning_rate": 0.00016594594594594595,
+      "loss": 0.8219,
       "step": 189
     },
     {
       "epoch": 1.71,
-      "learning_rate": 8.588588588588589e-06,
-      "loss": 1.5505,
+      "learning_rate": 0.00016576576576576578,
+      "loss": 0.9829,
       "step": 190
     },
     {
       "epoch": 1.72,
-      "learning_rate": 8.52852852852853e-06,
-      "loss": 1.3282,
+      "learning_rate": 0.00016558558558558558,
+      "loss": 0.7705,
       "step": 191
     },
     {
       "epoch": 1.73,
-      "learning_rate": 8.46846846846847e-06,
-      "loss": 1.4492,
+      "learning_rate": 0.0001654054054054054,
+      "loss": 0.8442,
       "step": 192
     },
     {
       "epoch": 1.74,
-      "learning_rate": 8.408408408408409e-06,
-      "loss": 1.4575,
+      "learning_rate": 0.00016522522522522523,
+      "loss": 0.8804,
       "step": 193
     },
     {
       "epoch": 1.75,
-      "learning_rate": 8.348348348348348e-06,
-      "loss": 1.4301,
+      "learning_rate": 0.00016504504504504506,
+      "loss": 0.7614,
       "step": 194
     },
     {
       "epoch": 1.76,
-      "learning_rate": 8.288288288288289e-06,
-      "loss": 1.4052,
+      "learning_rate": 0.00016486486486486486,
+      "loss": 0.8259,
       "step": 195
     },
     {
       "epoch": 1.77,
-      "learning_rate": 8.228228228228229e-06,
-      "loss": 1.7595,
+      "learning_rate": 0.0001646846846846847,
+      "loss": 1.1247,
       "step": 196
     },
     {
       "epoch": 1.77,
-      "learning_rate": 8.16816816816817e-06,
-      "loss": 1.5578,
+      "learning_rate": 0.00016450450450450452,
+      "loss": 0.8855,
       "step": 197
     },
     {
       "epoch": 1.78,
-      "learning_rate": 8.108108108108109e-06,
-      "loss": 1.4915,
+      "learning_rate": 0.00016432432432432435,
+      "loss": 0.9202,
       "step": 198
     },
     {
       "epoch": 1.79,
-      "learning_rate": 8.048048048048048e-06,
-      "loss": 1.3835,
+      "learning_rate": 0.00016414414414414415,
+      "loss": 0.9203,
       "step": 199
     },
     {
       "epoch": 1.8,
-      "learning_rate": 7.987987987987988e-06,
-      "loss": 1.3505,
+      "learning_rate": 0.00016396396396396395,
+      "loss": 0.8537,
       "step": 200
     },
     {
       "epoch": 1.81,
-      "learning_rate": 7.927927927927929e-06,
-      "loss": 1.5297,
+      "learning_rate": 0.00016378378378378378,
+      "loss": 0.8746,
       "step": 201
     },
     {
       "epoch": 1.82,
-      "learning_rate": 7.867867867867868e-06,
-      "loss": 1.6206,
+      "learning_rate": 0.0001636036036036036,
+      "loss": 1.0242,
       "step": 202
     },
     {
       "epoch": 1.83,
-      "learning_rate": 7.807807807807808e-06,
-      "loss": 1.4387,
+      "learning_rate": 0.00016342342342342344,
+      "loss": 0.8806,
       "step": 203
     },
     {
       "epoch": 1.84,
-      "learning_rate": 7.747747747747749e-06,
-      "loss": 1.5027,
+      "learning_rate": 0.00016324324324324324,
+      "loss": 0.8827,
       "step": 204
     },
     {
       "epoch": 1.85,
-      "learning_rate": 7.687687687687688e-06,
-      "loss": 1.468,
+      "learning_rate": 0.00016306306306306307,
+      "loss": 0.9695,
       "step": 205
     },
     {
       "epoch": 1.86,
-      "learning_rate": 7.6276276276276285e-06,
-      "loss": 1.4163,
+      "learning_rate": 0.0001628828828828829,
+      "loss": 0.9495,
       "step": 206
     },
     {
       "epoch": 1.86,
-      "learning_rate": 7.567567567567569e-06,
-      "loss": 1.225,
+      "learning_rate": 0.00016270270270270272,
+      "loss": 0.7122,
       "step": 207
     },
     {
       "epoch": 1.87,
-      "learning_rate": 7.507507507507507e-06,
-      "loss": 1.5404,
+      "learning_rate": 0.00016252252252252252,
+      "loss": 1.0111,
       "step": 208
     },
     {
       "epoch": 1.88,
-      "learning_rate": 7.447447447447448e-06,
-      "loss": 1.7203,
+      "learning_rate": 0.00016234234234234235,
+      "loss": 1.1255,
       "step": 209
     },
     {
       "epoch": 1.89,
-      "learning_rate": 7.387387387387388e-06,
-      "loss": 1.5402,
+      "learning_rate": 0.00016216216216216218,
+      "loss": 0.8789,
       "step": 210
     },
     {
       "epoch": 1.9,
-      "learning_rate": 7.327327327327328e-06,
-      "loss": 1.4962,
+      "learning_rate": 0.000161981981981982,
+      "loss": 0.9309,
       "step": 211
     },
     {
       "epoch": 1.91,
-      "learning_rate": 7.267267267267268e-06,
-      "loss": 1.4352,
+      "learning_rate": 0.0001618018018018018,
+      "loss": 0.9133,
       "step": 212
     },
     {
       "epoch": 1.92,
-      "learning_rate": 7.207207207207208e-06,
-      "loss": 1.4584,
+      "learning_rate": 0.0001616216216216216,
+      "loss": 0.9779,
       "step": 213
     },
     {
       "epoch": 1.93,
-      "learning_rate": 7.147147147147148e-06,
-      "loss": 1.3992,
+      "learning_rate": 0.00016144144144144144,
+      "loss": 0.8261,
       "step": 214
     },
     {
       "epoch": 1.94,
-      "learning_rate": 7.087087087087087e-06,
-      "loss": 1.4541,
+      "learning_rate": 0.00016126126126126127,
+      "loss": 0.8621,
       "step": 215
     },
     {
       "epoch": 1.95,
-      "learning_rate": 7.027027027027028e-06,
-      "loss": 1.3759,
+      "learning_rate": 0.0001610810810810811,
+      "loss": 0.7617,
       "step": 216
     },
     {
       "epoch": 1.95,
-      "learning_rate": 6.966966966966967e-06,
-      "loss": 1.4622,
+      "learning_rate": 0.0001609009009009009,
+      "loss": 0.9391,
       "step": 217
     },
     {
       "epoch": 1.96,
-      "learning_rate": 6.906906906906907e-06,
-      "loss": 1.4799,
+      "learning_rate": 0.00016072072072072073,
+      "loss": 0.8437,
       "step": 218
     },
     {
       "epoch": 1.97,
-      "learning_rate": 6.846846846846848e-06,
-      "loss": 1.4265,
+      "learning_rate": 0.00016054054054054056,
+      "loss": 0.9364,
       "step": 219
     },
     {
       "epoch": 1.98,
-      "learning_rate": 6.786786786786788e-06,
-      "loss": 1.3938,
+      "learning_rate": 0.00016036036036036038,
+      "loss": 0.8519,
       "step": 220
     },
     {
       "epoch": 1.99,
-      "learning_rate": 6.726726726726728e-06,
-      "loss": 1.5023,
+      "learning_rate": 0.00016018018018018018,
+      "loss": 0.8703,
       "step": 221
     },
     {
       "epoch": 2.0,
-      "learning_rate": 6.666666666666667e-06,
-      "loss": 1.2256,
+      "learning_rate": 0.00016,
+      "loss": 0.5854,
       "step": 222
     },
     {
       "epoch": 2.01,
-      "learning_rate": 6.606606606606607e-06,
-      "loss": 1.3189,
+      "learning_rate": 0.00015981981981981984,
+      "loss": 0.6418,
       "step": 223
     },
     {
       "epoch": 2.02,
-      "learning_rate": 6.546546546546547e-06,
-      "loss": 1.4738,
+      "learning_rate": 0.00015963963963963964,
+      "loss": 0.6642,
       "step": 224
     },
     {
       "epoch": 2.03,
-      "learning_rate": 6.486486486486487e-06,
-      "loss": 1.4976,
+      "learning_rate": 0.00015945945945945947,
+      "loss": 0.7597,
       "step": 225
     },
     {
       "epoch": 2.04,
-      "learning_rate": 6.426426426426427e-06,
-      "loss": 1.4145,
+      "learning_rate": 0.00015927927927927927,
+      "loss": 0.7035,
       "step": 226
     },
     {
       "epoch": 2.05,
-      "learning_rate": 6.366366366366366e-06,
-      "loss": 1.3549,
+      "learning_rate": 0.0001590990990990991,
+      "loss": 0.6098,
       "step": 227
     },
     {
       "epoch": 2.05,
-      "learning_rate": 6.3063063063063065e-06,
-      "loss": 1.4151,
+      "learning_rate": 0.00015891891891891893,
+      "loss": 0.6118,
       "step": 228
     },
     {
       "epoch": 2.06,
-      "learning_rate": 6.246246246246247e-06,
-      "loss": 1.6266,
+      "learning_rate": 0.00015873873873873876,
+      "loss": 0.6998,
       "step": 229
     },
     {
       "epoch": 2.07,
-      "learning_rate": 6.186186186186187e-06,
-      "loss": 1.4899,
+      "learning_rate": 0.00015855855855855856,
+      "loss": 0.6526,
       "step": 230
     },
     {
       "epoch": 2.08,
-      "learning_rate": 6.126126126126126e-06,
-      "loss": 1.2746,
+      "learning_rate": 0.0001583783783783784,
+      "loss": 0.6019,
       "step": 231
     },
     {
       "epoch": 2.09,
-      "learning_rate": 6.066066066066067e-06,
-      "loss": 1.3838,
+      "learning_rate": 0.00015819819819819822,
+      "loss": 0.5906,
       "step": 232
     },
     {
       "epoch": 2.1,
-      "learning_rate": 6.006006006006007e-06,
-      "loss": 1.3595,
+      "learning_rate": 0.00015801801801801804,
+      "loss": 0.6253,
       "step": 233
     },
     {
       "epoch": 2.11,
-      "learning_rate": 5.945945945945947e-06,
-      "loss": 1.3993,
+      "learning_rate": 0.00015783783783783785,
+      "loss": 0.6721,
       "step": 234
     },
     {
       "epoch": 2.12,
-      "learning_rate": 5.885885885885886e-06,
-      "loss": 1.2979,
+      "learning_rate": 0.00015765765765765767,
+      "loss": 0.5985,
       "step": 235
     },
     {
       "epoch": 2.13,
-      "learning_rate": 5.825825825825826e-06,
-      "loss": 1.424,
+      "learning_rate": 0.00015747747747747747,
+      "loss": 0.6828,
       "step": 236
     },
     {
       "epoch": 2.14,
-      "learning_rate": 5.765765765765766e-06,
-      "loss": 1.3078,
+      "learning_rate": 0.0001572972972972973,
+      "loss": 0.5669,
       "step": 237
     },
     {
       "epoch": 2.14,
-      "learning_rate": 5.7057057057057065e-06,
-      "loss": 1.4321,
+      "learning_rate": 0.00015711711711711713,
+      "loss": 0.7135,
       "step": 238
     },
     {
       "epoch": 2.15,
-      "learning_rate": 5.645645645645647e-06,
-      "loss": 1.3732,
+      "learning_rate": 0.00015693693693693693,
+      "loss": 0.5734,
       "step": 239
     },
     {
       "epoch": 2.16,
-      "learning_rate": 5.585585585585585e-06,
-      "loss": 1.5107,
+      "learning_rate": 0.00015675675675675676,
+      "loss": 0.7206,
       "step": 240
     },
     {
       "epoch": 2.17,
-      "learning_rate": 5.5255255255255255e-06,
-      "loss": 1.5524,
+      "learning_rate": 0.0001565765765765766,
+      "loss": 0.7708,
       "step": 241
     },
     {
       "epoch": 2.18,
-      "learning_rate": 5.465465465465466e-06,
-      "loss": 1.4516,
+      "learning_rate": 0.00015639639639639642,
+      "loss": 0.744,
       "step": 242
     },
     {
       "epoch": 2.19,
-      "learning_rate": 5.405405405405406e-06,
-      "loss": 1.2053,
+      "learning_rate": 0.00015621621621621622,
+      "loss": 0.5623,
       "step": 243
     },
     {
       "epoch": 2.2,
-      "learning_rate": 5.345345345345346e-06,
-      "loss": 1.459,
+      "learning_rate": 0.00015603603603603605,
+      "loss": 0.712,
       "step": 244
     },
     {
       "epoch": 2.21,
-      "learning_rate": 5.285285285285286e-06,
-      "loss": 1.3932,
+      "learning_rate": 0.00015585585585585588,
+      "loss": 0.5958,
       "step": 245
     },
     {
       "epoch": 2.22,
-      "learning_rate": 5.225225225225226e-06,
-      "loss": 1.3219,
+      "learning_rate": 0.0001556756756756757,
+      "loss": 0.7317,
       "step": 246
     },
     {
       "epoch": 2.23,
-      "learning_rate": 5.165165165165165e-06,
-      "loss": 1.3934,
+      "learning_rate": 0.0001554954954954955,
+      "loss": 0.7904,
       "step": 247
     },
     {
       "epoch": 2.23,
-      "learning_rate": 5.105105105105106e-06,
-      "loss": 1.3035,
+      "learning_rate": 0.0001553153153153153,
+      "loss": 0.6661,
       "step": 248
     },
     {
       "epoch": 2.24,
-      "learning_rate": 5.045045045045045e-06,
-      "loss": 1.3401,
+      "learning_rate": 0.00015513513513513514,
+      "loss": 0.6191,
       "step": 249
     },
     {
       "epoch": 2.25,
-      "learning_rate": 4.984984984984985e-06,
-      "loss": 1.3076,
+      "learning_rate": 0.00015495495495495496,
+      "loss": 0.6944,
       "step": 250
     },
     {
       "epoch": 2.26,
-      "learning_rate": 4.9249249249249255e-06,
-      "loss": 1.3364,
+      "learning_rate": 0.00015477477477477477,
+      "loss": 0.7057,
       "step": 251
     },
     {
       "epoch": 2.27,
-      "learning_rate": 4.864864864864866e-06,
-      "loss": 1.2897,
+      "learning_rate": 0.0001545945945945946,
+      "loss": 0.6131,
       "step": 252
     },
     {
       "epoch": 2.28,
-      "learning_rate": 4.804804804804805e-06,
-      "loss": 1.2572,
+      "learning_rate": 0.00015441441441441442,
+      "loss": 0.5626,
       "step": 253
     },
     {
       "epoch": 2.29,
-      "learning_rate": 4.7447447447447454e-06,
-      "loss": 1.3396,
+      "learning_rate": 0.00015423423423423425,
+      "loss": 0.6566,
       "step": 254
     },
     {
       "epoch": 2.3,
-      "learning_rate": 4.684684684684685e-06,
-      "loss": 1.4288,
+      "learning_rate": 0.00015405405405405405,
+      "loss": 0.7677,
       "step": 255
     },
     {
       "epoch": 2.31,
-      "learning_rate": 4.624624624624625e-06,
-      "loss": 1.3392,
+      "learning_rate": 0.00015387387387387388,
+      "loss": 0.7221,
       "step": 256
     },
     {
       "epoch": 2.32,
-      "learning_rate": 4.5645645645645645e-06,
-      "loss": 1.4231,
+      "learning_rate": 0.0001536936936936937,
+      "loss": 0.7011,
       "step": 257
     },
     {
       "epoch": 2.32,
-      "learning_rate": 4.504504504504505e-06,
-      "loss": 1.3351,
+      "learning_rate": 0.00015351351351351354,
+      "loss": 0.651,
       "step": 258
     },
     {
       "epoch": 2.33,
-      "learning_rate": 4.444444444444444e-06,
-      "loss": 1.4397,
+      "learning_rate": 0.00015333333333333334,
+      "loss": 0.7339,
       "step": 259
     },
     {
       "epoch": 2.34,
-      "learning_rate": 4.384384384384384e-06,
-      "loss": 1.4746,
+      "learning_rate": 0.00015315315315315314,
+      "loss": 0.7311,
       "step": 260
     },
     {
       "epoch": 2.35,
-      "learning_rate": 4.324324324324325e-06,
-      "loss": 1.4103,
+      "learning_rate": 0.00015297297297297297,
+      "loss": 0.799,
       "step": 261
     },
     {
       "epoch": 2.36,
-      "learning_rate": 4.264264264264265e-06,
-      "loss": 1.325,
+      "learning_rate": 0.0001527927927927928,
+      "loss": 0.5861,
       "step": 262
     },
     {
       "epoch": 2.37,
-      "learning_rate": 4.204204204204204e-06,
-      "loss": 1.4474,
+      "learning_rate": 0.00015261261261261262,
+      "loss": 0.7177,
       "step": 263
     },
     {
       "epoch": 2.38,
-      "learning_rate": 4.1441441441441446e-06,
-      "loss": 1.243,
+      "learning_rate": 0.00015243243243243243,
+      "loss": 0.6037,
       "step": 264
     },
     {
       "epoch": 2.39,
-      "learning_rate": 4.084084084084085e-06,
-      "loss": 1.3342,
+      "learning_rate": 0.00015225225225225225,
+      "loss": 0.595,
       "step": 265
     },
     {
       "epoch": 2.4,
-      "learning_rate": 4.024024024024024e-06,
-      "loss": 1.2141,
+      "learning_rate": 0.00015207207207207208,
+      "loss": 0.4945,
       "step": 266
     },
     {
       "epoch": 2.41,
-      "learning_rate": 3.9639639639639645e-06,
-      "loss": 1.389,
+      "learning_rate": 0.0001518918918918919,
+      "loss": 0.7155,
       "step": 267
     },
     {
       "epoch": 2.41,
-      "learning_rate": 3.903903903903904e-06,
-      "loss": 1.4364,
+      "learning_rate": 0.0001517117117117117,
+      "loss": 0.633,
       "step": 268
     },
     {
       "epoch": 2.42,
-      "learning_rate": 3.843843843843844e-06,
-      "loss": 1.4544,
+      "learning_rate": 0.00015153153153153154,
+      "loss": 0.698,
       "step": 269
     },
     {
       "epoch": 2.43,
-      "learning_rate": 3.7837837837837844e-06,
-      "loss": 1.4788,
+      "learning_rate": 0.00015135135135135137,
+      "loss": 0.7286,
       "step": 270
     },
     {
       "epoch": 2.44,
-      "learning_rate": 3.723723723723724e-06,
-      "loss": 1.3652,
+      "learning_rate": 0.0001511711711711712,
+      "loss": 0.8749,
       "step": 271
     },
     {
       "epoch": 2.45,
-      "learning_rate": 3.663663663663664e-06,
-      "loss": 1.2965,
+      "learning_rate": 0.000150990990990991,
+      "loss": 0.638,
       "step": 272
     },
     {
       "epoch": 2.46,
-      "learning_rate": 3.603603603603604e-06,
-      "loss": 1.4887,
+      "learning_rate": 0.0001508108108108108,
+      "loss": 0.7524,
       "step": 273
     },
     {
       "epoch": 2.47,
-      "learning_rate": 3.5435435435435437e-06,
-      "loss": 1.2402,
+      "learning_rate": 0.00015063063063063063,
+      "loss": 1.5031,
       "step": 274
     },
     {
       "epoch": 2.48,
-      "learning_rate": 3.4834834834834835e-06,
-      "loss": 1.265,
+      "learning_rate": 0.00015045045045045046,
+      "loss": 1.8588,
       "step": 275
     },
     {
       "epoch": 2.49,
-      "learning_rate": 3.423423423423424e-06,
-      "loss": 1.3136,
+      "learning_rate": 0.00015027027027027028,
+      "loss": 0.8453,
       "step": 276
     },
     {
       "epoch": 2.5,
-      "learning_rate": 3.363363363363364e-06,
-      "loss": 1.5174,
+      "learning_rate": 0.00015009009009009009,
+      "loss": 1.0263,
       "step": 277
     },
     {
       "epoch": 2.5,
-      "learning_rate": 3.3033033033033035e-06,
-      "loss": 1.3669,
+      "learning_rate": 0.00014990990990990991,
+      "loss": 0.7366,
       "step": 278
     },
     {
       "epoch": 2.51,
-      "learning_rate": 3.2432432432432437e-06,
-      "loss": 1.2474,
+      "learning_rate": 0.00014972972972972974,
+      "loss": 0.7533,
       "step": 279
     },
     {
       "epoch": 2.52,
-      "learning_rate": 3.183183183183183e-06,
-      "loss": 1.3761,
+      "learning_rate": 0.00014954954954954957,
+      "loss": 0.7375,
       "step": 280
     },
     {
       "epoch": 2.53,
-      "learning_rate": 3.1231231231231234e-06,
-      "loss": 1.3245,
+      "learning_rate": 0.00014936936936936937,
+      "loss": 0.8677,
       "step": 281
     },
     {
       "epoch": 2.54,
-      "learning_rate": 3.063063063063063e-06,
-      "loss": 1.3889,
+      "learning_rate": 0.0001491891891891892,
+      "loss": 0.689,
       "step": 282
     },
     {
       "epoch": 2.55,
-      "learning_rate": 3.0030030030030034e-06,
-      "loss": 1.2964,
+      "learning_rate": 0.00014900900900900903,
+      "loss": 0.6659,
       "step": 283
     },
     {
       "epoch": 2.56,
-      "learning_rate": 2.942942942942943e-06,
-      "loss": 1.4232,
+      "learning_rate": 0.00014882882882882883,
+      "loss": 1.3045,
       "step": 284
     },
     {
       "epoch": 2.57,
-      "learning_rate": 2.882882882882883e-06,
-      "loss": 1.5877,
+      "learning_rate": 0.00014864864864864866,
+      "loss": 0.8731,
       "step": 285
     },
     {
       "epoch": 2.58,
-      "learning_rate": 2.8228228228228234e-06,
-      "loss": 1.7226,
+      "learning_rate": 0.00014846846846846846,
+      "loss": 0.9237,
       "step": 286
     },
     {
       "epoch": 2.59,
-      "learning_rate": 2.7627627627627628e-06,
-      "loss": 1.3672,
+      "learning_rate": 0.0001482882882882883,
+      "loss": 0.7298,
       "step": 287
     },
     {
       "epoch": 2.59,
-      "learning_rate": 2.702702702702703e-06,
-      "loss": 1.3058,
+      "learning_rate": 0.00014810810810810812,
+      "loss": 0.5712,
       "step": 288
     },
     {
       "epoch": 2.6,
-      "learning_rate": 2.642642642642643e-06,
-      "loss": 1.5852,
+      "learning_rate": 0.00014792792792792795,
+      "loss": 0.8097,
       "step": 289
     },
     {
       "epoch": 2.61,
-      "learning_rate": 2.5825825825825827e-06,
-      "loss": 1.1589,
+      "learning_rate": 0.00014774774774774775,
+      "loss": 0.5713,
       "step": 290
     },
     {
       "epoch": 2.62,
-      "learning_rate": 2.5225225225225225e-06,
-      "loss": 1.4041,
+      "learning_rate": 0.00014756756756756758,
+      "loss": 0.7031,
       "step": 291
     },
     {
       "epoch": 2.63,
-      "learning_rate": 2.4624624624624628e-06,
-      "loss": 1.3072,
+      "learning_rate": 0.0001473873873873874,
+      "loss": 0.5795,
       "step": 292
     },
     {
       "epoch": 2.64,
-      "learning_rate": 2.4024024024024026e-06,
-      "loss": 1.3421,
+      "learning_rate": 0.00014720720720720723,
+      "loss": 0.6107,
       "step": 293
     },
     {
       "epoch": 2.65,
-      "learning_rate": 2.3423423423423424e-06,
-      "loss": 1.5669,
+      "learning_rate": 0.00014702702702702703,
+      "loss": 0.7417,
       "step": 294
     },
     {
       "epoch": 2.66,
-      "learning_rate": 2.2822822822822822e-06,
-      "loss": 1.4368,
+      "learning_rate": 0.00014684684684684686,
+      "loss": 0.7342,
       "step": 295
     },
     {
       "epoch": 2.67,
-      "learning_rate": 2.222222222222222e-06,
-      "loss": 1.1976,
+      "learning_rate": 0.00014666666666666666,
+      "loss": 0.6121,
       "step": 296
     },
     {
       "epoch": 2.68,
-      "learning_rate": 2.1621621621621623e-06,
-      "loss": 1.4212,
+      "learning_rate": 0.0001464864864864865,
+      "loss": 0.801,
       "step": 297
     },
     {
       "epoch": 2.68,
-      "learning_rate": 2.102102102102102e-06,
-      "loss": 1.3067,
+      "learning_rate": 0.00014630630630630632,
+      "loss": 0.5626,
       "step": 298
     },
     {
       "epoch": 2.69,
-      "learning_rate": 2.0420420420420424e-06,
-      "loss": 1.3672,
+      "learning_rate": 0.00014612612612612612,
+      "loss": 0.6167,
       "step": 299
     },
     {
       "epoch": 2.7,
-      "learning_rate": 1.9819819819819822e-06,
-      "loss": 1.3335,
+      "learning_rate": 0.00014594594594594595,
+      "loss": 0.606,
       "step": 300
     },
     {
       "epoch": 2.71,
-      "learning_rate": 1.921921921921922e-06,
-      "loss": 1.467,
+      "learning_rate": 0.00014576576576576578,
+      "loss": 0.7919,
       "step": 301
     },
     {
       "epoch": 2.72,
-      "learning_rate": 1.861861861861862e-06,
-      "loss": 1.535,
+      "learning_rate": 0.0001455855855855856,
+      "loss": 0.9143,
       "step": 302
     },
     {
       "epoch": 2.73,
-      "learning_rate": 1.801801801801802e-06,
-      "loss": 1.3377,
+      "learning_rate": 0.0001454054054054054,
+      "loss": 0.5416,
       "step": 303
     },
     {
       "epoch": 2.74,
-      "learning_rate": 1.7417417417417418e-06,
-      "loss": 1.5501,
+      "learning_rate": 0.00014522522522522524,
+      "loss": 0.9303,
       "step": 304
     },
     {
       "epoch": 2.75,
-      "learning_rate": 1.681681681681682e-06,
-      "loss": 1.5146,
+      "learning_rate": 0.00014504504504504506,
+      "loss": 0.7065,
       "step": 305
     },
     {
       "epoch": 2.76,
-      "learning_rate": 1.6216216216216219e-06,
-      "loss": 1.3974,
+      "learning_rate": 0.0001448648648648649,
+      "loss": 0.7287,
       "step": 306
     },
     {
       "epoch": 2.77,
-      "learning_rate": 1.5615615615615617e-06,
-      "loss": 1.4609,
+      "learning_rate": 0.0001446846846846847,
+      "loss": 0.6792,
       "step": 307
     },
     {
       "epoch": 2.77,
-      "learning_rate": 1.5015015015015017e-06,
-      "loss": 1.4186,
+      "learning_rate": 0.0001445045045045045,
+      "loss": 0.762,
       "step": 308
     },
     {
       "epoch": 2.78,
-      "learning_rate": 1.4414414414414416e-06,
-      "loss": 1.4664,
+      "learning_rate": 0.00014432432432432432,
+      "loss": 0.5965,
       "step": 309
     },
     {
       "epoch": 2.79,
-      "learning_rate": 1.3813813813813814e-06,
-      "loss": 1.3609,
+      "learning_rate": 0.00014414414414414415,
+      "loss": 0.6565,
       "step": 310
     },
     {
       "epoch": 2.8,
-      "learning_rate": 1.3213213213213214e-06,
-      "loss": 1.4355,
+      "learning_rate": 0.00014396396396396395,
+      "loss": 0.734,
       "step": 311
     },
     {
       "epoch": 2.81,
-      "learning_rate": 1.2612612612612613e-06,
-      "loss": 1.3592,
+      "learning_rate": 0.00014378378378378378,
+      "loss": 0.6371,
       "step": 312
     },
     {
       "epoch": 2.82,
-      "learning_rate": 1.2012012012012013e-06,
-      "loss": 1.3873,
+      "learning_rate": 0.0001436036036036036,
+      "loss": 0.67,
       "step": 313
     },
     {
       "epoch": 2.83,
-      "learning_rate": 1.1411411411411411e-06,
-      "loss": 1.3438,
+      "learning_rate": 0.00014342342342342344,
+      "loss": 0.6313,
       "step": 314
     },
     {
       "epoch": 2.84,
-      "learning_rate": 1.0810810810810812e-06,
-      "loss": 1.2623,
+      "learning_rate": 0.00014324324324324324,
+      "loss": 0.6081,
       "step": 315
     },
     {
       "epoch": 2.85,
-      "learning_rate": 1.0210210210210212e-06,
-      "loss": 1.4811,
+      "learning_rate": 0.00014306306306306307,
+      "loss": 0.7904,
       "step": 316
     },
     {
       "epoch": 2.86,
-      "learning_rate": 9.60960960960961e-07,
-      "loss": 1.2461,
+      "learning_rate": 0.0001428828828828829,
+      "loss": 0.6455,
       "step": 317
     },
     {
       "epoch": 2.86,
-      "learning_rate": 9.00900900900901e-07,
-      "loss": 1.2419,
+      "learning_rate": 0.00014270270270270272,
+      "loss": 0.6323,
       "step": 318
     },
     {
       "epoch": 2.87,
-      "learning_rate": 8.40840840840841e-07,
-      "loss": 1.5783,
+      "learning_rate": 0.00014252252252252253,
+      "loss": 0.779,
       "step": 319
     },
     {
       "epoch": 2.88,
-      "learning_rate": 7.807807807807808e-07,
-      "loss": 1.2684,
+      "learning_rate": 0.00014234234234234233,
+      "loss": 0.5011,
       "step": 320
     },
     {
       "epoch": 2.89,
-      "learning_rate": 7.207207207207208e-07,
-      "loss": 1.3713,
+      "learning_rate": 0.00014216216216216216,
+      "loss": 0.6007,
       "step": 321
     },
     {
       "epoch": 2.9,
-      "learning_rate": 6.606606606606607e-07,
-      "loss": 1.5109,
+      "learning_rate": 0.00014198198198198198,
+      "loss": 0.7067,
       "step": 322
     },
     {
       "epoch": 2.91,
-      "learning_rate": 6.006006006006006e-07,
-      "loss": 1.3602,
+      "learning_rate": 0.0001418018018018018,
+      "loss": 0.6962,
       "step": 323
     },
     {
       "epoch": 2.92,
-      "learning_rate": 5.405405405405406e-07,
-      "loss": 1.5054,
+      "learning_rate": 0.00014162162162162161,
+      "loss": 0.7823,
       "step": 324
     },
     {
       "epoch": 2.93,
-      "learning_rate": 4.804804804804805e-07,
-      "loss": 1.5096,
+      "learning_rate": 0.00014144144144144144,
+      "loss": 0.7142,
       "step": 325
     },
     {
       "epoch": 2.94,
-      "learning_rate": 4.204204204204205e-07,
-      "loss": 1.3165,
+      "learning_rate": 0.00014126126126126127,
+      "loss": 0.6356,
       "step": 326
     },
     {
       "epoch": 2.95,
-      "learning_rate": 3.603603603603604e-07,
-      "loss": 1.4575,
+      "learning_rate": 0.0001410810810810811,
+      "loss": 0.7062,
       "step": 327
     },
     {
       "epoch": 2.95,
-      "learning_rate": 3.003003003003003e-07,
-      "loss": 1.6184,
+      "learning_rate": 0.0001409009009009009,
+      "loss": 0.7607,
       "step": 328
     },
     {
       "epoch": 2.96,
-      "learning_rate": 2.4024024024024026e-07,
-      "loss": 1.2022,
+      "learning_rate": 0.00014072072072072073,
+      "loss": 0.6222,
       "step": 329
     },
     {
       "epoch": 2.97,
-      "learning_rate": 1.801801801801802e-07,
-      "loss": 1.4417,
+      "learning_rate": 0.00014054054054054056,
+      "loss": 0.7505,
       "step": 330
     },
     {
       "epoch": 2.98,
-      "learning_rate": 1.2012012012012013e-07,
-      "loss": 1.2554,
+      "learning_rate": 0.00014036036036036039,
+      "loss": 0.6673,
       "step": 331
     },
     {
       "epoch": 2.99,
-      "learning_rate": 6.006006006006006e-08,
-      "loss": 1.28,
+      "learning_rate": 0.0001401801801801802,
+      "loss": 0.6934,
       "step": 332
     },
     {
       "epoch": 3.0,
-      "learning_rate": 0.0,
-      "loss": 1.2557,
+      "learning_rate": 0.00014,
+      "loss": 0.5489,
       "step": 333
     },
     {
-      "epoch": 3.0,
-      "step": 333,
-      "total_flos": 1915204220928000.0,
-      "train_loss": 1.781525854233865,
-      "train_runtime": 119.0355,
-      "train_samples_per_second": 88.889,
-      "train_steps_per_second": 2.797
+      "epoch": 3.01,
+      "learning_rate": 0.00013981981981981982,
+      "loss": 0.4571,
+      "step": 334
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 0.00013963963963963964,
+      "loss": 0.4375,
+      "step": 335
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 0.00013945945945945947,
+      "loss": 0.4438,
+      "step": 336
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 0.00013927927927927927,
+      "loss": 0.4085,
+      "step": 337
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.0001390990990990991,
+      "loss": 0.5438,
+      "step": 338
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00013891891891891893,
+      "loss": 0.5038,
+      "step": 339
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 0.00013873873873873876,
+      "loss": 0.479,
+      "step": 340
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 0.00013855855855855856,
+      "loss": 0.405,
+      "step": 341
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 0.0001383783783783784,
+      "loss": 0.4181,
+      "step": 342
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 0.00013819819819819822,
+      "loss": 0.4513,
+      "step": 343
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 0.00013801801801801802,
+      "loss": 0.4685,
+      "step": 344
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 0.00013783783783783785,
+      "loss": 0.5291,
+      "step": 345
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 0.00013765765765765765,
+      "loss": 0.5497,
+      "step": 346
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 0.00013747747747747748,
+      "loss": 0.4366,
+      "step": 347
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.0001372972972972973,
+      "loss": 0.5815,
+      "step": 348
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 0.00013711711711711713,
+      "loss": 0.5056,
+      "step": 349
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 0.00013693693693693693,
+      "loss": 0.4975,
+      "step": 350
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 0.00013675675675675676,
+      "loss": 0.4601,
+      "step": 351
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 0.0001365765765765766,
+      "loss": 0.4328,
+      "step": 352
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 0.00013639639639639642,
+      "loss": 0.5189,
+      "step": 353
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 0.00013621621621621622,
+      "loss": 0.4054,
+      "step": 354
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 0.00013603603603603605,
+      "loss": 0.4604,
+      "step": 355
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 0.00013585585585585585,
+      "loss": 0.5539,
+      "step": 356
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 0.00013567567567567568,
+      "loss": 0.3773,
+      "step": 357
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.0001354954954954955,
+      "loss": 0.4631,
+      "step": 358
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 0.0001353153153153153,
+      "loss": 0.463,
+      "step": 359
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 0.00013513513513513514,
+      "loss": 0.4423,
+      "step": 360
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 0.00013495495495495497,
+      "loss": 0.5197,
+      "step": 361
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 0.0001347747747747748,
+      "loss": 0.6306,
+      "step": 362
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 0.0001345945945945946,
+      "loss": 0.431,
+      "step": 363
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 0.00013441441441441442,
+      "loss": 0.4192,
+      "step": 364
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 0.00013423423423423425,
+      "loss": 0.5175,
+      "step": 365
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 0.00013405405405405408,
+      "loss": 0.4623,
+      "step": 366
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 0.00013387387387387388,
+      "loss": 0.4332,
+      "step": 367
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.00013369369369369368,
+      "loss": 0.4969,
+      "step": 368
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 0.0001335135135135135,
+      "loss": 0.5908,
+      "step": 369
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 0.00013333333333333334,
+      "loss": 0.4597,
+      "step": 370
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 0.00013315315315315314,
+      "loss": 0.4516,
+      "step": 371
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 0.00013297297297297297,
+      "loss": 0.498,
+      "step": 372
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 0.0001327927927927928,
+      "loss": 0.4829,
+      "step": 373
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 0.00013261261261261263,
+      "loss": 0.4679,
+      "step": 374
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 0.00013243243243243243,
+      "loss": 0.5148,
+      "step": 375
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 0.00013225225225225226,
+      "loss": 0.4495,
+      "step": 376
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 0.00013207207207207208,
+      "loss": 0.4408,
+      "step": 377
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.0001318918918918919,
+      "loss": 0.4732,
+      "step": 378
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 0.00013171171171171171,
+      "loss": 0.4048,
+      "step": 379
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 0.00013153153153153154,
+      "loss": 0.4607,
+      "step": 380
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 0.00013135135135135134,
+      "loss": 0.457,
+      "step": 381
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 0.00013117117117117117,
+      "loss": 0.4446,
+      "step": 382
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 0.000130990990990991,
+      "loss": 0.5565,
+      "step": 383
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 0.0001308108108108108,
+      "loss": 0.452,
+      "step": 384
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 0.00013063063063063063,
+      "loss": 0.6298,
+      "step": 385
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 0.00013045045045045046,
+      "loss": 0.5399,
+      "step": 386
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 0.0001302702702702703,
+      "loss": 0.4797,
+      "step": 387
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.0001300900900900901,
+      "loss": 0.5867,
+      "step": 388
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 0.00012990990990990992,
+      "loss": 0.4448,
+      "step": 389
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 0.00012972972972972974,
+      "loss": 0.5132,
+      "step": 390
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 0.00012954954954954957,
+      "loss": 0.4374,
+      "step": 391
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 0.00012936936936936937,
+      "loss": 0.5119,
+      "step": 392
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 0.00012918918918918918,
+      "loss": 0.5887,
+      "step": 393
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 0.000129009009009009,
+      "loss": 0.477,
+      "step": 394
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 0.00012882882882882883,
+      "loss": 0.5135,
+      "step": 395
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 0.00012864864864864866,
+      "loss": 0.4852,
+      "step": 396
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 0.00012846846846846846,
+      "loss": 0.5327,
+      "step": 397
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.0001282882882882883,
+      "loss": 0.4726,
+      "step": 398
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 0.00012810810810810812,
+      "loss": 0.4805,
+      "step": 399
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 0.00012792792792792795,
+      "loss": 0.4577,
+      "step": 400
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 0.00012774774774774775,
+      "loss": 0.4823,
+      "step": 401
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 0.00012756756756756758,
+      "loss": 0.4779,
+      "step": 402
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 0.0001273873873873874,
+      "loss": 0.586,
+      "step": 403
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 0.00012720720720720723,
+      "loss": 0.4533,
+      "step": 404
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 0.00012702702702702703,
+      "loss": 0.5355,
+      "step": 405
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 0.00012684684684684684,
+      "loss": 0.581,
+      "step": 406
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.4933,
+      "step": 407
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.0001264864864864865,
+      "loss": 0.5883,
+      "step": 408
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 0.00012630630630630632,
+      "loss": 0.5552,
+      "step": 409
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 0.00012612612612612612,
+      "loss": 0.5098,
+      "step": 410
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 0.00012594594594594595,
+      "loss": 0.4585,
+      "step": 411
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 0.00012576576576576578,
+      "loss": 0.5726,
+      "step": 412
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 0.0001255855855855856,
+      "loss": 0.5016,
+      "step": 413
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 0.0001254054054054054,
+      "loss": 0.6272,
+      "step": 414
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 0.00012522522522522524,
+      "loss": 0.5357,
+      "step": 415
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 0.00012504504504504507,
+      "loss": 0.4562,
+      "step": 416
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 0.00012486486486486487,
+      "loss": 0.5279,
+      "step": 417
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.0001246846846846847,
+      "loss": 0.5381,
+      "step": 418
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 0.0001245045045045045,
+      "loss": 0.5124,
+      "step": 419
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 0.00012432432432432433,
+      "loss": 0.5242,
+      "step": 420
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 0.00012414414414414415,
+      "loss": 0.4864,
+      "step": 421
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 0.00012396396396396398,
+      "loss": 0.4408,
+      "step": 422
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 0.00012378378378378378,
+      "loss": 0.4686,
+      "step": 423
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 0.0001236036036036036,
+      "loss": 0.4583,
+      "step": 424
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 0.00012342342342342344,
+      "loss": 0.5698,
+      "step": 425
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 0.00012324324324324327,
+      "loss": 0.5193,
+      "step": 426
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 0.00012306306306306307,
+      "loss": 0.6061,
+      "step": 427
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.0001228828828828829,
+      "loss": 0.4556,
+      "step": 428
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 0.0001227027027027027,
+      "loss": 0.4494,
+      "step": 429
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 0.00012252252252252253,
+      "loss": 0.4351,
+      "step": 430
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 0.00012234234234234233,
+      "loss": 0.4312,
+      "step": 431
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 0.00012216216216216216,
+      "loss": 0.5729,
+      "step": 432
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 0.00012198198198198199,
+      "loss": 0.4375,
+      "step": 433
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 0.0001218018018018018,
+      "loss": 0.5225,
+      "step": 434
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 0.00012162162162162163,
+      "loss": 0.4699,
+      "step": 435
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 0.00012144144144144144,
+      "loss": 0.4862,
+      "step": 436
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 0.00012126126126126127,
+      "loss": 0.5477,
+      "step": 437
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00012108108108108109,
+      "loss": 0.5063,
+      "step": 438
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 0.00012090090090090092,
+      "loss": 0.4708,
+      "step": 439
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 0.00012072072072072073,
+      "loss": 0.3925,
+      "step": 440
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 0.00012054054054054053,
+      "loss": 0.5277,
+      "step": 441
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 0.00012036036036036036,
+      "loss": 0.5806,
+      "step": 442
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 0.00012018018018018017,
+      "loss": 0.4542,
+      "step": 443
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 0.00012,
+      "loss": 0.6045,
+      "step": 444
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 0.00011981981981981982,
+      "loss": 0.3265,
+      "step": 445
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 0.00011963963963963965,
+      "loss": 0.3141,
+      "step": 446
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 0.00011945945945945946,
+      "loss": 0.312,
+      "step": 447
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 0.00011927927927927929,
+      "loss": 0.3917,
+      "step": 448
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 0.0001190990990990991,
+      "loss": 0.4704,
+      "step": 449
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 0.00011891891891891893,
+      "loss": 0.3541,
+      "step": 450
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 0.00011873873873873875,
+      "loss": 0.401,
+      "step": 451
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 0.00011855855855855858,
+      "loss": 0.3266,
+      "step": 452
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 0.00011837837837837838,
+      "loss": 0.3155,
+      "step": 453
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 0.00011819819819819819,
+      "loss": 0.3531,
+      "step": 454
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 0.00011801801801801802,
+      "loss": 0.3775,
+      "step": 455
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 0.00011783783783783784,
+      "loss": 0.3703,
+      "step": 456
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 0.00011765765765765766,
+      "loss": 0.3814,
+      "step": 457
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 0.00011747747747747748,
+      "loss": 0.3786,
+      "step": 458
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 0.00011729729729729731,
+      "loss": 0.3343,
+      "step": 459
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 0.00011711711711711712,
+      "loss": 0.3516,
+      "step": 460
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 0.00011693693693693695,
+      "loss": 0.3729,
+      "step": 461
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 0.00011675675675675676,
+      "loss": 0.3886,
+      "step": 462
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 0.0001165765765765766,
+      "loss": 0.364,
+      "step": 463
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 0.00011639639639639641,
+      "loss": 0.3486,
+      "step": 464
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 0.00011621621621621621,
+      "loss": 0.3411,
+      "step": 465
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 0.00011603603603603604,
+      "loss": 0.4359,
+      "step": 466
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 0.00011585585585585585,
+      "loss": 0.336,
+      "step": 467
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 0.00011567567567567568,
+      "loss": 0.4453,
+      "step": 468
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.0001154954954954955,
+      "loss": 0.3626,
+      "step": 469
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 0.00011531531531531532,
+      "loss": 0.3608,
+      "step": 470
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 0.00011513513513513514,
+      "loss": 0.3803,
+      "step": 471
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 0.00011495495495495497,
+      "loss": 0.3862,
+      "step": 472
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 0.00011477477477477478,
+      "loss": 0.3639,
+      "step": 473
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 0.00011459459459459461,
+      "loss": 0.3345,
+      "step": 474
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 0.00011441441441441443,
+      "loss": 0.3343,
+      "step": 475
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 0.00011423423423423425,
+      "loss": 0.3943,
+      "step": 476
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 0.00011405405405405406,
+      "loss": 0.3802,
+      "step": 477
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 0.00011387387387387387,
+      "loss": 0.3523,
+      "step": 478
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 0.0001136936936936937,
+      "loss": 0.4398,
+      "step": 479
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 0.00011351351351351351,
+      "loss": 0.3851,
+      "step": 480
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 0.00011333333333333334,
+      "loss": 0.3755,
+      "step": 481
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 0.00011315315315315316,
+      "loss": 0.3402,
+      "step": 482
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 0.00011297297297297298,
+      "loss": 0.3227,
+      "step": 483
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 0.0001127927927927928,
+      "loss": 0.3282,
+      "step": 484
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 0.00011261261261261263,
+      "loss": 0.3621,
+      "step": 485
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 0.00011243243243243244,
+      "loss": 0.3507,
+      "step": 486
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 0.00011225225225225227,
+      "loss": 0.412,
+      "step": 487
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 0.00011207207207207209,
+      "loss": 0.5337,
+      "step": 488
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.00011189189189189189,
+      "loss": 0.4157,
+      "step": 489
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 0.0001117117117117117,
+      "loss": 0.411,
+      "step": 490
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 0.00011153153153153153,
+      "loss": 0.4512,
+      "step": 491
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 0.00011135135135135135,
+      "loss": 0.3734,
+      "step": 492
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 0.00011117117117117117,
+      "loss": 0.3837,
+      "step": 493
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 0.00011099099099099099,
+      "loss": 0.4167,
+      "step": 494
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 0.00011081081081081082,
+      "loss": 0.3796,
+      "step": 495
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 0.00011063063063063063,
+      "loss": 0.3891,
+      "step": 496
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 0.00011045045045045046,
+      "loss": 0.3086,
+      "step": 497
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 0.00011027027027027029,
+      "loss": 0.4173,
+      "step": 498
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 0.0001100900900900901,
+      "loss": 0.3275,
+      "step": 499
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 0.00010990990990990993,
+      "loss": 0.357,
+      "step": 500
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 0.00010972972972972972,
+      "loss": 0.4611,
+      "step": 501
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 0.00010954954954954955,
+      "loss": 0.3544,
+      "step": 502
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 0.00010936936936936936,
+      "loss": 0.3564,
+      "step": 503
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 0.00010918918918918919,
+      "loss": 0.3306,
+      "step": 504
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 0.000109009009009009,
+      "loss": 0.4079,
+      "step": 505
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 0.00010882882882882883,
+      "loss": 0.3906,
+      "step": 506
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 0.00010864864864864865,
+      "loss": 0.3863,
+      "step": 507
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 0.00010846846846846848,
+      "loss": 0.3378,
+      "step": 508
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.00010828828828828829,
+      "loss": 0.3556,
+      "step": 509
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 0.00010810810810810812,
+      "loss": 0.2989,
+      "step": 510
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 0.00010792792792792794,
+      "loss": 0.4635,
+      "step": 511
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 0.00010774774774774776,
+      "loss": 0.3909,
+      "step": 512
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 0.00010756756756756757,
+      "loss": 0.3124,
+      "step": 513
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 0.00010738738738738738,
+      "loss": 0.4026,
+      "step": 514
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 0.00010720720720720721,
+      "loss": 0.3955,
+      "step": 515
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 0.00010702702702702702,
+      "loss": 0.3776,
+      "step": 516
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 0.00010684684684684685,
+      "loss": 0.3324,
+      "step": 517
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 0.00010666666666666667,
+      "loss": 0.3567,
+      "step": 518
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.0001064864864864865,
+      "loss": 0.3758,
+      "step": 519
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 0.00010630630630630631,
+      "loss": 0.3658,
+      "step": 520
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 0.00010612612612612614,
+      "loss": 0.4016,
+      "step": 521
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 0.00010594594594594595,
+      "loss": 0.3351,
+      "step": 522
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 0.00010576576576576578,
+      "loss": 0.4547,
+      "step": 523
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 0.0001055855855855856,
+      "loss": 0.4222,
+      "step": 524
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 0.0001054054054054054,
+      "loss": 0.3836,
+      "step": 525
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 0.00010522522522522523,
+      "loss": 0.3061,
+      "step": 526
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 0.00010504504504504504,
+      "loss": 0.3326,
+      "step": 527
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 0.00010486486486486487,
+      "loss": 0.3756,
+      "step": 528
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.00010468468468468468,
+      "loss": 0.3527,
+      "step": 529
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 0.00010450450450450451,
+      "loss": 0.3688,
+      "step": 530
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 0.00010432432432432433,
+      "loss": 0.3322,
+      "step": 531
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 0.00010414414414414416,
+      "loss": 0.4191,
+      "step": 532
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 0.00010396396396396397,
+      "loss": 0.3586,
+      "step": 533
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 0.0001037837837837838,
+      "loss": 0.4528,
+      "step": 534
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 0.00010360360360360361,
+      "loss": 0.2971,
+      "step": 535
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 0.00010342342342342344,
+      "loss": 0.3338,
+      "step": 536
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 0.00010324324324324324,
+      "loss": 0.3805,
+      "step": 537
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 0.00010306306306306306,
+      "loss": 0.3958,
+      "step": 538
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.00010288288288288289,
+      "loss": 0.3995,
+      "step": 539
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 0.0001027027027027027,
+      "loss": 0.3661,
+      "step": 540
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 0.00010252252252252253,
+      "loss": 0.2927,
+      "step": 541
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 0.00010234234234234234,
+      "loss": 0.4004,
+      "step": 542
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 0.00010216216216216217,
+      "loss": 0.3619,
+      "step": 543
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 0.00010198198198198199,
+      "loss": 0.3478,
+      "step": 544
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 0.00010180180180180182,
+      "loss": 0.417,
+      "step": 545
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 0.00010162162162162163,
+      "loss": 0.4245,
+      "step": 546
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 0.00010144144144144146,
+      "loss": 0.319,
+      "step": 547
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 0.00010126126126126127,
+      "loss": 0.281,
+      "step": 548
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 0.00010108108108108108,
+      "loss": 0.3438,
+      "step": 549
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 0.00010090090090090089,
+      "loss": 0.3441,
+      "step": 550
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 0.00010072072072072072,
+      "loss": 0.3684,
+      "step": 551
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 0.00010054054054054053,
+      "loss": 0.3679,
+      "step": 552
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 0.00010036036036036036,
+      "loss": 0.3563,
+      "step": 553
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 0.00010018018018018018,
+      "loss": 0.4197,
+      "step": 554
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 0.0001,
+      "loss": 0.4142,
+      "step": 555
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 9.981981981981983e-05,
+      "loss": 0.2757,
+      "step": 556
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 9.963963963963965e-05,
+      "loss": 0.3093,
+      "step": 557
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 9.945945945945948e-05,
+      "loss": 0.2858,
+      "step": 558
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 9.927927927927928e-05,
+      "loss": 0.2819,
+      "step": 559
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 9.90990990990991e-05,
+      "loss": 0.2873,
+      "step": 560
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 9.891891891891892e-05,
+      "loss": 0.2807,
+      "step": 561
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 9.873873873873875e-05,
+      "loss": 0.3117,
+      "step": 562
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 9.855855855855856e-05,
+      "loss": 0.2886,
+      "step": 563
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 9.837837837837839e-05,
+      "loss": 0.2797,
+      "step": 564
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 9.81981981981982e-05,
+      "loss": 0.3165,
+      "step": 565
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 9.801801801801802e-05,
+      "loss": 0.3353,
+      "step": 566
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 9.783783783783784e-05,
+      "loss": 0.2902,
+      "step": 567
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 9.765765765765767e-05,
+      "loss": 0.2807,
+      "step": 568
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 9.747747747747748e-05,
+      "loss": 0.3003,
+      "step": 569
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 9.729729729729731e-05,
+      "loss": 0.2983,
+      "step": 570
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 9.711711711711712e-05,
+      "loss": 0.2716,
+      "step": 571
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 9.693693693693694e-05,
+      "loss": 0.3266,
+      "step": 572
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 9.675675675675677e-05,
+      "loss": 0.2585,
+      "step": 573
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 9.657657657657658e-05,
+      "loss": 0.2978,
+      "step": 574
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 9.639639639639641e-05,
+      "loss": 0.3214,
+      "step": 575
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 9.621621621621622e-05,
+      "loss": 0.2546,
+      "step": 576
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 9.603603603603604e-05,
+      "loss": 0.2801,
+      "step": 577
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 9.585585585585585e-05,
+      "loss": 0.2491,
+      "step": 578
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 9.567567567567568e-05,
+      "loss": 0.2667,
+      "step": 579
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 9.54954954954955e-05,
+      "loss": 0.2588,
+      "step": 580
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 9.531531531531533e-05,
+      "loss": 0.2887,
+      "step": 581
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 9.513513513513514e-05,
+      "loss": 0.2701,
+      "step": 582
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 9.495495495495496e-05,
+      "loss": 0.2838,
+      "step": 583
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 9.477477477477478e-05,
+      "loss": 0.3002,
+      "step": 584
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 9.45945945945946e-05,
+      "loss": 0.2738,
+      "step": 585
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 9.441441441441443e-05,
+      "loss": 0.2904,
+      "step": 586
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 9.423423423423424e-05,
+      "loss": 0.2691,
+      "step": 587
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 9.405405405405407e-05,
+      "loss": 0.3163,
+      "step": 588
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 9.387387387387387e-05,
+      "loss": 0.2756,
+      "step": 589
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 9.36936936936937e-05,
+      "loss": 0.2878,
+      "step": 590
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 9.351351351351351e-05,
+      "loss": 0.2815,
+      "step": 591
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 9.333333333333334e-05,
+      "loss": 0.3392,
+      "step": 592
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 9.315315315315316e-05,
+      "loss": 0.3289,
+      "step": 593
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 9.297297297297299e-05,
+      "loss": 0.3027,
+      "step": 594
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 9.279279279279279e-05,
+      "loss": 0.3511,
+      "step": 595
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 9.261261261261262e-05,
+      "loss": 0.2502,
+      "step": 596
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 9.243243243243243e-05,
+      "loss": 0.281,
+      "step": 597
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 9.225225225225226e-05,
+      "loss": 0.3211,
+      "step": 598
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 9.207207207207207e-05,
+      "loss": 0.2933,
+      "step": 599
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 9.18918918918919e-05,
+      "loss": 0.3106,
+      "step": 600
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 9.171171171171172e-05,
+      "loss": 0.2699,
+      "step": 601
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 9.153153153153153e-05,
+      "loss": 0.2849,
+      "step": 602
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 9.135135135135136e-05,
+      "loss": 0.2974,
+      "step": 603
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 9.117117117117118e-05,
+      "loss": 0.2942,
+      "step": 604
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 9.0990990990991e-05,
+      "loss": 0.2788,
+      "step": 605
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 9.081081081081082e-05,
+      "loss": 0.2658,
+      "step": 606
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 9.063063063063063e-05,
+      "loss": 0.3066,
+      "step": 607
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 9.045045045045045e-05,
+      "loss": 0.3215,
+      "step": 608
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 9.027027027027028e-05,
+      "loss": 0.2737,
+      "step": 609
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 9.009009009009009e-05,
+      "loss": 0.279,
+      "step": 610
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 8.990990990990992e-05,
+      "loss": 0.2721,
+      "step": 611
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 8.972972972972973e-05,
+      "loss": 0.2429,
+      "step": 612
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 8.954954954954955e-05,
+      "loss": 0.2799,
+      "step": 613
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 8.936936936936938e-05,
+      "loss": 0.2756,
+      "step": 614
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 8.918918918918919e-05,
+      "loss": 0.2578,
+      "step": 615
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 8.900900900900902e-05,
+      "loss": 0.3057,
+      "step": 616
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 8.882882882882884e-05,
+      "loss": 0.262,
+      "step": 617
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 8.864864864864866e-05,
+      "loss": 0.3209,
+      "step": 618
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 8.846846846846847e-05,
+      "loss": 0.2809,
+      "step": 619
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 8.82882882882883e-05,
+      "loss": 0.299,
+      "step": 620
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 8.810810810810811e-05,
+      "loss": 0.3359,
+      "step": 621
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 8.792792792792794e-05,
+      "loss": 0.3272,
+      "step": 622
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 8.774774774774775e-05,
+      "loss": 0.3184,
+      "step": 623
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 8.756756756756758e-05,
+      "loss": 0.3008,
+      "step": 624
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 8.738738738738738e-05,
+      "loss": 0.2611,
+      "step": 625
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 8.720720720720721e-05,
+      "loss": 0.3845,
+      "step": 626
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 8.702702702702702e-05,
+      "loss": 0.2434,
+      "step": 627
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 8.684684684684685e-05,
+      "loss": 0.3178,
+      "step": 628
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 8.666666666666667e-05,
+      "loss": 0.2643,
+      "step": 629
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 8.64864864864865e-05,
+      "loss": 0.3346,
+      "step": 630
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 8.630630630630631e-05,
+      "loss": 0.3385,
+      "step": 631
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 8.612612612612613e-05,
+      "loss": 0.2635,
+      "step": 632
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 8.594594594594595e-05,
+      "loss": 0.2559,
+      "step": 633
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 8.576576576576577e-05,
+      "loss": 0.2956,
+      "step": 634
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 8.55855855855856e-05,
+      "loss": 0.3279,
+      "step": 635
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 8.540540540540541e-05,
+      "loss": 0.2796,
+      "step": 636
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 8.522522522522523e-05,
+      "loss": 0.32,
+      "step": 637
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 8.504504504504504e-05,
+      "loss": 0.2425,
+      "step": 638
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 8.486486486486487e-05,
+      "loss": 0.3221,
+      "step": 639
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 8.468468468468469e-05,
+      "loss": 0.3018,
+      "step": 640
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 8.450450450450451e-05,
+      "loss": 0.2249,
+      "step": 641
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 8.432432432432433e-05,
+      "loss": 0.2763,
+      "step": 642
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 8.414414414414414e-05,
+      "loss": 0.2784,
+      "step": 643
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 8.396396396396397e-05,
+      "loss": 0.3659,
+      "step": 644
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 8.378378378378379e-05,
+      "loss": 0.2453,
+      "step": 645
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 8.360360360360362e-05,
+      "loss": 0.3197,
+      "step": 646
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 8.342342342342343e-05,
+      "loss": 0.3055,
+      "step": 647
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 8.324324324324326e-05,
+      "loss": 0.2642,
+      "step": 648
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 8.306306306306306e-05,
+      "loss": 0.2442,
+      "step": 649
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 8.288288288288289e-05,
+      "loss": 0.274,
+      "step": 650
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 8.27027027027027e-05,
+      "loss": 0.2701,
+      "step": 651
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 8.252252252252253e-05,
+      "loss": 0.2802,
+      "step": 652
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 8.234234234234235e-05,
+      "loss": 0.2709,
+      "step": 653
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 8.216216216216217e-05,
+      "loss": 0.3404,
+      "step": 654
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 8.198198198198198e-05,
+      "loss": 0.3093,
+      "step": 655
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 8.18018018018018e-05,
+      "loss": 0.3159,
+      "step": 656
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 8.162162162162162e-05,
+      "loss": 0.3564,
+      "step": 657
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 8.144144144144145e-05,
+      "loss": 0.2453,
+      "step": 658
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 8.126126126126126e-05,
+      "loss": 0.2488,
+      "step": 659
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 8.108108108108109e-05,
+      "loss": 0.3104,
+      "step": 660
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 8.09009009009009e-05,
+      "loss": 0.3627,
+      "step": 661
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 8.072072072072072e-05,
+      "loss": 0.3423,
+      "step": 662
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 8.054054054054055e-05,
+      "loss": 0.2948,
+      "step": 663
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 8.036036036036036e-05,
+      "loss": 0.2747,
+      "step": 664
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 8.018018018018019e-05,
+      "loss": 0.2977,
+      "step": 665
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 8e-05,
+      "loss": 0.2801,
+      "step": 666
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 7.981981981981982e-05,
+      "loss": 0.2422,
+      "step": 667
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 7.963963963963964e-05,
+      "loss": 0.2353,
+      "step": 668
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 7.945945945945946e-05,
+      "loss": 0.2646,
+      "step": 669
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 7.927927927927928e-05,
+      "loss": 0.2443,
+      "step": 670
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 7.909909909909911e-05,
+      "loss": 0.2291,
+      "step": 671
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 7.891891891891892e-05,
+      "loss": 0.2014,
+      "step": 672
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 7.873873873873874e-05,
+      "loss": 0.2264,
+      "step": 673
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 7.855855855855857e-05,
+      "loss": 0.2373,
+      "step": 674
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 7.837837837837838e-05,
+      "loss": 0.2223,
+      "step": 675
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 7.819819819819821e-05,
+      "loss": 0.2465,
+      "step": 676
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 7.801801801801802e-05,
+      "loss": 0.2812,
+      "step": 677
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 7.783783783783785e-05,
+      "loss": 0.2418,
+      "step": 678
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 7.765765765765765e-05,
+      "loss": 0.2626,
+      "step": 679
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 7.747747747747748e-05,
+      "loss": 0.2225,
+      "step": 680
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 7.72972972972973e-05,
+      "loss": 0.2215,
+      "step": 681
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 7.711711711711713e-05,
+      "loss": 0.2634,
+      "step": 682
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 7.693693693693694e-05,
+      "loss": 0.2405,
+      "step": 683
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 7.675675675675677e-05,
+      "loss": 0.2635,
+      "step": 684
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 7.657657657657657e-05,
+      "loss": 0.2163,
+      "step": 685
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 7.63963963963964e-05,
+      "loss": 0.2326,
+      "step": 686
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 7.621621621621621e-05,
+      "loss": 0.2794,
+      "step": 687
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 7.603603603603604e-05,
+      "loss": 0.2338,
+      "step": 688
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 7.585585585585586e-05,
+      "loss": 0.2364,
+      "step": 689
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 7.567567567567568e-05,
+      "loss": 0.2184,
+      "step": 690
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 7.54954954954955e-05,
+      "loss": 0.2297,
+      "step": 691
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 7.531531531531531e-05,
+      "loss": 0.2275,
+      "step": 692
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 7.513513513513514e-05,
+      "loss": 0.2431,
+      "step": 693
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 7.495495495495496e-05,
+      "loss": 0.2014,
+      "step": 694
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 7.477477477477479e-05,
+      "loss": 0.232,
+      "step": 695
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 7.45945945945946e-05,
+      "loss": 0.2394,
+      "step": 696
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 7.441441441441442e-05,
+      "loss": 0.252,
+      "step": 697
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 7.423423423423423e-05,
+      "loss": 0.2705,
+      "step": 698
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 7.405405405405406e-05,
+      "loss": 0.243,
+      "step": 699
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 7.387387387387387e-05,
+      "loss": 0.2024,
+      "step": 700
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 7.36936936936937e-05,
+      "loss": 0.2171,
+      "step": 701
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 7.351351351351352e-05,
+      "loss": 0.2162,
+      "step": 702
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 7.333333333333333e-05,
+      "loss": 0.2465,
+      "step": 703
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 7.315315315315316e-05,
+      "loss": 0.2256,
+      "step": 704
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 7.297297297297297e-05,
+      "loss": 0.2539,
+      "step": 705
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 7.27927927927928e-05,
+      "loss": 0.2265,
+      "step": 706
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 7.261261261261262e-05,
+      "loss": 0.266,
+      "step": 707
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 7.243243243243245e-05,
+      "loss": 0.227,
+      "step": 708
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 7.225225225225225e-05,
+      "loss": 0.2219,
+      "step": 709
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 7.207207207207208e-05,
+      "loss": 0.2071,
+      "step": 710
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 7.189189189189189e-05,
+      "loss": 0.247,
+      "step": 711
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 7.171171171171172e-05,
+      "loss": 0.2306,
+      "step": 712
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 7.153153153153153e-05,
+      "loss": 0.2531,
+      "step": 713
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 7.135135135135136e-05,
+      "loss": 0.2534,
+      "step": 714
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 7.117117117117116e-05,
+      "loss": 0.2102,
+      "step": 715
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 7.099099099099099e-05,
+      "loss": 0.2476,
+      "step": 716
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 7.081081081081081e-05,
+      "loss": 0.2229,
+      "step": 717
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 7.063063063063064e-05,
+      "loss": 0.217,
+      "step": 718
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 7.045045045045045e-05,
+      "loss": 0.2887,
+      "step": 719
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 7.027027027027028e-05,
+      "loss": 0.2373,
+      "step": 720
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 7.00900900900901e-05,
+      "loss": 0.2376,
+      "step": 721
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 6.990990990990991e-05,
+      "loss": 0.2175,
+      "step": 722
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 6.972972972972974e-05,
+      "loss": 0.2444,
+      "step": 723
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 6.954954954954955e-05,
+      "loss": 0.2278,
+      "step": 724
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 6.936936936936938e-05,
+      "loss": 0.2493,
+      "step": 725
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 6.91891891891892e-05,
+      "loss": 0.2598,
+      "step": 726
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 6.900900900900901e-05,
+      "loss": 0.2349,
+      "step": 727
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 6.882882882882882e-05,
+      "loss": 0.1987,
+      "step": 728
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 6.864864864864865e-05,
+      "loss": 0.2043,
+      "step": 729
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 6.846846846846847e-05,
+      "loss": 0.2178,
+      "step": 730
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 6.82882882882883e-05,
+      "loss": 0.2323,
+      "step": 731
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 6.810810810810811e-05,
+      "loss": 0.1981,
+      "step": 732
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 6.792792792792793e-05,
+      "loss": 0.2229,
+      "step": 733
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 6.774774774774775e-05,
+      "loss": 0.2083,
+      "step": 734
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 6.756756756756757e-05,
+      "loss": 0.2394,
+      "step": 735
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 6.73873873873874e-05,
+      "loss": 0.2321,
+      "step": 736
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 6.720720720720721e-05,
+      "loss": 0.2422,
+      "step": 737
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 6.702702702702704e-05,
+      "loss": 0.2542,
+      "step": 738
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 6.684684684684684e-05,
+      "loss": 0.2108,
+      "step": 739
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.2947,
+      "step": 740
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 6.648648648648648e-05,
+      "loss": 0.2397,
+      "step": 741
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 6.630630630630631e-05,
+      "loss": 0.2606,
+      "step": 742
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 6.612612612612613e-05,
+      "loss": 0.2134,
+      "step": 743
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 6.594594594594596e-05,
+      "loss": 0.2437,
+      "step": 744
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 6.576576576576577e-05,
+      "loss": 0.2596,
+      "step": 745
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 6.558558558558559e-05,
+      "loss": 0.3366,
+      "step": 746
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 6.54054054054054e-05,
+      "loss": 0.2668,
+      "step": 747
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 6.522522522522523e-05,
+      "loss": 0.2131,
+      "step": 748
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 6.504504504504504e-05,
+      "loss": 0.2526,
+      "step": 749
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 6.486486486486487e-05,
+      "loss": 0.2533,
+      "step": 750
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 6.468468468468469e-05,
+      "loss": 0.2435,
+      "step": 751
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 6.45045045045045e-05,
+      "loss": 0.2438,
+      "step": 752
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 6.432432432432433e-05,
+      "loss": 0.2876,
+      "step": 753
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 6.414414414414415e-05,
+      "loss": 0.2223,
+      "step": 754
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 6.396396396396397e-05,
+      "loss": 0.2464,
+      "step": 755
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 6.378378378378379e-05,
+      "loss": 0.2879,
+      "step": 756
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 6.360360360360362e-05,
+      "loss": 0.2551,
+      "step": 757
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 6.342342342342342e-05,
+      "loss": 0.2715,
+      "step": 758
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 6.324324324324325e-05,
+      "loss": 0.2768,
+      "step": 759
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 6.306306306306306e-05,
+      "loss": 0.2601,
+      "step": 760
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.288288288288289e-05,
+      "loss": 0.2163,
+      "step": 761
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.27027027027027e-05,
+      "loss": 0.1886,
+      "step": 762
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 6.252252252252253e-05,
+      "loss": 0.2579,
+      "step": 763
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 6.234234234234235e-05,
+      "loss": 0.2352,
+      "step": 764
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.216216216216216e-05,
+      "loss": 0.2469,
+      "step": 765
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 6.198198198198199e-05,
+      "loss": 0.2886,
+      "step": 766
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 6.18018018018018e-05,
+      "loss": 0.2168,
+      "step": 767
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 6.162162162162163e-05,
+      "loss": 0.2607,
+      "step": 768
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 6.144144144144145e-05,
+      "loss": 0.253,
+      "step": 769
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 6.126126126126126e-05,
+      "loss": 0.2079,
+      "step": 770
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 6.108108108108108e-05,
+      "loss": 0.2232,
+      "step": 771
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 6.09009009009009e-05,
+      "loss": 0.2432,
+      "step": 772
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 6.072072072072072e-05,
+      "loss": 0.2366,
+      "step": 773
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 6.0540540540540543e-05,
+      "loss": 0.2189,
+      "step": 774
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 6.0360360360360365e-05,
+      "loss": 0.253,
+      "step": 775
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 6.018018018018018e-05,
+      "loss": 0.2137,
+      "step": 776
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6e-05,
+      "loss": 0.2828,
+      "step": 777
+    },
+    {
+      "epoch": 7.01,
+      "learning_rate": 5.981981981981982e-05,
+      "loss": 0.1999,
+      "step": 778
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 5.9639639639639645e-05,
+      "loss": 0.1985,
+      "step": 779
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 5.9459459459459466e-05,
+      "loss": 0.183,
+      "step": 780
+    },
+    {
+      "epoch": 7.04,
+      "learning_rate": 5.927927927927929e-05,
+      "loss": 0.1728,
+      "step": 781
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 5.9099099099099096e-05,
+      "loss": 0.184,
+      "step": 782
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 5.891891891891892e-05,
+      "loss": 0.2041,
+      "step": 783
+    },
+    {
+      "epoch": 7.06,
+      "learning_rate": 5.873873873873874e-05,
+      "loss": 0.1711,
+      "step": 784
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 5.855855855855856e-05,
+      "loss": 0.1968,
+      "step": 785
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 5.837837837837838e-05,
+      "loss": 0.2036,
+      "step": 786
+    },
+    {
+      "epoch": 7.09,
+      "learning_rate": 5.8198198198198204e-05,
+      "loss": 0.2364,
+      "step": 787
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 5.801801801801802e-05,
+      "loss": 0.2074,
+      "step": 788
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 5.783783783783784e-05,
+      "loss": 0.2597,
+      "step": 789
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 5.765765765765766e-05,
+      "loss": 0.1965,
+      "step": 790
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 5.7477477477477484e-05,
+      "loss": 0.1815,
+      "step": 791
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 5.7297297297297305e-05,
+      "loss": 0.1948,
+      "step": 792
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 5.711711711711713e-05,
+      "loss": 0.2361,
+      "step": 793
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 5.6936936936936935e-05,
+      "loss": 0.2092,
+      "step": 794
+    },
+    {
+      "epoch": 7.16,
+      "learning_rate": 5.6756756756756757e-05,
+      "loss": 0.2038,
+      "step": 795
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 5.657657657657658e-05,
+      "loss": 0.1696,
+      "step": 796
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 5.63963963963964e-05,
+      "loss": 0.1944,
+      "step": 797
+    },
+    {
+      "epoch": 7.19,
+      "learning_rate": 5.621621621621622e-05,
+      "loss": 0.2115,
+      "step": 798
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 5.603603603603604e-05,
+      "loss": 0.1758,
+      "step": 799
+    },
+    {
+      "epoch": 7.21,
+      "learning_rate": 5.585585585585585e-05,
+      "loss": 0.1908,
+      "step": 800
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 5.567567567567567e-05,
+      "loss": 0.2106,
+      "step": 801
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 5.5495495495495494e-05,
+      "loss": 0.2048,
+      "step": 802
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 5.5315315315315316e-05,
+      "loss": 0.192,
+      "step": 803
+    },
+    {
+      "epoch": 7.24,
+      "learning_rate": 5.5135135135135144e-05,
+      "loss": 0.1656,
+      "step": 804
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 5.4954954954954966e-05,
+      "loss": 0.2072,
+      "step": 805
+    },
+    {
+      "epoch": 7.26,
+      "learning_rate": 5.4774774774774774e-05,
+      "loss": 0.1902,
+      "step": 806
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 5.4594594594594595e-05,
+      "loss": 0.2155,
+      "step": 807
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 5.441441441441442e-05,
+      "loss": 0.1983,
+      "step": 808
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 5.423423423423424e-05,
+      "loss": 0.19,
+      "step": 809
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 5.405405405405406e-05,
+      "loss": 0.2128,
+      "step": 810
+    },
+    {
+      "epoch": 7.31,
+      "learning_rate": 5.387387387387388e-05,
+      "loss": 0.1976,
+      "step": 811
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 5.369369369369369e-05,
+      "loss": 0.228,
+      "step": 812
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 5.351351351351351e-05,
+      "loss": 0.2673,
+      "step": 813
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 5.333333333333333e-05,
+      "loss": 0.218,
+      "step": 814
+    },
+    {
+      "epoch": 7.34,
+      "learning_rate": 5.3153153153153155e-05,
+      "loss": 0.1898,
+      "step": 815
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 5.2972972972972976e-05,
+      "loss": 0.1921,
+      "step": 816
+    },
+    {
+      "epoch": 7.36,
+      "learning_rate": 5.27927927927928e-05,
+      "loss": 0.1932,
+      "step": 817
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 5.261261261261261e-05,
+      "loss": 0.2193,
+      "step": 818
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 5.2432432432432434e-05,
+      "loss": 0.1821,
+      "step": 819
+    },
+    {
+      "epoch": 7.39,
+      "learning_rate": 5.2252252252252256e-05,
+      "loss": 0.2112,
+      "step": 820
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 5.207207207207208e-05,
+      "loss": 0.2047,
+      "step": 821
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 5.18918918918919e-05,
+      "loss": 0.1936,
+      "step": 822
+    },
+    {
+      "epoch": 7.41,
+      "learning_rate": 5.171171171171172e-05,
+      "loss": 0.2419,
+      "step": 823
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 5.153153153153153e-05,
+      "loss": 0.1825,
+      "step": 824
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 5.135135135135135e-05,
+      "loss": 0.2011,
+      "step": 825
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 5.117117117117117e-05,
+      "loss": 0.2126,
+      "step": 826
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 5.0990990990990994e-05,
+      "loss": 0.2394,
+      "step": 827
+    },
+    {
+      "epoch": 7.46,
+      "learning_rate": 5.0810810810810815e-05,
+      "loss": 0.2179,
+      "step": 828
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 5.063063063063064e-05,
+      "loss": 0.232,
+      "step": 829
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 5.0450450450450445e-05,
+      "loss": 0.2246,
+      "step": 830
+    },
+    {
+      "epoch": 7.49,
+      "learning_rate": 5.0270270270270267e-05,
+      "loss": 0.1931,
+      "step": 831
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 5.009009009009009e-05,
+      "loss": 0.1878,
+      "step": 832
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 4.9909909909909917e-05,
+      "loss": 0.2133,
+      "step": 833
+    },
+    {
+      "epoch": 7.51,
+      "learning_rate": 4.972972972972974e-05,
+      "loss": 0.1934,
+      "step": 834
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 4.954954954954955e-05,
+      "loss": 0.2257,
+      "step": 835
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 4.9369369369369375e-05,
+      "loss": 0.1962,
+      "step": 836
+    },
+    {
+      "epoch": 7.54,
+      "learning_rate": 4.9189189189189196e-05,
+      "loss": 0.1722,
+      "step": 837
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 4.900900900900901e-05,
+      "loss": 0.1857,
+      "step": 838
+    },
+    {
+      "epoch": 7.56,
+      "learning_rate": 4.882882882882883e-05,
+      "loss": 0.2043,
+      "step": 839
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 4.8648648648648654e-05,
+      "loss": 0.1977,
+      "step": 840
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 4.846846846846847e-05,
+      "loss": 0.2503,
+      "step": 841
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 4.828828828828829e-05,
+      "loss": 0.2246,
+      "step": 842
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 4.810810810810811e-05,
+      "loss": 0.2141,
+      "step": 843
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 4.792792792792793e-05,
+      "loss": 0.2114,
+      "step": 844
+    },
+    {
+      "epoch": 7.61,
+      "learning_rate": 4.774774774774775e-05,
+      "loss": 0.207,
+      "step": 845
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 4.756756756756757e-05,
+      "loss": 0.2045,
+      "step": 846
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 4.738738738738739e-05,
+      "loss": 0.1935,
+      "step": 847
+    },
+    {
+      "epoch": 7.64,
+      "learning_rate": 4.7207207207207214e-05,
+      "loss": 0.2299,
+      "step": 848
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 4.7027027027027035e-05,
+      "loss": 0.2157,
+      "step": 849
+    },
+    {
+      "epoch": 7.66,
+      "learning_rate": 4.684684684684685e-05,
+      "loss": 0.2246,
+      "step": 850
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.1902,
+      "step": 851
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 4.648648648648649e-05,
+      "loss": 0.2111,
+      "step": 852
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 4.630630630630631e-05,
+      "loss": 0.198,
+      "step": 853
+    },
+    {
+      "epoch": 7.69,
+      "learning_rate": 4.612612612612613e-05,
+      "loss": 0.2126,
+      "step": 854
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 4.594594594594595e-05,
+      "loss": 0.2093,
+      "step": 855
+    },
+    {
+      "epoch": 7.71,
+      "learning_rate": 4.5765765765765766e-05,
+      "loss": 0.2755,
+      "step": 856
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 4.558558558558559e-05,
+      "loss": 0.2028,
+      "step": 857
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 4.540540540540541e-05,
+      "loss": 0.2026,
+      "step": 858
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 4.5225225225225224e-05,
+      "loss": 0.1991,
+      "step": 859
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 4.5045045045045046e-05,
+      "loss": 0.2005,
+      "step": 860
+    },
+    {
+      "epoch": 7.76,
+      "learning_rate": 4.486486486486487e-05,
+      "loss": 0.1966,
+      "step": 861
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 4.468468468468469e-05,
+      "loss": 0.2325,
+      "step": 862
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 4.450450450450451e-05,
+      "loss": 0.1986,
+      "step": 863
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 4.432432432432433e-05,
+      "loss": 0.2288,
+      "step": 864
+    },
+    {
+      "epoch": 7.79,
+      "learning_rate": 4.414414414414415e-05,
+      "loss": 0.1883,
+      "step": 865
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 4.396396396396397e-05,
+      "loss": 0.2734,
+      "step": 866
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 4.378378378378379e-05,
+      "loss": 0.2244,
+      "step": 867
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 4.3603603603603605e-05,
+      "loss": 0.2134,
+      "step": 868
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 4.342342342342343e-05,
+      "loss": 0.2495,
+      "step": 869
+    },
+    {
+      "epoch": 7.84,
+      "learning_rate": 4.324324324324325e-05,
+      "loss": 0.2268,
+      "step": 870
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 4.306306306306306e-05,
+      "loss": 0.2232,
+      "step": 871
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 4.2882882882882885e-05,
+      "loss": 0.2142,
+      "step": 872
+    },
+    {
+      "epoch": 7.86,
+      "learning_rate": 4.2702702702702706e-05,
+      "loss": 0.1927,
+      "step": 873
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 4.252252252252252e-05,
+      "loss": 0.2356,
+      "step": 874
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 4.234234234234234e-05,
+      "loss": 0.1987,
+      "step": 875
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 4.2162162162162164e-05,
+      "loss": 0.203,
+      "step": 876
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 4.1981981981981986e-05,
+      "loss": 0.2034,
+      "step": 877
+    },
+    {
+      "epoch": 7.91,
+      "learning_rate": 4.180180180180181e-05,
+      "loss": 0.221,
+      "step": 878
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 4.162162162162163e-05,
+      "loss": 0.1952,
+      "step": 879
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 4.1441441441441444e-05,
+      "loss": 0.1798,
+      "step": 880
+    },
+    {
+      "epoch": 7.94,
+      "learning_rate": 4.1261261261261266e-05,
+      "loss": 0.2057,
+      "step": 881
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 4.108108108108109e-05,
+      "loss": 0.2084,
+      "step": 882
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 4.09009009009009e-05,
+      "loss": 0.2067,
+      "step": 883
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 4.0720720720720724e-05,
+      "loss": 0.2017,
+      "step": 884
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 0.1786,
+      "step": 885
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 4.036036036036036e-05,
+      "loss": 0.2265,
+      "step": 886
+    },
+    {
+      "epoch": 7.99,
+      "learning_rate": 4.018018018018018e-05,
+      "loss": 0.2221,
+      "step": 887
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4e-05,
+      "loss": 0.1592,
+      "step": 888
+    },
+    {
+      "epoch": 8.01,
+      "learning_rate": 3.981981981981982e-05,
+      "loss": 0.1533,
+      "step": 889
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 3.963963963963964e-05,
+      "loss": 0.201,
+      "step": 890
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 3.945945945945946e-05,
+      "loss": 0.2034,
+      "step": 891
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 3.927927927927928e-05,
+      "loss": 0.1815,
+      "step": 892
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 3.9099099099099105e-05,
+      "loss": 0.1729,
+      "step": 893
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 3.8918918918918926e-05,
+      "loss": 0.203,
+      "step": 894
+    },
+    {
+      "epoch": 8.06,
+      "learning_rate": 3.873873873873874e-05,
+      "loss": 0.1611,
+      "step": 895
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 3.855855855855856e-05,
+      "loss": 0.2099,
+      "step": 896
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 3.8378378378378384e-05,
+      "loss": 0.1435,
+      "step": 897
+    },
+    {
+      "epoch": 8.09,
+      "learning_rate": 3.81981981981982e-05,
+      "loss": 0.1857,
+      "step": 898
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 3.801801801801802e-05,
+      "loss": 0.2327,
+      "step": 899
+    },
+    {
+      "epoch": 8.11,
+      "learning_rate": 3.783783783783784e-05,
+      "loss": 0.1875,
+      "step": 900
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 3.765765765765766e-05,
+      "loss": 0.1824,
+      "step": 901
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 3.747747747747748e-05,
+      "loss": 0.1671,
+      "step": 902
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 3.72972972972973e-05,
+      "loss": 0.157,
+      "step": 903
+    },
+    {
+      "epoch": 8.14,
+      "learning_rate": 3.7117117117117115e-05,
+      "loss": 0.2169,
+      "step": 904
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 3.693693693693694e-05,
+      "loss": 0.1918,
+      "step": 905
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 3.675675675675676e-05,
+      "loss": 0.2088,
+      "step": 906
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 3.657657657657658e-05,
+      "loss": 0.1751,
+      "step": 907
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 3.63963963963964e-05,
+      "loss": 0.2023,
+      "step": 908
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 3.621621621621622e-05,
+      "loss": 0.1897,
+      "step": 909
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 3.603603603603604e-05,
+      "loss": 0.1792,
+      "step": 910
+    },
+    {
+      "epoch": 8.21,
+      "learning_rate": 3.585585585585586e-05,
+      "loss": 0.177,
+      "step": 911
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 3.567567567567568e-05,
+      "loss": 0.1938,
+      "step": 912
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 3.5495495495495496e-05,
+      "loss": 0.1987,
+      "step": 913
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 3.531531531531532e-05,
+      "loss": 0.1757,
+      "step": 914
+    },
+    {
+      "epoch": 8.24,
+      "learning_rate": 3.513513513513514e-05,
+      "loss": 0.2028,
+      "step": 915
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 3.4954954954954954e-05,
+      "loss": 0.1704,
+      "step": 916
+    },
+    {
+      "epoch": 8.26,
+      "learning_rate": 3.4774774774774776e-05,
+      "loss": 0.1984,
+      "step": 917
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 3.45945945945946e-05,
+      "loss": 0.2117,
+      "step": 918
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 3.441441441441441e-05,
+      "loss": 0.1817,
+      "step": 919
+    },
+    {
+      "epoch": 8.29,
+      "learning_rate": 3.4234234234234234e-05,
+      "loss": 0.2177,
+      "step": 920
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 3.4054054054054055e-05,
+      "loss": 0.1737,
+      "step": 921
+    },
+    {
+      "epoch": 8.31,
+      "learning_rate": 3.387387387387388e-05,
+      "loss": 0.1605,
+      "step": 922
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 3.36936936936937e-05,
+      "loss": 0.1568,
+      "step": 923
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 3.351351351351352e-05,
+      "loss": 0.1898,
+      "step": 924
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.1615,
+      "step": 925
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 3.3153153153153157e-05,
+      "loss": 0.1867,
+      "step": 926
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 3.297297297297298e-05,
+      "loss": 0.1879,
+      "step": 927
+    },
+    {
+      "epoch": 8.36,
+      "learning_rate": 3.279279279279279e-05,
+      "loss": 0.1822,
+      "step": 928
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 3.2612612612612615e-05,
+      "loss": 0.1782,
+      "step": 929
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 3.2432432432432436e-05,
+      "loss": 0.1718,
+      "step": 930
+    },
+    {
+      "epoch": 8.39,
+      "learning_rate": 3.225225225225225e-05,
+      "loss": 0.1944,
+      "step": 931
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 3.207207207207207e-05,
+      "loss": 0.2228,
+      "step": 932
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 3.1891891891891894e-05,
+      "loss": 0.1929,
+      "step": 933
+    },
+    {
+      "epoch": 8.41,
+      "learning_rate": 3.171171171171171e-05,
+      "loss": 0.1967,
+      "step": 934
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 3.153153153153153e-05,
+      "loss": 0.1918,
+      "step": 935
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 3.135135135135135e-05,
+      "loss": 0.1658,
+      "step": 936
+    },
+    {
+      "epoch": 8.44,
+      "learning_rate": 3.1171171171171174e-05,
+      "loss": 0.1878,
+      "step": 937
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 3.0990990990990995e-05,
+      "loss": 0.1712,
+      "step": 938
+    },
+    {
+      "epoch": 8.46,
+      "learning_rate": 3.081081081081082e-05,
+      "loss": 0.1775,
+      "step": 939
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 3.063063063063063e-05,
+      "loss": 0.1869,
+      "step": 940
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 3.045045045045045e-05,
+      "loss": 0.1943,
+      "step": 941
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 3.0270270270270272e-05,
+      "loss": 0.1863,
+      "step": 942
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 3.009009009009009e-05,
+      "loss": 0.1746,
+      "step": 943
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 2.990990990990991e-05,
+      "loss": 0.175,
+      "step": 944
+    },
+    {
+      "epoch": 8.51,
+      "learning_rate": 2.9729729729729733e-05,
+      "loss": 0.2085,
+      "step": 945
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 2.9549549549549548e-05,
+      "loss": 0.1638,
+      "step": 946
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 2.936936936936937e-05,
+      "loss": 0.2147,
+      "step": 947
+    },
+    {
+      "epoch": 8.54,
+      "learning_rate": 2.918918918918919e-05,
+      "loss": 0.1774,
+      "step": 948
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 2.900900900900901e-05,
+      "loss": 0.179,
+      "step": 949
+    },
+    {
+      "epoch": 8.56,
+      "learning_rate": 2.882882882882883e-05,
+      "loss": 0.1927,
+      "step": 950
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 2.8648648648648653e-05,
+      "loss": 0.1809,
+      "step": 951
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 2.8468468468468467e-05,
+      "loss": 0.168,
+      "step": 952
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 2.828828828828829e-05,
+      "loss": 0.1853,
+      "step": 953
+    },
+    {
+      "epoch": 8.59,
+      "learning_rate": 2.810810810810811e-05,
+      "loss": 0.2057,
+      "step": 954
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 2.7927927927927926e-05,
+      "loss": 0.1757,
+      "step": 955
+    },
+    {
+      "epoch": 8.61,
+      "learning_rate": 2.7747747747747747e-05,
+      "loss": 0.1706,
+      "step": 956
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 2.7567567567567572e-05,
+      "loss": 0.1854,
+      "step": 957
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 2.7387387387387387e-05,
+      "loss": 0.1873,
+      "step": 958
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 2.720720720720721e-05,
+      "loss": 0.1783,
+      "step": 959
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 0.1709,
+      "step": 960
+    },
+    {
+      "epoch": 8.66,
+      "learning_rate": 2.6846846846846845e-05,
+      "loss": 0.1766,
+      "step": 961
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.21,
+      "step": 962
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 2.6486486486486488e-05,
+      "loss": 0.196,
+      "step": 963
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 2.6306306306306306e-05,
+      "loss": 0.2047,
+      "step": 964
+    },
+    {
+      "epoch": 8.69,
+      "learning_rate": 2.6126126126126128e-05,
+      "loss": 0.1878,
+      "step": 965
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 2.594594594594595e-05,
+      "loss": 0.1713,
+      "step": 966
+    },
+    {
+      "epoch": 8.71,
+      "learning_rate": 2.5765765765765764e-05,
+      "loss": 0.169,
+      "step": 967
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 2.5585585585585586e-05,
+      "loss": 0.1749,
+      "step": 968
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 2.5405405405405408e-05,
+      "loss": 0.1968,
+      "step": 969
+    },
+    {
+      "epoch": 8.74,
+      "learning_rate": 2.5225225225225222e-05,
+      "loss": 0.173,
+      "step": 970
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 2.5045045045045044e-05,
+      "loss": 0.208,
+      "step": 971
+    },
+    {
+      "epoch": 8.76,
+      "learning_rate": 2.486486486486487e-05,
+      "loss": 0.1588,
+      "step": 972
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 2.4684684684684687e-05,
+      "loss": 0.1704,
+      "step": 973
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 2.4504504504504506e-05,
+      "loss": 0.1907,
+      "step": 974
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 2.4324324324324327e-05,
+      "loss": 0.1939,
+      "step": 975
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 2.4144144144144145e-05,
+      "loss": 0.1676,
+      "step": 976
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 2.3963963963963964e-05,
+      "loss": 0.2301,
+      "step": 977
+    },
+    {
+      "epoch": 8.81,
+      "learning_rate": 2.3783783783783785e-05,
+      "loss": 0.1836,
+      "step": 978
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 2.3603603603603607e-05,
+      "loss": 0.1972,
+      "step": 979
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 2.3423423423423425e-05,
+      "loss": 0.1865,
+      "step": 980
+    },
+    {
+      "epoch": 8.84,
+      "learning_rate": 2.3243243243243247e-05,
+      "loss": 0.168,
+      "step": 981
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 2.3063063063063065e-05,
+      "loss": 0.1778,
+      "step": 982
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 2.2882882882882883e-05,
+      "loss": 0.1863,
+      "step": 983
+    },
+    {
+      "epoch": 8.86,
+      "learning_rate": 2.2702702702702705e-05,
+      "loss": 0.1549,
+      "step": 984
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 2.2522522522522523e-05,
+      "loss": 0.1869,
+      "step": 985
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 2.2342342342342344e-05,
+      "loss": 0.2235,
+      "step": 986
+    },
+    {
+      "epoch": 8.89,
+      "learning_rate": 2.2162162162162166e-05,
+      "loss": 0.172,
+      "step": 987
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 2.1981981981981984e-05,
+      "loss": 0.2134,
+      "step": 988
+    },
+    {
+      "epoch": 8.91,
+      "learning_rate": 2.1801801801801803e-05,
+      "loss": 0.1754,
+      "step": 989
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 2.1621621621621624e-05,
+      "loss": 0.1436,
+      "step": 990
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 2.1441441441441442e-05,
+      "loss": 0.1933,
+      "step": 991
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 2.126126126126126e-05,
+      "loss": 0.2173,
+      "step": 992
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 2.1081081081081082e-05,
+      "loss": 0.1751,
+      "step": 993
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 2.0900900900900904e-05,
+      "loss": 0.177,
+      "step": 994
+    },
+    {
+      "epoch": 8.96,
+      "learning_rate": 2.0720720720720722e-05,
+      "loss": 0.1967,
+      "step": 995
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 2.0540540540540544e-05,
+      "loss": 0.1582,
+      "step": 996
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 2.0360360360360362e-05,
+      "loss": 0.1755,
+      "step": 997
+    },
+    {
+      "epoch": 8.99,
+      "learning_rate": 2.018018018018018e-05,
+      "loss": 0.1799,
+      "step": 998
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2e-05,
+      "loss": 0.2062,
+      "step": 999
+    },
+    {
+      "epoch": 9.01,
+      "learning_rate": 1.981981981981982e-05,
+      "loss": 0.1766,
+      "step": 1000
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 1.963963963963964e-05,
+      "loss": 0.1705,
+      "step": 1001
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 1.9459459459459463e-05,
+      "loss": 0.1559,
+      "step": 1002
+    },
+    {
+      "epoch": 9.04,
+      "learning_rate": 1.927927927927928e-05,
+      "loss": 0.1595,
+      "step": 1003
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 1.90990990990991e-05,
+      "loss": 0.1711,
+      "step": 1004
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 0.1782,
+      "step": 1005
+    },
+    {
+      "epoch": 9.06,
+      "learning_rate": 1.873873873873874e-05,
+      "loss": 0.1841,
+      "step": 1006
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 1.8558558558558558e-05,
+      "loss": 0.1725,
+      "step": 1007
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 1.837837837837838e-05,
+      "loss": 0.1974,
+      "step": 1008
+    },
+    {
+      "epoch": 9.09,
+      "learning_rate": 1.81981981981982e-05,
+      "loss": 0.1765,
+      "step": 1009
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 1.801801801801802e-05,
+      "loss": 0.1505,
+      "step": 1010
+    },
+    {
+      "epoch": 9.11,
+      "learning_rate": 1.783783783783784e-05,
+      "loss": 0.1642,
+      "step": 1011
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 1.765765765765766e-05,
+      "loss": 0.1547,
+      "step": 1012
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 1.7477477477477477e-05,
+      "loss": 0.163,
+      "step": 1013
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 1.72972972972973e-05,
+      "loss": 0.1567,
+      "step": 1014
+    },
+    {
+      "epoch": 9.14,
+      "learning_rate": 1.7117117117117117e-05,
+      "loss": 0.1563,
+      "step": 1015
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 1.693693693693694e-05,
+      "loss": 0.1349,
+      "step": 1016
+    },
+    {
+      "epoch": 9.16,
+      "learning_rate": 1.675675675675676e-05,
+      "loss": 0.1765,
+      "step": 1017
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 1.6576576576576578e-05,
+      "loss": 0.1638,
+      "step": 1018
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 1.6396396396396396e-05,
+      "loss": 0.16,
+      "step": 1019
+    },
+    {
+      "epoch": 9.19,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 0.2045,
+      "step": 1020
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 1.6036036036036036e-05,
+      "loss": 0.1729,
+      "step": 1021
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 1.5855855855855855e-05,
+      "loss": 0.1687,
+      "step": 1022
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 1.5675675675675676e-05,
+      "loss": 0.1639,
+      "step": 1023
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 1.5495495495495498e-05,
+      "loss": 0.1709,
+      "step": 1024
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 1.5315315315315316e-05,
+      "loss": 0.1806,
+      "step": 1025
+    },
+    {
+      "epoch": 9.24,
+      "learning_rate": 1.5135135135135136e-05,
+      "loss": 0.192,
+      "step": 1026
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 1.4954954954954956e-05,
+      "loss": 0.1636,
+      "step": 1027
+    },
+    {
+      "epoch": 9.26,
+      "learning_rate": 1.4774774774774774e-05,
+      "loss": 0.1797,
+      "step": 1028
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 1.4594594594594596e-05,
+      "loss": 0.1976,
+      "step": 1029
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 1.4414414414414416e-05,
+      "loss": 0.1657,
+      "step": 1030
+    },
+    {
+      "epoch": 9.29,
+      "learning_rate": 1.4234234234234234e-05,
+      "loss": 0.178,
+      "step": 1031
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 1.4054054054054055e-05,
+      "loss": 0.1709,
+      "step": 1032
+    },
+    {
+      "epoch": 9.31,
+      "learning_rate": 1.3873873873873874e-05,
+      "loss": 0.1604,
+      "step": 1033
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 1.3693693693693693e-05,
+      "loss": 0.1804,
+      "step": 1034
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 1.3513513513513515e-05,
+      "loss": 0.1661,
+      "step": 1035
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.1542,
+      "step": 1036
+    },
+    {
+      "epoch": 9.34,
+      "learning_rate": 1.3153153153153153e-05,
+      "loss": 0.1807,
+      "step": 1037
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 1.2972972972972975e-05,
+      "loss": 0.1976,
+      "step": 1038
+    },
+    {
+      "epoch": 9.36,
+      "learning_rate": 1.2792792792792793e-05,
+      "loss": 0.162,
+      "step": 1039
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 1.2612612612612611e-05,
+      "loss": 0.173,
+      "step": 1040
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 1.2432432432432435e-05,
+      "loss": 0.1679,
+      "step": 1041
+    },
+    {
+      "epoch": 9.39,
+      "learning_rate": 1.2252252252252253e-05,
+      "loss": 0.1732,
+      "step": 1042
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 1.2072072072072073e-05,
+      "loss": 0.1774,
+      "step": 1043
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 1.1891891891891893e-05,
+      "loss": 0.1579,
+      "step": 1044
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 1.1711711711711713e-05,
+      "loss": 0.1631,
+      "step": 1045
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 1.1531531531531532e-05,
+      "loss": 0.1698,
+      "step": 1046
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 1.1351351351351352e-05,
+      "loss": 0.1702,
+      "step": 1047
+    },
+    {
+      "epoch": 9.44,
+      "learning_rate": 1.1171171171171172e-05,
+      "loss": 0.1565,
+      "step": 1048
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 1.0990990990990992e-05,
+      "loss": 0.1443,
+      "step": 1049
+    },
+    {
+      "epoch": 9.46,
+      "learning_rate": 1.0810810810810812e-05,
+      "loss": 0.1805,
+      "step": 1050
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 1.063063063063063e-05,
+      "loss": 0.1462,
+      "step": 1051
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 1.0450450450450452e-05,
+      "loss": 0.1723,
+      "step": 1052
+    },
+    {
+      "epoch": 9.49,
+      "learning_rate": 1.0270270270270272e-05,
+      "loss": 0.1684,
+      "step": 1053
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 1.009009009009009e-05,
+      "loss": 0.1866,
+      "step": 1054
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 9.90990990990991e-06,
+      "loss": 0.1562,
+      "step": 1055
+    },
+    {
+      "epoch": 9.51,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 0.1614,
+      "step": 1056
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 9.54954954954955e-06,
+      "loss": 0.1592,
+      "step": 1057
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 9.36936936936937e-06,
+      "loss": 0.1659,
+      "step": 1058
+    },
+    {
+      "epoch": 9.54,
+      "learning_rate": 9.18918918918919e-06,
+      "loss": 0.1632,
+      "step": 1059
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 0.1697,
+      "step": 1060
+    },
+    {
+      "epoch": 9.56,
+      "learning_rate": 8.82882882882883e-06,
+      "loss": 0.1893,
+      "step": 1061
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 0.1768,
+      "step": 1062
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 8.46846846846847e-06,
+      "loss": 0.1828,
+      "step": 1063
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 8.288288288288289e-06,
+      "loss": 0.1687,
+      "step": 1064
+    },
+    {
+      "epoch": 9.59,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 0.1662,
+      "step": 1065
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 7.927927927927927e-06,
+      "loss": 0.1681,
+      "step": 1066
+    },
+    {
+      "epoch": 9.61,
+      "learning_rate": 7.747747747747749e-06,
+      "loss": 0.1521,
+      "step": 1067
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 7.567567567567568e-06,
+      "loss": 0.158,
+      "step": 1068
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 7.387387387387387e-06,
+      "loss": 0.1679,
+      "step": 1069
+    },
+    {
+      "epoch": 9.64,
+      "learning_rate": 7.207207207207208e-06,
+      "loss": 0.158,
+      "step": 1070
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.1688,
+      "step": 1071
+    },
+    {
+      "epoch": 9.66,
+      "learning_rate": 6.846846846846847e-06,
+      "loss": 0.1714,
+      "step": 1072
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.1693,
+      "step": 1073
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 0.1465,
+      "step": 1074
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 6.306306306306306e-06,
+      "loss": 0.1723,
+      "step": 1075
+    },
+    {
+      "epoch": 9.69,
+      "learning_rate": 6.126126126126126e-06,
+      "loss": 0.1704,
+      "step": 1076
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 5.945945945945946e-06,
+      "loss": 0.143,
+      "step": 1077
+    },
+    {
+      "epoch": 9.71,
+      "learning_rate": 5.765765765765766e-06,
+      "loss": 0.1696,
+      "step": 1078
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 5.585585585585586e-06,
+      "loss": 0.1596,
+      "step": 1079
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.2013,
+      "step": 1080
+    },
+    {
+      "epoch": 9.74,
+      "learning_rate": 5.225225225225226e-06,
+      "loss": 0.191,
+      "step": 1081
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 5.045045045045045e-06,
+      "loss": 0.177,
+      "step": 1082
+    },
+    {
+      "epoch": 9.76,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.1849,
+      "step": 1083
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 4.684684684684685e-06,
+      "loss": 0.1549,
+      "step": 1084
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 0.1875,
+      "step": 1085
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 0.1544,
+      "step": 1086
+    },
+    {
+      "epoch": 9.79,
+      "learning_rate": 4.1441441441441446e-06,
+      "loss": 0.1816,
+      "step": 1087
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 3.963963963963964e-06,
+      "loss": 0.1707,
+      "step": 1088
+    },
+    {
+      "epoch": 9.81,
+      "learning_rate": 3.783783783783784e-06,
+      "loss": 0.1575,
+      "step": 1089
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 0.1607,
+      "step": 1090
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 3.4234234234234234e-06,
+      "loss": 0.1771,
+      "step": 1091
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.1688,
+      "step": 1092
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 3.063063063063063e-06,
+      "loss": 0.1689,
+      "step": 1093
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 2.882882882882883e-06,
+      "loss": 0.1711,
+      "step": 1094
+    },
+    {
+      "epoch": 9.86,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.1696,
+      "step": 1095
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 2.5225225225225225e-06,
+      "loss": 0.1863,
+      "step": 1096
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 2.3423423423423424e-06,
+      "loss": 0.1648,
+      "step": 1097
+    },
+    {
+      "epoch": 9.89,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.1883,
+      "step": 1098
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 1.981981981981982e-06,
+      "loss": 0.1593,
+      "step": 1099
+    },
+    {
+      "epoch": 9.91,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 0.1602,
+      "step": 1100
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.2187,
+      "step": 1101
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 1.4414414414414416e-06,
+      "loss": 0.1863,
+      "step": 1102
+    },
+    {
+      "epoch": 9.94,
+      "learning_rate": 1.2612612612612613e-06,
+      "loss": 0.1656,
+      "step": 1103
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.1939,
+      "step": 1104
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 0.1407,
+      "step": 1105
+    },
+    {
+      "epoch": 9.96,
+      "learning_rate": 7.207207207207208e-07,
+      "loss": 0.1753,
+      "step": 1106
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.1542,
+      "step": 1107
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 3.603603603603604e-07,
+      "loss": 0.1609,
+      "step": 1108
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 1.801801801801802e-07,
+      "loss": 0.1702,
+      "step": 1109
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 0.1462,
+      "step": 1110
+    },
+    {
+      "epoch": 10.0,
+      "step": 1110,
+      "total_flos": 6398509737984000.0,
+      "train_loss": 0.5453134933838973,
+      "train_runtime": 627.0712,
+      "train_samples_per_second": 56.246,
+      "train_steps_per_second": 1.77
     }
   ],
-  "max_steps": 333,
-  "num_train_epochs": 3,
-  "total_flos": 1915204220928000.0,
+  "max_steps": 1110,
+  "num_train_epochs": 10,
+  "total_flos": 6398509737984000.0,
   "trial_name": null,
   "trial_params": null
 }